Aarhus School of Business, Aarhus University PhD Thesis PER

Aarhus School of Business, Aarhus University
PhD Thesis
ENTERPRISE SYSTEMS AND INSTITUTIONS
Theorizing about Enterprise Systems in Organizations Using Institutional Theory – A Case Study Approach
PER SVEJVIG – 2010
The thesis has, on 2nd June 2010 been accepted for defense
PREFACE
I have about 25 years’ experience from practice where I mainly worked as a project
manager and managed projects of varying size, the vast majority of which were IS/ITrelated. Through this work I have gradually learned that managing change and social
issues is crucial for the successful implementation of organizational initiatives. A focus
on managing change thus increasingly became a central part of my project management
approach, whereby I empirically sought to find strategies and models that could address
social and organizational issues. In a very large SAP/R3 enterprise systems project in
the late 1990s, with more than 100 participants, numerous deficiencies were revealed in
our approach to managing change, and there appeared to be a profound need for theoretical approaches and models to support this process.
That was the beginning of an intense literature search on the subject in a self-study,
which was later replaced by an academic study with which I achieved a master’s degree
in IT and organizations. The master’s study was part-time and during my journey I
worked with models targeting social considerations, as part of my consultancy work.
This gave me solid practical experiences and the opportunity to combine theory and
practice. Anyway, I was at a crossroads in 2007: should I return to being a full-time
project manager or continue my theoretical journey? I chose the latter to satisfy my curiosity and gain a deeper understanding of the issues related to managing change by the
implementation of information systems.
This was the outset of my PhD journey, during which I initially focused on implementation and post-implementation issues but, after several diversions, I ended up applying
institutional theory to the management, implementation and use of enterprise systems
from a larger perspective. I have realized that managing change forms pieces of a much
broader puzzle with many other elements like institutional structures and processes. My
focus and understanding have thus changed considerably during my PhD journey. The
journey has been like a mental marathon with many intellectual challenges substantiated
in this PhD thesis, which I hope you find interesting.
ii
ACKNOWLEDGEMENTS
Conducting a PhD study is far from an easy task and is not carried out in isolation. It
requires much support and help from many people and organizations.
I have been very lucky to have two interested, creative and knowledgeable supervisors
for my last two years as a PhD student. Andrea Carugati, your helpfulness, sincerity and
dedication have been very important “to keep the project going” and our academic discussions have served as fruitful inspiration for my research. Our relationship has turned
into a partnership, and I really appreciate your human-to-human understanding. You
said some time ago “I have learned that to be a supervisor is in fact to be a sparring
partner,” and you have exceedingly fulfilled that role. Jan Pries-Heje, we have known
each other for many years, and you are the man focusing on “getting the job done” – we
truly share this sense of urgency to create lasting results! I have enjoyed working with
you and learning from your long academic experience, whether in paper writing, applying for research grants or advancing in the academic world. Your well-chosen advice
has been highly valuable in helping me steer my research. Pernille Kræmmergaard, you
were the one to push me into this PhD study, and were my first-year supervisor. You are
the driver and the entrepreneur challenging and developing people’s intellectual capital
and I am very grateful to you for pushing me into the academic sphere. Tina Blegind
Jensen, you are the person, beyond my supervisors, who has had the most influence on
my research, and you have served as an excellent “extra supervisor.” You are always
ready to help by giving good advice and taking part in constructive dialogues, which I
am very thankful for. It has furthermore been a pleasure to write three papers with you
and our productive collaboration appears “to make good sense.”
I will express my sincere gratitude to the people at SCANDI (a pseudonym for the case
settings in this study) for their openness and willingness to incorporate me into their
business activities: first, the initial contact person, the project manager of the project I
followed for two years, and then the consultant taking over the job from her; second, the
many project participants, users and managers involved in the project including suppliers with their consultants; finally, accounts clerks, purchasers and managers from logis-
iii
tic and finance departments and beyond. This has been very important help and is a
foundation stone for this research.
My early research, after the first year of study, was evaluated by Geoff Walsham and
Morten Thannig Vendelø. Your timely comments were very much to the point, and I
have truly aimed to follow your advice. I highly appreciate your help.
Colleagues from the IS research group at our Department of Management should also
be thanked for giving moral support and saving me from teaching and other obligations
in the second part of my PhD study.
Producing the final result, this PhD thesis, is a master challenge and I am grateful for
the people who found time to support me in this closing process: Christian Koch (chapters 1, 2 and 3, but especially chapter 2 about enterprise systems), Tina Blegind Jensen
(chapter 5), Anna Holm, Jan Pries-Heje and Andrea Carugati (the entire PhD thesis).
I would also like to thank the committee for taking an interest in my thesis.
Special thanks to my adult children, Birgitte and Morten, family and friends for supporting me and accepting my opting out of some social activities and obligations during this
study.
Last, but not least, my dear wife Jytte for tremendous patience and faithful support in
my PhD study. Your attitude was in the end what made this study possible!
Per Svejvig, March 2010.
iv
SAMMENDRAG PÅ DANSK
Virksomhedssystemer har vundet mere og mere indpas i både den private og den offentlige sektor i løbet af det seneste årti. De har været på markedet siden begyndelsen af
halvfemserne, som en løsning til den stigende tendens til globalisering, fusioner og opkøb, og til at optimere og effektivisere virksomheder. I dag er det sådan, at stort set ingen virksomheder kan leve uden disse systemer, hvor de må betegnes som en de facto
standard.
Virksomhedssystemer (også betegnet standardsystemer) kan defineres som store organisatoriske systemer, baseret på software-pakkeløsninger, der gør det muligt for organisationen at automatisere og integrere en omfattende del af forretningsprocesserne, og understøtte online, integreret behandling af data på tværs af hele organisationen. Softwarepakkeløsningerne leveres som generiske "halvfabrikata" fra leverandører som SAP og
Oracle, til virksomheder der kan tilpasse systemerne, til deres egne krav og behov, vha.
opsætninger og tilretninger. På trods af muligheder for tilpasninger, så påtvinger virksomhedssystemerne sin egen logik på virksomhedens strategi, kultur og organisation og det er meget usandsynligt, at der er en perfekt pasform mellem systemet og organisationen.
Implementering af virksomhedssystemer er ofte kompleks på grund af at det indebærer:
(1) integration og standardisering af data på tværs af hele organisationen, (2) indførelse
af "best practice" forretningsprocesser, som organisationen skal tilpasse sig til, (3) overholdelse af stramme tidsplaner og (4) deltagelse af mange interessenter. De medfører
ofte store organisatoriske konsekvenser med både muligheder og risici. Nogle virksomheder har opnået store fordele, mens andre har oplevet problematiske implementeringer
pga. brugermodstand, mangel på support fra topledelsen, dårlig pasform mellem system
og organisation, og mange andre årsager. En del virksomheder har endvidere stærkt
overvurderet fordelene ved systemerne, og har oplevet dem som rigide, så de hæmmer
fremtidige optimeringer i organisationen.
En væsentlig årsag til de problematiske implementeringer og / eller manglende realiseringer af fordele (nytteværdi), kunne være en overdreven fokus på ledelsesmæssige og
tekniske forhold, hvor instrumentelle løsninger anses for hensigtsmæssige og tilstrækkev
lige, hvilket betyder at organisatoriske og menneskelige aspekter nedtones eller helt
overses. Det kan have store organisatoriske konsekvenser i form af alvorlige driftsforstyrrelser ved idriftsættelse og langvarige driftsproblemer efterfølgende.
Den store udbredelse af virksomhedssystemer i organisationer kombineret med de mange udfordringer og problemer i forbindelse med ledelse, implementering og anvendelse
af systemerne, betyder at det er et yderst relevant emne for både praksis og forskning.
Meget forskning har derfor været helliget ledelse, implementering og anvendelse af
virksomhedssystemer, hvor der dog har været en overvægt på ledelsesmæssige og tekniske forhold, der synes at forenkle de sociale sammenhænge i moderne virksomheder,
fremhæve instrumentelle løsninger og nedtone menneskelige aspekter.
En af måderne til at imødegå den ulige fokus på ledelsesmæssige og tekniske forhold, er
at anvende institutionel teori, som adresserer de organisatoriske og menneskelige aspekter i højere grad. Institutionel teori bidrager med en mere strukturel og systemisk forståelse for, hvordan teknologier (såsom virksomhedssystemer) indgår i komplekse sociale,
økonomiske og politiske netværk, som er indbyrdes afhængige, og hvordan teknologierne er formet af bredere institutionelle strukturer og processer. På trods af de interessante muligheder som institutionel teori tilbyder, anvendes teoriapparatet sjældent inden
for IS forskningen. Den manglende brug af dette lovende teoretiske perspektiv, antyder
at de mangler, der er identificeret i forskningen og problemer der ses i praksis, kan
adresseres ved en systematisk anvendelse af teorien. Jeg har derfor valgt institutionel
teori, som den vigtigste teoretiske ramme i denne afhandling.
Formålet med afhandlingen er således at teoretisere om virksomhedssystemer i organisationer ved hjælp af institutionel teori, som leder hen til følgende overordnede forskningsspørgsmål: Hvordan former institutionelle strukturer og processer ledelsen, implementeringen og anvendelsen af virksomhedssystemer?
Et fortolkende casestudie i en højteknologisk virksomhed
Til at besvare forskningsspørgsmålet har jeg valgt en fortolkende casestudietilgang.
Fortolkende forskning forsøger at forstå fænomener gennem de betydninger, som vi
mennesker tildeler dem, og adgangen til virkeligheden, er gennem sociale konstruktioner såsom sprog og fælles betydninger. Fortolkende forskning anses for at være særligt
vi
velegnet, når kontekstuelle forhold skal betragtes (som i institutionel analyse), og der
indgår et kompliceret samspil mellem mennesker, ideer og institutioner. Metoden har
fokus på at indsamle fyldige og detaljerede beskrivelser om menneskers handlinger og
holdninger, for at forstå hvordan og hvorfor de handler og tænker, som de gør.
Det empiriske grundlag for afhandlingen er et længerevarende feltstudie hos SCANDI
(pseudonym), en skandinavisk virksomhed inden for forsyningssektoren med mere end
10.000 ansatte, som producerer og sælger højteknologiske ydelser og produkter.
SCANDI besluttede i 1996 at implementere virksomhedssystemet Oracle E-Business
Suite, og har siden da været igennem flere opgraderinger og implementeringer. SCANDI valgte i sommeren 2007 at igangsætte en større re-implementering af deres Oracle
system, blandt andet for at få en mere standardiseret løsning og en fremtidssikret platform. Løsningen blev idriftsat i januar 2009, og fungerer nu i normal drift efter en længere stabiliseringsfase. Mit feltarbejde startede tilbage i januar 2008 og fortsatte indtil
udgangen af 2009, hvor jeg har interviewet mange brugere, ledere, projektdeltagere og
konsulenter, observeret mange projektmøder og andre aktiviteter, samt haft adgang til
omfattende dokumentation om såvel projektet som virksomheden generelt. Jeg har haft
tæt kontakt til selve projektet samt indkøbsafdelingen og finansafdelingen, hertil kommer mere perifert andre afdelinger i SCANDI samt leverandører knyttet til projektet.
Afhandlingen består af et omslag på syv kapitler og fem artikler, hvoraf de fire artikler
er empirisk funderet, og her findes en kontekstuel beskrivelse af SCANDI, det historiske forløb med Oracle E-business suiten (1996 -2009) og konkrete oplevelser med
Oracle systemet i henholdsvis indløbsafdelingen og finansafdelingen.
Resultater og implikationer
Resultatet af dette studie består af to konceptuelle modeller og en række specifikke resultater i relation til beslutninger, implementering og anvendelse af virksomhedssystemer. De to konceptuelle modeller kan anvendes til at beskrive, fortolke, undersøge og
analysere institutionelle strukturer og processer i organisationer, der anvender virksomhedssystemer. Den første model indeholder en række grundlæggende institutionelle
elementer såsom isomorfisme, rationaliserede myter og institutionel logik samt muligheden for multi-niveau (makro, meso og mikro) og multi-teori analyser. Den konceptuelle ramme blev udvidet med en ”dual strukturel teknologimodel”, hvor der sættes fokus
vii
på institutionelle forhold såvel hos leverandørorganisationen, kundeorganisationen og
virksomhedssystemet. De specifikke resultater fra studiet er: (1) at rationelle og institutionelle forklaringer sameksisterer og supplerer hinanden i forbindelse med outsourcing
beslutninger; (2) at der er et gensidigt samspil mellem makro institutionelle strukturer
og mikro meningsdannende processer, såsom ændringer i institutionelle logikker fra
"tilret virksomhedssystemet til forretningsprocesserne" til "tilret forretningsprocesserne
til virksomhedssystemet" og endelig; (3) at institutionelle processer i form af deinstitutionalisering og institutionalisering spiller en vigtig rolle, når der skiftes fra et
system til et andet system. De to institutionelle processer overlapper hinanden, og initieres af en række pres og praksisser, der medvirker til skiftet, fx at virksomhedssystemet
videreudvikles til et ”multi-alting” system, hvilket reducerer behovet for tilretninger.
Studiet bidrager især til forskningen inden for ”sociale studier af virksomhedssystemer”,
men der kan dog i et vist omfang trækkes paralleller til andre områder i IS feltet såsom
skræddersyede systemer (egenudviklet eller kontraheret) og mere generiske standardprodukter (fx standard PC software). De konceptuelle modeller giver et overblik over
anvendelsesmulighederne af institutionel teori inden for virksomhedssystemer, som kan
være en hjælp til dels at komme i gang med denne teoretiske ramme, og kan være med
til at identificere potentielle nye forskningsområder. De fleste studier af virksomhedssystemer med anvendelse af institutionel teori, bruger organisationen som analyseenhed
(dvs. et meso niveau), hvilket også gælder dette studie. Til gengæld peges der i afhandlingen på, at studier på makro niveau (fx industrisegmenter) eller mikro niveau (fx individuelle brugere af et specifikt modul) stort set ikke findes, og dermed er mulige fremtidige forskningsområder. Et sidste potentiel forskningsområde er at anvende selve virksomhedssystemet som analyseenhed og undersøge ontologiske og epistemologiske perspektiver, hvor systemet metaforisk kan opfattes som et jernbur, et byggesæt eller måske noget helt tredje. Afslutningsvis skal det nævnes at studiet også kan anvendes af
ledere og medarbejdere til at komplementere den ”almindelige” teknisk rationelle forståelse af ledelse, implementering og anvendelse af virksomhedssystemer, og derved
medvirke til at sætte institutionelle strukturer og processer på dagsordenen i praksis.
viii
TABLE OF CONTENTS
1 INTRODUCTION ........................................................................................................ 1 1.1 RESEARCH METHOD ............................................................................................ 4 1.2 THESIS STRUCTURE AND ABSTRACTS OF PAPERS ................................................ 7 2 THE MANAGEMENT, IMPLEMENTATION AND USE OF ENTERPRISE SYSTEMS:
CONCEPTS AND PERSPECTIVES ............................................................................... 13 2.1 ENTERPRISE SYSTEMS CONCEPTS...................................................................... 13 2.2 TWO CONTRASTING PERSPECTIVES OF ENTERPRISE SYSTEMS .......................... 18 2.3 THE MANAGERIAL AND TECHNICAL UNDERSTANDING OF ENTERPRISE SYSTEMS
.......................................................................................................................... 18 2.4 THE SOCIAL STUDY OF ENTERPRISE SYSTEMS .................................................. 22 2.5 TOWARDS RECONCILING THE CAMPS ................................................................ 28 3 INSTITUTIONAL THEORY ........................................................................................ 30 3.1 INTRODUCTION .................................................................................................. 30 3.2 INSTITUTIONS, INSTITUTIONALIZATION AND ORGANIZATIONS .......................... 32 3.3 INSTITUTIONAL AND COMPETITIVE PRESSURES LEADING TO ISOMORPHISM ..... 34 3.4 RATIONALIZED MYTHS ..................................................................................... 35 3.5 MULTIPLE LEVELS IN INSTITUTIONAL THEORY BRIDGING MACRO AND MICRO
STRUCTURES ..................................................................................................... 36 3.6 INSTITUTIONAL LOGICS ..................................................................................... 37 3.7 INSTITUTIONAL PROCESSES AND CHANGE ......................................................... 39 3.8 CRITIQUE OF INSTITUTIONAL THEORY .............................................................. 41 3.9 SUMMARY OF INSTITUTIONAL CONCEPTS AND IMPLICATIONS FOR ENTERPRISE
SYSTEMS RESEARCH ......................................................................................... 42 4 BASIC PHILOSOPHICAL ASSUMPTIONS ................................................................... 45 4.1 INTRODUCTION .................................................................................................. 45 4.2 ONTOLOGICAL AND EPISTEMOLOGICAL PERSPECTIVES ON CONSTRUCTIVISM .. 46 4.3 SOCIAL CONSTRUCTIONISM ............................................................................... 48 4.4 HUMAN AGENCY ............................................................................................... 52 4.5 SUMMARY OF BASIC PHILOSOPHICAL ASSUMPTIONS ........................................ 52 5 RESEARCH PROCESS............................................................................................... 54 5.1 INTRODUCTION .................................................................................................. 54 5.2 INTERPRETIVE RESEARCH ................................................................................. 56 5.3 CONDUCTING INTERPRETIVE CASE STUDY RESEARCH ...................................... 58 5.4 SELECTING THE CASE AND ENTERING THE FIELD ............................................. 61 ix
5.5 BRIEF ABOUT THE EMPIRICAL CONTEXT ........................................................... 63 5.6 DATA COLLECTION ........................................................................................... 66 5.7 DATA ANALYSIS................................................................................................ 77 5.8 EVALUATING INTERPRETIVE RESEARCH............................................................ 83 5.9 REFLECTIONS ON THE RESEARCH PROCESS ....................................................... 85 6 THEORIZING ABOUT ENTERPRISE SYSTEMS IN ORGANIZATIONS USING
INSTITUTIONAL THEORY ........................................................................................ 88 6.1 THE DIFFUSION AND USE OF INSTITUTIONAL THEORY IN ENTERPRISE SYSTEMS
RESEARCH ......................................................................................................... 89 6.2 THE DEVELOPMENT OF CONCEPTUAL MODELS TO STUDY INSTITUTIONAL
STRUCTURES AND PROCESSES ........................................................................... 89 6.3 DRAWING SPECIFIC IMPLICATIONS FOR THE MANAGEMENT, IMPLEMENTATION
AND USE OF ENTERPRISE SYSTEMS ................................................................... 93 7 CONCLUSION ........................................................................................................ 100 7.1 IMPLICATIONS FOR RESEARCH......................................................................... 102 7.2 IMPLICATIONS FOR PRACTICE .......................................................................... 106 7.3 BRIEF SUMMARY ............................................................................................. 108 APPENDIX A – KANT’S TRANSCENDENTAL IDEALISM ................................................... 109 APPENDIX B – SOCIAL CONSTRUCTIONISM VERSUS SOCIAL CONSTRUCTIVISM ............. 113 APPENDIX C – EXAMPLE OF AN INTERVIEW GUIDE ....................................................... 114 APPENDIX D – TEMPLATE FOR FIELD NOTES IN PARTICIPANT OBSERVATION ............... 118 APPENDIX E – EXAMPLES OF DOCUMENTS .................................................................... 120 APPENDIX F – CODING IN NVIVO .................................................................................. 121 REFERENCES .................................................................................................................. 122 PAPER 1: USING INSTITUTIONAL THEORY IN ENTERPRISE SYSTEMS RESEARCH –
DEVELOPING A CONCEPTUAL MODEL FROM A LITERATURE REVIEW................... 141
PAPER 2: ENTERPRISE SYSTEMS OUTSOURCING “BEHIND THE CURTAIN” –
A CASE STUDY SHOWING HOW RATIONAL AND INSTITUTIONAL EXPLANATIONS
CO-EXIST AND COMPLEMENT EACH OTHER ........................................................... 180
PAPER 3: ENTERPRISE SYSTEM ADAPTATION: A COMBINATION OF INSTITUTIONAL
STRUCTURES AND SENSEMAKING PROCESSES ...................................................... 209
PAPER 4: MAKING SENSE OF ENTERPRISE SYSTEMS IN INSTITUTIONS: A CASE STUDY
OF A WELL-ADAPTED SYSTEM ............................................................................. 227
PAPER 5: MAKING NEW SYSTEMS IS BREAKING OLD SYSTEMS – A CASE STUDY
ABOUT PRACTICES FOR DEINSTITUTIONALIZING AN ENTERPRISE SYSTEM ........... 264
x
LIST OF FIGURES
FIGURE 1: CONCEPTUAL MODEL OF THE RESEARCH PROCESS ............................................ 6 FIGURE 2: POSITIONING THE PAPERS TO ES RESEARCH ...................................................... 8 FIGURE 3: THE PRODUCTION AND REPRODUCTION OF SOCIAL REALITY ........................... 51 FIGURE 4: OUTLINE OF THE RESEARCH PROCESS.............................................................. 55 FIGURE 5: TIMELINE OF THE IMPLEMENTATION AND USE OF THE ORACLE E-BUSINESS
SUITE ..................................................................................................................... 64 FIGURE 6: FIELD NOTES FROM PARTICIPANT OBSERVATION, JULY 2008 .......................... 73 FIGURE 7: FREQUENCY DISTRIBUTION OF DATA COLLECTION ACTIVITIES ........................ 77 FIGURE 8: EXPLICIT DATA ANALYSIS ACTIVITIES ............................................................. 80 FIGURE 9: EXAMPLE OF NVIVO CODING .......................................................................... 82 FIGURE 10: CONCEPTUAL MODEL FOR USING INSTITUTIONAL THEORY IN ES RESEARCH . 91 FIGURE 11: THE DUAL STRUCTURAL MODEL OF TECHNOLOGY FOR ENTERPRISE SYSTEMS 93 FIGURE 12: MODEL OF THE DUAL DEINSTITUTIONALIZATION AND INSTITUTIONALIZATION
PROCESS (ADAPTED FROM PAPER 5)........................................................................ 98 FIGURE 13: RICH PICTURE OF THE CONTRIBUTION FROM THIS RESEARCH PROCESS ........ 101 FIGURE 14: TEMPLATE FOR PARTICIPANT OBSERVATION ............................................... 118 xi
LIST OF TABLES
TABLE 1: THE MANAGEMENT, IMPLEMENTATION AND USE OF ENTERPRISE SYSTEMS ...... 17 TABLE 2: IDEAS AND ARGUMENTS WITHIN THE MANAGERIAL AND TECHNICAL
UNDERSTANDING OF ES ......................................................................................... 19 TABLE 3: IDEAS AND ARGUMENTS WITHIN THE SOCIAL STUDY OF ES .............................. 23 TABLE 4: KEY FEATURES OF INSTITUTIONAL THEORY RELATED TO ES RESEARCH .......... 44 TABLE 5: EPISTEMOLOGICAL AND ONTOLOGICAL ASSUMPTIONS IN THIS THESIS ............. 47 TABLE 6: THE SPECIFIC USE OF THEORY IN THIS STUDY ................................................... 60 TABLE 7: DATA COLLECTION OVERVIEW ......................................................................... 68 TABLE 8: EVALUATION OF THE STUDY............................................................................. 85 TABLE 9: RATIONAL AND INSTITUTIONAL EXPLANATIONS FOR ES OUTSOURCING (EXCERPT
FROM PAPER 2) ....................................................................................................... 94 TABLE 10: FINDINGS RELATED TO THE DUAL STRUCTURAL MODEL OF TECHNOLOGY
(EXCERPT FROM PAPER 4) ....................................................................................... 97 TABLE 11: EXAMPLE OF CODING IN NVIVO ................................................................... 121 xii
1 INTRODUCTION
Enterprise systems (ES) have been a major trend in both the private and public sectors
over the past decade. They have been on the market since the beginning of the nineties
(Jacobs and Weston 2007) as a solution to the growing tendency for globalization, mergers and acquisitions (Chang et al. 2003) and as a way to optimize and improve business operation (Häkkinen and Hilmola 2008). ES “… has become the software that no
business can live without” (Wailgum 2009b) and “standardized packaged solutions account for the bulk of systems used today” (Pollock and Williams 2009: 4) – they are the
de facto standard in many organizations.
Seddon et al. (2003) define ES as large-scale organizational systems, built around
packaged enterprise systems software, enabling an organization to automate and
integrate a comprehensive part of its business processes, to share common data and
practices and to produce and access information in real time. ES target private
organizations, but also public organizations like hospitals (Sia and Soh 2007) and
municipalities (Caccia and Steccolini 2006). Packaged ES software is generic “semifinished products” from vendors like SAP and Oracle delivered to user organizations
that tailor the products to their own needs (Brehm and Markus 2000; Seddon et al.
2003). Davenport (1998: 122) expresses the consequences of ES in the much cited line:
“An enterprise system imposes its own logic on a company’s strategy, culture and
organization” – and it is very unlikely that there is a perfect fit between the ES and the
organization.
The implementation of ES is often complex due to enterprise-wide integration and data
standardization, adoption to “best-practice” business models with re-engineering of
business processes, compressed schedules and, finally, the participation of a large number of stakeholders (Soh et al. 2000: 47). ES often trigger major organizational changes
and at the same time introduce high risk with a potential high reward (Chae and Lanzara
2006: 100; Markus 2004). Some companies have gained an important increase in productivity and speed (Häkkinen and Hilmola 2008), while others have experienced failureprone ES implementations (Grabski et al. 2003; Sumner 2003) due to users’ resistance
(Grabski et al. 2003; Sumner 2003), lack of senior management support (Sumner 2003),
1
misalignment between the ES and the organization (Sia and Soh 2007) and many other
reasons. Still others have highly overestimated the value of ES (Davenport 1998; Robbins-Gioia 2002) and realized that the benefits did not materialize (Lindley et al. 2008).
Lindley et al. (2008) furthermore argue that an “ES already in use” might prevent future
optimizations in the organization, due the rigidity built into the systems.
A major reason for failure-prone implementations and/or lack of benefits (Davenport
1998) might be the focus on managerial and technical issues where instrumental solutions are considered superior and sufficient, ignoring implementation and integration
problems (Dillard and Yuthas 2006). This is in accordance with my own experience as a
project manager of numerous projects and from evaluation of many projects, that social
considerations are downplayed or even overlooked, resulting in poor ability to manage
change (see also Panorama Consulting Group 2010). This might have severe consequences such as operational disruptions at go-live and hampered business operation
afterwards (Markus et al. 2000).
The widespread penetration of ES in organizations combined with the many challenges
and problems associated with the management, implementation and use of ES implies
that it is a highly important area of concern for both practice and academia. Much research has therefore been devoted to ES implementation and use in general as well as
alignment between the organization and the ES in particular, but, as argued by Pollock
and Williams (2009) and others (Berente 2009; Boudreau and Robey 2005; Lamb and
Kling 2003), the research around “Enterprise Systems has been unevenly developed and
unhelpfully fragmented between rather narrow (e.g. managerial or technical) perspectives” (Pollock and Williams 2009), which appears to simplify the social settings of
modern enterprises, emphasize instrumental solutions and downplay social considerations.
However, one way to overcome the managerial and technical understanding and address
the social and organizational aspects is to use institutional theory with its ability to “develop a more structural and systemic understanding for how technologies [such as ES]
are embedded in complex interdependent social, economic, and political networks, and
how they are consequently shaped by such broader institutional influences” (Orlikowski
and Barley 2001: 154) and with its capability to deal with the logics that ES imposes on
2
organizations (Gosain 2004). Despite the advantages hinted at by Orlikowski and Barley (2001), IS researchers rarely adopt an institutional perspective (Berente 2009; Orlikowski and Barley 2001; Weerakkody et al. 2009), and when they do it is a narrow use
not exploiting the potential of institutional theory (Currie 2009). The lack of comprehensive use of this promising theoretical lens hints that the gaps identified in the ES
literature and the problems experienced in practice may be filled by applying this theory
systematically to the analysis of a relevant case.
I have therefore chosen institutional theory (Greenwood et al. 2008b) as the main theoretical framework for the following specific reasons: first, ES are particularly well
suited to institutional analysis (Gosain 2004; Sia and Soh 2007) as they are intended to
integrate processes across an entire organization (Davenport 1998) and organizations
typically lie at a locus of multiple, often inconsistent and conflicting, institutions, where
the institutional analysis of ES can prove fertile for investigating different outcomes
within an organization and across organizations (Berente 2009: 9). Second, “Existing
studies both downplay the influence of technology supply and often overlook the influence of the broader historical setting on the unfolding of the technology” (Pollock and
Williams 2009: 9), and this can be addressed by institutional theory with its contextual
focus combined with the multi-level and multi-stakeholder approach (Currie 2009). Finally, institutional theory is in its infancy in IS research where “the use … is comparatively narrow and limited when considering its wider use in organizational studies”
(Weerakkody et al. 2009: 354), which opens unexplored research avenues that this thesis can address.
But how can institutional structures and processes enrich our understanding of the management, implementation and use of ES? Institutional theory addresses the social, historical and contextual influences on the management, implementation and use of ES
(adapted from Jensen et al. 2009) and how these systems are objects for and carriers of
institutional structures (Gosain 2004). Institutional theory attempts to describe the deeper and more resilient aspects of how institutions are created, maintained, changed and
dissolved (Scott 2004; 2008), and deals with the pervasive influence of institutions on
human behavior including the processes by which structures, e.g. rules, routines and
norms, guide social behavior. Institutions are multi-faceted, durable, resilient social
3
structures, made up of symbolic elements, social activities and material resources (Currie 2009; Scott 2001: 48-50). The management, implementation and use of ES are embedded in these wider institutional structures and take part in the reciprocal interactions
with institutional processes shaping our decisions, actions and understandings.
There are interesting issues at stake here. How can we study institutional processes and
structures that are relevant to ES research? How are management decisions about ES
shaped by institutional processes? In which ways do institutional structures impact on
the use of ES? How do institutional structures shape implementation processes when
replacing an old system with a new system? These are all relevant questions and there
are good reasons to suspect that institutional structures and processes play an important
role in the management, implementation and use of ES, which guide the overall purpose
and research question of this thesis. The purpose is thus to theorize about enterprise
systems in organizations using institutional theory, and to answer the overall research
question: How do institutional structures and processes shape the management, implementation and use of enterprise systems? In order to position and clarify the purpose
and research question some words about the research method are appropriate to outline
the approach, scope and limitations of this thesis.
1.1 RESEARCH METHOD
To answer the overall research question, I have adopted a contextualized, interpretive
research approach (Pettigrew 1990; Walsham 2002; 2006). Interpretive research attempts to understand phenomena through the meanings that people assign to them
(Myers and Avison 2002), and access to reality is through social constructions such as
language, consciousness and shared meanings (Berger and Luckmann 1966; Myers and
Avison 2002). Interpretive research is considered particularly suitable when the context
has to be considered (as in institutional analysis), and the “often-complicated relationship between people, ideas and institutions” (Travers 2001: 18). Interpretive research is
founded on social constructions, which fits well with organizational institutionalism
(Greenwood et al. 2008b), the branch of institutional theory applied in this thesis, due to
their common origin (e.g. Berger and Luckmann 1966; Schutz 1967; see also Scott
2008: 42-44), and is therefore suitable for studying institutional structures and
processes. Studying ES and institutional structures/processes in the interpretive tradition
4
implies understanding and interpreting the life world of people as well as ambiguity and
contradictions in this life world (Kvale 2007) – the goal is not to generate truth or social
laws, but to make credible and trustworthy interpretations and understandings available
as “consultable records” (Walsham 2002: 103). The interpretive approach encourages
openness in the research process as you do not precisely know where the journey will
go and what you will experience during the journey (Kvale 2007; Walsham 2002), and
the overall research question is deliberately broad to facilitate this openness in the research process.
I have selected a case study approach (Stake 2005; Yin 2003) as it is considered appropriate when the study has a descriptive and exploratory focus (Marshall and Rossman
1989: 78). Case studies can be valuable for generating an understanding of reality (Yin
2003), allowing rich descriptions and empirical evidence from people in situ in natural
organizations (Myers 2009). Thus, an appropriate approach is an in-depth study that a
single case provides, and it has been termed a “revelatory case” (Yin 2003).
An often-cited limitation of the single case study method is its lack of generalizability
as the data collected are typically specific to a particular situation at a particular point in
time. However, single case studies are generalizable to theoretical propositions (Flyvbjerg 2006; Schwandt 2007; Silverman 2005; Walsham 2002; Yin 2003) and the results
in this thesis, based on Walsham’s (2002; 2006) types of generalization, are used to develop concepts and draw specific implications. These generative mechanisms identified
for the management, implementation and use of ES should be viewed as tendencies that
interpret or explain past data, but are only partially predictive for future situations (Walsham 2002: 110).
The realm of practice explored in this thesis is SCANDI, a Scandinavian high-tech
company with more than 10 000 employees. It belongs to the utility industry segment
where it produces and sells high-tech services. SCANDI performs primary activities
like logistics, operations, sales and marketing services as well as supporting activities
like administration, human resource management, procurement, information systems
services etc. The first company in SCANDI was established in the late 1890s, and the
company today is a result of a merger between several companies. SCANDI decided in
1996 to implement Oracle E-Business Suite, also known as Oracle Financials (James
5
and Seibert 1999), and has been through several upgrades and implementations since
then. A major reimplementation project was started in summer 2007 and the reimplemented ES was launched in January 2009. My fieldwork started back in January 2008
and continued until the end of 2009. SCANDI was chosen because: (1) it had a comprehensive ES installation with more than 3000 users including many organizational and
technical facets, (2) it was possible to follow a major project over a longer period including implementation, launch and post-implementation (longitudinal perspective) and,
finally, (3) SCANDI was very open and collaborative towards the research project,
which fertilized the ground for extensive fieldwork.
Interviews, participant observations and access to numerous documents were used as
data collection methods in a longitudinal study during my two years’ fieldwork (2008–
2009), combined with historical reconstructions. The data analysis followed the interpretive tradition (Walsham 2002; 2006) using hermeneutics (Myers 2009), where concepts from institutional theory were used as sensitizing devices (Patton 2002: 452-462)
to support the coding and analysis process.
Below is a conceptual model that aims to capture and reconstruct the essential part of
the research process presented in this thesis (adapted from Checkland 1985; Jackson
2000: 12-14):
Figure 1: Conceptual model of the research process
6
The figure illustrates the relationship between elements in the research process. However, the interpretation of the figure should be cautious as it tends to model the research
process as more rational and linear than it was intended to be and than how it turned out
to be.
The research process is elaborated in chapter 5, and the result of my research is documented in five papers, which will be very briefly described in the next section.
1.2 THESIS STRUCTURE AND ABSTRACTS OF PAPERS
The thesis contains a cover part and a collection of five papers. The cover part intends
to synthesize the research described in the papers, but the cover part seeks also to complement the five papers with related and relevant issues. Conference and journal papers
are often focused and “cut into shape,” meaning background information, reflections on
the research process and many other things are left out, whereas the cover part aim to
fill in the gaps and provide a coherent presentation of my research.
The cover part consists of seven chapters. The next chapter presents some existing perspectives on the management, implementation and use of ES, which is the area of concern in this thesis. The chapter that follows presents an overview of institutional theory,
which is mainly a synthesis and consolidation of the theory presented in the various
papers. This is followed by two chapters of respectively basic philosophical assumptions and the research process, where the latter chapter presents interpretive research
methodology, data collection and analysis methods. The chapter also provides a brief
overview of the SCANDI case study and finishes with an evaluation of and reflections
on the research process. The next chapter synthesizes the findings across the papers.
The last chapter concludes the thesis and describes implications for research and practice followed by a very concise summary.
7
The collection of papers follows directly the cover part chapters with appendices. The
Venn diagram below shows how the papers map into ES research (the area of concern)
using the distinction between the management, implementation and use of ES:
Figure 2: Positioning the papers to ES research
The figure shows one conceptual paper outside the Venn diagram and four empirical
papers inside. The titles, statuses and abstracts of the five papers are described below:
Paper 1: Svejvig, P. “Using institutional theory in enterprise systems research – Developing a conceptual model from a literature review.”
Status: The paper is ready for a second revision and resubmission to the Scandinavian
Journal of Information Systems.
Abstract: This paper sets out to examine the use of institutional theory as a conceptually rich lens to study social issues of enterprise systems (ES) research. More precisely,
the purpose is to categorize current ES research using institutional theory to develop a
conceptual model that advances ES research. Key institutional features are presented
such as isomorphism, rationalized myths, and bridging macro and micro structures, and
institutional logics and their implications for ES research are discussed. Through a literature review of 180 articles, of which 18 papers are selected, we build a conceptual
model that advocates multi-level and multi-theory approaches and applies newer institutional aspects such as institutional logics. The findings show that institutional theory in
ES research is in its infancy and adopts mainly traditional institutional aspects like iso-
8
morphism, with the organization as the level of analysis, and in several cases it is complemented by structuration theory and other theories.
Paper 2: Svejvig, P. and J. Pries-Heje. “Enterprise systems outsourcing behind the curtain – A case study showing how rational and institutional explanations co-exist and
complement each other.”
Status: Earlier versions of this paper have been presented at two conferences: (1) “SIM
Academic Workshop”, 13 December 2008, an ancillary meeting to “International Conference on Information Systems ICIS 2008”, 14 to 17 December, Paris, France, and (2)
“The Enterprise Information Systems International Conference on Research and Practical Issues of EIS (CONFENIS)”, 28 to 30 October 2009, Gyor, Hungary. The conference paper from CONFENIS 2009 is furthermore published in a special issue of the
Professional Journal of the Scientific and Educational Forum on Business Information
Systems (Svejvig and Pries-Heje 2009). The paper has also been presented at several
seminars for practitioners and the research results are briefly mentioned in a newspaper
article (Ramskov 2009). The paper presented in this thesis is an essential extension of
the earlier-published version and has been submitted to Information Technology &
People.
Abstract: Outsourcing is now a feasible means for enterprise systems (ES) cost savings,
but does however increase the complexity of coordination substantially when many organizations are involved. We set out to study ES outsourcing in a large Scandinavian
high-tech organization, SCANDI, a case setting with many interorganizational partners,
trying to answer the question: Why does SCANDI engage in these very complex outsourcing arrangements? To answer this question we have analyzed documents, observed
meetings and gathered data from interviews in four parts of SCANDI. The first data
analysis found just the rational frontstage cost-saving explanation; but then, with a more
careful analysis focusing on institutional factors, other backstage explanations “behind
the curtain” were uncovered, such as management consultants with a “best practice”
agenda, people promoting outsourcing, thereby being promoted themselves, and a belief
in outsourcing as a “silver bullet”: a recipe to success, solving everything.
9
Paper 3: Svejvig, P. and T. B. Jensen (2009). “Enterprise system adaptation: A combination of institutional structures and sensemaking processes,” AMCIS – Americas Conference on Information Systems 2009, San Francisco, 2009.
Status: The paper elaborates on some earlier work that I co-authored with Tina Blegind
Jensen and Annemette Kjærgaard, where we combine institutional theory and sensemaking – this is published in the Journal of Information Technology (Jensen et al.
2009). The paper was presented at AMCIS 2009.
Abstract: In this paper we set out to investigate how an Enterprise System (ES) adaptation in a Scandinavian high-tech organization, SCANDI, can be understood using a
combination of institutional and sensemaking theory. Institutional theory is useful in
providing an account for the role that the social and historical structures play in ES
adaptations, and sensemaking can help us investigate how organizational members
make sense of and enact ES in their local context. Based on an analytical framework,
where we combine institutional theory and sensemaking theory to provide rich insights
into ES adaptation, we show: 1) how changing institutional structures provide a shifting
context for the way users make sense of and enact ES, 2) how users’ sensemaking
processes of the ES are played out in practice, and 3) how sensemaking reinforces institutional structures.
Paper 4: Svejvig, P. and T. B. Jensen. “Making sense of enterprise systems in institutions: A case study of a well-adapted system.”
Status: Paper 4 was intended to be an extended version of paper 3 (same as Svejvig and
Jensen 2009), but evolved into a different paper utilizing “the structural technology
model” (Orlikowski 1992) together with institutional theory and sensemaking theory.
An earlier version of the paper has been accepted for the Academy of Management Annual Meeting, August 6-10, 2010. Montréal, Canada (Svejvig and Jensen forthcoming).
The paper presented in this thesis has been submitted to European Journal of Information Systems.
Abstract: Contrary to previous research that provides numerous accounts of failure
prone enterprise systems (ES) adaptations in organizations, empirical data from an ES
adaptation in a Scandinavian high-tech company, SCANDI, shows how the system was
10
highly integrated, accepted by its users, and well-aligned to the work processes. It is
therefore natural to ask: Why is the enterprise system so well-adapted in SCANDI and
what can we learn from this case study? Building on concepts from institutional theory
and sensemaking theory, we present three sets of reasons for the well-adapted system:
(1) “a rationalized myth” about an efficient ES that will create effective work practices
travels from a national to a local level; (2) a long transition process from “match to current business processes” towards “match to standard package”; and (3) the users in practice find it easy to adjust to the ES and reinforce existing structures. We present the lessons learned from the study such as: first customize then un-customize, be prepared for
a long term adaptation process, and consider the match between the users and the system. We discuss theoretical and practical implications.
Paper 5: Svejvig, P. and A. Carugati. “Making new systems is breaking old systems –
A case study about practices for deinstitutionalizing an enterprise system.”
Status: An extended abstract of the paper was presented at the “Organizations and Society in Information Systems” (OASIS) 2009 Workshop, 15 December 2009, Phoenix,
Arizona, USA (Svejvig and Carugati 2009). An earlier version of the paper has been
accepted for the European Group for Organizational Studies (EGOS) 2010, June 28 to
30, Lisbon, Portugal (Svejvig and Carugati forthcoming). The version in this thesis is
being prepared for journal submission in the near future.
Abstract: This paper investigates the process of implementation of new enterprise systems (ES) as the process of decommissioning of the old system takes place. Decommissioning is a vastly overlooked and oversimplified process in IS research and the role of
the pressures and practices leading to the abandonment of the practices supported and
embedded in the old system is not well understood. To answer the research question we
use institutional theory as a lens to make sense of the case study of SCANDI, a large
Scandinavian high-tech organization, in the process of implementing a new ES after 13
years’ use of their first ES. By drawing on institutional theory we seek to understand the
pressures and practices related to making a new system and breaking the old system in a
multi level analysis. Through the institutional lens we examine intended and unintended
practices deployed by SCANDI to deinstitutionalize their current system and replace it
with a new one. The analysis shows that deinstitutionalization is inseparable from insti11
tutionalization, some practices impact both processes, while other practices only apply
to either deinstitutionalization or institutionalization. Comparing with previous experiences found in the literature we conclude that the knowledge of the pertinence of the
practices to either process is vital in the successful implementation of deeply embedded
and pervasive systems like ES.
The papers are referenced as paper 1; paper 2 etc. in the cover part. The papers have
been reformatted so they follow much the same style as the cover part.
The cover part can be read separate, but I recommend reading the papers before reading
chapters 6 and 7 in the cover part.
12
2 THE MANAGEMENT, IMPLEMENTATION AND USE
OF ENTERPRISE SYSTEMS: CONCEPTS AND PERSPECTIVES
ES research is diverse, comprehensive and interdisciplinary, the area spanning IS, computer science, accounting, operation management and several other research disciplines
(Kraemmergaard and Schlichter 2009). ES has been on the market since the beginning
of the nineties (Jacobs and Weston 2007), but academic research into ES first took off
in the late nineties (Lorenzo 2004; Ramiller et al. 2008), ostensibly started by Davenport’s paper about mega packages (1996) followed by his seminal paper “Putting the
enterprise into the enterprise system” (1998) (see also Klaus et al. 2000). The topic has
been on the agenda in journals and at conferences since then (Esteves and Bohorquez
2007; Ramiller et al. 2008). However, forerunners to ES, such as accounting systems
and material resource planning systems (MRP, MRP-II), can be traced back to the
1960s (Pollock and Williams 2009: 22-32; Scarbrough et al. 2008).
The purpose of this chapter is to distill relevant elements from this broad research area
by means of (1) clarifying the ES concepts used in this thesis and (2) presenting two
contrasting perspectives of ES. This is not a formal literature review and the chapter
does not seek to do justice to the many streams of ES research, but instead takes a focused approach to develop an integrated understanding of ES concepts and perspectives
based on current research.
2.1 ENTERPRISE SYSTEMS CONCEPTS
Enterprise systems are large-scale organizational systems, built around packaged enterprise systems software, enabling an organization to automate and integrate a comprehensive part of its business processes, to share common data and practices and to produce and access information in real time (Seddon et al. 2003). ES promise to make a
seamless flow of information across all the parts of an organization, crossing geographic, functional and organizational boundaries (Dillard and Yuthas 2006). ES can be characterized as “multi-everything” artifacts – multi-industry, multi-function, multiprocesses etc. (Kholeif et al. 2008: 20). The systems are designed to manage business
13
systems globally, drive compliant business processes worldwide and deploy countryspecific capabilities to operate anywhere in the world (Oracle 2008). Industry-specific
versions are available for many industries such as financial (banking, insurance), public
services (health care, public administration), manufacturing (consumer products, pharmaceuticals) and service (utility, retail) (SAP 2009).
The ES market is enormous and accounted for US$ 50 billion in 2008, when SAP and
Oracle dominated the market with a share of respectively 31% and 17% according to
AMR research (D’Aquila et al. 2009). ES implementation costs might be 5 times the
cost of the software and services or even more (Scheer and Habermann 2000), so the
total spending might approximate US$ 250 billion in 2008, indicating huge investments
in ES. ES investments top the list for corporate IT spending according to Forrester Research, who asked 400 North American and European ES decision makers (fourth quarter 2008), and only 1% of the respondents plan to decrease their ES investment (Wailgum 2009a) despite the global financial crisis and recession. The survey shows the dependency and criticality of ES for companies.
“ES have fundamentally changed the way business is done in many industries” (Dillard
and Yuthas 2006: 205). When industry leaders implement an ES, competitors have a
propensity to follow along, often with the same ES vendor (Davenport 1998: 126), as
well as upstream and downstream business partners in the supply chain. Organizations
are also pressured to implement ES by professional associations, consultants, accounting firms etc. (Dillard and Yuthas 2006; see also Swanson and Ramiller 1997), that is,
institutional pressures from field and society (Scott 2008).
The most important category of ES is enterprise resource planning systems (ERP systems) with other categories such as customer relationship management systems (CRM
systems), supply chain management systems (SCM systems), human capital management systems (HCM systems), product life-cycle management systems (PLM systems),
data warehousing, decision support and others (Jacobson et al. 2007; Seddon et al.
2003). Davenport (1998) states that an ERP system consists of a central database as a
backbone connected to various application modules such as financial applications, manufacturing applications, inventory and supply applications, sales and delivery applications, service applications and human resource applications.
14
ES software is, unlike tailor-made software, not designed for the precise needs and requirements of an organization, but instead a semi-finished product that the organization
has to tailor to its needs. Tailoring could be configuration (setting up parameters) and/or
customization (adding non-standard features to the software by programming) (Seddon
et al. 2003). Organizations are advised to go for vanilla implementations (i.e. configuring without customizations) by consultants (Hildebrand 2009) and academia (Parr and
Shanks 2003; Seddon et al. 2003) or at least to minimize customizations as much as
possible (Beatty and Williams 2006) – vanilla implementations have thus been institutionalized advice and practice. The reason is that vanilla ES implementations are easier,
cheaper (Fitz-Gerald and Carroll 2003) and quicker (Willis and Willis-Brown 2002) and
furthermore reduce future maintenance and upgrading costs (Hildebrand 2009). However, the consequence might be major changes to current practices (see also Soh et al.
2000) and, as Willis and Willis-Brown argue (2002: 36), the outcome is standardized
practices that only “minimally meet the needs of the unique business operation.”
So, although ES ships as “complete, though flexible, ready to implement solution[s]”
(Soh and Sia 2004: 376), they do come with built-in assumptions (Seddon et al. 2003)
often referred to as best practices based on industry standards anticipated to be the most
rational and progressive business processes available. Organizations have to replace
their current practices with these best practices through a re-engineering process (Dillard and Yuthas 2006: 206), particularly with vanilla implementations. This might be a
problematic process depending on the difference between the current practices and the
best practices embedded in the ES, and some organizations find it very difficult to accept the standardized solutions. The problems with misfit or misalignment are well
treated in the IS literature (e.g. Kien and Soh 2003; Soh et al. 2000; Sumner 2009; Wei
et al. 2005).
Organizations are in addition challenged to understand the many options supported by
ES, and to choose the most optimal solution (Seddon et al. 2003). This is difficult due to
the high complexity of “multi-everything ES software” and the pace at which new functionality is added. Further to this, any ES implementation is a trade-off between tailoring the ES (configurations and possible customizations) and re-engineering the practices, and every choice has its short-term and long-term consequences for the organization
15
in question, whether the preferred mix is a clean vanilla solution, a highly customized
solution or something in between on this continuum (see also Parr and Shanks 2000).
Company-wide ES implementations often trigger major organizational changes and at
the same time introduce high risk with a potential high reward (Chae and Lanzara 2006:
100; Markus 2004). Some companies have gained an important increase in productivity
and speed (Häkkinen and Hilmola 2008), while others have highly overestimated the
value of ES (Davenport 1998; Robbins-Gioia 2002). Many organizations have collectively spent hundreds of billions of dollars on ES implementations and are asking (Seddon et al. 2003: 4): How can we gain greater benefits from our investment? When will
our investment pay off? Did our investment pay off? These questions are of course very
important and have been taken up by various researchers, consultants and others who
describe ES benefits as multidimensional (e.g. Davenport et al. 2002; Deloitte Consulting 1999; Hawking et al. 2004; Ross and Vitale 2000) – examples are (1) operational
improvements in management decision making, financial management etc., (2) business
and IT headcount reductions, (3) inventory and cycle time reduction and (4) support for
strategic goals like new products or new channels. However, these benefits are seldom
achieved in the first implementation (first wave), and a second-wave implementation is
needed to harvest the benefits of ES-enabled practices according to Deloitte Consulting
(1999). Brynjolfsson et al. (2010) furthermore state “that it typically takes between five
to seven years for major IT investments, like ERP systems, to deliver substantial returns” (cited in Wailgum 2009b), so organizations have to be very patient when dealing
with ES.
Before finalizing the conceptualization of ES it is relevant to consider the future of ES.
There are several technological trends for ES, for instance cloud computing; Software
as a Service (SaaS) or on-demand software to enable alternative delivery models where
user organizations outsource the operation and maintenance of their ES (faster release of
new features, less IT operation complexity etc.); embedded analytics to cope with the
dramatic growth of data in ES; and a new software star called “socialytic business applications” combining social collaboration software (Web 2.0) with analytics software,
addressing new ways of conducting business (Pollock and Williams 2009: 48-50; Wailgum 2009a; b; c). However, talking about next-generation ES in technological terms is
16
insufficient and mainly adhering to the rhetoric spread by ES vendors and other opinion
makers in the ES community – a complementary view is expressed by industry consultant Reed who sums up what ES customers are really looking for (cited in Wailgum
2009c):
Empower me. Give me the tools to create differentiating processes that allow me to define myself from my competitors. And make sure that it’s easier for me to do, so I don’t have to hire 100 programmers. Give me the
building blocks to put that together quickly, so that it’s just humming in the
background, and leave me free to focus on what makes us better than other
companies. That’s what customers are expecting now and really want.
This statement refers to the discussion about tailoring, alignment and how to obtain
competitive advantage from ES, which is a recurrent challenge.
The management, implementation and use of ES are the next concepts to elaborate
upon. These concepts reflect processes, activities or decisions associated with ES. The
table below exemplifies these concepts (inspired by Esteves and Bohorquez 2007; Gulledge and Simon 2005; Henfridsson 1999; Lorenzo 2004; Ross et al. 2003; Silver et al.
1995):
Concept
Examples of processes, activities and decisions
Management of ES
•
•
•
•
•
Work carried out through others:
Project management and change management
IT governance and IT strategy activities
Organization of business and IT resources (sourcing strategies)
Performance management (harvest benefits)
Implementation of ES
•
•
•
•
•
•
Life-cycle activities before the launch of ES:
Adoption (decision to acquire ES)
Acquisition (vendor assessment and selection)
Design
Configuration and customization, re-engineering practices
Preparation for launch
Use of ES
•
•
•
•
•
•
Life-cycle activities after the launch of ES:
Introduction (from the launch until ES is stabilized)
Adaptation (the ongoing adaptation between the ES and the organization)
Evolution (continuous improvement)
Maintenance (patches and bug fixes)
Retirement (replacing ES with another system or version)
Table 1: The management, implementation and use of enterprise systems
17
The examples in the table are by no means exhaustive or comprehensive, and serve only
as an explanation of how the concepts are understood in this context. The concepts are
furthermore not distinct from each other. Implementation and use are often elements of
life-cycle models (e.g. Markus and Tanis 2000; Ross and Vitale 2000; Somers and Nelson 2004), but the concepts are applied in a more inclusive way in this thesis, as presented above in the table.
To summarize: implementation embraces all the activities carried out before an ES is
launched; use covers all the activities after the ES is launched; and management includes all the activities where the work is performed through others during the implementation and use of ES.
2.2 TWO CONTRASTING PERSPECTIVES OF ENTERPRISE SYSTEMS
Pollock and Williams (2009: 5) argue that social science research on ES is largely gathered around two contrasting perspectives. The first is “the managerial or technical
understanding of ES” focusing on economic efficiency and improved financial performance (Dillard and Yuthas 2006). ES are broadly viewed as transformative technologies
with universal logics, which can be applied extensively across organizations to bring
widespread change. This perspective has been contrasted with a second perspective labeled “the social study of enterprise systems,” where scholars have advanced situated
and localist explanations for ES. The focus is on uniqueness in the structure and practices in the organization, which often counteract the adaptation of standardized ES
(adapted from Pollock and Williams 2009: 5). The following sections will elaborate on
the two perspectives and point to selected literature.
2.3 THE MANAGERIAL AND TECHNICAL UNDERSTANDING OF
ENTERPRISE SYSTEMS
The managerial and technical understanding of ES is evidenced in trade publications,
consultancy reports, supply-dominated information (vendor white papers, brochures,
manuals etc.) and much academic literature (adapted from Dillard and Yuthas 2006;
Pollock and Williams 2009).
Below is a table that provides an overview of themes with ideas/arguments and selected
references within the managerial and technical understanding of ES:
18
Themes
Selected references (trade Ideas/arguments
publications, consultancy
reports, vendor information and academic literature)
ES as generic systems
(scope, technology, history
etc.)
(Davenport 1998; 2000;
Jacobs and Weston 2007;
Klaus et al. 2000; Seddon et
al. 2003)
ES are generic systems, which can be recycled across similar organizations and now
increasingly also across different industry
sectors or organizational forms.
ES are thus viewed as transformative technologies containing universal logics that can
be applied to all kinds of organizations.
ES offer greater certainty and rational control that provides a solution to the challenges and problems that organizations are
facing.
ES are an instrument for achieving managerial/organizational goals based on rational
choice and control.
Methodologies, life-cycle
models, tools and techniques for ES
(Bancroft et al. 1998; Markus and Tanis 2000;
O'Leary 2000; Parr and
Shanks 2000; Ptak and
Schragenheim 2003; Ross
et al. 2003)
Critical success factors and (Finney and Corbett 2007;
risk factors
Parr and Shanks 2003;
Sumner 2003)
Successful implementation
of ES
(Beatty and Williams 2006;
Hildebrand 2009; Scheer
and Habermann 2000)
Gaining value from ES
(Davenport et al. 2002;
Davenport et al. 2004; Deloitte Consulting 1999;
Hawking et al. 2004; Ross
and Vitale 2000; Shang and
Seddon 2003; Ward 2006;
Willis and Willis-Brown
2002)
Positioning of ES by vendors
(Oracle 2008; SAP and
Capgemini 2009)
ES market (sizing, vendor
ranking etc.)
(D’Aquila et al. 2009; Jacobson et al. 2007)
There is a strong managerial focus on specifying guidelines for successful implementation.
The literature is framed in strictly technical
or managerial terms where contextual circumstances, behavioral assumptions underlying ES implementations and the architecture of packages are given little attention.
ES are drivers of organizational change
anticipated to enhance economic efficiency,
improve financial performance and bring
organizational improvement in general.
The managerial and technical understanding
is shaped by the statements and rhetorics of
technology supply.
Table 2: Ideas and arguments within the managerial and technical understanding of ES
The table repeats many references from the previous section about ES concepts, which
indicate that it is biased towards the managerial and technical understanding of ES. This
fairly heterogeneous research stream covers a wide range of assumptions and ideas.
19
First, the literature presents ES as generic systems, which can be recycled across similar
organizations and now increasingly also across different industry sectors or organizational forms (e.g. from the private to the public sector). ES are thus viewed as transformative technologies containing universal logics that can be applied to all kinds of organizations (Bendoly and Jacobs 2005; O'Leary 2000) based on the way companies
operate in general (Davenport 1998). Jacobs and Weston (2007: 363) explain that “Generic ERP software packages are already increasingly tailored to specific market segments … Preconfigured software modules incorporating best practices and standard
business processes will simplify future implementations,” and this illustrates the understanding of the universal ES with best practices and standard business processes that can
be diffused to different market segments.
Second, there is a strong managerial focus in the literature on specifying guidelines for
successful implementation like enterprise systems’ experience cycle linked to recipe for
success (Markus and Tanis 2000 e.g. table 10.3) best practices for successfully implementing an ERP upgrade (Beatty and Williams 2006) and general methodologies, tools
and techniques for ES implementations (Bancroft et al. 1998; O'Leary 2000; Ptak and
Schragenheim 2003). A related set of literature presents lists of critical success factors
or risk factors associated with ES implementations based on literature reviews (Finney
and Corbett 2007), a combination of reviews of literature and interviews with ERP implementation practitioners (Parr and Shanks 2003) and case studies (Sumner 2003). The
literature is concerned with all the issues related to ES implementations addressing
technical aspects, working practices, operating procedures and human competences including the re-engineering of the organization with new structures and procedures. The
literature is framed in strictly technical or managerial terms where contextual circumstances, behavioral assumptions underlying ES implementations and the architecture of
packages are given little attention (Kallinikos 2004: 12).
Third, ES are drivers of organizational change (Davenport 2000) anticipated to enhance
economic efficiency, improve financial performance (Dillard and Yuthas 2006: 204)
and bring organizational improvement in general. ES vendors of course promote the
benefits of their ES whether they are streamlining a global operation to reduce operating
costs (Oracle 2008) or changing a business to accelerate sales (SAP and Capgemini
20
2009). The value theme is taken up by consultancy companies, with Deloitte Consulting
(1999) advocating for a second wave to maximize the value of ERP and in the same
vain Accenture stating “that many leading organizations are revisiting the original
promise of enterprise solutions, and taking actions now to realize the value they need
from the systems” (Davenport et al. 2002: 5). The second-wave approach was taken up
by researchers in the book “Second-Wave Enterprise Resource Planning Systems – Implementing for Effectiveness” (Shanks et al. 2003) and by other researchers (Ward
2006; Willis and Willis-Brown 2002) as the impetus for (finally) realizing the benefits
of ES.
Fourth, ES offers greater certainty and rational control that provides a solution to the
challenges and problems that organizations are facing. This is an instrumental view of
technology as a subject (tool) for achieving managerial/organizational goals based on
rational choice and control (Davenport 1998; O'Leary 2000). “ERP systems represent
powerful mechanisms for structuring organizational context and practices” (Dillard and
Yuthas 2006: 204) as a rational tool for managers to achieve economic efficiency and
effectiveness or other organizational goals. Some organizations have used ES to make
the culture less autonomous or to streamline operating practices across geographically
dispersed units and achieve tighter coordination (and control) throughout their business
(Davenport 1998). Technology is viewed as an exogenous causal factor (independent
variable) with calculative logic causing a specific outcome (dependent variable) of
achieving the goals (Pollock and Williams 2009).
Fifth, the managerial and technical understanding is shaped by the statements and rhetorics of technology supply. Technological practice is ahead of research in the ES domain (Pollock and Williams 2009: 53-56, 69), which means that there is a flow of
statements and rhetoric from ES vendors, industry analysts and consultants to researchers (which is sometimes adopted fairly uncritically) (see also Baskerville and Myers
2009). The ERP concept coined by the Gartner Group in the early 1990s (Jacobs and
Weston 2007) and the second-wave concept (Deloitte Consulting 1999) are two examples of topics where some management researchers and others tend to align with the
statements and rhetorics of technology supply. The rhetorics of technology supply are
persistent and the primary motive is to promote a particular discourse on how the world
21
could be (or should be) and induce the discourse (Pollock and Williams 2009: 54-55)
from the ES opinion formers (vendors, industry analysts, consultants) to ES consumers
(decision makers, users) and the ES research community – that is, alignment of interest
(Latour 1987). It is desirable, perhaps inevitable, to follow the technological pathways
stipulated by the discourse.
2.4 THE SOCIAL STUDY OF ENTERPRISE SYSTEMS
The term “social study of enterprise systems” is borrowed and adapted from Pollock
and Williams (2009: 8), and embraces studies where “the social” forms an essential part
of the understanding and interpreting of ES. These studies generally contrast the managerial and technical understanding by focusing on situated and localist explanations of
ES, on organizations’ uniqueness in structure and practices and the wider historical,
contextual and processual circumstances.
While the managerial and technical understanding of ES is imprinted by a rather prescriptive orientation (Kallinikos 2004) and “tends to be based on a … weak empirical
base” (Pollock and Williams 2009: 6), the social study of ES is to a large extent analytic
and exploratory, providing rich lenses for viewing and explaining ES in empirical settings.
The research stream is furthermore characterized by applying theoretical frameworks
such as actor network theory (ANT) (Elbanna 2008; Hanseth and Braa 1998; Quattrone
and Hopper 2006), science and technology studies (Koch 2001; Pollock and Williams
2009), structuration theory (Jack and Kholeif 2008; Pozzebon 2000; Volkoff 1999),
institutional theory (Gosain 2004; Soh and Sia 2004; Svejvig and Jensen 2009), sensemaking (Boersma and Kingma 2005; Jensen and Aanestad 2007) and other social theories.
The table below shows some of the arguments and ideas expressed by this research
stream:
Themes
Selected references
Ideas/arguments
ES as enabler and constrainer
(Kallinikos 2004; Quattrone ES influence patterns of cognition, action
and Hopper 2006)
and communication, but they do not definitely determine organizational actors’ be-
22
Themes
Selected references
Ideas/arguments
havior. They have profound effects on the
structuring of work practices and enable as
well as constrain human action.
ES as technology of power
and empowerment
(Elmes et al. 2005; Sia et al. ES serves as a mechanism for increased
2002; Tang et al. 2000)
power of management as the global, crossfunctional and real-time information is ideal
for increasing control.
However, users might also experience
greater empowerment because they obtain
more and broader visibility in operations
with real-time information enabling them to
be more responsive and autonomous.
Drift instead of control
(Ciborra and Failla 2001;
ES implementations can only be controlled
Elbanna 2008; Hanseth and to some extent and drift is inherent.
Braa 1998; Hanseth et al.
Drift might be useful for keeping commit2001)
ment among users or even necessary to
ensure user acceptance of ES.
The consequences of drift might be problematic for organizations because the unanticipated outcomes impede current or future
business operations and/or reduce the anticipated benefits.
ES are initially flexible but
turn into rigid inflexible
systems
(Hanseth and Braa 1998;
Lindley et al. 2008)
ES ship as multi-everything, flexible and
configurable systems, but solidify into rigid
inflexible systems once they are set (i.e.
configured and possibly customized).
This prevents or delays future development
in organizations and increases the cost of
change and innovation.
ES resist change instead of driving change.
The complexity of misalignment
(Berente 2009; Elbanna
2008; Gosain 2004; Hanseth and Braa 1998; Kholeif
et al. 2008; Soh and Sia
2004; Svejvig and Jensen
2009)
Misalignment is a fundamental and complex
problem associated with ES.
Studies point to some strategies to reduce
gaps and misalignment: (1) a high degree of
tailoring, (2) accept the drift of the solution
and benefits and/or (3) accept loose couplings.
Tailor-made software might be the solution
for certain institutional contexts in order to
overcome misalignment.
Table 3: Ideas and arguments within the social study of ES
The table above presents some of the themes in the social study of ES, and the following will detail these starting with basic assumptions about ES.
23
The basic epistemological and ontological assumptions associated with the managerial
and technical understanding of ES are not well articulated in the relevant literature, but
are embedded implicitly. This perspective seems to understand “ES as an object with
predefined structures enabling predefined actions by organizational actors,” alias a form
of technological determinism. Social studies of ES move away from technological determinism to social constructionism or the relational stance expressed by ANT (see also
Cordella and Shaikh 2006; Hanseth and Monteiro 1997). One mainstream position within social constructionism is the interpretive stance, which could be illustrated by Orlikowski’s (1992) “interpretive flexibility of technology” where users interpret and enact
IT differently – that is, multiple interpretations of the same object (Quattrone and Hopper 2006). However, as Kallinikos (2004: 10) argues, ES moves a step further in shaping human agency, contrary to some information systems (e.g. groupware, decision
support), as ES have a profound effect on the structuring of work and how they enable
or constrain organizational actors, so the interpretive flexibility might be an imperfect
description as it tends to suggest that anything (or everything) goes, which is certainly
not the case with ES. Quattrone and Hopper (2006: 220) propose a more adequate description: “Objects [like ES] should be more than social construction (to avoid relativism whereby anything goes and humans are king) and less than realism (to avoid reifying objects and ignoring interpretive flexibility),” and the theorization should be within
the two extremes. The quotation above can be translated into a less elastic account
(adapted from Kallinikos 2004: 10):
ES influence patterns of cognition, action and communication, but they do
not definitely determine organizational actors’ behavior. They have profound effects on the structuring of work practices and enable as well as
constrain human action.
Hanseth and Monteiro (1997) argue that the intermediate position between technological determinism and social constructionism expressed by “IT enables and constrains” is
too crude and informs us that ANT will bring us a more detailed understanding of the
complicated relationship between IT and its use, because ANT describes which and how
actions are enabled and constrained. However, the account above appears to capture a
mainstream understanding of ES across the literature in this research stream while the
24
relational stance associated with ANT is more peripheral, although relevant and innovative.
ES are, from the managerial perspective, implemented to achieve some organizational
goals and enforce some standard practices across the organization. They are implemented to tighten management control (Sia et al. 2002). However, studies point out that
the outcomes of ES implementations are ambivalent (Elmes et al. 2005; Sia et al. 2002).
A study of a global organization (ACRO) with 20 000 employees producing and selling
high-precision industrial products and undergoing a multi-year, multi-phased ES implementation showed that the ES at ACRO served as a mechanism for increased organizational control, as the ES vendors advocated and advertised, but the users at the same
experienced greater empowerment (Elmes et al. 2005). The users were empowered because they obtained more and broader visibility in operations with real-time information, enabling them to be more responsive and autonomous. The ES also acts as a panopticon (cf. Foucault 1977) for management as the global, cross-functional and realtime information is ideal for increasing control and surveillance by management. Elmes
et al. (2005) propose the concept of panoptic empowerment as an apparent paradox related to ES implementations involving power and empowerment. Sia et al. (2002) describe a similar study in a Singaporean hospital where the ES implementation facilitated
an increase in both control and empowerment. However, the management consciously
resisted the empowerment by reinstituting the power lost through the ES implementation. Both cases indicate that ES are a technology of power and control while the empowerment perspective is more ambiguous and can be disabled or enabled under certain
circumstances.
Although ES are implemented to achieve organizational goals and enforce standard
practices across the organization, some studies shows that drift is inherent in ES implementations. An ERP project in a large multinational reputable food and beverage company (with more than 25 business units) had the objective of implementing a single system across its units, replacing local systems, but the project drifted, due to user pressure,
to an ERP system that interfaces to many other systems instead of replacing them (Elbanna 2008). Norsk Hydro, a diversified Norwegian company, had the same goal, “to
have one shared universal solution” with their SAP system, but the solution drifted into
25
variants at each site (different tailoring) and the SAP solution changed from one coherent common system into a complex heterogeneous infrastructure (Hanseth and Braa
1998). The consequences of drift might be problematic as the case with Norsk Hydro
shows, because it is very expensive to upgrade and develop the heterogeneous infrastructure, but drift can also be useful to continue the commitment and contributions
from organizational actors (Elbanna 2008) or even a necessary price to pay “[t]he customization … seems to have been the price to pay to enroll local users and managers
into the project, and to counter to the objections that SAP was a significant step back in
functionality compared to existing local systems” (Hanseth and Braa 1998: 194).
The Norsk Hydro study presents another related problem. The SAP system ships as a
multi-everything, flexible and configurable system, but when the configuration is set
(and, even worse, customizations implemented) then SAP solidifies to concrete figurative speaking. Although it is technically possible to change the solidified heterogeneous
infrastructure it is very complicated in practice (resulting in long schedules and high
costs for re-engineering projects) (Hanseth and Braa 1998). Lindley et al. (2008) point
out the same problem with the lack of flexibility in ES, which impedes change and innovation by increasing costs or delays in implementations. This rigidity could even lead
to the rejection of many cost-saving or efficiency-enhancing projects because they are
not profitable. So, the consequences are severe for organizations and sometimes the
direct reverse of the promises from the ES opinion formers (vendors, consultants, industry analysts etc.), although in fairness it should be stated that the same audience loudly
warns against customizations (e.g. Beatty and Williams 2006; Hildebrand 2009), which
are one of the major “concrete creators” with the argument that vanilla implementations
are much easier to upgrade. Davenport (2000) has also addressed the inflexibility of ES
and he acknowledges the critique but asks “what is the alternative?” The old silo legacy
system was probably even more rigid and “the flexible object-oriented highly modular
system” is still hypothetical dream-ware – although customers demand these flexible
solutions (see the quotation from industry consultant Reed on page 17). Regardless of
the reasons, the understanding amongst organizations and users might be that ES resist
change instead of driving change due to the rigidity in the ES (see also Nash 2010).
26
The challenges with drift in ES implementation and inflexibility in ES are related to the
debate about misalignment. Misalignment is basically a misfit between the existing
practices in the organization and the best practices embedded in the ES, and the managerial objective is to re-engineer the practices and tailor the ES to achieve a tight coupling, which is important for integration and control (Sumner 2009). However, the gaps
can be quite great (Soh et al. 2000); a high degree of tailoring is one approach to reduce
the gaps and drift is another (unplanned) way emerging from implementation and use.
These strategies, alone or combined, might lead to the unwanted solidified concrete ES,
and to avoid this the preferred managerial advice is often to re-engineer the existing
practices (Beatty and Williams 2006; Hildebrand 2009) or “putting the enterprise into
the Enterprise System” (Davenport 1998). A third strategy is to accept a loose coupling
between the organization and the system, but this compromises the goals of ES implementations of tight integration and control (Sumner 2009). Alignment strategies are
filled with dilemmas, compromises and difficult decisions. A study of ERP implementation in three hospitals in Singapore by Soh and Sia (2004) illustrates the misalignment
problem. The ERP system was developed for the European and US markets, where the
institutional context at the society level for healthcare is marked by being either highly
subsidized (the European market) or paid for by healthcare insurance (the US market),
and the ERP system was inscribed (Latour 1987) with this logic, which is contrary to
the tradition in Singapore, where a complicated co-payment calculation depending on
bed-class etc. is widely implemented with invoices sent to both the patient and the state
for a stay in hospital. This is an example of a clash between the Western and Singaporean institutional contexts, shaping the implementation and use of the ERP package in
the three hospitals. Another study indicates many difficulties of the implementation of
ERP in highly regularized organizations in Egypt, a transitional country, because the
organizations have to obey certain laws and regulations, which are hard to build into the
ERP systems. The recommendation is to adopt tailor-made software in these kinds of
organizations (Kholeif et al. 2008), which is in stark contrast to the promises of global
systems with country-specific capabilities (Oracle 2008). Both studies are non-Western,
so is the problem that ES are designed in a Western context for Western organizations?
That is probably one reason and indicates that the “multi-everything ES” of course has
its limitations, but another study shows that the American institution NASA had similar
27
problems, and they solved the gaps between the existing practices in the organization
and the best practices supporting the ES by accepting a loose coupling; the study points
out the fact that a loose coupling can be vital to stabilize an ES (Berente 2009). Finally,
the two studies about the food and beverage company and Norsk Hydro indicate similar
alignment challenges (Elbanna 2008; Hanseth and Braa 1998). Resolving misalignment
is thus a fundamental and complex problem associated with the management, implementation and use of ES.
2.5 TOWARDS RECONCILING THE CAMPS
The previous sections have presented two contrasting perspectives on ES. Such descriptions of two camps tend to be more rigid and undifferentiated than the underlying literature deserves, and have a tendency to present the perspectives as black or white, right or
wrong – in short, dichotomous – and my presentation is no exception! This could lead
to an unfruitful polarization on equal terms with the “hard” and “soft” approaches to
research in the IS field (Fitzgerald and Howcroft 1998). However, a more viable approach is to reconcile the two camps. The concept of polarity can help advance this reconciliation and can be summarized briefly as “All magnets have both a north and a
south pole. Neither can exist without the other – if the north pole section of a magnet is
removed … a new magnet is created with both north and south pole … these poles exist
not in isolation of each other, but by virtue of each other” (Fitzgerald and Howcroft
1998: 323). The same kind of thinking can be applied in this context. The two camps
are not entrenched and isolated opposites, but are mutually dependent and overlapping,
supporting a viable debate. Much of the critical discourse in the social study of ES has
grown up in opposition to the managerial and technical understanding (Pollock and Williams 2009: 7). The social studies call attention to the complexities, difficulties and
struggles of ES (e.g. Hanseth and Braa 1998; Soh and Sia 2004), and sometimes give
the impression that it is a wonder that ES extends at all (Pollock and Williams 2009: 7).
The enormous growth in ES packaged software indicates (D’Aquila et al. 2009; Jacobson et al. 2007) that ES do extend and that a number of organizations must have some
degree of success with ES despite all the challenges, and it is relevant to ask why the ES
was a success, which factors contributed to the success and how benefits can be
achieved, to mention a few questions. The studies, reports about how to achieve a suc-
28
cessful implementation (e.g. Beatty and Williams 2006; Parr and Shanks 2003) and gain
value from ES (e.g. Deloitte Consulting 1999; Ross and Vitale 2000) etc., have their
raison d’être with the managerial and technical focus on the management, implementation and use of ES. It is thus essential to understand how the two camps contribute to
research and practice including their strengths and weaknesses, and move towards a
differentiated, nuanced and reconciled view!
This thesis utilizes institutional theory as the main theoretical framework, as argued in
the introduction. The branch of institutional theory (Greenwood et al. 2008b) that I use
applies to the social study of ES with its roots in social constructionism (Berger and
Luckmann 1966). This does not mean that I abandon the managerial and technical understanding of ES, but on the contrary that this study calls for a sociological approach to
answer the overall research question. The next chapter will describe institutional theory
in detail and its implications for ES research.
29
3 INSTITUTIONAL THEORY
“It is my strong conviction that institutional theory provides the most promising and
productive lens for viewing organizations in contemporary society” (Scott 2008: viii).
This chapter consolidates the discussion of institutional theory across the collection of
papers and complements it with an additional review of relevant literature to provide a
more in-depth presentation than presented in the papers. Central parts of this chapter
overlap with paper 1 but, in order to increase the readability and present a coherent account, I have chosen this approach. The chapter starts with two introductory sections
about institutional theory, followed by the presentation of five key features of institutional theory. A critique of institutional theory is then discussed, and the chapter finishes with a summary of institutional theory related to ES research.
3.1 INTRODUCTION
In discussing institutional theory it should be emphasized at the outset that it is a highly
complex and diverse theory (Currie 2009) spanning economics, political science and
sociology (Scott 2008) rather than a theory specific to enterprise systems or information
systems. The focus in this thesis will be on organizational institutionalism used in organization theory and sociology. Organizational institutionalism addresses the overall
question: What does the institutional perspective tell us about organizational behavior?
(Greenwood et al. 2008a: 1).
Institutional theory attempts to describe the deeper and more resilient aspects of how
institutions are created, maintained, changed and dissolved, and deals with the pervasive
influence of institutions on human behavior including the processes by which structures,
e.g. rules, routines, norms and belief systems, guide social behavior (Scott 2004; 2008).
Institutional theory has responded to empirical anomalies where “what we observe in
the world is inconsistent with the ways in which contemporary theories ask us to talk”
(March and Olsen 1984: 747). Studies of organizational change are often hard to fit into
the rational-actor or functional theories (Barley and Tolbert 1997; Orlikowski and Barley 2001; Powell and DiMaggio 1991; Scott 2008), and these persistent findings from
empirical studies provoke us to complement the rational theories (as technical contin30
gency or strategic choice) with alternative models better explaining organizational observations, and to move away from such atomistic accounts of social processes to a
wider understanding of institutional arrangements (DiMaggio and Powell 1991: 3) emphasizing the contextual, historical and processual aspects in which organizational actions take place (adapted from Currie 2009).
Institutional thinking has been around for well over a century and different historical
roots are amongst others Karl Marx, Emile Durkheim, Max Weber, Alfred Schutz and
Talcott Parsons. They all formed early understandings about institutions in a sociological context from normative elements like folkways, mores, rules and norms to cognitive
elements such as shared knowledge and belief systems (Scott 2008, chapter 1). Berger
and Luckman (1966) provided a link from this earlier work and later work of organizational scholars (Scott 2008) with their conceptual framework for analyzing the social
construction of everyday life habits, routines and institutionalization (Berger and Luckmann 1966) (see also section 4.3). However, most of the early work on institutions between 1880 and the mid-twentieth century paid little attention to organizations (Scott
2008), including Berger and Luckmann’s work (1966).
Institutional arguments began to be connected with organizational studies in the 1950s
by Robert K. Merton and his students, particularly Philip Selznick. Selznick draws on
Merton’s work that some consequences of actions are planned and others are unanticipated, as social action is not context-free but is constrained and shaped by the context.
Particularly significant are the constraints on action that arise from commitments enforced by institutionalization (Scott 2008: 20-23; Selznick 1949; 1957). Talcott Parson
argued that wider normative structures in the society legitimate organizations, and Herbert Simon put forward that value assumptions, cognitive frames and rules impact on
individuals’ behavior (Scott 2008: 23-26). Silverman (1971) attacked Parson’s and
Selznick’s structural–functional frameworks and focused on meaning systems, arguing
that meanings do not operate only in minds but also as objective facts residing in social
institutions – the environment is the “source of meanings for the members of organizations” (Scott 2008: 42). Two seminal papers were released in 1977 that introduced the
modern organizational institutionalism (Greenwood et al. 2008a) and appeared to be
very influential (Scott 2008). Articles by Meyer and Rowan (1977) and Zucker (1977)
31
built on Berger and Luckmann’s work on institutions and institutionalization. Meyer
and Rowan (1977) embraced views of institutions as complexes of cultural roles from a
macro perspective while Zucker studied the micro foundations of institutions with the
power of cognitive belief guiding the behavior of individuals (Scott 2008: 42-44). The
two seminal papers were followed by other influential articles such as those by DiMaggio and Powell (1983) and Meyer and Scott (1983) focusing on the macro (environmental) perspective, where the former discussed isomorphism (structural similarity) and the
latter took the stance that all organizations are shaped by both technical and institutional
forces.
The literature in this condensed historical presentation has made a substantial contribution to our conception of modern organizational institutionalism.
This introduction has deliberately avoided using the term “new institutional theory”
(Powell and DiMaggio 1991), because it is often understood as opposite to old institutional theory (DiMaggio and Powell 1991 see table 1.1 page 13) and as implicit replacement of the old with the new. Greenwood and Hinings (1996: 1048) argued that
old and new institutional theory have to be combined in order to understand radical
changes in organizations (see also Hirsch and Lounsbury 1997; Selznick 1996), and ES
implementations are about radical or major changes (e.g. Davenport 1996). ES research
focuses on change, organizations and unanticipated consequences, which are fundamental elements of old institutional theory, but can be extended to persistence (stability),
environment (field, sector or society) and unreflective activity, which are the core of
new institutional theory (DiMaggio and Powell 1991) – this underlines the relevance of
combining old and new institutional theory. This combined view is either referred to as
institutional theory or organizational institutionalism in this thesis.
The next section will describe essential institutional concepts.
3.2 INSTITUTIONS, INSTITUTIONALIZATION AND ORGANIZATIONS
Institutions are multi-faceted, durable, resilient social structures, made up of symbolic
elements, social activities and material resources (Currie 2009; Scott 2008: 48-50). Jepperson (1991: 143) denotes an institution as an “organized, established, procedure,”
which implies structure (organized), history (established) and persistence (procedure) –
32
these procedures become “the rules of the game” and an example is a standard operating
procedure in an organization. North (1990: 4-5) uses in the same vein the game analogy,
and presents an important, although simplified, distinction between organizations and
institutions: institutions are the rules of the games and organizations are the players.
Institutions exist at a variety of levels and take many forms: examples of institutions are
human rights, societies, contracts, families, handshakes and belief systems like Buddhism (adapted from Jepperson 1991). These institutions are interwoven and nested at all
levels from world systems at the macro level (e.g. the United Nations World Food Programme) to interpersonal interactions at the micro level (e.g. charity) – and they are
potentially contradictory (Friedland and Alford 1991).
Institutions are “inhabited by people and their interactions” (Hallett and Ventresca
2006: 213), and cannot be isolated from the logic of actual life. They constrain and empower people’s behavior, as institutions offer guidelines and resources for taking actions as well as imposing restrictions on action (Scott 2008: 48-50). Institutions are both
the result of past actions and the context in which ongoing actions occur (Scott 1995:
52), which means that they can be perceived as “properties” (state) of an existing order,
but also as a “process” (Scott 2008: 50) including the processes of institutionalization
and deinstitutionalization through incremental and radical changes (Tolbert and Zucker
1996).
Institutionalization is a process and to institutionalize is to infuse with value beyond the
technical requirements of the task at hand (Selznick 1957: 16-17). Institutionalization is
also a property (Zucker 1977), which happens when actions are repeated and given
shared meanings by actors (Berger and Luckmann 1966; Scott 2008, chapter 6), whereby the institution becomes stable and durable (Currie 2009). Both institutions and institutionalizations can be conceived as properties and processes.
Institutions and organizations: What are the organizational implications of institutions?
If we broadly understand institutions as the rules of the game, then an organization
plays multiple games at the same time. Organizations operate in pluralistic institutional
contexts and the internal functioning reflects the larger systems themselves (Kraatz and
Block 2008). This results in an organization that: (1) may have multiple institutionally
given identities, (2) may be the structural embodiment of multiple logics, (3) may be
33
legitimated by multiple mythologies and (4) takes for granted very different beliefs and
values – in short multiple things to multiple people. The organizational pluralism may
potentially lead to fragmentation, incoherence, conflict, goal ambiguity and organizational instability (Kraatz and Block 2008: 244). This portrays the organizational arena
for the management, implementation and use of ES.
The central institutional concepts presented will be applied in the subsequent sections –
although mainly implicitly. These following sections will elaborate on five key features
of institutional theory relevant to ES research. The five key features are isomorphism,
rationalized myths, bridging macro and micro structures, institutional logics and institutional processes and change.
3.3 INSTITUTIONAL AND COMPETITIVE PRESSURES LEADING TO
ISOMORPHISM
The modern organizational institutionalism introduced in the late 1970s focused on culture and cognition, where taken-for-granted rules lead to isomorphism in the formal
structures of the organization, and organizations had to conform to society for legitimacy (Meyer and Rowan 1977; Zucker 1977). DiMaggio and Powell (1983) “moved” the
focus on isomorphism from the society level to the organizational field level with coercive, normative and cognitive institutional pressures leading to isomorphism, which is
nowadays part of many institutional analyses. Isomorphism means “a constraining
process that forces one unit in a population to resemble other units that face the same set
of environmental conditions” (DiMaggio and Powell 1983: 149) or is simply expressed
as structural similarity. Liang et al. (2007) argue that cognitive, coercive and normative
institutional pressures impact on the assimilation of enterprise systems, for instance the
normative pressure in an organizational field, where suppliers, customers, consultants
and professional associations collectively assess and endorse IS innovations (Swanson
and Ramiller 1997), shaping the implementation and assimilation of enterprise systems
by providing institutional norms that guide top managers (Liang et al. 2007).
Isomorphism is an important consequence of both competitive and institutional pressures (Scott 2008), and one of the challenges of using institutional theory is to distinguish between the two kinds of pressures. Competitive pressures assume system ratio-
34
nality, often used in ES research (Dillard and Yuthas 2006), which emphasizes market
competition where organizations compete for resources and customers, and is closely
related to the technical environment where products and services are expected to be
produced in an effective and efficient way (Scott and Meyer 1991), but “organizations
compete not just for resources and customers, but for political power and institutional
legitimacy, for social as well as economic fitness” (DiMaggio and Powell 1983: 150).
Competitive and institutional pressures “live side by side” and we shall avoid dichotomous explanations, where e.g. social explanations exclude techno-rational explanations (adapted from Greenwood et al. 2008a: 32), and instead acknowledge that social
situations, such as ES in organizations, consist of interdependent non-rational and rational elements (adapted from Scott 2008). It is therefore difficult empirically to distinguish between these explanations, being reinforced by the fact that institutional explanations strive to appear technical in nature (Scott and Meyer 1991) as a disguise. Greenwood et al. (2008a) state that institutional theory is well suited to being juxtaposed with
other theories. This is taken up by Vitharana and Dharwadkar in their paper about IS
outsourcing (2007), in which they present competitive pressures explained by transaction cost theory and institutional pressures explained by institutional theory. The study
shows how organizational analyses can cover both rational and non-rational elements.
The next section about rationalized myths elaborates on the entangledness.
3.4 RATIONALIZED MYTHS
Rationalized myths are rationalized and impersonal rules that bind the various different
organizations through the belief in its legitimacy (Meyer and Rowan 1977). They are a
key theme related to institutional isomorphism where organizations have to conform to
these rationalized myths in order to be “proper” organizations (Boxenbaum and Jonsson
2008). Institutionalized products, services, techniques, regulatory systems, public opinions, professional standards, etc. act as powerful rationalized myths exerting institutional pressures on organizations in multiple and complex ways. Rationalized myths
may develop in organizations, where organizational actors believe that their responses to
these multiple pressures are aimed at organizational efficiency, but they are more aimed
at achieving legitimacy for the organization (Meyer and Rowan 1977). Alvarez (2002)
examined the role of myths in an ERP implementation. The old legacy system was
35
deinstitutionalized by creating a story of “performance crisis” and a myth-making
process took place “constructing the new ERP system as an integrated system,” which
was aligned with the overall organizational goals of the organization, but the benefit of
the integration was not supported by objectively testable facts. The rationalized myth
thus legitimized the ERP implementation, “and the story-making process served to align
the technology with ideal organizational values” (Alvarez 2002: 82). The case study by
Alvarez also shows the deinstitutionalization process of the old legacy system followed
by the reinstitutionalization process of the new integrated ERP system (Greenwood et
al. 2002; Scott 2008; Tolbert and Zucker 1996), and that narratives can support the institutionalization process (see also Hedman and Borell 2004), which can be a relevant
“technique” in practical ERP implementations.
3.5 MULTIPLE LEVELS IN INSTITUTIONAL THEORY BRIDGING
MACRO AND MICRO STRUCTURES
Institutional and competitive pressures are often exerted by the society and the organizational field on the organization, where the organizational field is defined as “those organizations that, in the aggregate, constitute a recognized area of institutional life: key
suppliers, resource and product consumers, regulatory agencies, and other organizations
that produce similar services or products” (DiMaggio and Powell 1983: 148). Scott
(2008) argues that it is beneficial to look at multiple levels in a given study in order to
enrich the understanding in institutional analysis, and this is exactly one of the powerful
features of institutional theory with its ability to operate at varying levels ranging from
the society, organizational field and organization to the individual actor level (Scott
2008: 85-90). What is likewise important is the reciprocal interaction between levels,
where macro structures in society are bridged by organizational fields to micro structures in organizations or even “down” to the individual actor level. Institutional creation
and diffusion thus happen, where top-down processes allow higher-level structures to
shape the structure and action of lower levels, while bottom-up processes shape, reproduce and change the context within which they operate (Scott 2008: 190-195). Scott’s
argument is mirrored by Currie (2009), who encourages IS researchers to work with
multiple levels and multiple stakeholders as this is the mainstay of institutional theory.
36
Soh et al. (2000) point out that the area of accounting and finance is moving towards a
global standardization because of international accounting standards initiated and promulgated by standard boards. “International Financial Reporting Standards” (IFRS) (Tsakumis et al. 2009) are such global standards. The IFRS were mandated by the European
Union in 2002 and have currently been adopted by about 100 countries, and others like
the US have a program in place to converge national standards with the IFRS (Tsakumis
et al. 2009). The IFRS are a global institution diffusing from the global level to the local
level through top-down processes. However, a study in Egypt shed light on the conflict
between the globalization and localization of management accounting standards. There
was resistance to the global best practices built into the ERP system and alternative local practices were perceived as more appropriate (misalignment). The study further
questioned the use of ERP systems to promote Anglo-American management accounting techniques (Kholeif et al. 2008: 241). This study shows how standards travel between multiple levels (see also Czarniawska and Joerges 1996), in this case top-down,
but also how organizations respond to the customization or use of custom software to
reinforce local practices (Kholeif et al. 2008: 241).
The next section will further advance how macro and micro levels can be combined by
using institutional logics, which can be understood as a social mechanism (Hedström
and Swedberg 1996) mediating the top-down and bottom-up processes.
3.6 INSTITUTIONAL LOGICS
There has been a strong focus on isomorphism within institutional theory (Greenwood
et al. 2008a), but this focus has changed nowadays and it is no longer so much on isomorphism, whether in society or within the organizational field, but more on the effects/processes of different, often conflicting, institutional logics on individuals and
organizations. “Institutional logics shape rational, mindful behavior, and individual and
organizational actors have some hand in shaping and changing institutional logics”
(Thornton and Ocasio 2008: 100), where institutional logics can be defined as “the way
a particular social world works” (Thornton and Ocasio 2008: 101). Institutional logics
link institution and action (see also Barley and Tolbert 1997) and provide a bridge between macro-structural perspectives (DiMaggio and Powell 1983; Meyer and Rowan
1977) and micro-process approaches (Zucker 1991). Multiple institutional logics are
37
“available” for organizations and individuals (Scott 2008), and the embedded agency in
institutional logics presupposes partial autonomy for individuals and organizations
(Thornton and Ocasio 2008), so actions, decisions and outcomes are a result of the interaction between an individual agency and an institutional structure (Friedland and Alford 1991; Thornton and Ocasio 2008: 103-104). Some IS researchers have addressed
institutional logics related to information systems or enterprise systems (Berente 2009;
Currie and Guah 2007; Gosain 2004). Gosain (2004) argues that a mismatch between
institutional logics in an enterprise system and the incumbent institutional logics in an
organization can lead to institutional misalignment. Varying degrees of mismatch between institutional logics in enterprise systems and organizations can lead to varying
degrees of institutional misalignment, which again can have problematic consequences
like resistance against the new enterprise system. Berente (2009) shows how conflicting
institutional logics are solved by loose coupling of practices with NASA’s ES. Contradictory institutional logics often guide organizational practices, given that organizations
are playing multiple games with multiple rules. ES are implemented across these practices and some are congruent with the best practices in the ES, enabling tight coupling
(after some adaptation), while other practices are incongruent, often resulting in loose
coupling (although promoting stability in an ES implementation). The study furthermore indicates that the stability achieved by loose coupling at the local level can enable
greater integration and control at the overall process level and thereby meet the objective of tight control and integration often associated with ES implementations. Other
researchers discuss misalignment between enterprise systems and organizations, which
is similar to the accounts of Gosain and Berente, although they do not use the “institutional logic” concept directly (Sia and Soh 2007; Soh and Sia 2004).
The concept of institutional misalignments presented by Gosain can be used to emphasize several aspects of institutional logics. First, Fligstein (2001: 100) criticized institutional theory for considering organizational actors to be passive recipients or “cultural
dopes,” using readily available scripts provided by government, professionals or other
institutional carriers to structure their actions. However, applying institutional logics
counters this critique, where an individual agency plays an important role in selecting
and changing institutional logics in the working practices (Thornton and Ocasio 2008),
so the users of an enterprise system might adopt the embedded institutional logics in ES
38
and then change the incumbent organizational institutional logics to fit “the ES logics.”
Thus, institutional misalignment is reduced, whatever consequences this may have, but
anyway this implies an agency from the organizational actors, who are guided by interest, power and opportunism. Second, the changes in institutional logics are part of (or
are) the institutional/organizational changes taking place in an organization, for instance
by implementing an enterprise system that could be designated a “precipitating technological jolt” for starting a change process (Greenwood et al. 2002: 60) (this will be elaborated in the next section). We can thus analyze the process and stages of change using
“institutional logics as a method of analysis” (Thornton and Ocasio 2008: 109-111).
Finally, the institutional logics perspective provides an approach to bridging macro and
micro perspectives, where e.g. the institutional logics “built into” enterprise systems
from “the original U.S./Western European development context” (macro perspective)
are then used for e-procurement by a purchaser in a Singaporean defense organization
(micro perspective) (Sia and Soh 2007).
The move away from focusing on institutional pressures, leading to isomorphism, to the
effects of institutional logics seems to be promising not least in enterprise systems research, because it is a way to “open” the enterprise system artifact (see also Orlikowski
and Iacono 2001) and to understand how micro processes unfold with ES as “institutions are reproduced through the everyday activities of individuals” (Powell and Colyvas 2008: 277).
3.7 INSTITUTIONAL PROCESSES AND CHANGE
The process perspective of institutions and institutionalization as briefly presented in the
introductory sections of this chapter has only been touched on implicitly in sections
about key features of institutional theory, whereas this section elaborates on how and
why institutions are created, maintained, changed and dissolved.
Scott (2008, chapter 6) discusses three underlying mechanisms for the process of institutionalization of social systems – based on increasing returns, increasing commitments
and increasing objectification. The latter will be taken up in this context as an expanded
version of Berger and Luckmann’s (1966) concept of objectification (see also section
4.3). Tolbert and Zucker (1996) propose a multistage model of institutional processes
39
consisting of innovation, habitualization, objectification and sedimentation. The institutional processes happen both intra- and inter-organizationally. Organizations innovate,
advancing new ideas, solutions and practices in response to changes in political, technological or market conditions. Some of these innovations prove more viable and become habituated within and between organizations, leading to “theorizing.” Theorizing
is the development and specification of abstract categories and the formulation of why
and how the innovation is effective. Objectification is the movement to a more permanent and widespread status that involves the development of some degree of social consensus among organizational actors concerning the value of a structure and an increasing adoption by organizations on the basis of that consensus. Sedimentation is the
process by which the innovation is perpetuated across several generations and diffused
to relevant potential adopters (Scott 2008: 125-128; Tolbert and Zucker 1996: 180-184).
However, innovations are not only created but also changed and dissolved, for instance
when an ES is replacing an old legacy system. This is addressed by Greenwood et al.’s
(2002) multistage model for institutional change. Change can be theorized as consisting
of several institutional processes (stages) starting with precipitating jolts initiating the
change followed by deinstitutionalization, preinstitutionalization, theorization, diffusion
and (re)institutionalization. Most of the phases mirror the work by Tolbert and Zucker,
but the deinstitutionalization phase is additional and indicates that the incumbent practices (innovations, enterprise systems etc.) have to be deinstitutionalized in order for
new practices to be institutionalized (see also Oliver 1992, who presents a detailed discussion about deinstitutionalization). Very few studies within ES research, and few in
IS research, appear to focus explicitly on the details of institutional processes, with
some exceptions (Alvarez 2001; 2002; Mattila et al. 2009; Nicholson and Sahay 2009;
Silva and Backhouse 1997; 2003; Sæbø et al. 2008), but that is also an opportunity for
studies as addressed in paper 5.
This section finalizes the discussion about five key features of institutional theory related to ES research. I will argue that these features are important for understanding the
institutional structures and processes shaping ES, which has been substantiated in the
papers (paper 1 … paper 5) in this thesis, although institutional logics appear to be particularly advantageous because they integrate macro/micro and structure/agency.
40
3.8 CRITIQUE OF INSTITUTIONAL THEORY
Critiques have been raised of institutional theory and this section gives a brief overview
of some of this criticism and how it is addressed in this thesis.
First, institutional theory is concerned with structural similarity among organizations as
a result of institutional pressures from the organizational field or society – as stated in
seminal papers (DiMaggio and Powell 1983; Meyer and Scott 1983; Meyer and Rowan
1977; Zucker 1977). This consistency across organizations is common in institutional
analysis and gives the impression that institutional theory is mainly about isomorphism
and stability (Lounsbury 2007), neglecting institutional change (Scott 2008).
Second, organizations are often treated as singular, homogenous actors as the bulk of
institutional literature focuses on macro-level phenomena (field, sector or society) with
the organization as the smallest unit. It makes sense in certain macro-level studies to
rely on a reductionistic understanding of organizations (Berente 2009: 21), but studies
of individual organizations involving ES typically imply a richer institutional view
(Friedland and Alford 1991).
The first and second critiques are addressed by institutional logics and with the combination of old and new institutional theory being a premise in this thesis. I acknowledge
the importance of isomorphism and stability in institutional theory, but argue that it has
to be complemented by an understanding of the multiple concurrent responses organizations have to multiple institutional pressures (organizational pluralism).
Third, the institutional understanding of how a pool of social ideas, instrumental orientations and schemes, exogenous to organizations embedded in a rationalized environment, is carried and translated to specific patterns in specific organizations through diffusion and adoption is inadequate and somewhat idealistic. This leaves out important
questions such as “why do some patterns diffuse relatively unchanged while others are
highly changed to unrecognizability?” and “why do some patterns survive while others
fail to do so?” (Hasselbladh and Kallinikos 2000). Institutionalization has to “abandon
the bird’s eye view of the field and come closer to the social and cognitive means and
procedures underlying rationalized beliefs and schemes of action” (Hasselbladh and
41
Kallinikos 2000: 700). This critique is only to some extent taken up in this thesis by
focusing on the organizational level with meso-micro institutional analysis.
Fourth, organizations and organizational actors are seen as passive recipients or “cultural dopes” who use readily available scripts provided by the government, professionals
or other institutional carriers to structure their actions (Fligstein 2001). Other scholars
(e.g. DiMaggio 1988) criticize institutional theory in the same vein for lacking explanatory power for why actors act as they do and what interests motivate them (e.g. to adopt
technology). Institutional logics and organizational pluralism (multiple games with multiple rules) bring in agency, and mitigate this critique.
To finish this discussion about critique it is appropriate to cite Donaldson (1995: 128),
who delivered a striking critique of institutional theory:
The institutional theory of organizations is one of the most prominent examples of the present-day tendency in American organization theory to depict organizations as irrational and duplicitous, as composed of myths and
symbols, and as creatures of the normative and cultural system of society.
Yet these theoretical claims remain mainly as grandiloquent assertions and
rhetoric, at variance with the findings of empirical research. Scott (1987)
has written of the institutional theory of organization as being in its adolescence. It is unlikely to enter adulthood.
Donaldson’s critique was raised in 1995 and institutional theory has developed very
much like institutional logics since then, which bring back agency to actors. However,
his principal discussion about what is rational and what is institutional is still a very
relevant and complex topic, although my conception is that any social situation consists
of interdependent non-rational and rational elements (Scott 2008: 217–218): the challenge is thus to distinguish between the two categories if they are dichotomous.
3.9 SUMMARY OF INSTITUTIONAL CONCEPTS AND IMPLICATIONS
FOR ENTERPRISE SYSTEMS RESEARCH
As shown in the previous sections, institutional theory has a number of attractions to
offer to enterprise systems research, especially as a way to look beyond the managerial
and technical understanding of ES so often embraced in enterprise systems research,
42
and it emphasizes social considerations to complement technical considerations. Below
is a table with a summary of the key features of institutional theory presented in the
previous sections augmented with their possible implications for ES research:
Institutional
and competitive
pressures leading to isomorphism
Key features of institutional theory
Implications for enterprise systems research
Organizations face both competitive and
institutional pressures leading to isomorphism (structural similarity).
Researchers should look beyond rational
explanations to institutional explanations
with regard to understanding the management, implementation and use of ES.
Institutional pressures could be coercive,
normative and cognitive.
Social situations consist of interdependent
non-rational and rational elements.
Rationalized
myths
Rationalized myths related to technology
are technical procedures, accounting, personnel selection or data processing. Such
institutionalized techniques establish an
organization as appropriate, rational and
modern, quite apart from its possible efficiency (Meyer and Rowan 1977).
Institutional pressures shape the implementation, use and management of ES, for
instance decisions to adopt a specific ES.
We are surrounded by rationalized myths
in enterprise systems research, whether it
is the ES itself that is a rationalized myth
or “best practices” like Business Process
Re-engineering (BPR), Total Quality Management (TQM) etc. embedded in ES or
the implementation and use process of ES.
Rationalized myths can be used as techniques in ES implementations.
Multiple levels
of analysis
bridging macro
and micro
structures
Institutional theory can be applied at varying levels of analysis ranging from the
society, organizational field and organization to the individual actor level.
ES research can be performed at different
levels, for instance at the organizational
field level examining the diffusion of specific enterprise systems, or at the organizational level understanding institutional
Top-down processes allow higher-level
misalignment between enterprise systems
structures to shape the structure and action
and organizations.
of lower levels, while the bottom-up
processes shape, reproduce and change the ES research can also take advantage of
context within which they operate.
combining micro and macro perspectives
where the institutional macro context
shapes the management, implementation
and use of ES in an organization’s micro
practices executed by actors.
Institutional
logics
Institutional logics are a set of material
practices and symbolic constructions linking institution and action, and they provide
a bridge between macro-structural perspectives and micro processes.
Enterprise systems embed institutional
logics, which are inscribed during development and implementation. The institutional logics in the ES constrain the use
process (Gosain 2004).
Institutional logic is the way a particular
social world works.
Institutional logics can be used to “open”
the ES and thereby counteract the tendency
to black box the IT artifact in ES/IS research.
Institutional logics are a promising theoretical lens for understanding the interaction
between enterprise systems and organiza-
43
Key features of institutional theory
Implications for enterprise systems research
tions both statically (structures) and dynamically (processes).
Institutional
processes and
change
Institutional processes and change can be
explained by multistage models:
The detailed processes for institutionalization and deinstitutionalization are relevant
for studying the ES implementation
(a) Innovation, habitualization, objecprocess (e.g. from an old to a new system).
tification and sedimentation
(b) Precipitating jolts initiating the
The stabilization and routinization of ES
change, deinstitutionalization,
after launch could also be studied as an
preinstitutionalization, theorizainstitutionalization process
tion, diffusion and
(re)institutionalization.
Table 4: Key features of institutional theory related to ES research
The table above highlights key features of institutional theory, which offers a distinctive
perspective on organizations, enterprise systems and their interplay that is highly relevant for enterprise systems research. However, this chapter also illuminates the complexity, ambiguity and diversity associated with institutional theory, so it is both an opportunity and a challenge to use institutional theory for instance to provide conceptual
clarity (Currie 2009).
The next chapter will present the basic philosophical assumptions in this research
project, which are linked to institutional theory through the work of Berger and Luckmann (1966).
44
4 BASIC PHILOSOPHICAL ASSUMPTIONS
4.1 INTRODUCTION
The purpose of this chapter is to elaborate and explain the underlying basic philosophical assumptions in this research project. These basic philosophical assumptions are
based upon my view and belief about (1) “what is the nature of knowable things?” (ontology), (2) “what do we know, and how do we know it?” (epistemology) (Brier 2005;
Burrell and Morgan 1979; Lübcke 2004) and, finally, (3) “how do humans make choices and act upon them?” (human agency) (Burr 2003: 201). The chapter provides neither
a detailed discussion of philosophy in science in general nor an elaboration of the highly
complex and equivocal variants of constructivism. The intention is thus solely to present
the basic philosophical assumptions for this research project so others can understand
and evaluate the underlying premises in relation to the project.
Fundamental assumptions and beliefs are paradigms or worldviews of their holders,
which are defined as follows by Guba and Lincoln (1994: 107):
… a set of basic beliefs (or metaphysics) that deals with ultimates or first
principles. It represents a worldview that defines, for its holders, the nature
of the “world”, the individual’s place in it, and the range of possible relationships to that world and its parts.
The intellectual source for my basic philosophical assumptions and beliefs about physical and social reality are based on Kant’s transcendental idealism, which could be summarized in the following famous quotation about mind and experience:
Though all our knowledge begins with experience, it by no means follows
that all arises out of experience. (Kant 1781 (2007): 25)
A more comprehensive evaluation of Kant’s transcendental idealism is presented in
appendix A. I am not a dogmatic Kantian, but I find some useful patterns in his thinking
congruent with mine, which have influenced my view of social constructionism presented in this chapter.
45
The chapter is laid out as follows. In the next section, different perspectives on constructivism are discussed, and my view is positioned. This is followed by a detailed description of social constructionism based on Berger and Luckmann’s work (1966). Then
a discussion about human agency is presented and the chapter concludes with a summary. There is progression in this chapter starting with abstract philosophical thoughts
moving towards more concrete research assumptions.
4.2 ONTOLOGICAL AND EPISTEMOLOGICAL PERSPECTIVES ON
CONSTRUCTIVISM
Constructivism has ancestors in philosophy where points of view are called idealism
(transcendental idealism, German idealism etc.). Kant is a central figure in idealism and
his position can be viewed as the original form of constructivism, although contemporary constructivism differs on several points, for instance Kant is individualistic with the
“Transcendental I” while modern constructivism generally is collectivistic, which
means that the role of constructor is typically ascribed to a community of actors (Collin
2003: 13-15; Fuglsang and Bitsch Olsen 2004: 356).
Constructivism covers a wide range of ideas and conceptions, and can most likely best
be classified as rather equivocal (Martínez-Delgado 2002: 841), and with a huge variety
of terms describing perspectives, which might be more or less similar (Geelan 1997:
17).
Collin (2003) presents three dimensions related to constructivism, which are described
below:
Dimension I: Epistemological constructivism and ontological constructivism are two
main categories within constructivism, and can be defined as follows:
(1) Epistemological constructivism views “knowledge” as something that individuals
and groups construct from their own choices, perhaps in interaction with non-humans.
Acceptance, not “truth” is the key step in making something knowledge (Wink 2006:
113).
(2) Ontological constructivism views both “reality” and “knowledge about reality” as a
construction, constructed by individuals, groups or others (e.g. episteme and non-
46
humans) (adapted from Collin 2003: 23; Johannessen and Olaisen 2005: 1271). Epistemological constructivism might be seen as embedded in ontological constructivism,
because knowledge is a construction in both cases.
Dimension II: We can further distinguish between physical reality and social reality.
Nature and physical artifacts, constructed by man, both belong to the physical reality –
natural science is engaged in physical reality, but is also part of everyday life practice.
Social reality is broadly speaking constituted by human interactions and the intentions
that humans ascribe to these interactions – social science and humanities are engaged in
this domain as is everyday life practice (Collin 1997; Collin 2003; Wenneberg 2002).
Wenneberg (2002: 177-182) refers to this dimension as a distinction between nature and
culture and argues that the two reality spheres presuppose each other.
Dimension III: We can finally ask “who is the constructor?” Some constructivists will
say researchers (scientific knowledge), and others will mention ordinary individuals and
groups (everyday life practice). To these must be added the more abstract constructors
like Foucault’s episteme, actor–network theory’s non-human actors and others (Collin
2003).
My position is that social reality does not exist independently of human cognition and is
therefore a social construction. Therefore, the knowledge about the social reality is also
a social construction – that is, ontological constructivism. Physical reality exists independently of human cognition, but our knowledge about the physical reality is a social
construction – that is, epistemological constructivism. The position is elaborated in the
table below (adapted from Fitzgerald and Howcroft 1998: 319):
Physical reality – nature and manmade physical artifacts
Epistemological
assumptions
Social reality – culture
Social constructionism: Our scientific and everyday knowledge about the
physical and social reality is a social construction
Ontological assump- Realist: The physical reality consists
tions
of pre-existing hard, tangible structures that exist independently of human cognition
Relativist: The social reality consists
of multiple realities and exists as subjective constructions of mind
Category of construc- Epistemological constructivism
tivism
Ontological constructivism
Table 5: Epistemological and ontological assumptions in this thesis
47
Constructivism has so far been treated at a superior level mainly aimed at discussing
fairly abstract ontological and epistemological assumptions. The next section will discuss social constructionism in more detail, coming closer to more concrete research
assumptions.
4.3 SOCIAL CONSTRUCTIONISM
This section seeks to narrow the understanding and use of constructivism in this thesis.
Geelan (1997) limits the earlier discussion about “who the constructor is” to a personal–
social dimension where personal constructivism means that the construction of meaning
and social reality takes place through individual cognitive processes, while social constructionism (or social constructivism, see appendix B for elaboration) implies that
meaning and social reality are constructed through social processes and interaction
(Young and Collin 2004: 375).
The premise in this research project is that meaning and social reality are socially constructed, which is social constructionism. This is more an appropriate premise than a
“deep” philosophical belief as I acknowledge the debate about the constructor to be both
complex and comprehensive (Collin 2003; Geelan 1997). I will avoid unfolding this
here, but will be content with what I find plausible, namely that some kind of intersubjectivity exists where shared meanings are constructed through interactions with each
other.
One of the most influential and explicit thematizing of social construction (social constructionism) can be found in Berger and Luckmann’s book “The Social Construction of
Reality” (1966), in which they present a conceptual framework that makes it possible to
analyze the social construction of everyday life habits, routines and institutionalization.
It is about analyzing the reality of everyday life and the knowledge that guides actions
in everyday life, and only peripherally knowledge from theoretical perspectives (scientific knowledge). The basis for knowledge in everyday life is “the objectivations of subjective processes (and meanings) by which the intersubjective common-sense world is
constructed” (Berger and Luckmann: 33-34).
48
Berger and Luckmann’s approach to the process of social construction makes up a dialectical frame integrating two opposing views in the form of structure and action: that
society is created by humans, but society also influences humans, or, very much to the
point: “Society is a human product. Society is an objective reality. Man is a social product” (Berger and Luckmann: 79); this is then a dialectical process.
The social construction of reality can be explained as a process consisting of four stages: subjectification, externalization, objectification and internalization. Subjectification
is the process whereby we create our own experiences and can be explained by intentionality (Arbnor and Bjerke 1997). Consciousness is always intentional, and our consciousness is directed at some object, whether this object belongs to the external physical world or is inside our minds (Berger and Luckmann 1966: 34). This means that our
mental acts are intentional (based on Husserl philosophy in Arbnor and Bjerke 1997:
40).
The process of externalization is where the individual subjective experiences are made
externally available and shared by others through language and other communicative
means (e.g. facial mien); and they can be perceived, negotiated, scrutinized, reshaped
etc. in order to reach a shared understanding of social reality (intersubjectivity) (Berger
and Luckmann 1966).
These shared understandings can become typified through an objectification process
(taken-for-granted elements) (Arbnor and Bjerke 1997). Berger and Luckmann explain
objectification like this:
Objectification ... manifests itself in products of human activity that are
available both to their producers and to other men as elements of a common
world (1966: 49).
Objectification extends beyond face-to-face situations by creating some enduring elements. Humans create symbols, which are things that can carry meaning beyond here
and now (Burr 2003: 185). Objectification can be divided into two phases: institutionalization and legitimization. Institutionalization happens when actors share common typifications of some habitualized actions, and actors themselves are typified (Berger and
Luckmann 1966: 72). Cooking in a restaurant – a typified action – is carried out by a
49
cook – a typified actor. The institution emerges of this typification (Arbnor and Bjerke
1997: 179), so the restaurant is an institution. Institutions are a product of a historical
process (Berger and Luckmann 1966: 72). Legitimization is a second-order objectification explaining and justifying institutional order, and is needed when institutional order
has to be transmitted to new members. Institutions are explained through cognitive validity of objectivated common knowledge, for instance what is right and wrong in an
institution, and justified by normative values like “a good man takes care of his family”
(Berger and Luckmann 1966: 111).
The last stage is internalization, which means that the objectivated world is retrojected
(thrown back) to human consciousness during a socialization process (Berger and
Luckmann 1966: 78-79). Berger and Luckmann distinguish between primary socialization, which an individual undergoes in childhood, thereby becoming a member of society, and secondary socialization as the internalization of institutional “sub-worlds”
(1966: 150, 158), that is, for instance, becoming an employee of a company or a member of a sports club. We accept and become members of an established sub-world,
where others already live, during the internalization (socialization) process (Arbnor and
Bjerke 1997: 180). Confer also with Schutz’s (2005) multiple realities and life world
concept.
The four-stage model described above is illustrated in the figure below, and two concepts are added: production and reproduction of social reality (adapted from Arbnor and
Bjerke 1997: 182; Norén 1998: 172):
50
Figure 3: The production and reproduction of social reality
Organizational theory based on social constructionism has circled about the objectified
reality and how this reality is “produced” (created) and “reproduced” (maintained). This
is illustrated in the figure above as two processes with orthogonal timelines. The vertical timeline is the process where the social reality is produced and changed, and the horizontal line shows the process where the social reality is maintained and stabilized. The
concepts of time and process are exposed as straight linear in the figure, but should instead be understood as dialectical and iterative (Norén 1998: 171-172). An individual
human being can simultaneously externalize his or her actions into the social world, and
at the same time internalize the objectified reality experienced (Berger and Luckmann
1966: 149).
The figure above is a summary of this section about social constructionism with the
four-stage model embedding two processes concerning the production and reproduction
of social reality. The next section will deal with human agency, that is “how do humans
make choices and act upon them?” (Burr 2003: 201).
51
4.4 HUMAN AGENCY
The discussion about human agency implies several questions about the relationship
between individuals and institutions (or more general society) (adapted from Burr 2003:
182):
1. Do individuals have the power to change themselves, their roles in an institution
(organization) and their life stories?
2. Do they hold a capacity to change the institution they work in (or more generally
the society they live in)?
3. Are individual human beings determined by institutional structures that are
beyond their control?
4. Are “some” individuals (for instance managers or other influential persons) able
to change the institutional structures?
These are demanding and complex questions about human agency, and they will be discussed from two opposing sociological points of view as presented by Burrell and Morgan (1979). Human agency is either (1) deterministic, indicating that individual human
beings respond mechanistically and even deterministically to the external world, or (2)
voluntaristic, meaning that individuals have a free will and shape the external world.
Determinism and voluntarism could be seen as two extreme opposing views of human
agency.
Berger and Luckmann are positioned somewhere in between the two extreme views.
The interaction between humans and society (institutions) is a dialectical process in
which society is a human product and man is a social product (1966: 79). The person
has agency constructing the society (institutions) and is constrained by the society –
“the person is as much constructing as constructed” (Burr 2003: 183, 186-187).
Thus, the answers to the questions introducing this section are both yes and no. In
theory, we can act as we like, but in practice we are constrained by the structures imposed on us (Boudreau and Robey 2005: 15), and this is aligned with my stance.
4.5 SUMMARY OF BASIC PHILOSOPHICAL ASSUMPTIONS
This chapter contains a considerable amount of philosophical discussions, and it is appropriate in this final section to summarize the essential parts about my view and belief.
52
First, the social reality consists of multiple realities and exists as subjective constructions of mind. The knowledge about the social reality is a social construction – that is,
ontological constructivism. Physical reality exists independently of human cognition,
but our knowledge about the physical reality is a social construction – that is, epistemological constructivism. Second, social constructionism implies that intersubjectivity
exists where shared meanings are constructed through social processes and interactions.
Finally, human agency involves a dialectical process in which we are constructing and
being constructed.
These are the basic philosophical assumptions that this research project rests on, which
in short is social constructionism. The assumptions are aligned with institutional theory
via Berger and Luckmann’s work. The next chapter will present the research process
and how it relates to the assumptions presented in this chapter.
53
5 RESEARCH PROCESS
5.1 INTRODUCTION
Embarking on a research project involves a lot of choices and decisions for the researcher. On the one hand the journey can go nearly everywhere, but on the other hand
there are some boundaries and limitations related to (1) the researcher’s belief, preferences, knowledge and experience; (2) practical circumstances like access to case study
companies and funding; and (3) institutional frames guiding the research such as research traditions, politics of research and ethics. It is among these many choices and
boundaries that a specific research journey will pass off. The organization of this chapter and the representation show to some extent a rational, well-considered and carefully
planned and executed research process, but the journey has also been impinged with
messy situations, unanticipated activities and several diversions. It is important to distinguish between “practiced logic” and “reconstructed logic.” The practiced logic is the
actual path taken while the reconstructed logic is the reconstruction to present a rational
and clear account (Arbnor and Bjerke 2009: 49-50; Kaplan 1964: 3-11). The reconstructed logic has to appear as consistent and coherent, that is, “methodological congruence,” which means that the purpose, research questions, basic philosophical assumptions, theoretical lenses, empirical settings and research methods are all interrelated, so
the study appears as a cohesive whole rather than as fragmented, isolated parts (adapted
from Creswell 2007: 42) (although methodological congruence is objected to by Feyerabend (1975), who argues that “anything goes”).
This chapter addresses the more practical part of the research process concerning research design, data collection, data analysis, evaluation and reflection. The chapter furthermore aims at providing transparency and clarity about how the research process has
been carried out. This includes how my understandings, interpretations and conclusions
were formed in close interaction with the field, discussions with colleagues and by consulting literature. It is my intention to open the “black box” with these descriptions and
reflections and provide the reader with a foundation to understand how the thesis came
about and thereby to evaluate and follow my way of thinking.
54
The figure below outlines the reconstructed research process (inspired by Kragh 2007;
Myers 2009):
Figure 4: Outline of the research process
The figure above shows the various research activities and their linear relationship (unbroken arrows) as well as the iterative nature of the research process (dashed arrows on
the right). The first part of the research design has been described in previous chapters
(basic philosophical assumptions, research purpose, research questions and theoretical
framework) while this chapter will discuss the other activities. Some arguments from
the research method section in the introductory chapter will be repeated as they are central to the overall understanding of the research process and to presenting a coherent
account.
55
5.2 INTERPRETIVE RESEARCH
The basic philosophical assumptions summarized as social constructionism guide the
overall research approach, and this implies a qualitative endeavor (Schwandt 2000) delimiting the research methodology to qualitative research, described as:
Qualitative research is a situated activity that locates the observer in the
world. It consists of a set of interpretive, material practices that make the
world visible. The practices transform the world. They turn the world into
series of representations, including field notes, interviews, conversations,
photographs, recordings, and memos to the self … This means that qualitative researchers study things in their natural settings, attempting to make
sense of them, or interpret, phenomena in terms of the meaning people
bring to them (Denzin and Lincoln 2005: 3).
Interpretive research (Walsham 2002; 2006) is occasionally used as a synonym for all
qualitative research, according to Schwandt (2007:160-161), which he argues blurs important distinctions, and he proposes instead three epistemological stances within qualitative research: “interpretivism,” “hermeneutics” and “social constructionism”
(Schwandt 2000) (several other categorizations of qualitative inquiry exist (e.g. Creswell 2007; Myers 2009; Patton 2002 see comparison on page 79)). Interpretivism implies that meaning is reproduced or reconstructed from the original meaning of action,
while social constructionism has the conception that human beings do not find or discover meaning but that they construct or make meaning (Schwandt 2000: 193, 197).
Social constructionism is often combined with interpretivism, despite the epistemological differences argued by Schwandt, which is also the stance I take in this thesis. I see
social constructionism as an overall paradigm or worldview and interpretivism (interpretive research) operating at a less philosophical level that is closer to practical research (Creswell 2007: 19-21, 23-25).
In his seminal papers and book Walsham (1993; 2002; 2006) advocates for interpretive
studies in IS research when the focus is on organizational issues involving information
systems and human action. He states:
56
Interpretive methods of research start from the position that our knowledge
of reality, including the domain of human action, is a social construction by
human actors and that applies equally to researchers. Thus there is no objective reality which can be discovered by researchers and replicated by
others, in contrast to the assumptions of positivist science. Our theories
concerning reality are ways of making sense of the world and shared meanings are a form of intersubjectivity rather than objectivity. Interpretivism is
thus an epistemological position, concerned with approaches to the understanding of reality and asserting that all such knowledge is necessarily a social construction (Walsham 1993: 5).
Interpretive research implies a double interpretation with subject–subject relation to its
field of study, which means that interpretive researchers are subjects and they are just as
much interpreters of social situations as the people being studied (also subjects) (Myers
2009: 39). Geertz (1973) expresses this in a very nice way, as cited in Walsham (2006:
320): “What we call our data are really our own constructions of other people’s constructions of what they and their compatriots are up to.” Meanings are what constitute
the facts, and meanings are in a given context, which defines the situation and makes it
what it is (Myers 2009: 39-40). Meanings are constructed from real people in real situations undertaken in real-world settings (adapted from Hirschheim and Klein 2003: 259).
Interpretive research can take advantage of a contextualized approach (Walsham 1993)
linking context, content and process concerning transformation and change (Pettigrew
1985; 1987) and institutional theory corresponds well with the contextualized approach
in the kind of study I undertake, with its focus on context (society, sector, field and organization) and institutionalization processes. The contextualized approach can thus be
used to bridge institutional theory to practical research.
I have based my research on the interpretive tradition inspired by the contextualistic
approach linking context, content and process. I have studied the context for SCANDI
and to a much lesser extent Oracle from public documentation such as newspaper and
magazine articles as well as internal SCANDI documents, but also from interviewing
former and present SCANDI and Oracle employees. The historical course has gained
considerable attention, for instance SCANDI’s shift from monopoly to competitive
57
market in the 1990s as this is reflected in people’s conception of SCANDI and how the
ES has developed. The changes to content were considered amongst others the drastic
reduction of customizations in the ES, and how this was understood and interpreted by
accounts clerks and purchasers in their local settings (finance department and purchasing department). I aimed to understand the life world of the employees (mainly their
working life) and why and how they acted as they did in their local settings. I strived to
be open in my approach and “to let people speak for themselves.” I was conscious that I
was constructing meanings as an active and involved researcher drawing on long practical experience with ES in organizations. The scenery for these discussions has been
the overall context (more or less implicitly) but also the broader longer-term processual
perspective from the first installation of Oracle E-Business Suite in 1996 until today,
and especially the reimplementation process from 2007 to 2009, which I followed from
the beginning of 2008 until the end of 2009.
5.3 CONDUCTING INTERPRETIVE CASE STUDY RESEARCH
Walsham (1993; 2002) recommends in-depth interpretive case study research as a guide
to performing interpretive research in his earlier writings, but in a later paper he adds
ethnography, action research and critical research as relevant to IS research based on the
interpretive approach (Walsham 2006).
I have chosen interpretive case study research because it is appropriate when the study
has a descriptive and exploratory focus (Marshall and Rossman 1989: 78). Case studies
can be valuable for generating an understanding of reality (Yin 2003) and constructing
rich descriptions and empirical evidence (Myers 2009). Thus, an appropriate approach
is an in-depth study that a single case provides, and it has been termed a “revelatory
case” (Yin 2003). Case study research could be defined as:
An empirical enquiry that investigates a contemporary phenomenon within
its real-life context, especially when the boundaries between phenomenon
and context are not clearly evident (Yin 2003: 13).
Another definition is suggested by Myers (2009: 76):
58
Case study research in business uses empirical evidence from one or more
organizations where an attempt is made to study the subject matter in context. Multiple sources of evidence are used, although most of the evidence
comes from interviews and documents.
My study fits well with the definitions above: the subject matter or phenomenon studied
is the management, implementation and use of ES in organizations. This is a contemporary phenomenon where the boundaries between ES, organizational actors and the organization (and its environment) are not distinct – this is even more prevalent when we
study ES from an organizational or social perspective, as this study does.
The overall research question of “How do institutional structures and processes shape
the management, implementation and use of enterprise systems?” is furthermore consistent with Yin’s (2003: 9) recommendation about choosing a case study approach when
“a how or why question is being asked about a contemporary set of events, over which
the investigator has little or no control.”
The purpose of case study research is to use empirical evidence from people in situ in
natural organizations to make an original contribution to knowledge (Myers 2009: 73).
An interpretive case study allows us to gain insight into multifaceted meanings and
complex processes associated with the management, implementation and use of ES
whose dimensions and structures are not yet completely understood (adapted from Jensen 2007: 74).
Once the research method has been chosen, in this situation interpretive case study research, more steps follow. First, one or more organizations have to be selected based on
some criteria and boundary conditions relevant to the subject matter to be studied (Miles
and Huberman 1994). Second, appropriate data collection techniques have to be chosen,
which support the objective of the research and are possible to carry out in the case organization. Multiple sources are often used in case study research (Creswell 2007).
Third, once the data have been collected, the next step is to analyze the data and use the
analysis to write up the findings to make a contribution to knowledge (Myers 2009).
The mentioned steps will be dealt with at length in the coming sections together with
my interpretations, decisions and conclusions during the research process.
59
A key question for researchers concerns the role of theory in their research, regardless
of the research tradition. Walsham (2002) presents three distinct examples of the use of
theory based on Eisenhardt (1989), and these are presented in the table below together
with the specific way that I have used institutional theory in this study:
Use of theory
The specific use of theory in this study
As an initial guide to the design Institutional theory has guided the overall research design to
and data collection
include a focus on the context embedding the management,
implementation and use of ES, and to consider multiple levels
in data collection.
Institutional theory has furthermore guided specific issues such
as interview guides.
As part of an iterative process
of data collection and analysis
The use of institutional theory has framed the data collection
process, but has also sharpened my senses and “capacity for
surprises.”
Institutional theory has been used broadly in the data analysis
process as a sensitizing device.
As a final product of research
Institutional theory is an integral part of the final product of
this research theorizing about ES in organizations.
Table 6: The specific use of theory in this study
The specific use of theory in the table above will be elaborated subsequently. My immediate approach to entering the field was to follow the reimplementation project and to
understand how the ES were used in different departments, with a rather micro-level
focus on process and content. However, institutional theory guided me to complement
this approach with a context-based perspective (involving the meso-macro level), which
implied: (1) interview guides were prepared to include context-specific questions to
discuss the institutional structures framing the interviewees’ life world; (2) listening and
questioning in participant observations and informal talks were also framed to consider
contextual issues and (3) additional sources such as strategy documents, project documents, press releases, newspaper articles, white papers etc. were included to construct a
richer understanding of the contextual situation, the institutional structures and the historical course.
The use of institutional theory has framed the data collection process, but has also sharpened my senses and “capacity for surprises” (McCracken (1988) cited in Jensen (2007:
76)) – theory has been an enabler and constrainer, and how you study the world deter-
60
mines what you learn about the world (Patton 2002: 125). Outsourcing decisions related
to institutional circumstances (paper 2) were one of these surprises emerging from the
fieldwork, and were not considered a topic for this study initially. I have tried “to preserve … [some] degree of openness to the field data, and a willingness to modify initial
assumptions” (Walsham 2002: 105), for instance by following the dynamics of the
reimplementation project at a practical level and allowing for openness in interviews.
Institutional theory has broadly been used in the data analysis process as a sensitizing
device (Klein and Myers 1999: 75-76; Patton 2002: 452-462) such as searching for taken-for-granted beliefs, ideas diffused to the organization and many other things (cf. section 5.7). The aim of this thesis is to theorize about ES in organizations using institutional theory, which metaphorically speaking means that institutional theory is not a
scaffold that can be removed after the construction process, but on the contrary is an
integral part of the building (Walsham 2002: 105).
A final topic in this section is generalization from interpretive research. A typical claim
from a positivistic stance is that “you can’t generalize from a single case study,” but this
is opposed by Walsham (Walsham 2002; 2006) and others (Flyvbjerg 2006; Lee and
Baskerville 2003; Silverman 2005; Yin 2003). Generalizing from a particular case to
construct a theory is called analytic generalization (Schwandt 2007: 5), where theory
means something as the development of models or concepts, drawing specific implications and contributions of rich insights (Walsham 2002). These theories are not systematic, unified causal explanations of social phenomena (Schwandt 2007: 5), but instead:
… tendencies, which are valuable in explanations of past data but are not
wholly predictive for future situations … The generalization … should …
be seen as explanations of particular phenomena derived from empirical interpretive research in specific IS settings, which may be valuable in the future in organizations and contexts. (Walsham 2002: 110)
I have so used analytic generalization to theorize about ES in organizations.
5.4 SELECTING THE CASE AND ENTERING THE FIELD
Selecting case(s) in the interpretive tradition is not about sampling logic. Confidence
levels and intervals are meaningless and using three or four cases is not better than one
61
case as a case study is not an instance in a larger random sample (see also the discussion
about generalization in the previous section). The plausibility of the case is more important than its design, and presenting a complete, interesting and believable (e.g. with sufficient evidence) story (Myers 2009: 82-85).
The vehicle for interpretive research is often in-depth case studies with frequent visits to
the field over an extended time period (Walsham 2002: 101). I was looking for an organization where I could conduct such an in-depth case study concerning the management, implementation and use of ES with the following specific requirements, which
have been defined based on my practical experience and the literature (e.g. Shanks et al.
2003). First, it should have a comprehensive and companywide ES, preferably a highend system like SAP or Oracle where I expected to find many complex organizational
and technical facets. Second, I wanted to follow a major change process over a longer
period in its real-life settings before and after the launch of an ES. This is based on the
assumption that the dynamics of institutionalization are best studied when a major
change happens, allowing the researcher to use a “before and after” research design
(Barley and Tolbert 1997: 103-104). This requirement created a critical schedule dependency between the empirical work and my research project. Third, the organization
should be represented in Denmark due to practical geographical considerations and the
available funding, but the organization might be part of an international organization.
Fourth, the organization should be prepared for a long-term engagement (a year or even
more), open to the research process and willing to incorporate the researcher into its
business activities (participant observations, access to confidential documents etc.).
These were quite demanding requirements!
SCANDI was the second organization I contacted and they accepted the requirements. I
had some previous experience with SCANDI as a consultant several years ago, which
meant that I had some knowledge about the company (products, business, culture etc.).
SCANDI had started a major ES reimplementation project (RE-ES project) in summer
2007 and the launch was planned for May 2008, which fitted very well with my research schedule. SCANDI was furthermore very open and collaborative towards the
research project, which fertilized the ground for extensive fieldwork. My initial contact
person was the project manager of the RE-ES project.
62
When I entered the field in January 2008 I was promised the opportunity to follow the
reimplementation project and its stakeholders, but neither SCANDI nor I could know
how the major change process would progress. “It was real-time research” and I could
only strive to construct a complete, interesting and believable story based on my fieldwork. The next section will elaborate on the empirical context.
5.5 BRIEF ABOUT THE EMPIRICAL CONTEXT
The purpose of this section is to give a brief overview of the empirical context while the
details can be found in papers 2, 3, 4 and 5.
SCANDI is a Scandinavian company with more than 10 000 employees, which produces and sells high-tech services. The first company in SCANDI was established in the
late 1890s, and the company today is a result of a merger between several companies.
SCANDI and its predecessors operated for many years as territorial companies in a
highly regulated context, meaning that customers were only allowed to buy utility services from them. This monopoly situation changed in the 1990s when the Scandinavian
countries decided to deregulate and liberalize the utility market. SCANDI now faced
higher competition although only from a few strong competitors.
The low level of competition before liberalization has marked SCANDI’s internal
context of structure and culture (Pettigrew 1987). A consultant describes SCANDI as:
“… a supertanker that does not have all the needed engines to react promptly enough or
all the engines are not started simultaneously … or said more directly you cannot lay off
more than a [certain] number of employees and SCANDI has a huge backlog.” This
statement indicates one of the business challenges that SCANDI faces, where a truce
between trade unions and SCANDI limits its maneuvering. At the same time the
company benefits from its many years of monopoly by having a big market share and
owning a considerable part of the utility infrastructure in the present market, so the shift
from monopoly to competition implies both opportunities and challenges.
In 1996 SCANDI decided to implement Oracle E-Business Suite (OEBS), also known
as Oracle Financials (James and Seibert 1999). The figure on the next page shows the
timeline for the implementation and use of the enterprise system, and the connection to
major OEBS releases from Oracle Corporation.
63
Year
Financial System
(FinSys)
Supply Chain
System (SCS)
Oracle
E-Business Suite
Research
Oracle Financials
Launched in the US
1989
1990
1991
1992
International version
1993
1994
1995
1996
Launch FinSys
1997
1998
1999
Release 11i launched
2000
2001
Start SCS project
2002
Launch SCS
2003
2004
Upgrade of FinSys
Upgrade of SCS
Major conversion
of legacy system
2005
2006
2007
Start of RE-ES project
Release 12 launched
Research period
2008
2009
Launch of RE-ES project
Figure 5: Timeline of the implementation and use of the Oracle E-Business Suite
The following describes the events illustrated in the figure above except for the OEBS
timeline (see the description in paper 4). The financial system (FinSys) was the first
enterprise system to be launched in SCANDI in 1996 based on one of the earliest
international versions of OEBS. This was installed to have an up-to-date and common
system in the merged SCANDI. FinSys covered financial management including the
general ledger, fixed assets, receivables and payables etc. with around 40 heavy users.
SCANDI decided in 1999 to enhance its use of standard software packages to include
the supply chain as well. Subsequently, the Supply Chain System (SCS) was introduced
and operational from 2002 and employed by more than 100 users for purchasing,
64
inventory and other logistic functionalities. Another 3000 employees used SCS on an ad
hoc basis much like an online shopping cart solution to register purchase order
requisitions. FinSys and SCS were integrated with more than 40 other systems via
middleware integration or simple batch-oriented integration, which meant that they
were part of a fairly complex technical information infrastructure. The SCS system had
a major extension in 2004 when a logistic legacy system was closed down and the
functionality transferred to SCS. Both FinSys and SCS were upgraded in 2003 to OEBS
release 11i. FinSys and SCS were technically implemented as two separate enterprise
systems based on OEBS.
OEBS has evolved into a customized standard system since its first implementation in
1996. About 400 customizations were made throughout the years of implementation,
leading to difficulties in upgrading the system and locking SCANDI into an old ES
architecture – an untenable situation that had to be changed. Consequently, a
reimplementation project (RE-ES project) started in summer 2007 covering both FinSys
and SCS functionality with the motto “one finance and supply chain.” The purpose of
the project was primarily to: 1) reduce the number of customizations from 400 to about
150, 2) implement standard release 12 functionality, 3) optimize standard business
processes, 4) improve the use of standard functionality, 4) modernize the application
architecture and 5) reduce the IT cost by approximately 40%.
The RE-ES project was considered a technical reimplementation project with very few
benefits for the business. It was a one-to-one implementation that did not allow new
functionality although release 12 offered a number of possibilities. The implementation
approach of the RE-ES project was halfway between a “complete replacement of a
legacy system” and “a technical upgrade” (see also Parr and Shanks 2000).
The RE-ES project was very turbulent due to several shifts in outsourcing partners and
higher complexity in the project than expected, which caused several delays, but in
January 2009 the system finally went live.
The post-implementation phase has been a struggling phase, but the reimplemented ES
has slowly stabilized since the launch in January 2009 and a discussion with users indicates that they perceived the system as “in normal operation” in autumn 2009.
65
5.6 DATA COLLECTION
Data are not discovered and collected in my understanding, but are instead generated
and/or constructed (Schwandt 2007: 128-129). This follows the social constructivistic
stance, and reflects my understanding of the involved researcher (Walsham 2006). Anyway, to avoid confusion data collection is the title of this section and will be the terminology that I use subsequently to be compliant with mainstream methodology literature
(e.g. Creswell 2007; Gray 2004; Myers 2009; Patton 2002; Yin 2003).
Data collection and data analysis are not distinct processes in qualitative research, but
overlapping and carried out iteratively. Nevertheless, in this thesis, they are discussed
separately and presented in a linear fashion starting with data collection in this section
and then data analysis in the next section, as this is a useful and logical way of understanding the two activities (Myers 2009: 165).
Interpretive research is concerned with generating “thick descriptions,” a term borrowed
from anthropologist Clifford Geertz (Schwandt 2007; Walsham 2002), which means
providing rich insights and descriptions of human, social and organizational aspects of
organizations (Myers 2009: 92) such as circumstances, meanings, intentions, strategies,
motivations and so on (Schwandt 2007: 296). In this study it is rich insights into multifaceted meanings and complex processes associated with the management, implementation and use of ES.
Some recommend the use of multiple data collection sources in case study research
(Creswell 2007; Myers 2009; Yin 2003), although most of the evidence might come
from interviews and documents (Myers 2009). I decided at the outset to use interviews,
participant observations and documents. Interviews and participant observations were
aimed to gain rich insights into aspects of the organization with a focus on process and
content at the micro-level, while documents (public and internal) were expected to provide contextual and historical understanding at a more meso-macro level. However, all
the sources contributed with insights into all the analytical areas and levels in practice.
Additional sources were added during the research process, some because they were
perceived as valuable (the focus group interview) and others more as experiments that
might be relevant (process measurements and video recordings). I have collected a good
66
deal more data than I have used in the papers presented in this thesis, as it was difficult
to predict which data would be relevant in later stages and how the major change
process would pass off as it was real-time research, but there might also be an opportunity for future uses of the data.
The project management team for the RE-ES project helped to gain access to the rest of
the SCANDI organization and external parties (outsourcing vendors/suppliers). After
the introduction to the present use of the Oracle E-Business Suite and the RE-ES
project, four study areas were selected in cooperation with the project management
team: (1) the RE-ES project, (2) the finance department, (3) the purchasing department
(supply chain procurement) and finally (4) a cross-organizational user group (supply
chain requesters). The four areas were very different and imply contact with IT, business and consultants and with managers, project participants and daily users of the system, broadly embracing the management, implementation and use of ES. Oracle was
added as a fifth study area much later in the research process (see paper 4). The interviewees were selected with the help of managers, the project management team and
others. I had free access to the people involved in the RE-ES project while access to
employees in departments had to be agreed with the local management. The initial list
of interviewees was expanded during the research process mainly as a result of specific
events, consequences and decisions in the organization but also by snowball sampling
(interviewing one person leads to another) (Myers 2009; Patton 2002).
Below is a table showing the data collection methods, departments, roles and numbers
of instances:
Data Collection
Methods
SCANDI
Project
Group
Number of
persons
SCANDI
Finance
SCANDI
Supply
Chain
Procurement
SCANDI
Supply
Chain
Requester
Number of interviews, meetings etc.
Semi-structured
Interviews
- Project participants
4
4
- Finance manager
1
1
- Finance super user
1
1
67
Oracle
Data Collection
Methods
SCANDI
Project
Group
Number of
persons
SCANDI
Finance
SCANDI
Supply
Chain
Procurement
SCANDI
Supply
Chain
Requester
Oracle
Number of interviews, meetings etc.
- Accounts clerks
3
6
- Logistic managers
3
6
- Purchasers
3
6
- Logistic super users
2
1
- Ad hoc users
5
- Consultants
3
1
4
3
Short unstructured
phone interviews
- Across positions
20
3
8
9
4
Focus group
interview
- Finance employees
3
1
Participant
observations
- Project meetings
18
- Purchase meeting
1
- Super user meetings
2
Video recordings
- Ad hoc users
1
1
- Purchasers
1
2
3
6
Process
measurements
- Purchasers
Documents
Plans, reports, minutes, presentations; press releases and Internet information
from SCANDI; Oracle information (e.g. www.oracle.com); news articles;
magazine reports etc.
Table 7: Data collection overview
The data collection methods in the table above will be detailed and discussed in the following.
Semi-structured interviews
Interviews lie at the heart of interpretive research and are used in most interpretive studies as an important way to understand, interpret and construct the life world of infor-
68
mants in the field (Kvale 1997; Schutz 2005; Walsham 2006). I see the interview as a
social construction of knowledge about the informant’s life world taking place in the
conversation between the informant and the researcher (adapted from Kvale 1997).
I prepared about ten interview guides (see the example in appendix C), which took place
as an iterative process in connection with interviewing. Some of the interview guides
were drawn from the literature (enterprise systems, institutional theory etc.) and my
practical understanding of ES in organizations while others targeted specific events,
decisions, findings etc. The interview guides were designed to enable me to conduct
semi-structured interviews in order to focus on specific themes that are neither too
tightly nor too loosely controlled (Kvale 1997).
The interviews were conducting in the following way. (1) I read a checklist, “mode of
understanding in a qualitative research interview” (Kvale 2007: 10-14), before starting
the interview as mental preparation. (2) I started by introducing the purpose of the interview, advising the informant about anonymity and asking for consent to the digital recording of the interview. The actual flow of the interview appeared in the interview
guide, but it is important to emphasize that the structure was distantly followed as it was
a semi-structured interview, and I was open “to diversion of the interview into pathways
… not originally considered as part of the interview” (Gray 2004: 217). Field notes
were taken during the interview as a supplement to the digital recording and as backup
to the digital recording. (3) I spent some time after the interview reflecting upon what
was said, making additional field notes (Kvale 1997: 133) and checking that the digital
recording was satisfactory.
The first interview was carried out in January 2008 and the last in October 2009. The
duration of the semi-structured interviews varied from 30 minutes to 2 hours. Most of
the interviews were digitally recorded except for the three Oracle interviews. The majority of the semi-structured interviews (27 out of 33) were transcribed either by me or
by a research assistant. The Oracle interviews were written up immediately after the
interviews. A few transcriptions were sent to informants as agreed for comments. Below
is a transcription excerpt from an interview with a purchaser in April 2008 (translated
into English):
69
Discussion related to institutional theory (procedures, norms, cultures etc.)
Per: Please look at this situation and go with the thought experiment that a new employee enters the purchasing department; what would the newcomer experience? Purchaser: She would notice that we probably do not have the same work practices even if we work in the same department (dissimilarity). After a few days, I believe that the person, if she asked for – not just by making a purchase order for it we'll do the same – but asked to enter something in the system so we would come up with different answers even though we should do the work in the same way. Per: How has this dissimilarity come into existence? Purchaser: I think it is precisely because there are so many who have been here for many years and has something to do with our luggage, and also something we have learned the system by ourselves. R12 [new version of OEBS] is the first time that we have been trained in the new sys‐
tem ... it was self‐taught last time … Per: What kind of standard operating procedures do you have? Purchaser: For example, if they say in an article how you calculate the min–max, then you'll get seven different proposals from different people … Per: Okay, you don’t have a manual that says anything about the stock calculation points. Purchaser: We have now. Per: Do you use the manual for stock calculation points? Purchaser: It is probably not all that use the procedure – and I do not – we have had such an easy stock calculation procedure in a spreadsheet, but I do not always use the proposal from the spreadsheet (laughing) – it is something that I know better … Per: Does the dissimilarity give you some challenges in your everyday work? Purchaser: Yes, sometimes when we have to overlap each other … We work together in pairs so we always have one resource on duty also during vacation etc. … Per: Do you discuss the dissimilarity in your department? Purchaser: This is something which is accepted … We talk about it in the department in order to clarify issues which come from this dissimilarity. Per: Does it also give rise to conflicts? Purchaser: Yes it does, which I do not fully understand as my colleague and I work perfectly to‐
gether, but we also work differently than the other teams … I cannot understand if you are away for three weeks and told how to overlap your colleague, and then the person says that I'm not doing the way I was told – instead the one on the job should say that for your sake I have fol‐
lowed your working style and avoid … a mega‐conflict … Per: What else might a newcomer notice? Purchaser: A newcomer would notice ... we have a good climate. Per: Can you be more specific? Purchaser: At a time I had many updates to the system and I got 17 full pages, then a couple of colleagues immediately came and helped me – that is helpfulness … Per: What about trust? Purchaser: It depends on what team you ask – some have it with "this is my article" and those do not touch … Per: We might call it territories. Purchaser: Yes but we have sat together for so long, so we know who it is … Per: How long have you been working together in the purchasing department? Purchaser: We have then been brought together from three different departments … 70
The transcription on the previous page shows a dialogue about institutional elements
such as procedures, norms and cultures.
Two user groups (the finance department and the purchasing department) were interviewed before (March 2008–April 2008) and after (June 2009–September 2009) the
launch of the reimplemented ES in January 2009. This was in order to understand and
interpret the changes in the ES, working practices and more intangible elements such as
“perception of the ES” and to follow institutional processes (e.g. paper 5 elaborates on
deinstitutionalization and reinstitutionalization processes) (see also Barley and Tolbert
1997).
Short unstructured phone interviews
In order to follow the change processes at SCANDI, I carried out 24 short unstructured
interviews from January 2009 (the launch of the reimplemented ES) to April 2009. The
overall purpose of these unstructured interviews was to understand how and when the
ES was perceived to be stabilized by the informants (Ross et al. 2003: 108-109) as this
was part of the before and after research design. The interviews were very loosely designed with the following three high-level questions: (1) How would you describe the
situation right now, (2) what are the biggest challenges and (3) what about performance? The duration of the interviews was 10 to 30 minutes, and I just called the informant without a preceding appointment. The interviews were transcribed immediately
after the interview based on the field notes. The outcome of these interviews was partly
an understanding of the stabilization process, but also emerging discussions with informants about other topics like apparently strange management decisions etc. The unstructured interviews were a good opportunity to “catch the dynamics of the process.”
Below is a transcription excerpt from a short unstructured phone interview with a super
user from logistics one week after the launch in January 2009 (translated into English):
71
Per: How would you describe the situation right now?
Super user: It is a bit hectic, but it's starting to get better here in the second week after go live, but there are obviously still some challenges. Per: What are the hurdles? Super user: (1) To resolve known bugs – there are approx. 60 errors including 10 critical priority 1 errors, but [Supplier] has really to speed up the debugging process, (2) to stabilize production, i.e. have executed orders and delivered products, (3) users must learn to do things differently … (4) ordering of commodity products should be easy enough, even if there really have been major changes [but appears anyway to be challenging] … Per: What are the biggest challenges? Super user: (1) To solve the known bugs including separating system errors and user errors, (2) be sure to get the priority right of errors and (3) [Supplier] to come up to speed. They work hard on the case and make a good effort, but there will obviously be many things that are to be deployed. Per: What about performance? Super user: During the first week, there has been some loss of performance. Focus group interview
When my colleague and I worked on paper 3 and paper 4, we realized that it would be
appropriate to present our findings to the informants that I interviewed about the topics
in these papers. A focus group interview with three finance employees was arranged
(October 2009) to obtain a collective view of our findings (Myers 2009). The colleague
and I took the informants through a PowerPoint presentation, which initiated the discussions, and the interview took about two hours. The interview was digitally recorded and
transcribed. This is a useful approach to involve informants in the data analysis process
and as a kind of triangulation.
Participant observations
Observation is watching other people from the outside, like participating in an annual
meeting or watching people in a street, while participant observation is when you not
only observe people but also participate to some extent by talking and interacting with
people from the inside (Myers 2009: 137-139).
My first experience as a participant observer took place in January 2008. The way in
which I took field notes was similar to being the keeper of the minutes, which I have
performed numerous times as a project manager. I did not consider the participant observation from a research methodological point of view, which is completely contrary to
what I did with the interviews as these were carefully planned with interview guides,
72
but during the course I realized that this is not an appropriate approach. I consulted the
literature on observation, participant observation and ethnography (Bailey 2007; Brewer
2000; Creswell 2007; Gray 2004; Silverman 2001). The outcome, after several revisions, was a pre-printed square paper sheet ready for field notes. The figure below
shows an example of field notes from a project status meeting in July 2008 when the
template was used:
Figure 6: Field notes from participant observation, July 2008
The template in the figure has a title bar and three columns labeled “time & category,”
primary observation (direct field notes) and finally the right-most column for preliminary data analysis (no title) – all the columns were used for field notes during the participant observation. As shown in the figure, data collection is entangled with data analysis
in participant observation as a lot of thinking goes on when you are in a two- or threehour meeting and the template is designed to capture preliminary analysis with four
predefined categories: (1) reflection and recall, (2) ideas and inferences, (3) experiential
data like impressions and personal feelings and finally (4) forward planning. Refer to
appendix D for a detailed description of the template.
73
Participant observations in anthropology often means staying in the field for 12 months
or more, while IS research is typically much shorter (adapted from Myers 2009). I participated in various meetings with a duration of two to four hours and a few all day –
these took place from January 2008 to June 2009 although the most intensive period
was the first year in the field (2008). Meetings are suitable for gaining compressed key
information about discussions, decisions, plans, politics, conflicts and motives, and then
you can obtain more details, if needed, by informal talks, interviews, documents etc. I
was invited to participate in project meetings from the very beginning while participation in the other meetings was initiated by key informants.
The most important outcomes of the participant observations were: (1) to build trust
with project participants and become accepted, which are important for gaining genuine
discussions with informants, (2) to establish a network fertilizing the ground for the ongoing fieldwork and easy access to informants and information and (3) to understand
the dynamics, the pulse concerning the management, implementation and use of the ES
as well as obtaining relevant background information. All this has been important for
conducting the interviews, collecting documents and writing papers.
The field notes from participant observations have not been written up except for a very
few summaries and notes. This is a possible drawback and against advice (Myers 2009:
146), but a lot of other information was available from these meetings (minutes, presentations, reports, plans etc.) in addition to my field notes, so I will argue that in the given
situation it was a sensible approach and a way to optimize limited time.
Video recordings and process measurements
Video recordings and process measurements were added as experiments as they might
contribute relevant insights – they will be described briefly. Both methods were used in
a before and after design. Video recordings are a suitable way to document and understand micro social practices (Pink 2007; Silverman 2001) and they were used to record
specific scripts such as “update order confirmation” (a series of screen images). Process
measurements were time recordings of a specific process like “register purchase order”
measured by stopwatch. They were measured to gather performance data, which could
74
be relevant to understanding changes in processes – and quantitative data can be valid
inputs to interpretive studies (Walsham 2006: 323).
The data have not been used so far, due to the overall prioritization in the research
project, but they are available for future use – maybe to investigate micro-level issues in
institutional analysis (Powell and Colyvas 2008), although this is still an immature idea
that needs further consideration. Video recordings and process measurements belong to
the broad category of documents, which will be discussed subsequently.
Documents
Documents provide evidence that can be used to construct a richer picture than can be
obtained from interviews and participant observation and they “can be as significant as
speech in social action” (Myers 2009: 152-153). Documents are “any symbolic representation that can be recorded or retrieved for analysis” (Altheide 1996: 2) and can be
stored in a digital file on a computer. They might exist only in hard copy but can be
stored in a digital format (text, audio, pictures, video or records in a database) (adapted
from Myers 2009: 154). Examples of documents are letters, emails, minutes of meetings, written reports, project plans, requirement specifications, presentations and newspaper articles (Yin 2003: 85-88) but also pictures, photos, videos, films, television programs, interactive websites, software, video games, blogs, interview transcripts and
much more. Documents can be categorized in a number of ways such as personal (letters, diaries), private (produced by an organization for internal use, e.g. minutes of
meetings) and public (annual reports, press releases) (adapted from Myers 2009: 154158; Schwandt 2007: 75).
I have gathered a large amount of internal documents (private) as well as public documents in the research process. Internal documents include plans, reports, minutes of
meetings, presentations and emails while public documents comprise press releases,
Internet information from SCANDI, Oracle information (e.g. www.oracle.com), news
articles, magazine reports and many others (see appendix E). I have furthermore had
online access to a shared document store and content management system (Microsoft
SharePoint) for the RE-ES project, so I could search for and download project docu-
75
ments and other information as needed, which has been a valuable source in the research process.
The documents have been used for a variety of purposes: (1) to serve as relevant background information, (2) to learn about specific strategic, technical and organizational
concepts, (3) to enrich discussions at interviews and participant observations, (4) to be
used in data analysis and the findings (e.g. excerpts from documents). An excerpt from
an email is shown below (translated into English) to illustrate the value of documents:
From: [Name removed]
Sent: 4 May 2009 09:16 To: Per Svejvig Subject: Status of the reimplementation project Hi Per, Generally, I think that [the system] is working well, but there are still some issues: - There are problems with some entries, with the result that there are some customers who have problems with ordering some of the goods. - There are some orders that go into error status, and they must be resolved directly with [outsourcing vendor] - We have to contact the main support if we have erroneous orders where we need to cancel order lines. The email is an easy way to forward the status to me, and also serves as a document that
can be used in the data analysis as well as empirical evidence in final products (papers).
The graph below summarizes the data collection activities with a histogram of the frequency distribution of semi-structured interviews, short unstructured phone interviews
and participant observations:
76
LAUNCH
12
Semi-structured interviews
10
Short unstructured phone
interviews
8
Participant observations
6
4
2
Dec-09
Nov-09
Oct-09
Sep-09
Aug-09
Jul-09
Jun-09
May-09
Apr-09
Mar-09
Feb-09
Jan-09
Dec-08
Nov-08
Oct-08
Sep-08
Aug-08
Jul-08
Jun-08
May-08
Apr-08
Mar-08
Feb-08
Jan-08
0
Figure 7: Frequency distribution of data collection activities
The graph indicates that some data collection activities have been more intense in some
periods than others, and the fieldwork can be divided into major phases:
-
Jan. 08 to May 08: Before-launch semi-structured interviews
May 08: Planned launch of system (but delayed several times)
Jun. 08 to Dec. 08: Following the project (participant observations)
Jan. 09: Actual launch of the reimplemented system
Jan. 09 to Apr. 09: Stabilization of the system after launch (short interviews)
Jun. 09 to Sep. 09: After-launch semi-structured interviews
Oct. 09: Oracle semi-structured interviews
This completes the discussion about data collection activities and the associated methods, which has briefly touched on data analysis activities, whereas the next section
will discuss the analysis activities in detail.
5.7 DATA ANALYSIS
Qualitative data analysis aims to transform data into something that is meaningful
(Myers 2009). There are many ways to analyze and interpret qualitative data, such as
content analysis, series of events, analytic induction, phenomenological analysis, her-
77
meneutics, semiotics, narrative analysis and others (Creswell 2007; Myers 2009; Patton
2002). Some of these methods have already been used in previous sections, e.g. the
timeline of implementation and the use of OEBS (figure 5) are a “series of events.”
The hermeneutic method is a fundamental principle in some branches of interpretive
research (Klein and Myers 1999); it is a methodology associated with constructivism
(Guba and Lincoln 2005) and suitable for studying unclear and possible contradictory
organizational issues and events (Myers 2009), which makes it appropriate for this
study.
Hermeneutics is both an underlying philosophy and a specific method (Myers 2009;
Schwandt 2007) and it is the latter that is referred to in this context, although the philosophical underpinnings have some resemblance to the social constructivistic stance (see
however Schwandt (2000)). The main purpose of hermeneutics is to understand what
people say and do, and why. Hermeneutics is about interpreting text and text analogues;
Taylor (1976: 153) says that:
Interpretation in the sense relevant to hermeneutics is an attempt to make
clear, to make sense of an object of study. This object must, therefore, be a
text, or a text-analogue, which in some way is confused, incomplete,
cloudy, seemingly contrary – in one way or another, unclear. The interpretation aims to bring to light an underlying coherence or sense.
The concept “text-analogue” could be anything that can be treated as text, such as culture, organization or use of an enterprise system in a purchasing department. Texts also
include social action, conversations and non-verbal communications. The hermeneutic
circle is a fundamental concept in hermeneutics and refers to the dialectical idea that
understanding the text as a whole is meant by making sense of the individual parts of
the text and understanding individual parts is by having some sense of the whole. There
is constant movement from the whole to the part and back to the whole as a dialectical
process. Prejudice and preconceptions are a necessary starting point for our understanding and hermeneutics has the maxim “no knowledge without foreknowledge.” Prior
knowledge is not a hindrance (or problematic bias as suggested by a positivistic stance)
but a “help” to become clearer about a meaning of an object (text). However, as re-
78
searchers, we must be aware that our own views and biases have a significant impact on
how we interpret the world and construct meaning (Myers 2009: chapter 14; Schwandt
2007: 133-137).
I will use paper 2 to describe the hermeneutic method in a more concrete way. During
my fieldwork I became gradually aware that outsourcing was one the most important
management issues in the RE-ES project. Three outsourcing vendors were involved in
the project with several shifts of responsibilities from one vendor to another during the
project course. There were many tensions and conflicts in the project due to the complex and changing outsourcing set-up, which were manifested at the weekly project
meetings in which I often participated (participant observation). I wondered and asked
the question “why did SCANDI engage in these very complex outsourcing arrangements?” The official answer was cost savings, but that was not consistent with what I
observed in the organization and with accounts from informants. In this scenario the
SCANDI organization with the RE-ES project and its outsourcing vendors form a “textanalogue” representing the whole. The parts are the interview transcripts, documents,
field notes from participant observations and many impressions (lots of pages of texts
and preconceptions in my mind). These parts were analyzed piece by piece and connected to institutional concepts (see paper 2 for more details about coding etc.), where I
mentally moved from and back to the whole and the parts until the explanations made
sense to me, and later in the process also to the co-author of the paper. I arranged additional interviews and informal talks (more parts) to discuss my findings and preconceptions to gain a more complete and clear understanding of the whole and to produce the
story presented in paper 2. The hermeneutic method was used implicitly in the data
analysis (and data collection) and in hindsight I can “see” the congruence between what
I did and the hermeneutic method.
The hermeneutic method is to some extent embedded in more explicit data analysis activities, which are illustrated in the figure below:
79
Select
data
Theorize
Discover
strategies
Work with
data
•
Select data (documents, texts and text
analogues) to be analyzed (Myers 2009)
•
Discover strategies for seeing and naming
concepts and categories in the data (coding, annotating, memoing, linking and
modeling) (Bazeley 2007)
•
Work with data according to strategy (e.g.
coding) (Bazeley 2007)
•
Theorize to construct concepts and models, generate theory, draw specific implications and/or contribute to a rich insight
(Walsham 2002)
Figure 8: Explicit data analysis activities
The activities are highly iterative and dialectical; they are not numbered as there is no
predefined sequence or starting point. The data analysis might involve further data collection activities. Selecting data may imply a preliminary reading or skimming of interview transcripts, field notes or other documents to sort important from unimportant data
(Jensen 2007: 105). This initial selection and sorting process will typically create some
initial themes, ideas, categories or patterns that can be used to discover suitable strategies for working with the data. Working with data can start by being “close to the data”
and “let the data tell their own story” (Patton 2002: 457), such as reading data very carefully line by line, assigning codes to the data, linking data elements and so forth (inductive approach). “Codes are tags or labels for assigning units of meaning to the descriptive or inferential information compiled during a study” (Miles and Huberman 1994:
56). Codes are identified and updated (renamed, merged, deleted etc.) during the careful
reading process. The codes might be grouped, linked and categorized into more abstract
codes, e.g. pattern codes, axial codes, selective codes (Miles and Huberman 1994: 5576; Myers 2009: 110-111). These more abstract codes can then be used to theorize (generalize) to develop models or create rich insights into phenomena. Theory has been
used in data analysis (see also section 5.3), sometimes as coding categories like coercive, normative and cognitive institutional forces to apply a more deductive approach
assigning units of texts to these predefined categories derived from theory. In other situations theory has been used to analyze patterns or accounts in the data, starting with
inductive analysis followed by a more deductive approach using institutional theory and
other theories as sensitizing devices (Klein and Myers 1999: 75-76; Patton 2002: 452462); this bears some resemblance to grounded theory in the coding process, but not as
80
an overall research method (Myers 2009). Theory has in both approaches played an important role as part of the iterative data analysis process (Walsham 2002) involving inductive and deductive analysis.
The detailed data analysis activities have to some extent been different for the four empirical papers (2, 3, 4 and 5), but to be more specific, the data analysis for paper 4 will
be presented including the use of qualitative data analysis software NVivo (Bazeley
2007). This paper was started by the observation and interpretation that FinSys at
SCANDI appeared to be very well adapted, which is contrary to numerous accounts of
failure-prone ES, so we asked the question “why is the ES [FinSys] so well adapted at
SCANDI, and what can we learn from this?” The selection of data was fairly
straightforward, centered on the finance department but also as an iterative process
where additional interviews (Oracle interviews and the focus group interview) and documents were collected and used in the analysis. Interview transcripts and very few documents were imported to NVivo and the major part of the codes were generated by asking the research question “why is the ES so well adapted …?” (Myers 2009: 110-111).
NVivo was thus used to organize data and to construct interpretive codes. An example
from the coding process is shown in the figure below:
81
Transcription (Danish) Coding Figure 9: Example of NVivo coding
Figure 9 shows two columns, the left column containing the transcription of an interview with an accounts clerk in June 2009, and the right column showing the associated
coding strips for this small piece of transcription.
Both authors were involved in the coding process separately (see appendix F for a list of
NVivo codes). NVivo was used in the first part of the data analysis process while theorizing was outside NVivo. We used the codes to develop more abstract themes drawing
on institutional theory and sensemaking theory as sensitizing devices, for example “a
rationalized myth about effective work practices and an efficient enterprise system.”
The major findings were summarized in a table and the lessons learned specified in paper 4. The additional interviews and dual coding process served as triangulation. Triangulation in interpretive research does not imply that different sources or methods necessarily converge towards the same meaning but instead that genuine meaning can be constructed best by viewing it from different vantage points (Schwandt 2007: 297-298).
82
5.8 EVALUATING INTERPRETIVE RESEARCH
The goodness or quality criteria of interpretive case study research cannot be judged by
the conventional benchmark of rigor (validity, reliability and objectivity) as used in positivistic research (Guba and Lincoln 2005), but on the contrary by trustworthiness, credibility, transferability and confirmability (Denzin and Lincoln 2005), to construct a
complete, interesting and believable story (Myers 2009).
Klein and Myers (1999) suggested a set of principles for the evaluation of interpretive
field studies (in-depth case studies and ethnographic research based on hermeneutics),
which will be adopted in this section because it directly relates to IS research although
other evaluation principles are available from other research areas (e.g. Creswell 2007:
chapter 10; Golden-Biddle and Locke 1993; Kvale 1995; Patton 2002: chapter 9). The
principles for evaluation consist of seven explicit criteria that can be used to evaluate
interpretive research. The principles are not bureaucratic rules of conduct nor do all the
principles apply in every situation, but their systematic approach intends to improve and
evaluate the quality of interpretive research in IS (Klein and Myers 1999: 70-71).
The scope for the evaluation will be the thesis as a whole (i.e. cover part and papers),
and the evaluation is presented in the table below together with a summary of the seven
principles (Klein and Myers 1999):
No.
Summary of principles
Evaluation of principles in this study
1.
The Fundamental Principle of the Hermeneutic Circle
The hermeneutic method has been used implicitly and thoroughly in this study, but is
first discovered as an explicit method in hindsight (see section 5.7).
This principle suggests that all human understanding is achieved by iterating between
considering the interdependent meanings of
parts and the whole that they form. This principle of human understanding is fundamental
to all the other principles.
2.
The Principle of Contextualization
Requires critical reflection of the social and
historical background of the research setting,
so that the intended audience can see how the
current situation under investigation emerged.
3.
The Principle of Interaction Between the
Researchers and the Subjects
Requires critical reflection on how the re-
83
Contextualization is a main principle used
throughout this study based on Pettigrew
(1985; 1987) (see section 5.2 and case study
descriptions in papers 2, 3, 4 and 5).
One of the main purposes of this chapter 5 on
the research process has been to open the
“black box” and invite the readers to evaluate
No.
Summary of principles
Evaluation of principles in this study
search materials (or “data”) were socially
constructed through the interaction between
the researchers and participants.
and follow my way of thinking.
While it is intended to be explicit in this chapter, it is largely missing from the empirical
papers 2, 3, 4 and 5.
4
The Principle of Abstraction and Generali- Institutional theory and other theories have
explicitly been used in all the phases of the
zation
research process to support abstraction and
Requires relating the idiographic details regeneralization.
vealed by the data interpretation through the
application of principles one and two to theoretical, general concepts that describe the
nature of human understanding and social
action.
5
The Principle of Dialogical Reasoning
6
The Principle of Multiple Interpretations
7
The Principle of Suspicion
“The most fundamental point is that the researcher should make the historical intellecRequires sensitivity to possible contradictions
tual basis of the research [basic philosophical
between the theoretical preconceptions guidassumptions] as transparent to the reader and
ing the research design and actual findings
himself ...” (Klein and Myers 1999: 76).
(“the story that the data tell”) with subsequent
cycles of revision.
I assume that the readers are well aware of
my assumptions (cf. chapter 4) and how they
have influenced the research process in various ways. However, the dialogue between the
text and the interpreter is not discussed, for
instance how the hermeneutic process has
evolved during the research process (see e.g.
six stage process summarized by Lee 1991:
356).
Several papers touch on multiple interpretations, e.g. paper 2 “cost savings versus instituRequires sensitivity to possible differences in
tional explanations of outsourcing decisions”
interpretations among the participants as they
and paper 5 “purchasers perceive the value of
are typically expressed in multiple narratives
release 12 differently where seven out of nine
or stories of the same sequence of events
purchasers prefer the new release.”
under study. Similar to multiple witness acMultiple interpretations emerge in various
counts even if all tell it as they saw it.
situations and are presented in some of the
papers, but this area could have been cultivated more in the research process, by specifically searching for multiple interpretations.
Requires sensitivity to possible “biases” and
systematic “distortions” in the narratives
collected from the participants.
Informants’ views and other views from documents are not taken at face value and critical
thinking has been applied. The story in paper
2 is built around the theater metaphor, where
we have frontstage and backstage explanations and where we aimed to go beyond
frontstage surface meanings (Klein and Myers
1999: 78).
Furthermore, most of the papers are written
with co-authors and they contribute critical
thinking and challenge possible biases and
84
No.
Summary of principles
Evaluation of principles in this study
distortions.
However, it has on some occasions been difficult to chase the latent distortions, which
could be illustrated by papers 3 and 4: the
accounts clerks expressed that they find their
working conditions suitable and that FinSys
suits their needs. My colleague and I thought
the statements were too rosy, but despite
further interviews and a focus group interview, we are still left with some doubts of a
trustworthy narrative.
Table 8: Evaluation of the study
The table above more or less speaks for itself, so instead of discussing the seven principles (the parts) further, it is appropriate to reflect on this evaluation as a whole. The
seven principles could with advantage have been used much earlier in the research
process to support the design and conduct of data collection and data analysis. However,
they can still be applied to papers 2, 4 and 5 at least from an evaluation perspective. The
next section will continue the reflections on the research process.
5.9 REFLECTIONS ON THE RESEARCH PROCESS
Life must be lived forward, but can only be understood backwards (Søren Kierkegaard,
1813–1855).
The actual path taken in the research process is somewhat different from the reconstructed logic presented in the major part of this thesis. This chapter’s final section aims
to reveal some of the diversions I have taken, some of the reflections I have made and
how these have formed my current understanding of the study area. Such self-reflexive
and self-revealing accounts of the research process (Schultze 2000) “are necessarily
partial, and we choose what we wish to confess” (Walsham 2005: 111) and this account
is no exception although I aim to let the readers into the essential issues.
I started this research project in 2007 by consulting various pieces of IS literature on
implementation. This was later narrowed to the post go live period of enterprise systems
designated “the maiden voyage of enterprise systems.” This stabilization period typically has an extent of 4 to 12 months (Ross et al. 2003: 108-109). This appeared to be a
relevant area to study based on my incipient understanding from literature and practical
85
experience. Several concepts and theories were consulted about routinization, stabilization, adaptation, adoption, assimilation and institutionalization (e.g. Cooper and Zmud
1990; Gallivan 2001; Henfridsson 2000; Markus and Tanis 2000; Ross et al. 2003; Silva
and Backhouse 1997). Institutional theory was selected because it seemed to provide a
conceptually rich body of work to analyze and understand complex social phenomena
(Scott 2001) such as the stabilization process after go live, and it was well aligned with
my paradigmatic view of social constructionism. My awareness of institutional theory
came from a PhD course in May 2007 where W. Richard Scott presented an overview of
the theory and suggested the theory might be applicable to my research, although it was
late in 2007 before I decided to follow his advice after consulting several other theories.
The more specific reasons for selecting institutional theory as presented in the introduction were constructed much later. Despite the arguments for selecting institutional
theory, it is nevertheless subjective and based on the researcher’s experiences, preferences and particularly interest (adapted from Walsham 2006). As I became more acquainted with institutional theory and institutionalization processes during 2008, I increasingly realized that my empirical scope was too narrow, and it was cemented with
advice from colleagues who commented on my early work, especially Geoff Walsham,
who wrote to me that “institutionalization processes are complex and major shifts often
take place over years rather than months” – i.e. there was apparently an inconsistency
between the duration of the post go live period and the theory chosen to study this phenomenon. I therefore took the decision to broaden my empirical scope to study the management, implementation and use of ES and apply a longer time period than originally
anticipated. A consequence of this change was a move away from the micro-level focus
to multi-level considerations including the meso-macro level. This shift in focus thus
enabled a broader use of institutional theory and planted the seeds to analyze the empirical data in other ways than planned – it furthermore formed my overall understanding
of thinking in multiple levels and their interactions and better exploiting the potential of
institutional theory (Currie 2009).
The decision to work with the “before and after design” (Barley and Tolbert 1997) related to the launch of ES was decided before I entered the field, and the planned research approach is presented elsewhere (Svejvig 2008). However, the importance of this
design was reduced when I broadened the scope and increased the time period although
86
I kept the before and after interviews (see the major phases of field work on page 77).
Some of the results appear in paper 5, where the subjective user perception before and
after the launch of the ES is presented based on the IS success model (DeLone and
McLean 2003). However, the full potential from the before and after research design
has still to materialize. I have reflected on this issue, and there might be several reasons
such as (1) a lack of details in the data to compare before and after situations, (2) the
data analysis has focused on other issues (video recordings and process measurements
are not analyzed so far), (3) the use of institutional theory has “guided” me elsewhere or
(4) the institutional changes associated with the reimplementation project were too
small really to expose the changes. It is anyhow something that can be considered in
future research and the actual course can be seen as part of the emergent nature of the
research process with shifts in focus and intensity (Henfridsson 1999: 12-17).
This completes the discussion about the research process, throughout which I have tried
to open the black box and provide transparency to the reader in line with the principle of
interpretive research and particularly the principle of interaction between the research
and the subjects as well as dialogical reasoning (Klein and Myers 1999).
87
6 THEORIZING ABOUT ENTERPRISE SYSTEMS IN
ORGANIZATIONS USING INSTITUTIONAL THEORY
This chapter sets out to present the findings and results from the five papers, and explores how institutional theory can be used to theorize about enterprise systems in organizations. Theorizing implies activities “like abstracting, generalizing, relating, selecting, explaining, synthesizing, and idealizing” (Weick 1995: 389), leading to findings
and results, which in this thesis embrace conceptual modeling and drawing specific implications. These should be viewed as valuable explanations of past data that are not
entirely predictive for future situations (Walsham 2002: 110) (see also section 5.3).
The five papers included in this thesis are:
1. Using institutional theory in enterprise systems research – Developing a conceptual model from a literature review
2. Enterprise systems outsourcing “behind the curtain” – A case study showing
how rational and institutional explanations co-exist and complement each other
3. Enterprise system adaptation: A combination of institutional structures and
sensemaking processes
4. Making sense of enterprise systems in institutions: A case study of a welladapted system
5. Making new systems is breaking old systems – A case study about practices for
deinstitutionalizing an enterprise system
The findings and results in the individual papers are focused on specific research question(s) and therefore to some extent fragmented, although ES and institutional theory
are recurrent in all the papers. I have attempted to synthesize the results across the papers and produce a more coherent presentation, but also to add and complement results
where it makes sense, particularly drawing from the cover part. As stated in the introduction, I recommend reading the papers before reading this chapter, as the presentation
is rather succinct.
The chapter is divided into three subsections starting with an overview of the diffusion
and use of institutional theory in ES research; the next two sections present the development of conceptual models and draw specific implications.
88
6.1 THE DIFFUSION AND USE OF INSTITUTIONAL THEORY IN ENTERPRISE SYSTEMS RESEARCH
A relevant starting point for this thesis was to understand the diffusion and use of institutional theory in ES research through a literature review. Three scholarly databases
were searched and yielded 180 articles, of which only 18 papers fulfilled the criteria and
were selected for the literature review. The findings show that institutional theory in ES
research is in its infancy and adopts mainly traditional institutional aspects like isomorphism, with the organization as the level of analysis, and in several cases complemented
by structuration theory and other theories. A consequence of this infancy is that there
are many unexplored research avenues available, whereby we can widen ES research to
embrace complex social situations of the management, implementation and use of ES
(paper 1).
My results from ES research are furthermore consistent with results in IS research (literature survey by Weerakkody et al. 2009) and recommendations from Currie (2009) to
use institutional theory in a wider perspective in IS research. However, we will probably
see an increase of publications in the IS field fertilized amongst others by a special issue
on institutional theory in the Journal of Information Technology (volume 24, issue 4,
December 2009), which includes a paper I have co-authored (Jensen et al. 2009). That
might have a knock-on effect on the diffusion and use of institutional theory in ES research.
6.2 THE DEVELOPMENT OF CONCEPTUAL MODELS TO STUDY INSTITUTIONAL STRUCTURES AND PROCESSES
The management, implementation and use of ES imply three basic elements such as
enterprise systems, organizations and actors and their reciprocal interactions. The three
elements have been discussed thoroughly in earlier sections, but are captured here to
remind the reader about the conception of these basic elements in this thesis:
Enterprise systems are large-scale organizational systems, built around packaged enterprise systems software, enabling an organization to automate and integrate a comprehensive part of its business processes, to share common data and practices, and to produce and access information in real time. Enterprise systems influence patterns of cog-
89
nition, action and communication, but they do not definitely determine organizational
actors’ behavior. They have profound effects on the structuring of work practices and
enable as well as constrain human action (see pages 13 and 22).
Organizations are the players and institutions are the rules of the game, and organizations play multiple games at the same time. Organizations operate in pluralistic institutional contexts and the internal functioning reflects the larger systems themselves. This
results in an organization that: (1) may have multiple institutionally given identities, (2)
may be the structural embodiment of multiple logics, (3) may be legitimated by multiple
mythologies and (4) may take for granted very different beliefs and values – in short,
multiple things to multiple people (see page 33).
Actors (human actors) can do as they like, but are in practice constrained by the structures imposed on them; they are as much constructing as being constructed (see section
4.4).
Enterprise systems, organizations and actors are part of much ES research, but institutional theory can contribute with further elements, as will be shown later. I have proposed four requirements for how to use institutional theory in ES research derived from
a literature review (paper 1):
1. It must include core features of institutional theory like institutional and competitive isomorphism and rationalized myths, but also newer features such as institutional logics and institutional processes.
2. It should support multi-level analysis bridging macro and micro structures, and thereby implicitly address a multi-stakeholder approach, all being strong features of
institutional theory.
3. It must contain multi-theory elements to gain from juxtaposing institutional theory
with other theories.
4. It must incorporate the ES artifact in order to be specific about the technology.
90
A conceptual model was developed in the quest to fulfill the four requirements and is
presented in the figure below:
Figure 10: Conceptual model for using institutional theory in ES research
The purpose of the model is to provide an analytical model to advance both theoretical
and empirical ES research using institutional theory.
The model is divided into three frames: institutional theory, phenomenon of study and
juxtaposed theories. First, the institutional theory frame catches many of the institutional elements considered in this thesis, but the elements shown should not be taken as
fixed and complete, but on the contrary as a menu to choose from, with the possibility
to add other dishes (i.e. an open model). The different elements can be used at different
levels, and there is no direct correspondence between the location of elements and the
levels of analysis presented in the phenomenon of study frame. Second, the juxtaposed
theories frame is a placeholder to visualize the possibility of juxtaposing other theories
with institutional theory, where these theories could attach at different levels of analysis
in the phenomenon of study frame. Finally, the phenomenon of study frame presents the
multi-level approach often used in institutional theory (Scott 2008). The three basic
elements organization, enterprise system and actor have similarities to Orlikowski’s
91
structural model of technology (1992) and seek to give the ES artifact a salient role in
the organization. The arrows are conceptual links between the elements in the frame,
and they are not meant to be causal, but instead descriptive and exploratory relationships to aid theoretical discussions and empirical analyses (adapted from Fligstein 2001:
115).
Many of the features in the conceptual model have been used in subsequent papers. All
the papers (2, 3, 4 and 5) apply a multi-level approach although the unit of analysis is
the organizational level as in most institutional papers in IS research (Currie 2009; Weerakkody et al. 2009) and ES research (paper 1). Three of the papers (2, 3 and 4) make
use of the multi-theory approach.
However, the model overlooks an essential circumstance as stated by Pollock and Williams (2009: 4):
[Enterprise systems] only work to the extent they are adapted by user organizations … Thus, according to many sociologists it is “users” and
“adopting organizations” that should be studied … [therefore] recent research on information systems has become somewhat unbalanced … [and]
there is not … a comprehensive understanding of the … supplier organizations.
This theme is taken up in paper 4, where we propose a dual structural model of technology that addresses both the supplier and the user organization, as shown below (adapted
from Brehm and Markus 2000; Orlikowski 1992):
92
Figure 11: The dual structural model of technology for enterprise systems
The figure shows that the design process in the supplier organization results in an ES as
a semi-finished product. The system embeds best practices in the form of business
process reference models and it stands out as “a complete, though flexible, ready to
implement solution” (Soh and Sia 2004: 376) crossing the border to the customer
organization. The system is then configured and customized through another design
process depending on the perceived gap between the requirements and the functionality
provided by the semi-finished product. The dual design processes followed by the use
process are repeated as new releases of the system are produced (Brehm and Markus
2000). Both supplier and customer organizations have institutional properties, which
reflect internal functioning and the wider environment (field, sector or society).
There are other conceptual models presented in the papers but they are either variants of
the models above (paper 3) or address more specific issues (paper 5).
6.3 DRAWING SPECIFIC IMPLICATIONS FOR THE MANAGEMENT,
IMPLEMENTATION AND USE OF ENTERPRISE SYSTEMS
The conceptual models presented in the previous section are able to provide an overview and to support understanding, interpretation, exploration and analysis of institutional structures and processes. This section will move from concepts to specific implications (Walsham 2002).
93
Intertwined rational and institutional factors impacting on outsourcing decisions
We do not always make rational decisions and neither do organizations. Any social situation consists of interdependent non-rational and rational elements (Scott 2008: 217218). ES outsourcing decisions are shaped by both competitive pressures (rational elements) and institutional pressures (non-rational elements), which are furthermore intertwined (paper 2). The examination of outsourcing events in paper 2 by two theoretical frameworks (institutional theory and transaction cost theory) resulted in several rational and institutional explanations for outsourcing; below is an excerpt of these from
paper 2:
Rational explanations
“Behind the curtain” institutional
explanations
Economic
explanations
Cost savings in order to elicit a
healthy and attractive company for
shareholders – to make SCANDI
saleable
Management consultants “bring”
benchmark numbers for companies to
SCANDI, which is “travel of best
practices”
Technical
explanations
The ERP area is not business critical Outsourcing becomes an institutionaand therefore suitable for outsourcing lized technical practice, i.e. a habitual
choice
Symbolic–
interpretive
explanations
Outsourcing is a way to force a cultural change from a “highly institutionalized culture of dissimilarity” to a
“more agile and streamlined company”
Outsourcing is seen as prestigious in
SCANDI. Employees are promoted if
they have mastered outsourcing
Outsourcing is a silver bullet – a recipe for success
Table 9: Rational and institutional explanations for ES outsourcing (excerpt from paper 2)
One important implication of these findings is the value of the complementary learning
from applying not only a classic rational perspective, but also an institutional perspective. Each model provides a different, but still useful domain of learning. The two perspectives have progressed not only in terms of new concepts, but also by shifting the
domain of learning. The old rational model remains useful and important in developing
an understanding of outsourcing decisions. Newer institutional work has neither refined
nor supplanted the older work, but rather augmented and extended it.
94
Rationalized myths and enterprise systems
Paper 2 presents the rationalized myth that ES outsourcings are cost savings. SCANDI
undertook five major outsourcing events from 2005 to 2009, which affected ES implementation and operation. The prevailing argument for these outsourcing decisions was
cost saving, but is it really cost saving? If we take a transaction cost perspective (Williamson 1981), it is fairly obvious that the coordination cost among the many interorganizational partners is considerable (also designated coordination complexity), and if we
add to this the cost of the several transitions, then we end up with high total coordination costs that have to be compensated for by lower production costs in order to have
optimal economic efficiency – and there is no warrant for this claim. The cost for the
RE-ES project has increased by factor 5 from €1.6 million to €8 million during the
project course, which is a large increase even for an IS/IT project, covering both production and coordination costs. It makes sense to suppose that the coordination cost
would have been lower with fewer outsourcing vendors or a totally insourced project,
which is supported by informants and the fact that SCANDI has considered insourcing
some resources again. This is further warranted by industry analysts who point out that
multi-sourcing has high transitioning costs and requires 2 to 3 times more oversight than
working with a single partner (Overby 2010) (paper 2).
Other rationalized myths are presented, such as “effective work practices and an efficient enterprise system” (papers 3 and 4) and “the uncustomized enterprise system as
the most efficient system” (paper 4).
The value of combining institutional theory and sensemaking
IS researchers are advised to learn from organization studies (Orlikowski and Barley
2001) and this advice was followed in a series of papers in which we combine institutional theory and sensemaking theory (Jensen et al. 2009; paper 3; paper 4). Weber et al.
(2006) have used the combination in organization studies, which we adapted to the IS
context, and conclude in our first paper (Jensen et al. 2009: 350):
[W]e rely on both institutional and sensemaking theories. Institutional
theory provides explanations of the outcomes of institutional pressures and
logics on the [ES] implementation by having its main focus on the macro
95
level. Sensemaking theory directs our attention to the micro-level processes
and how organisational actors’ cognition and situated actions are made collective and reified through social construction processes … Both theories
are well-established and proven theories that offer valuable explanations of
[ES] implementation. However, based on our analysis, we allege that by
combining the theories, we attain even more valuable explanations of our
research phenomenon.
The multi-theory approach with institutional theory and sensemaking was subsequently
used in papers 3 and 4.
Paper 3 examined the potential of combining the two theories and applied it to the financial department in SCANDI, showing the reciprocal interaction between macro-level
institutional structures and micro-level sensemaking processes. First, changing institutional structures from monopoly to competition meant that the institutional pressure to
comply with governmental regulation was exchanged for competitive pressure with a
focus on cost and downsizing whereby the ES became a coercive institutional force in
order to streamline the information infrastructure to enable cost savings. Second, the
rationale of an effective ES as a rationalized myth was accepted by the accounts clerks,
which was developed along with the shift from monopoly to competition. They enacted
the use of the effective ES as part of their daily work practices and this served to construct their identity as productive employees. Finally, the enactment of the rather fixed
and cemented ES in practice reinforces the existing institutional structures.
We developed the multi-theory approach even further in paper 4 by applying the dual
structural model (see figure 11), which resulted in the findings in the table below (excerpt from paper 4):
Oracle Corporation
(Supplier organization)
SCANDI
(Customer organization)
Institutional
Properties
Oracle adapts and reinforces the
rationalized myth about “the efficient uncustomized ES” and uses marketing
communication to diffuse the message and
persuade customers
SCANDI has experienced a shift from
monopoly to competition, which means a
focus on more effective and streamlined
work practices
Enterprise
From a US-specific system to a “multieverything” global, highly integrated
First a highly customized version adhering
to the institutional logic “match to current
96
Oracle Corporation
(Supplier organization)
System
Human
Actors
SCANDI
(Customer organization)
(globalization process) and flexible
system with no need for customizations
business processes”
[Not investigated in this study]
Finance employees belong to a small and
harmonic group
Then after the RE-ES project a much less
customized version coming closer to the
institutional logic of “match to standard
package”
Long adaptation process from 1996 to
2009
Table 10: Findings related to the dual structural model of technology (excerpt from paper 4)
We asked the question “Why is the ES so well adapted in the financial department at
SCANDI …” and we present three sets of reasons for the successful adaptation that relate to: (1) The institutional properties where a rationalized myth about effective work
practices and an efficient ES travels from a national to a local level; (2) The nature of
the enterprise system with a long transition process from “match to current business
processes” towards “match to standard package”; and (3) The human actors’ enactment
of the ES in practice where existing structures are reinforced. We present three lessons
learned from the study: first customize then un-customize, be prepared for a long-term
adaptation process and consider the match between the users and the system.
We claim that there is much value in combining institutional theory and sensemaking
theory and we might only have cut the first turf with the three papers (Jensen et al.
2009; Paper 3; Paper 4). We therefore suggest other researchers to leverage our work to
their own purposes.
On the deinstitutionalization and institutionalization practices of enterprise systems reimplementation
The previous empirical papers (2, 3 and 4) are mainly about the management and use of
ES while paper 5 focuses on implementation. We decided to examine details about
deinstitutionalization processes as they appear to be vastly overlooked and asked the
following question: “What practices can be enacted to deinstitutionalize an existing enterprise system and fertilize the ground for a new one?”
We performed a multi-level analysis that shows the diversity and entanglement of institutionalization and deinstitutionalization practices, intended and unintended, aimed at
97
breaking the old system and fertilizing the ground for the new system. A major challenge in this study is to disentangle this practice and shed light on the specific sphere of
influence. Some are clearly deinstitutionalization pressures/practices as “release 11 is
not supported by vendor,” which clearly erodes the old system as this will become very
problematic in the longer run. Others are mainly institutionalization pressures/practices
fertilizing the ground for the new system, e.g. “enhanced business functionality in release 12.” Other pressures act in both ways like “reduce IT cost,” an example of pressure/practice to deinstitutionalize the old system and institutionalize the new system as
the cost reduction is not possible without the consolidation and un-customization part in
the RE-ES project. Specifying the practices affecting one process or the other is valuable because actions can be planned and sequenced. The model in the figure below
shows a few important pressures/practices leading to the deinstitutionalization of the old
system and the institutionalization of the new system as well as practices influencing
both deinstitutionalization and institutionalization.
Figure 12: Model of the dual deinstitutionalization and institutionalization process (adapted from paper 5)
We finally asked what we can learn from this case study. First, making new systems is
breaking old systems! The deinstitutionalization process is inseparable from the institu-
98
tionalization process, and both processes have to happen in order to implement the new
system. Second, the two processes overlap with each other, which is contrary to Greenwood et al.’s (2002) linear stages of institutional change, although their model addresses
the organizational field level while our model targets the organizational level, which
might imply differences. However, Hinings, Greenwood and colleagues (2004) have in
a newer model, “the dynamics of change,” also stated that the processes of de- and
reinstitutionalization pass off in parallel. Third, deinstitutionalization should take place
for both the old system and the old structures and practices as well as the institutionalization targeting both the new system and the new structures and practices. Fourth, the
pressures/practices for deinstitutionalization might also influence the institutionalization
process. Institutionalization of the new system/new practices might indirectly act as
pressures for deinstitutionalization.
This completes the summary of findings and implications across the papers, where we
have theorized by means of conceptual models and specific implications. The next
chapter will conclude the thesis.
99
7 CONCLUSION
This study has aimed to answer the overall research question “How do institutional
structures and processes shape the management, implementation and use of enterprise
systems?” The answer to the overall research question is the contribution substantiated
in the five papers and the cover part. In short, the contribution is two conceptual models
to describe and explore institutional structures and processes in organizations using ES.
The specific implications from the study are: (1) rational and institutional explanations
co-exist and complement each other in outsourcing decisions; (2) there is a reciprocal
interaction between macro-level institutional structures and micro-level sensemaking
processes such as changes in context from monopoly to competition and changing institutional logics from “match to current business processes” to “match to standard package” and, finally, (3) institutional processes such as deinstitutionalization and institutionalization play an important role when decommissioning old systems and cultivating
new systems and the processes overlap.
The rich picture (Checkland 1999) presented in figure 13, on the next page, summarizes
this contribution in a condensed way, and will be discussed subsequently.
The study was initiated by a literature review of how institutional theory has been used
in ES research, which indicated that institutional theory in ES research is its infancy
(paper 1). Several requirements for studying institutional structures and processes in ES
research were laid down, and a conceptual model was developed covering selected institutional elements, multi-level, multi-theory and multi-stakeholder characteristics (paper
1). The conceptual apparatus was enhanced with a dual structural model of technology
to emphasize that ES are delivered by supplier organizations to customer organizations
(paper 4). This makes up the fundamental lenses to study institutional structures and
processes for ES in organizations symbolized by the “eye” in figure 13. One of the major challenges in the RE-ES project was the multi-sourcing arrangements, and by means
of a multi-theory approach utilizing institutional theory and transaction cost theory, we
showed that the outsourcing decisions were based on both rational explanations (e.g.
cost savings) and institutional explanations (e.g. outsourcing as a silver bullet – a recipe
for success) and how they complement each other (paper 2).
100
Figure 13: Rich picture of the contribution from this research process
101
The multi-theory approach is furthermore used in paper 3 and paper 4, where we
showed the reciprocal interaction between macro-level institutional structures and micro-level sensemaking processes. The institutional context has changed from monopoly
to competition, which has facilitated the standardization of the ES. The rationalized
myth about the efficient ES travelled from society or the organizational field and was
accepted by the accounts clerks (paper 3). The next step was to involve the technology
supply, applying the dual structural model, which showed that the shift from a USspecific ES to a multi-everything ES had a considerable impact on the customer organization, and has been contributory factor to the shift in institutional logics from “match
to current business processes” to “match to standard package” together with the logics
of standardization and reducing customizations. Several lessons were learned, such as
first customize then un-customize the system, adaptation is a long-term process and the
importance of a good match between the ES and the users (paper 4). Deinstitutionalization and institutionalization processes have been examined related to the shift from
“current ES and practices” (release 11i) to “future ES and practices” (release 12). Some
of the lessons learned are that deinstitutionalization is inseparable from institutionalization and pressures/practices might impact on both processes (paper 5). The rich picture
in figure 13 emphasizes that institutional structures and processes play an important role
in the management, implementation and use of ES to address social considerations.
The remaining part of this chapter is devoted to implications for research and implications for practice, which can serve as stepping stones to future appropriations in both
domains. The chapter finishes with a very concise summary.
7.1 IMPLICATIONS FOR RESEARCH
This study has contributed to ES research by theorizing about ES in organizations using
institutional theory. Several topics and issues have been addressed in the thesis, mainly
derived from and in concert with the empirical context except for the literature review in
paper 1. However, it leaves other areas untouched and brings up new questions – some
of these will be discussed subsequently.
102
Implications for ES research
The thesis as a whole and particularly paper 1 can serve as a broad theoretical foundation for future research on ES using institutional theory, because it provides a comprehensive overview of key elements in institutional theory and its implications for ES research. This is especially relevant for IS researchers who are newcomers to institutional
theory. ES research using institutional theory is in its infancy and opens up quite a few
unexplored research avenues. An underutilized area is to apply the actor/sub-group level
or organizational field level as the unit of analysis instead of the organizational level,
which is the default unit of analysis in IS and ES research (see also Weerakkody et al.
2009). We can learn from organization studies (Orlikowski and Barley 2001) as there
are plenty of organizational field studies such as institutional change in health-care organizations (Scott et al. 2000), environmentalism and the U.S. chemical industry
(Hoffman 1999) and transforming from press to e-media in the business press field
(Mazza and Pedersen 2004), and add to these some studies within ES/IS research (e.g.
Currie and Guah 2007; Knutsen and Lyytinen 2008; Reimers 2003). One suggestion
could be to study the diffusion and use of ES (general, SAP, Oracle etc.) in a specific
field, sector, national or cross-national level. A further example is Pollock and Williams
(2009), who present The biography of the enterprise-wide system or how SAP conquered the world with the theoretical mainstay in science and technology studies (STS);
although carefully written and very enlightening, the topic might be enriched from an
institutional perspective. The actor/sub-group level studies are very different from the
organizational field level studies as there is limited research on micro processes in organization studies (Powell and Colyvas 2008) and ES research in particular (paper 1).
Powell and Colyvas (2008: 276) argue “that much analytical purchase can be gained by
developing a micro-level component of institutional analysis” and sensemaking is a
useful building block for micro-level perspectives in multi-theory analyses with institutional theory (Glynn 2008; Powell and Colyvas 2008: 282-284; Weber and Glynn
2006), which can be used to study ES adaptation(s) including misalignments. Our own
papers have to some extent involved the sub-group level as part of the multi-level studies we carried out (Jensen et al. 2009; paper 3; paper 4), but this could be extended to
other settings with the micro level as the unit of analysis.
103
I have argued that the ES artifact should be part of the conceptual model for studying
ES using institutional theory (see figure 10 and paper 1) in order to be specific about the
technology (Orlikowski and Iacono 2001), and we have made an attempt to be specific
about technology in papers 3, 4 and 5. However, the ES artifact is still fairly blackboxed in these papers and we have “difficulties in grasping the inner structure of the
technology artifact” (Czarniawska 2009: 50) as the ontology of ES is an intriguing and
complex topic. Gosain (2004) posits that ES are objects for institutional forces in the
design phase and then become carriers of institutional forces in the use phase, and he
furthermore labels the ES as the new iron cage drawing on Weber’s metaphor of rationality. Despite a convincing message this could be questioned. First, if we take the premise that ES are semi-finished products with best practices and built-in institutional
logics (paper 4) then they must be carriers of institutional forces from the very beginning when they enter the customer organization, and tailoring cannot undo the embedded institutional logics and the very presuppositions on which the ES rests (Kallinikos
2004). Second, the use process is assumed to be in the iron cage as ES are carriers of
institutional logics, but the use of ES develops over time, whether it is loose coupling
(Berente 2009), tailoring (customization, configuration etc.), changing/adding interfaced
systems (bolt-ons and more self-contained applications) or other kinds of adaptations
(Pollock and Williams 2009: 41-45). That is why ES cannot unequivocally be classified
as iron cages, but may be more like assembly kits with both rigidity and malleability.
The ES at SCANDI was heavily customized over time with the need-to-have elements
and nice-to-have adornments, but was later uncustomized through the RE-ES project.
The ontological status of the ES could be described as an assembly kit as iron cage appears to be the wrong label, expressing structures that are too fixed. It is probably an
organizational construction (such as a strategic choice) whether the ES turns into an iron
cage or an assembly kit. Both characteristics might nevertheless be endogenous properties of ES if we ontologically understand ES as multiple objects themselves (Quattrone
and Hopper 2006). The discussion reveals the complexity with the ontology of ES, and
it is my contention that this topic is relevant for further research, using institutional
analysis to investigate the materiality of ES and with the ES as the unit of analysis.
Misalignment has been described as a fundamental and complex problem in many studies (e.g. Hong and Kim 2002; Sia and Soh 2007; Wei et al. 2005) (see also section
104
2.4). Many strategies have been suggested to reduce misalignment, such as a high degree of tailoring (Hanseth and Braa 1998), drift of solution (Elbanna 2008), loose coupling (Berente 2009) as well as, the most radical, to abandon ES and use tailor-made
software instead (Kholeif et al. 2008). We cannot uncover significant misalignment at
SCANDI either in the finance department or in the purchasing department, even with
the drastic reduction in customizations from 400 to 150, and we state several explanations in our papers, which will not be repeated here (cf. papers 3, 4 and 5). However,
more overall questions can be brought up, such as “what is the degree of misalignment?” and “how is the degree of misalignment developing over time?” There has been
a mutual adaptation between the organization and the ES at SCANDI, but that was after
several upgrade projects and the major reimplementation project, so it is probably too
optimistic for organizations to expect a well-adapted ES after the first implementation,
and on the contrary accept that a number of iterations are needed before achieving a
well-adapted enterprise system. Thus, it might be beneficial to study the dynamics of
misalignment and follow longer-term adaptation processes using institutional logics and
other institutional instruments in order to theorize about the topic.
Implications for IS research and beyond
The findings described in the previous chapter 6 and the papers ought to have some kinship with IS in general, and it is tempting to generalize from ES to IS, but we have to be
very cautious and understand the differences between ES and other categories of IS.
First, ES has a number of particular characteristics, such as: they are semi-finished
products, implementations are organizational wide or even interorganizational wide
involving cross-functional business processes and many stakeholders, they are part of a
wider community (vendors, industry analysts, consultants, user groups etc.) and they
have lower asset specificity than much tailor-made software (easier to outsource). Nevertheless, there are contributions that can be applied to the broader IS context, especially
to categories of IS that have some resemblance to ES. The first category is “application
product software used in an organizational context.” Application software implies software used by end-users (contrary to system software) and “product software is defined
as a packaged configuration of software components … with auxiliary materials, which
is released for and traded in a specific market” with the principle “make one, sell many”
105
(Xu and Brinkkemper 2007). The other category is “tailor-made software used in an
organizational context” either provided in-house or by a contractor (outsourcing vendor)
(Xu and Brinkkemper 2007: 533-534). The conceptual model for using institutional
theory in ES research (figure 10) is fairly general and it is plausible that ES can be exchanged in the model with application product software and tailor-made software used
in an organizational context. The dual structural model of technology for ES (figure 11)
is more limited, where ES can be exchanged with application product software used in
an organizational context involving a supplier organization that delivers software to a
number of customer organizations.
Second, the value of combining institutional theory and sensemaking is applicable in an
IS context to study adaptation but also other processes like the development, implementation and management of IS, where it makes sense to combine macro-level institutional
structures and micro-level sensemaking processes (papers 3 and 4).
Third, the study of deinstitutionalization and institutionalization practices/pressures (paper 5) can be relevant in IS contexts where one system is replaced by another system.
However, the complexity of replacing ES can be much higher than replacing simple PC
software, but both scenarios involve institutional processes and the systems are expected to be part of the furniture in the long run (Silva and Backhouse 1997).
Fourth, the multi-sourcing issues, motives for outsourcing and not least the consequences presented in paper 2 appear to be applicable to other settings such as most
kinds of IS outsourcings, and maybe even more broadly to decision making in organizations beyond the IS context as organizations and individuals do not always make rational decisions (Scott 2008: 217-218), as paper 2 has underlined.
7.2 IMPLICATIONS FOR PRACTICE
The management, implementation and use of ES in practice have become “typified”
with the managerial and technical understanding of ES presented in chapter 2 (see also
Dillard and Yuthas 2006). The ES community is imprinted by “the rhetorics of technology supply” (Pollock and Williams 2009) embedding this understanding. Managers,
users, developers, consultants and other stakeholders appear to be brought up in this
tradition in business, management, software engineering and computer science although
106
sociological perspectives also sneak in. Applying institutional theory to ES offers practitioners conceptual tools for understanding and explaining complex scenarios in organizations involving ES (adapted from Currie 2009) to complement and contrast the typified managerial and technical understanding. We need to train practitioners in sociological terms and communicate institutional thinking to a broader audience in the trade
press and other media for practitioners. This is an overall consideration for practice and
the following will point to more specific implications.
Organizations and actors are encircled by rationalized myths, a few of which have been
elaborated and discussed in this thesis, such as “ES outsourcing is cost saving” and “the
uncustomized ES is the most efficient system.” Actors have a tendency, possibly unreflective, to jump on the bandwagon and adopt the mythologies. ES are fads or a fashion
in the meaning “a … collective belief in IS research and practice, disseminated by fashion setters, that a technique or technology leads to rational IS innovation” (Baskerville and Myers 2009: 649) and the myths are most likely emerged and brought out from
this collective belief cultivated by the fad setters. However, practice should take a critical position regarding these myths. The first step for practitioners is to be aware of the
myths, which is not easy as making choices such as selecting a vanilla implementation
strategy seems natural and legitimate, and becomes a standard organizational response
(adapted from Gosain 2004: 174). “The high triumph of institutional thinking is to make
the institution completely invisible” (Douglas 1986: 98), so the awareness process is
about making myths visible. The next step is to deconstruct the myth and make a more
reflective informed choice, and so be able to resist the bandwagon effect in the case of
rejecting the myth.
A renewed view on tailoring (configuration and/or customization) and adaptation
processes is needed in practice. Customization is seen as a nuisance and the ES community can easily present several convincing arguments for configuration only and vanilla
implementations. However, such an implementation strategy might seriously hamper
the adaptation process or even be impossible to implement (Hanseth and Braa 1998).
Managers and other key stakeholders are encouraged to take a more nuanced view of
tailoring, adaptation and their relationship. One approach to cultivate the issues is to
107
carry out action research (Mumford 2001) or collaborative practice research (Mathiassen 2002), where we engage in the intervention together with practice.
Finally, a personal experience from practice will be described. I have had the opportunity to discuss the findings from the ES outsourcing study in paper 2 at two practitioners’
workshops. Both theoretical frameworks, institutional theory and transaction cost
theory, were presented together with the findings. The theories and findings had great
appeal to the audience and resulted in fruitful discussions about motives for outsourcing
and decision making in general. This could remind us that applied research, such as ES
research, has academia and practitioners as its audience. Practice is sensitive to social
studies of ES despite its esoteric appearance, which can then be used to contrast the
possible managerial and technical understanding.
7.3 BRIEF SUMMARY
The purpose of this thesis was to theorize about enterprise systems in organizations using institutional theory. ES are widespread in both private and public organizations, and
provide potential benefits as well as high risk. The many challenges and problems associated with ES imply that they are a highly important topic for both practice and academia and much research has therefore been devoted to the topic. However, this research
has been dominated by a managerial and technical understanding, where social considerations are downplayed or even overlooked. The thesis addresses this issue and
presents an attempt to expand the knowledge about the social study of ES using institutional theory. An interpretive case study was conducted in SCANDI over a two-year
period (2008–2009), studying the management, implementation and use of their Oracle
E-Business Suite and particularly following a reimplementation project. The contributions are conceptual models to study ES in organizations and draw specific implications
about adaptation, customizations and decision making.
This brief summary rounds off the written work. The thesis with the cover part and the
papers is the tangible result, but there are many intangible outcomes from this challenging, exciting and intense learning process, which I described as a mental marathon in the
preface. The end of a three-year research journey is coming closer, although I still have
to cross the finishing line!
108
APPENDIX A – KANT’S TRANSCENDENTAL IDEALISM
Without … knowledge of Kant, one can scarcely even be said to have a
knowledge of the history of modern philosophy! (Schacht 1984: 222)
It is difficult, if not impossible, to position Kant’s transcendental idealism in a brief appendix, and it deserves a much more comprehensive and detailed representation than is
possible here. The purpose of this appendix is to highlight some key understandings of
Kant’s theory of perception centered on transcendental idealism, but to leave out the
more “long-haired” philosophical arguments and discussions in order to keep it relatively short. This appendix is mainly based on Schacht’s (1984: chapter VII) interpretation
of Kant and complemented by Kant’s original work (1781 (2007); 1783 (1997); 1783
(2007)).
Metaphysics has traditionally investigated principles of reality transcending any particular science. Metaphysics sought to conceive the world, the soul and God – the knowledge of the fundamental natures of things – through the use of pure reason, because
empirical knowledge based on sensing lacked the certainty that they sought. The rationalists did not question that the pure reason was able to understand the fundamental
nature of things, and operate independently of our senses. Kant did not question this
until late in his life. It was Hume’s attack on metaphysics that awakened Kant, so he
understood that there was a problem. Hume demonstrated irrefutably that it was absolute impossible for (pure) reason to think a priori concerning the connection of cause
and effect (causality), but also had problems explaining this connection by sense experience only (Husted and Lübcke 2001: 127-144; Kant 1781 (2007); 1783 (2007);
Schacht 1984). Kant repositioned this problem:
The question was not whether the concept of cause was right, useful, and
even indispensable for our knowledge of nature … but whether the concept
could be thought by a reason a priori … and consequently whether it possessed an inner truth, independent of all experience … This was Hume’s
problem. (Kant cited in Schacht 1984: 223)
109
Kant started to philosophize over this problem and realized that the connection between
cause and effect was not the only concept for which the understanding thinks the connection of things a priori, but rather that metaphysics consists of the whole of such concepts (“categories of understanding” from metaphysical deduction). Thus, Kant arrived
at a different conclusion from Hume – namely that metaphysics is not meaningless although chastened. The speculation about the nature of things in themselves (world, soul
and God) is beyond the phenomena of perceptual experience, devoid of meaning and
could not be ascribed knowledge. Metaphysics for Kant is synthetic and a priori – no
empirical or posteriori knowledge of metaphysics is possible. Metaphysics can only
yield essential structures of the phenomena we experience, and the structures of the
mind they assume, but not the existence and natures of objects, which are beyond possible experience (Kant 1781 (2007); 1783 (2007); Schacht 1984). Kant’s metaphysical
view could lead to the interpretation that there are no things in themselves, but this is
not the case as Kant argues:
Though we cannot know these objects as things in themselves, we must yet
be in a position at least to think them as things in themselves; otherwise we
should be landed in the absurd conclusion that there can be appearance
without anything that appears. (Kant cited in Schacht 1984: 226)
The quotation emphasizes one of the main issues treated in the “Critique of Pure Reason” (Kant 1781 (2007)) where Kant attempts to bridge the gap between rationalism and
empiricism. He argues concerning the pure reason that we cannot know all the things
that the rationalists thought we could, but we can know much more than the empiricist
thought we could – thereby “bridging the gap.” He rejects the rationalist assumption
that reality corresponds to the determinations of human reason, but on the other hand
argues that the structure of the empirical reality must reflect the categories of understanding (Schacht 1984).
Kant’s method has been known as the transcendental method. It consists of stating various situations of experience, which are irrefutable, and then investigating what they
presuppose (Schacht 1984), i.e. how could they be understood. The method is used in
his treatment of transcendental problems like “how is pure mathematics possible?” and
110
“how is natural science possible?” (Kant 1783 (1997)), and “he convinced us” that mathematics and natural science are both synthetic and a priori (Schacht 1984).
Kant argues that experience is not simply an unstructured stream of consciousness, but
instead a unified structure of empirical knowledge. This stems from the fact that the
mind is not merely passive but active, and structures our experience so that the synthetic
a priori judgments of mathematics and natural science apply within the possible experience and present a unified structure to that experience. Kant continues on this road
explaining the nature of the mind and thereby forms his metaphysical system (Schacht
1984). This conception leads to the famous quotation about mind and experience:
Though all our knowledge begins with experience, it by no means follows
that all arises out of experience. (Kant 1781 (2007): 25)
Our experience exists not in our stream of consciousness, but in our judging that things
are this and that. The metaphysical system is a necessary condition for this judgment,
and this metaphysical account of mind was called transcendental idealism, which
makes sense of these conditions for the possibility of experience. Our knowledge is limited to the experience by our forms of sensibility (time and space) and categories of
understanding (ding für uns), and we cannot understand the nature as they are in themselves (ding an sich) (Larmore 2003: 261; Lübcke 2004: 227-234; Schacht 1984).
My view of mind and world developed before Kant “came” to my knowledge, but reading and pursuing an understanding of Kant made me realize that there is some congruence between his view and my view. The implicit understanding I had of mind and
world has to some extent been explicated by Kant and could be summarized as:
•
The physical reality or universe exists independently of human cognition,
thought and speech processes, i.e. ontological realism, but our knowledge about
the physical reality is a social construction thereby rejecting epistemological
realism in the understanding that objective cognition of an independent reality is
possible (Becker and Niehaves 2007: 202; Collin 2003: 24; Fuglsang and Bitsch
Olsen 2004: 81-82).
•
The social reality or universe is a social construction. Multiple social realities
exist as subjective constructions of the mind (ontological relativist position)
111
(Fitzgerald and Howcroft 1998: 319), and one perceived reality presents itself
for the individual human being. The knowledge about the social reality, including everyday life common-sense knowledge, is negotiated, institutionalized and
maintained through social interaction (Berger and Luckmann 1966: 35).
A final comment! Kant’s transcendental idealism or more general theory of perception
has had an enormous influence on today’s thinking, but has also been widely criticized
and debated. Schelling wrote to Hegel on 6 January 1795: “Kant has given us conclusions, but premises are still lacking. And who can understand the conclusions without
the premises?” The quotation was part of the philosophical debate at that time “to
ground philosophical theory of mind and world upon one ultimate principle” (foundational principle), which was different from Kant’s position (Larmore 2003: 262).
Kant has described “space and time as forms of sensibility,” which are known independently of experience, and as the framework in which all appearances occur (Schacht
1984: 232). Space is identical to the Euclidian geometry according to Kant, but this understanding is in conflict with Einstein’s theory of relativity where space could best be
described with non-Euclidian geometry (Husted and Lübcke 2001: 138-139), so Kant’s
proposition could be questioned. However, we still find Newtonian physics useful despite the fact that it has been disproved by Einstein and others, and the same argument
might be applied to Kant. I am not a dogmatic Kantian, but I find some useful patterns
in his thinking congruent with mine, which this appendix has intended to unfold.
112
APPENDIX B – SOCIAL CONSTRUCTIONISM VERSUS
SOCIAL CONSTRUCTIVISM
The terms social constructivism and social constructionism are closely related concepts
and are often used interchangeably (Kragh 2007: 14 footnote). Papert (1991) is cited in
Leonardi and Barley (2008: 168-169): “[He] uses constructivism to refer to the cognitive processes by which people construct unique understandings and interpretations of
the world. Constructionism, on the other hand, involves communicative acts in which
multiple people, through their interaction with one another, make the world in common.
Constructivism highlights subjectivity, while constructionism concerns the intersubjective.”
I have chosen social constructionism because Berger and Luckmann (1966) discuss
social construction of knowledge as an outcome of intersubjective processes (multiple
people). However, the distinction is hardly that important in this context as long as the
underlying premises are understood. Refer to Burr (2003:19-20), Leonardi and Barley
(2008:168-171), Papert (1991) and Patton (2002: 96-103) for a more elaborated discussion of the two related concepts.
The distinction between social constructionism and social constructivism is less clear in
Danish literature, where social constructionism is often translated as socialkonstruktivisme in social sciences (Kragh 2007: 14 footnote) – examples are Andersen et al.
(2007: 234), Fuglsang and Bitsch Olsen (2004) and Wenneberg (2002).
113
APPENDIX C – EXAMPLE OF AN INTERVIEW GUIDE
Below is the interview guide in which some concepts are made anonymous (organization, enterprise system “ES” is the current version and NEW-ES is the reimplemented
version of the enterprise system):
ID Heading
Research
questions
Interview questions/topics
Scale Theoretical anc(1–10) horing
Name
Time
Make an appointment with the interviewee and explain briefly what it is
all about. Important: ask the interviewee to find a meeting room where we
can talk without interruption.
Send out an electronic notification of
the meeting with a short description
of my role and what the interview is
about. This is the start of informed
consent. (Kvale 1997:118-120)
Read the aspects of the qualitative
research interview as a kind of mental
mind setter. (Kvale 1997:41-42)
(Kvale 1997)
Introduction
and interview
procedure
Short presentation of the main points:
• Information about the interview: duration, recording and anonymity
• Purpose of the interview and the research project as a whole …
(Kvale 1997)
Voluntary
participation
in the interview
Have you voluntarily decided to participate in this interview?
Before the
interview
1.
2.
3.
1
Do you have any questions before we
start?
2
Personal data
2.01
Name (row has not been removed in order
to ensure consistency with previous interviews).
2.02
Sex:
2.03
(Age – estimate made by PS):
2.04
Work experience (number of years):
2.05
Employed at SCANDI (number of years):
2.06
Worked with ES (number of years):
2.07
Background: education, work experience
114
Yes or (Kvale 1997:118No
120)
(Kvale 1997:132)
ID Heading
Research
questions
Interview questions/topics
Scale Theoretical anc(1–10) horing
etc.
3
Work tasks at
SCANDI
3.01
Which role do you play in ES?
3.02
What tasks do you carry out?
3.04
Specifically discuss work task XX
4
Organizational performance
How effectively do ES function in your
daily work?
4.01 Use of the
system
How much of your working time directly
involves the use of ES?
4.02 User satisfaction
How satisfied are you with ES in your
daily work? (on a scale from 1 to 10
where 10 is best – please comment on
this)
1–10
(DeLone and
McLean 2002)
4.03 User
friendliness
How user friendly do you find ES?
1–10
(on a scale from 1 to 10 where 10 is best –
please comment on this)
(DeLone and
McLean 2002)
4.04 Information
quality
What is the quality of the information in
1–10
the system? (on a scale from 1 to 10 where
10 is best – please comment on this)
• Is the data content relevant and
adequate?
• Do you receive data in due time?
• Is the quality of the data OK?
(DeLone and
McLean 2002)
4.05 System quality
What is the quality of the system? (on a
scale from 1 to 10 where 10 is best –
please comment on this)
• User friendliness (please see
above)
• Consistent in use
• Response time
• System error
• Consistent user interface
• Maintenance
(DeLone and
McLean 2002)
4.06 Service quality
How is the service on the system? (on a
1–10
scale from 1 to 10 where 10 is best –
please comment on this)
• Support/service desk
• Training in the system
• Supervision of the system and action in connection with errors
(DeLone and
McLean 2002;
Stylianou and Kumar 2000)
4.07 Process diagram
Description of the process and specification of who is talking to whom
• Process diagram and if possible
swim lane diagram with the in-
(Barley and Tolbert
1997:105-106)
115
(DeLone and
McLean 2002)
1–10
ID Heading
Research
questions
Interview questions/topics
Scale Theoretical anc(1–10) horing
terested parties
4.08 Performance
for a business
process
5
Determine/discuss performance in connection with a specific business process
• e.g. when setting up a new supplier (both manual time consumption and the system’s time
consumption)
• Please notice that the order of
4.07 and 5.01 can be changed according to what seems most logical during the interview
Describe the framework of the present
work situation
Institutional
theory
Ensure specificity, i.e. obtain concrete
examples instead of general considerations
5.01 Habits (open
discussion,
make sure that
you obtain the
story about
their life
world)
5.02 Interpretation
of the process
(Kvale 1997:41)
Imagine that a new employee becomes a
member of your group – please answer the
following questions:
a. What would the new employee notice?
b. What should the new employee learn?
c. What would the new employee regard
as special?
The partici- •
pant’s own •
interpreta•
tions of what
is happening
in the
process
•
5.03 Regulative
mechanism
•
•
•
•
5.04 Normative
mechanism
•
•
•
5.05 Cognitive mechanism
(Melville et al.
2004)
•
How is the process interpreted?
Could you have acted differently?
Would you have acted differently if
you were to decide?
(Barley and Tolbert
1997:105-106)
What procedures (rules) do you have
to follow when you use ES?
To what extent are those procedures
formal or informal?
How does the company follow up on
the use of the procedures?
What happens if you do not follow
the procedures?
(Scott 2001:35-37)
What attitudes, values and norms
characterize your work in the department?
Do you share these norms in the entire group?
Or are there differences?
Does ES influence attitudes, norms
and values – and how?
(Scott 2001:37-40)
Sense- (Difficult question …)
making? • How would you describe the culture?
116
(Barley and Tolbert
1997; Berger and
ID Heading
(Symbols in
the form of
words, signs
and gestures)
Research
questions
•
•
Cognitive
maps?
Scripts?
Interview questions/topics
•
•
6
(the interviewee will have to define
the question him/herself, i.e. the culture at SCANDI, the culture in the
department” etc.)
How would you describe your work
situation? (Metaphors, analogies, stories, anecdotes, plays).
How would you describe your work
with ES? (Metaphors, analogies, stories, anecdotes, plays).
Luckmann 1966;
Scott 2001:40-45)
Expectations
of the reimplementation
project
6.01
What are your expectations of NEW-ES?
6.02
How do you think that the NEW-ES will
affect your work?
• In the first three months
• In the long run
7
Scale Theoretical anc(1–10) horing
Open questions
To provide an opportunity for exceeding
the semi-structured interviews’ limitations
7.01
Are there other conditions that you would
like to mention in connection with our talk
today?
7.02
Are there questions or topics that you
would have liked me to ask or talk about?
7.03
Would it be useful for me to discuss with
other persons the topics that we have already discussed?
8
Thank you for
your participation and
explain about
the further
process
Briefly explain the further process:
• This is measuring point 1
• I will be back immediately after
the launch of NEW-ES
• Repeat the conditions about anonymity
9
After the interview
Spend at least 10 minutes on reflection
and write down essential observations
117
(Creswell
2007:129-134)
(Kvale 1997:133)
APPENDIX D – TEMPLATE FOR FIELD NOTES IN
PARTICIPANT OBSERVATION
The figure below shows the template for participant observation:
Figure 14: Template for participant observation
The figure above is divided into a number of frames, and here follows a short explanation of these frames.
n Header information: This is essential information about the observation such as
“type of observation”, “place, date & time” and “page” number.
o Primary observation – chronological log (direct field notes): This frame can be used
much in the same way as taking minutes or other ways of note taking (e.g. Angrosino
2007; Myers 2009).
p Time & category: Time is used to gain a temporal understanding of how the time is
distributed across the topics in the observation. Category is used to assign descriptive
codes (Miles and Huberman 1994: 57).
118
q No title: This section is used for initial data analysis, which could be performed during the participant observation or afterwards. Four predefined categories are suggested:
(A) reflection and recall, (B) ideas and inferences as a kind of pre-analysis of data, (C)
experiential data like impressions and personal feelings and, finally, (D) forward planning, which is proposed actions to be undertaken (Gray 2004: 244-250). The categories
A, B, C & D are used as a reminder of some possible data analysis activities to take on
during and after the participant observation.
119
APPENDIX E – EXAMPLES OF DOCUMENTS
Below is a list of internal and public documents. The list is not exhaustive but aims to
present the major types of documents used in the research project.
Internal documents (private documents from SCANDI)
-
-
Project description (including business case and other documents)
Minutes of meetings
Stakeholder analysis
Emails
Newsletters
Presentations about:
o Project description, project organization, roles and responsibilities
o Status of project groups
o The story about how the project came about
o Process descriptions (swim lane diagrams)
o Performance test approach
o Meeting presentations (e.g. steering committee meetings)
o Strategy and policy presentations
o Enterprise system user survey
Organizational charts (printed from the intranet)
Schedules, plans (MS-Project plans)
User manuals (e.g. management of local warehouses)
Screen images (business intelligence system connected to enterprise system)
Service level agreements
Issues and risks matrices (Excel spreadsheet)
Oracle-specific project documents (e.g. CV.010 Conversion Requirements and
Strategy … and TE.010 Testing Requirements and Strategy)
List of customizations (Excel spreadsheet)
Public documents (SCANDI, Oracle and outsourcing vendors)
-
Annual reports
Press releases
Company history
Internet information (e.g. www.oracle.com)
Articles in newspapers
Articles in trade press (e.g. Computerworld)
White papers and reports from consultancy and analyst companies (Deloitte Consulting, Forrester Research etc.)
120
APPENDIX F – CODING IN NVIVO
The table below shows the codes from the coding of paper 4 (coding list exported from
NVivo to a table):
01 Why is the ES so w ell-adapted in the financial department at SCANDI
A secure w orld
1
1
26-10-2009 08:07
Decoupling of decentral users w ith pre-systems like KIWI for invoice handling and travel expenses
1
1
26-10-2009 08:03
Everyone makes an effort to get w ork done
1
1
26-10-2009 08:06
Fairly quick stabilization after R12 go live
6
13
13-08-2009 13:12
Familiar w ith FinSys
1
1
26-10-2009 08:05
Few bugs in FinSys (spring 2008)
1
1
13-08-2009 11:45
Few or no suggestions for functional improvement of R12
6
7
13-08-2009 13:30
Few users in Finance Department (5-10 heavy users and 30 users in total)
2
3
13-08-2009 11:39
Finance employees are a harmonic group
7
8
13-08-2009 12:57
Finance employees have been in Finance for many 10, 20 or even more
6
8
11-08-2009 10:39
FinsSys is perceived as the hub of 'the' universe i.e. Very important system
2
3
13-08-2009 11:46
FinSys is perceived to w ork w ell
1
3
13-08-2009 12:32
Firing people can cause a forced adaptive attitude among employees
1
1
14-08-2009 10:17
Job fits w ell w ith personal w orking style
1
1
14-08-2009 08:56
Keep customers satisfied and happy
1
1
26-10-2009 08:06
Local module responsible has both IT and business know ledge
1
1
26-10-2009 08:02
Optimize procedures
1
1
26-10-2009 08:06
Peer review , control, police
1
1
26-10-2009 08:07
Perceived improvements after RE-ES launch
6
10
14-08-2009 09:14
Quality of w ork
1
1
26-10-2009 08:06
Revisions and regulations (norms, rules, procedures)
1
1
26-10-2009 08:06
Routines (it has alw ays been like that)
1
1
26-10-2009 08:07
Started w ith many customizations w hich have decreased since
7
13
11-08-2009 10:44
Support employees at vendor know the system very w ell
1
1
13-08-2009 13:31
The several delays in launch of RE-ES project have increased the quality
1
1
13-08-2009 13:42
Time The first FinSys version w as implemented in 1996
3
5
11-08-2009 10:38
We are in the same boat
3
3
13-08-2009 13:38
Well established support' organization
4
5
14-08-2009 08:39
Well-defined w ork procedures roles
1
1
26-10-2009 08:07
Working Practice w ell established
7
18
13-08-2009 12:31
Information quality
7
7
13-08-2009 12:45
Service quality
7
8
13-08-2009 12:47
System quality
7
7
13-08-2009 12:46
User satisfaction
7
7
13-08-2009 12:40
User-Friendliness and Ease of Use
7
7
13-08-2009 12:43
02 Delone and McLean IS Success Model
03 Other coding miscellaneous
Decreased efficiency w ith R12
5
11
14-08-2009 16:17
Expectations of re-implementation process (stabilization)
1
1
14-08-2009 16:17
Freezing of 11i before launch of R12 caused a w ork backlog
1
1
14-08-2009 16:17
Loose coupling of systems thorugh batch interfaces
4
4
14-08-2009 16:17
Shift from local language to English has caused problems
2
4
14-08-2009 16:17
Very complex IT infrastructure w ith 40 plus integrations makes it a rigid structure
3
3
14-08-2009 16:17
Table 11: Example of coding in NVivo
121
REFERENCES
Altheide, D. (1996). Qualitative Media Analysis. Thousand Oaks: Sage Publications
Inc.
Alvarez, R. (2001). "It was a great system": Face-work and the Discursive Construction
of Technology during Information Systems Development. Information Technology & People 14(4): 385-405.
Alvarez, R. (2002). The Myth of Integration: A Case Study of an ERP Implementation.
In F. F. H. Nah (Ed.), Enterprise Resource Planning Solutions and Management.
Hershey: Idea Group Inc., 63-88.
Andersen, H., T. Brante and O. Korsnes (2007). Leksikon i sociologi. København: Akademisk Forlag.
Angrosino, M. (2007). Doing Ethnographic and Observational Research. London: Sage
Publications Ltd.
Arbnor, I. and B. Bjerke (1997). Methodology for Creating Business Knowledge. Thousand Oaks: Sage Publications Inc.
Arbnor, I. and B. Bjerke (2009). Methodology for Creating Business Knowledge. London: Sage Publications Ltd.
Bailey, C. A. (2007). A Guide to Qualitative Field Research. Thousand Oaks: Pine
Forge Press.
Bancroft, N. H., H. Seip and A. Sprengel (1998). Implementing SAP R/3: How to Introduce a Large System into a Large Organization. Greenwich: Manning.
Barley, S. R. and P. S. Tolbert (1997). Institutionalization and Structuration: Studying
the Links between Action and Institution. Organization Studies 18(1): 93-117.
Baskerville, R., L. and M. D. Myers (2009). Fashion Waves in Information Systems
Research and Practice. MIS Quarterly 33(4): 647-662.
Bazeley, P. (2007). Qualitative Data Analysis with NVivo. London: Sage Publications
Ltd.
Beatty, R. C. and C. D. Williams (2006). ERP II: Best Practices for Successfully Implementing an ERP Upgrade. Communications of the ACM 49(3): 105-110.
Becker, J. and B. Niehaves (2007). Epistemological Perspectives on IS Research: A
Framework for Analysing and Systematizing Epistemological Assumptions. Information Systems Journal 17(2): 197-214.
Bendoly, E. and F. R. Jacobs. (2005). Strategic ERP extension and use. Retrieved 16.
December, 2009, from http://site.ebrary.com/lib/royallibrary/Doc?id=10110292.
122
Berente, N. (2009). Conflicting Institutional Logics and The Loose Coupling of Practice
with NASE's Enterprise Information System. Department of Information Systems, Case Western Reserve University, Cleveland. PhD Thesis.
Berger, P. L. and T. Luckmann (1966). The Social Construction of Reality. A Treatise in
the Sociology of Knowledge. New York: Doubleday.
Boersma, K. and S. Kingma (2005). From means to ends: The transformation of ERP in
a manufacturing company. Journal of Strategic Information Systems 14(2): 197219.
Boudreau, M. C. and D. Robey (2005). Enacting Integrated Information Technology: A
Human Agency Perspective. Organization Science 16(1): 3-18.
Boxenbaum, E. and S. Jonsson (2008). Isomorphism, Diffusion and Decoupling. In R.
Greenwood, C. Oliver, K. Sahlin and R. Suddaby (Eds.), The SAGE Handbook
of Organizational Institutionalism. London: Sage Publications, 78-98.
Brehm, L. and M. L. Markus (2000). The Divided Software Life Cycle of ERP Packages. Proceedings of 1st Global Information Technology Management (GITM)
World Conference, Memphis, Tennessee.
Brewer, J. D. (2000). Ethnography. Philadelphia: Open University Press
Brier, S. (2005). Informationsvidenskabsteori. Frederiksberg: Samfundslitteratur.
Brynjolfsson, E. and A. Saunders (2010). Wired for Innovation. London: The MIT
Press.
Burr, V. (2003). Social Constructionism. Hove, East Sussex: Routledge.
Burrell, G. and G. Morgan (1979). Sociological Paradigms and Organisational Analysis. Elements of the Sociology of Corporate Life. London: Heinemann Educational.
Caccia, L. and I. Steccolini (2006). Accounting change in Italian local governments:
What's beyond managerial fashion? Critical Perspectives on Accounting 17(23): 154-174.
Chae, B. and G. F. Lanzara (2006). Self-destructive Dynamics in Large-scale Technochange and Some Ways of Counteracting it. Information Technology & People
19(1): 74-97.
Chang, K.-c., A. Gold and W. Kettinger (2003). The Extent of Enterprise System Adoption in Companies: A Multiple Theoretical Perspective. AMCIS 2003 Proceedings. Paper 56.
Checkland, P. (1985). From Optimizing to Learning: A Development of Systems Thinking for the 1990s. Journal of the Operational Research Society 36(9): 757-767.
123
Checkland, P. (1999). Soft Systems Methodology: A Thirty Year Retrospective. Chichester: John Wiley & Sons, Ltd.
Ciborra, C. U. and A. Failla (2001). Infrastructure as a Process: The Case of CRM in
IBM. In C. U. Ciborra (Ed.), From Control To Drift - The Dynamics of Corporate Information Infrastructures Oxford: Oxford University Press, 105-124.
Collin, F. (1997). Social Reality. London: Routledge.
Collin, F. (2003). Konstruktivisme. Frederiksberg: Samfundslitteratur.
Cooper, R. B. and R. W. Zmud (1990). Information Technology Implementation Research: A Technological Diffusion Approach. Management Science 36(2): 123139.
Cordella, A. and M. Shaikh (2006). From Epistemology to Ontology: Challenging the
Constructed "Truth" of ANT. Working Paper Series no. 143, Department of Information Systems, London School of Economics and Political Science.
Creswell, J. W. (2007). Qualitative Inquiry and Research Design. Thousand Oaks: Sage
Publications Inc.
Currie, W. (2009). Contextualising the IT artefact: Towards a Wider Research Agenda
for IS using Institutional Theory. Information Technology & People 22(1): 6377.
Currie, W. L. and M. W. Guah (2007). Conflicting Institutional Logics: A National Programme for IT in the Organisational Field of Healthcare. Journal of Information
Technology 22(3): 235-247.
Czarniawska, B. (2009). How Institutions are Inscribed in Technical Objects and What
it may mean in the case of the Internet. In F. Contini and G. F. Lanzara (Eds.),
ICT and Innovation in the Public Sector - European Studies in the Making of EGovernment. Houndmills, Basingstoke, Hampshire: Palgrave Macmillan, 49-65.
Czarniawska, B. and B. Joerges (1996). Travels of Ideas. In B. Czarniawska and G.
Sevón (Eds.), Translating Organizational Change. Berlin: Walter de Gruyter,
13-48.
D’Aquila, M., J. Freyermuth, S. Jacobson, J. Shepherd, N. Tohamy, M. N. Rizza, M.
Burkett, D. Aquino and C. Fletcher (2009). The Global Enterprise Application
Market Sizing Report, 2008–2013 AMR Research, Inc, Document number
AMR-R-20495.
Davenport, T. H. (1996). Holistic Management of Megapackage Change: The Case of
SAP. AMCIS 1996 Proceedings.
Davenport, T. H. (1998). Putting the Enterprise into the Enterprise System. Harvard
Business Review 76(4): 121-131.
124
Davenport, T. H. (2000). Mission Critical: Realizing the Promise of Enterprise Systems.
Boston: Harvard Business School Press.
Davenport, T. H., J. Harris and S. Cantrell (2002). The Return of Enterprise Solutions:
The Director's Cut. Accenture Institute for High Performance Business Research
Report.
Davenport, T. H., G. Harris Jeanne and S. Cantrell (2004). Enterprise Systems and Ongoing Process Change. Business Process Management Journal 10(1): 16-26.
Deloitte Consulting (1999). ERP's Second Wave: Maximizing the Value of ERPEnabled Processes. Deloitte Consulting. New York: Deloitte Consulting LLC.
DeLone, W. H. and E. R. McLean (2002). Information Systems Success Revisited. System Sciences, 2002. HICSS. Proceedings of the 35th Annual Hawaii International Conference on System Sciences: 2966-2976.
DeLone, W. H. and E. R. McLean (2003). The DeLone and McLean Model of Information Systems Success: A Ten-year Update. Journal of Management Information
Systems 19(4): 9-30.
Denzin, N. K. and Y. S. Lincoln (2005). Introduction. In N. K. Denzin and Y. S. Lincoln (Eds.), The Sage Handbook of Qualitative Research. Thousand Oaks: Sage
Publications Inc, 1-32.
Dillard, J. F. and K. Yuthas (2006). Enterprise Resource Planning Systems and Communicative Action. Critical Perspectives on Accounting 17(2-3): 202-223.
DiMaggio, P. and W. W. Powell (1991). Introduction. In W. W. Powell and P. DiMaggio (Eds.), The New Institutionalism in Organizational Analysis. Chicago: The
University of Chicago Press, 1-40.
DiMaggio, P. J. (1988). Interest and Agency in Institutional Theory. In L. G. Zucker
(Ed.), Institutional Patterns and Organizations: Culture and Environment.
Cambridge, Massachusetts: Ballinger Publishing Company, 3-21.
DiMaggio, P. J. and W. W. Powell (1983). The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields. American Sociological Review 48(2): 147-160.
Donaldson, L. (1995). American Anti-Management Theories of Organization: A Critique of Paradigm Proliferation. Cambridge: Cambridge University Press.
Douglas, M. (1986). How Institutions Think. New York: Syracuse University Press.
Eisenhardt, K. M. (1989). Building Theories from Case Study Research. Academy of
Management Review 14(4): 532-550.
Elbanna, A. R. (2008). Strategic Systems Implementation: Diffusion through Drift.
Journal of Infomation Technology 23(2): 89-96.
125
Elmes, M. B., D. M. Strong and O. Volkoff (2005). Panoptic Empowerment and Reflective Conformity in Enterprise Systems-Enabled Organizations. Information and
Organization 15(1): 1-37.
Esteves, J. and V. Bohorquez (2007). An Updated ERP Systems Annotated Bibliography: 2001-2005. Communications of the Association for Information Systems
19(Article 18): 386-447.
Feyerabend, P. K. (1975). Against Method. London: Verso Books.
Finney, S. and M. Corbett (2007). ERP Implementation: A Compilation and Analysis of
Critical Success Factors. Business Process Management Journal 13(3): 329-347.
Fitz-Gerald, L. and J. Carroll (2003). The Role of Governance in ERP System Implementation. Australasian Conference on Information Systems, Perth, Western
Australia.
Fitzgerald, B. and D. Howcroft (1998). Towards Dissolution of the IS Research Debate:
From Polarization to Polarity. Journal of Information Technology 13(4): 313326.
Fligstein, N. (2001). Social Skill and the Theory of Fields. Sociological Theory 19(2):
105-125.
Flyvbjerg, B. (2006). Five Misunderstandings About Case-Study Research. Qualitative
Inquiry 12(2): 219-245.
Foucault, M. (1977). Discipline and Punish: The Birth of the Prison. London: Penguin
Books.
Friedland, R. and R. R. Alford (1991). Bringing Society Back In: Symbols, Practices,
and Institutional Contradictions. In W. W. Powell and P. DiMaggio (Eds.), The
New Institutionalism in Organizational Analysis. Chicago: University of Chicago Press, 232-263.
Fuglsang, L. and P. Bitsch Olsen (2004). Videnskabsteori i Samfundsvidenskaberne. På
Tværs af Fagkulturer og Paradigmer. Frederiksberg: Roskilde Universitetsforlag.
Gallivan, M. J. (2001). Organizational Adoption and Assimilation of Complex Technological Innovations: Development and Application of a New Framework.
ACM SIGMIS Database 32(3): 51-85.
Geelan, D. R. (1997). Epistemological Anarchy and the Many Forms of Constructivism.
Science & Education 6(1-2): 15-28.
Geertz, C. (1973). The Interpretation of Cultures: Selected Essays. New York: Basic
Books.
126
Glynn, M. A. (2008). Beyond Constraint: How Institutions Enable Identities. In R.
Greenwood, C. Oliver, K. Sahlin and R. Suddaby (Eds.), The SAGE Handbook
of Organizational Institutionalism. London: Sage Publications, 413-430.
Golden-Biddle, K. and K. Locke (1993). Appealing Work: An Investigation of How
Ethnographic Texts Convince. Organization Science 4(4): 595-616.
Gosain, S. (2004). Enterprise Information Systems as Objects and Carriers of Institutional Forces: The New Iron Cage. Journal of the Association for Information
Systems 5(4): 151-182.
Grabski, S. V., S. A. Leech and B. Lu (2003). Enterprise System Implementation Risks
and Controls. In P. B. Seddon, L. Willcocks and G. Shanks (Eds.), Second-Wave
Enterprise Resource Planning Systems: Implementing for Effectiveness. Cambridge: Cambridge University Press, 135-156.
Gray, D. E. (2004). Doing Research in the Real World. London: Sage Publications.
Greenwood, R. and C. R. Hinings (1996). Understanding Radical Organizational
Change: Bringing together the Old and the New Institutionalism. Academy of
Management Review 21(4): 1022-1054.
Greenwood, R., C. Oliver, K. Sahlin and R. Suddaby (2008a). Introduction. In R.
Greenwood, C. Oliver, K. Sahlin and R. Suddaby (Eds.), The SAGE Handbook
of Organizational Institutionalism. London: Sage Publications, 1-46.
Greenwood, R., C. Oliver, K. Sahlin and R. Suddaby (2008b). The SAGE Handbook of
Organizational Institutionalism. London: Sage Publications.
Greenwood, R., R. Suddaby and C. R. Hinings (2002). Theorizing Change: The Role of
Professional Associations in the Transformation of Institutionalized Fields.
Academy of Management Journal 45(1): 58-80.
Guba, E. G. and Y. S. Lincoln (1994). Competing Paradigms in Qualitative Research. In
N. K. Denzin and Y. S. Lincoln (Eds.), The Sage Handbook of Qualitative Research. Thousand Oaks: Sage Publications Inc, 105-117.
Guba, E. G. and Y. S. Lincoln (2005). Paradigmatic Controversies, Contradictions, and
Emerging Confluences. In N. K. Denzin and Y. S. Lincoln (Eds.), The Sage
Handbook of Qualitative Research. Thousand Oaks: Sage Publications Inc, 191215.
Gulledge, T. and G. Simon (2005). The Evolution of SAP Implementation Environments: A Case Study from a Complex Public Sector Project. Industrial Management & Data Systems 105(6): 714-736.
Hallett, T. and M. J. Ventresca (2006). Inhabited Institutions: Social Interactions and
Organizational Forms in Gouldner's Patterns of Industrial Bureaucracy. Theory
and Society 35(2): 213-236.
127
Hanseth, O. and K. Braa (1998). Technology as Traitor: Emergent SAP Infrastructure in
a Global Organization. ICIS 1998 Proceedings. Paper 17.
Hanseth, O., U. Ciborra Claudio and K. Braa (2001). The control devolution: ERP and
the side effects of globalization. Database for Advances in Information Systems
32(4): 34-46.
Hanseth, O. and E. Monteiro (1997). Inscribing behaviour in information infrastructure
standards. Accounting, Management and Information Technologies 7(4): 183211.
Hasselbladh, H. and J. Kallinikos (2000). The Project of Rationalization: A Critique and
Reappraisal of Neo-Institutionalism in Organization Studies. Organization Studies 21(4): 697-720.
Hawking, P., A. Stein and S. Foster (2004). Revisiting ERP Systems: Benefit Realization. Proceedings of the 37th Annual Hawaii International Conference on System Sciences, 2004.
Hedman, J. and A. Borell (2004). Narratives in ERP systems evaluation. Journal of Enterprise Information Management 17(4): 283-290.
Hedström, P. and R. Swedberg (1996). Social Mechanisms: An Introductory Essay. In
P. Hedström and R. Swedberg (Eds.), Social Mechanisms - An Analytical Approach to Social Theory. Cambridge: Cambridge University Press, 1-31.
Henfridsson, O. (1999). IT-Adaptation as Sensemaking: Inventing New Meaning for
Technology in Organizations. Department of Informatics, Umeå University,
Umeå. PhD Thesis.
Henfridsson, O. (2000). Ambiguity in IT Adaptation: Making Sense of First Class in a
Social Work Setting. Information Systems Journal 10(2): 87-104.
Hildebrand, C. (2009). The Value of Sticking with Vanilla. Profit Online. Retrieved
25th March, 2009, from
http://www.oracle.com/profit/smb/122808_ziegele_qa.html.
Hinings, C. R., R. Greenwood, T. Reay and R. Suddaby (2004). Dynamics of Change in
Organizational Fields. In M. S. Poole and A. H. V. d. Ven (Eds.), Handbook of
Organizational Change and Innovation. New York: Oxford University Press
Inc., 304-323.
Hirsch, P. and M. Lounsbury (1997). Ending the Family Quarrel: Toward a Reconciliation of "Old" and "New" Institutionalisms. American Behavioral Scientist 40(4):
406.
Hirschheim, R. and H. K. Klein (2003). Crisis in the IS Field? A Critical Reflection on
the State of the Discipline. Journal of the Association for Information Systems
4(5): 237-294.
128
Hoffman, A. J. (1999). Institutional Evolution and Change: Environmentalism and the
U.S. Chemical Industry. Academy of Management Journal 42(4): 351-371.
Hong, K.-K. and Y.-G. Kim (2002). The Critical Success Factors for ERP Implementation: An Organizational Fit Perspective. Information & Management 40(1): 2540.
Husted, J. and P. Lübcke (2001). Politikens filosofihåndbog. København: Politikens
Forlag.
Häkkinen, L. and O.-P. Hilmola (2008). ERP Evaluation during the Shakedown Phase:
Lessons from an After-sales Division. Information Systems Journal 18(1): 73100.
Jack, L. and A. Kholeif (2008). Enterprise Resource Planning and a contest to limit the
role of management accountants: A strong structuration perspective. Accounting
Forum 32(1): 30-45.
Jackson, M. C. (2000). Systems Approaches to Management. New York: Kluwer Academic / Plenum Publishers.
Jacobs, F. R. and F. C. T. Weston, Jr. (2007). Enterprise Resource Planning (ERP) - A
brief History. Journal of Operations Management 25(2): 357-357.
Jacobson, S., J. Shepherd, M. D’Aquila and K. Carter (2007). The ERP Market Sizing
Report, 2006–2011 AMR Research, Inc, Document number AMR-R-20495.
James, D. and G. H. Seibert (1999). Oracle Financials Handbook: Planning and Implementing the Oracle Financial Applications Suite. Berkeley: Oracle Press Edition from Osborne.
Jensen, T. B. (2007). IS Adoption from a User Perspective. A Case Study of How Health
Care Professionals Experience and Make Sense of an Electronic Patient Record
Adoption. Department of Business Studies, Aarhus School of Business, Aarhus
University, Århus. PhD Thesis.
Jensen, T. B., A. Kjaergaard and P. Svejvig (2009). Using Institutional Theory with
Sensemaking Theory: A Case Study of Information System Implementation in
Healthcare. Journal of Information Technology 24(4): 343-353.
Jensen, T. B. and M. Aanestad (2007). How Healthcare Professionals "Make Sense" of
an Electronic Patient Record Adoption. Information Systems Management 24(1):
29-43.
Jepperson, R. (1991). Institutions, institutional effects, and institutionalism. In W. P.
Walter and P. J. Dimaggio (Eds.), The New Institutionalism in Organizational
Analysis. Chicago: University of Chicago Press 143-163.
129
Johannessen, J.-A. and J. Olaisen (2005). Systemic philosophy and the philosophy of
social science - Part I: Transcendence of the naturalistic and the anti-naturalistic
position in the philosophy of social science. Kybernetes 34(7): 1261-1277.
Kallinikos, J. (2004). Deconstructing information packages: Organizational and behavioural implications of ERP systems. Information Technology & People 17(1):
8-30.
Kant, I. (1781 (2007)). Critique of Pure Reason. The Project Gutenberg Etext. Retrieved
26th November, 2007, from http://www.gutenberg.org/etext/4280.
Kant, I. (1783 (1997)). Prolegomena to Any Future Metaphysics. Cambridge: Cambridge University Press.
Kant, I. (1783 (2007)). Prolegomena - til enhver fremtidig metafysik, der skal kunne
optræde som videnskab. Frederiksberg: Det lille Forlag.
Kaplan, A. (1964). The Conduct of Inquiry - Methodology of Behavioral Science. San
Francisco: Chandler Publishing Company.
Kholeif, A. O., M. G. Abdel-Kader and M. J. Sherer (2008). Enterprise Resource Planning: Implementation and Management Accounting Change in a Transitional
Country. Basingstoke: Palgrave Macmillan.
Kien, S. S. and C. Soh (2003). An Exploratory Analysis of the Sources and Nature of
Misfits in ERP Implementations Second-Wave Enterprise Resource Planning
Systems: Implementing For Effectiveness. Cambridge: Cambridge University
Press, 374-387.
Klaus, H., M. Rosemann and G. G. Gable (2000). What is ERP? Information Systems
Frontiers 2(2): 141-162.
Klein, H. K. and M. D. Myers (1999). A Set of Principles for Conducting and Evaluating Interpretive Field Studies in Information Systems. MIS Quarterly 23(1): 6793.
Knutsen, L. A. and K. Lyytinen (2008). Messaging Specifications, Properties and Gratifications as Institutions: How Messaging Institutions shaped Wireless Service
Diffusion in Norway and Japan. Information and Organization 18(2): 101-131.
Koch, C. (2001). Enterprise resource planning. Journal of Organizational Change
Management 14(1): 64-78.
Kraemmergaard, P. and B. R. Schlichter (2009). A Comprehensive Literature Review of
the ERP Research Field. CIM Working Paper No. 1, Vol. 1. Aalborg University,
Aalborg.
Kragh, H. (2007). Implementing Internet-Enabled ICT in Export Marketing Systems:
Effect on Activity Organization. Department of Management, Aarhus School of
Business, Aarhus University, Aarhus. PhD Thesis.
130
Kraatz, M. S. and E. S. Block (2008). Organizational Implications of Institutional Pluralism. In R. Greenwood, C. Oliver, K. Sahlin and R. Suddaby (Eds.), The SAGE
Handbook of Organizational Institutionalism. London: Sage Publications, 243275.
Kvale, S. (1995). The Social Construction of Validity. Qualitative Inquiry 1(1): 19-40.
Kvale, S. (1997). Interview - en introduktion til det kvalitative forskningsinterview.
København: Hans Reitzels Forlag.
Kvale, S. (2007). Doing Interviews. Los Angeles: Sage Publications.
Lamb, R. and R. Kling (2003). Reconceptualizing Users as Social Actors in Information
Systems Research. MIS Quarterly 27(2): 197-235.
Larmore, C. (2003). Back to Kant? No Way. Inquiry: An Interdisciplinary Journal of
Philosophy 46(2): 260-271.
Latour, B. (1987). Science in action: How to follow scientists and engineers through
society. Cambridge, Massachusetts: Harvard University Press.
Lee, A. S. (1991). Integrating Positivist and Interpretive Approaches to Organizational
Research. Organization Science 2(4): 342-365.
Lee, A. S. and R. L. Baskerville (2003). Generalizing generalizability in information
systems research. Information Systems Research 14(3): 221-243.
Leonardi, P. M. and S. R. Barley (2008). Materiality and change: Challenges to building
better theory about technology and organizing. Information and organization.
18(3): 159-176.
Liang, H., N. Saraf, H. Qing and X. Yajiong (2007). Assimilation of Enterprise Systems: The Effect of Institutional Pressures and the Mediating Role of Top Management. MIS Quarterly 31(1): 59-87.
Lindley, J. T., S. Topping and L. T. Lindley (2008). The hidden Financial Costs of ERP
Software. Managerial Finance 34(2): 78-90.
Lorenzo, O. (2004). A comprehensive review of enterprise system (ES) research. Academia. Revista Latinoamericana de Administración 33(2): 34-51.
Lounsbury, M. (2007). A Tale of Two Cities: Competing Logics and Practice Variation
in the Professionalizing of Mutual Funds. Academy of Management Journal
50(2): 289-307.
Lübcke, P. (2004). Politikens filosofi leksikon. København: Politikens Forlag.
March, J. G. and J. P. Olsen (1984). The New Institutionalism: Organizational Factors
in Political Life. The American Political Science Review 78(3): 734-749.
131
Markus, M. L. (2004). Technochange Management: Using IT to drive Organizational
Change. Journal of Information Technology 19(1): 4-20.
Markus, M. L., S. Axline, D. Petrie and S. C. Tanis (2000). Learning from Adopters'
Experiences with ERP: Problems Encountered and Success Achieved Journal of
Information Technology 15(4): 245-265.
Markus, M. L. and C. Tanis (2000). The Enterprise Systems Experience-From Adoption
to Success. In R. W. Zmud (Ed.), Framing the Domains of IT Research: Glimpsing the Future Through the Past. Cincinnati, Ohio: Pinnaflex Educational Resources, Inc., 173–207.
Marshall, C. and G. B. Rossman (1989). Designing Qualitative Research. Newbury
Park, California: Sage Publications, Inc.
Martínez-Delgado, A. (2002). Radical constructivism: Between realism and solipsism.
Science Education 86(6): 840-855.
Mathiassen, L. (2002). Collaborative practice research. Information Technology &
People 15(4): 321-345.
Mattila, M., J. Nandhakumar, P. Hallikainen and M. Rossi (2009). Emerging Role of
Enterprise System in Radical Organizational Change. Proceedings of JAIS
Theory Development Workshop, Sprouts: Working Papers on Information Systems, 9(46). http://sprouts.aisnet.org/9-46.
Mazza, C. and J. S. Pedersen (2004). From Press to E-media? The Transformation of an
Organizational Field. Organization Studies 25(6): 875-896.
McCracken, G. (1988). The Long Interview. Newbury Park, California: Sage Publications, Inc.
Melville, N., K. Kraemer and V. Gurbaxani (2004). Information Technology and Organizational Performance: An Integrative Model of IT Business Value. MIS Quarterly 28(2): 283-322.
Meyer, J. and W. Scott (1983). Organizational Environments: Ritual and Rationality.
Newbury Park, California: Sage Publications, Inc.
Meyer, J. W. and B. Rowan (1977). Institutionalized Organizations: Formal Structure as
Myth and Ceremony. American Journal of Sociology 83(2): 340-363.
Miles, M. B. and A. M. Huberman (1994). Qualitative Data Analysis: An Expanded
Sourcebook. Thousand Oaks: Sage Publications Inc.
Mumford, E. (2001). Advice for an Action Researcher. Information Technology &
People 14(1): 12-27.
Myers, M. D. (2009). Qualitative Research in Business & Management. London: Sage
Publications.
132
Myers, M. D. and D. Avison (2002). An Introduction to Qualitative Research in Information Systems. In M. D. Myers and D. Avison (Eds.), Qualitative Research in
Information Systems - A Reader. London: Sage Publications, 3-12.
Nash, K. S. (2010). ERP: How and Why You Need to Manage It Differently. CIO magazine. Retrieved 17th February, 2010, from
http://www.cio.com.au/article/334137.
Nicholson, B. and S. Sahay (2009). Deinstitutionalization in the Context of Software
Exports Policymaking in Costa Rica. Journal of Information Technology 24(4):
332-342.
Norén, L. (1998). Tolkande Företagsekonomisk Forskning: En Metodbok. Lund: Studentlitteratur.
North, D. C. (1990). Institutions, Institutional Change and Economic Performance.
Cambridge: Cambridge University Press.
O'Leary, D. E. (2000). Enterprise Resource Planning Systems: Systems, Life Cycle,
Electronic Commerce, and Risk. Cambridge: Cambridge University Press.
Oliver, C. (1992). The Antecedents of Deinstitutionalization. Organization Studies
13(4): 563-589.
Oracle (2008). Meet the Challenges of Globalization. Oracle E-Business Suite, Brochure, 2008, Document number C16140-01.
Orlikowski, W. J. (1992). The Duality of Technology: Rethinking the Concept of Technology in Organizations. Organization Science 3(3): 398-427.
Orlikowski, W. J. and S. R. Barley (2001). Technology and Institutions: What can Research on Information Technology and Research on Organizations Learn from
each other. MIS Quarterly 25(2): 145-165.
Orlikowski, W. J. and C. S. Iacono (2001). Research Commentary: Desperately Seeking
the 'IT' in IT Research - A Call to Theorizing the IT Artifact. Information Systems Research 12(2): 121-135.
Overby, S. (2010). IT Outsourcing: Don't Get Caught in Multi-Sourcing's Costly Trap.
CIO magazine. Retrieved 20th January, 2010, from
http://www.cio.com/article/512867.
Panorama Consulting Group (2010). 2010 ERP Report, Part One in a Series. Panorama
Consulting Group.
Papert, S. (1991). Situating Constructionism. In S. Papert and I. Harel (Eds.), Constructionism. Norwood, NJ: Ablex Publishing, 1-12.
Parr, A. and G. Shanks (2003). Critical Success Factors Revisited: A Model for ERP
Project Implementation. In P. B. Seddon, L. Willcocks and G. Shanks (Eds.),
133
Second-Wave Enterprise Resource Planning Systems. Cambridge: Cambridge
University Press, 196-219.
Parr, A. N. and G. Shanks (2000). A Taxonomy of ERP Implementation Approaches.
Proceedings of the 33rd Hawaii International Conference on System Sciences,
Hawaii.
Patton, M. Q. (2002). Qualitative Research & Evaluation Methods. Thousand Oaks:
Sage Publications Inc.
Pettigrew, A. M. (1985). Contextualist Research and the Study of Organizational
Change Processes. In E. Mumford, R. Hirschheim, G. Fitzgerald and A. T.
Wood-Harper (Eds.), Research Methods in Information Systems. Amsterdam:
Elsevier Science Publishers B.V., 53-78.
Pettigrew, A. M. (1987). Context and Action in the Transformation of the Firm. The
Journal of Management Studies 24(6): 649-671.
Pettigrew, A. M. (1990). Longitudinal Field Research on Change: Theory and Practice.
Organization Science 1(3): 267-292.
Pink, S. (2007). Doing Visual Ethnography. Images, Media and Representation in Research. London: Sage Publications.
Pollock, N. and R. Williams (2009). Software and Organizations: The Biography of the
Enterprise-wide System or How SAP Conquered the World. Milton Park, Abingdon, Oxon: Routledge.
Powell, W. W. and J. A. Colyvas (2008). Microfoundations of Institutional Theory. In
R. Greenwood, C. Oliver, K. Sahlin and R. Suddaby (Eds.), The SAGE handbook of organizational institutionalism. London: Sage Publications, 276-298.
Powell, W. W. and P. DiMaggio (1991). The New Institutionalism in Organizational
Analysis. Chicago: The University of Chicago Press.
Pozzebon, M. (2000). Combining a Structuration Approach with a Behavioral-Based
Model to Investigate ERP Usage. AMCIS 2000 Proceedings. Paper 124.
Ptak, C. A. and E. Schragenheim (2003). ERP: Tools, Techniques and Applications for
Integrating the Supply Chain. London: St .Lucie Press.
Quattrone, P. and T. Hopper (2006). What is IT?: SAP, Accounting, and Visibility in a
Multinational Organisation. Information & Organization 16(3): 212-251.
Ramiller, N. C., E. B. Swanson and W. Ping (2008). Research Directions in Information
Systems: Toward an Institutional Ecology. Journal of the Association for Information Systems 9(1): 1-22.
Ramskov, J. (2009). It-udvikling outsources i blinde. Ingeniøren. Retrieved 3rd December, 2009, from http://ing.dk/artikel/104499.
134
Reimers, K. (2003). Developing Sustainable B2B E-Commerce Scenarios in the Chinese Context: A Research Proposal. Electronic Markets 13(4): 261 - 270.
Robbins-Gioia. (2002). ERP Survey Results Point to Need For Higher Implementation
Success. Robbins-Gioia Press Release. Retrieved 20th August, 2007, from
http://www.robbinsgioia.com/news_events/012802_erp.aspx.
Ross, J., M. R. Vitale and L. P. Willcocks (2003). The Continuing ERP Revolution:
Sustainable Lessons, New Modes of Delivery. In P. B. Seddon, L. Willcocks
and G. Shanks (Eds.), Second-Wave Enterprise Resource Planning Systems: Implementing For Effectiveness. Cambridge: Cambridge University Press, 102132.
Ross, J. W. and M. R. Vitale (2000). The ERP Revolution: Surviving vs. Thriving. Information Systems Frontiers 2(2): 233-241.
SAP. (2009). Industries. Retrieved 6th December, 2009, from
http://www.sap.com/industries/index.epx.
SAP and Capgemini (2009). Time to Change New Thoughts on Supporting Business
Change Fast and Flexibly. SAP, Document number 50 095 228 (09/05).
Scarbrough, H., M. Robertson and J. Swan (2008). Developing the Processual Analysis
of Institutionalization: The Case of Resource Planning Systems Innovation.
Academy of Management Proceedings.
Schacht, R. (1984). Classical Modern Philosophers. Boston: Routledge & Kegan Paul.
Scheer, A.-W. and F. Habermann (2000). Making ERP a success. Communications of
the ACM 43(4): 57-61.
Schultze, U. (2000). A Confessional Account of an Ethnography about Knowledge
Work. MIS Quarterly 24(1): 3-41.
Schutz, A. (1967). Phenomenology of the Social World. Evanston, Illinois: Northwestern University Press.
Schutz, A. (2005). Hverdagslivets Sociologi. En Tekstsamling. København: Hans Reitzel Forlag.
Schwandt, T. A. (2000). Three Epistemological Stances for Qualitative Inquiry: Interpretivism, Hermeneutics, and Social Constructionism, Handbook of Qualitative
Research. Thousand Oaks: Sage Publications, Inc., 189-214.
Schwandt, T. A. (2007). The Sage Dictionary of Qualitative Inquiry. Thousand Oaks:
Sage Publications.
Scott, W. R. (1987). The Adolescence of Institutional Theory. Administrative Science
Quarterly 32(4): 493-512.
135
Scott, W. R. (1995). Institutions and Organizations: Theory and Research. Thousands
Oaks: Sage Publications.
Scott, W. R. (2001). Institutions and Organizations. Thousands Oaks: Sage Publications
Scott, W. R. (2004). Institutional Theory: Contributing to a Theoretical Research Program. In K. G. Smith and M. A. Hitt (Eds.), Great Minds in Management: The
Process of Theory Development. Oxford: Oxford University Press, 460-485.
Scott, W. R. (2008). Institutions and Organizations: Ideas and Interests. Thousands
Oaks: Sage Publications.
Scott, W. R. and J. W. Meyer (1991). The Organization of Societal Sectors: Propositions and Early Evidence. In W. W. Powell and P. DiMaggio (Eds.), The New
Institutionalism in Organizational Analysis. Chicago: University of Chicago
Press, 108-140.
Scott, W. R., M. Ruef, P. J. Mendel and C. A. Caronna (2000). Institutional Change and
Healthcare Organizations: From Professional Dominance to Managed Care.
Chicago: University of Chicago Press.
Seddon, P. B., L. Willcocks and G. Shanks (2003). Introduction: ERP - The Quiet
Revolution? In P. B. Seddon, L. Willcocks and G. Shanks (Eds.), Second-Wave
Enterprise Resource Planning Systems: Implementing for Effectiveness. Cambridge: Cambridge University Press, 1-19.
Selznick, P. (1949). TVA and the Grass Roots: A study in the Sociology of Formal Organization. Berkeley: University of California Press.
Selznick, P. (1957). Leadership in Administration: A Sociological Interpretation. New
York: Harper & Row.
Selznick, P. (1996). Institutionalism "Old" and "New". Administrative Science Quarterly 41(2): 270-277.
Shang, S. and P. B. Seddon (2003). A Comprehensive Framework for Assessing and
Managing the Benefits of Enterprise Systems: The Business Managers Perspective, Second-Wave Enterprise Resource Planning Systems: Implementing for Effectiveness. Cambridge: Cambridge University Press, 74-101.
Shanks, G., P. B. Seddon and L. Willcocks (2003). Second-Wave Enterprise Resource
Planning Systems: Implementing for Effectiveness. Cambridge: Cambridge University Press.
Sia, S. K. and C. Soh (2007). An Assessment of Package-Organisation Misalignment:
Institutional and Ontological Structures. European Journal of Information Systems 16(5): 568-583.
136
Sia, S. K., M. Tang, C. Soh and F. Boh Wai (2002). Enterprise resource planning (ERP)
systems as a technology of power: Empowerment or panoptic control? Database
for Advances in Information Systems 33(1): 23-37.
Silva, L. and J. Backhouse (1997). Becoming Part of the Furniture: The Institutionalization of Information Systems. In A. S. Lee, J. Liebenau and J. I. DeGross (Eds.),
Information systems and qualitative research. proceedings of the IFIP TC8 WG
8.2 International Conference on Information Systems and Qualitative Research,
31st May-3rd June 1997. Philadelphia, Pennsylvania, USA: Chapman & Hall,
389-414.
Silva, L. and J. Backhouse (2003). The Circuits-of-Power Framework for Studying
Power in Institutionalization of Information Systems. Journal of the Association
for Information Systems 4(6): 294-336.
Silver, M. S., M. L. Markus and C. M. Beath (1995). The Information Technology Interaction Model: A Foundation for the MBA Core Course. MIS Quarterly 19(3):
361-390.
Silverman, D. (1971). The Theory of Organisations: A Sociological Framework. London: Heinemann.
Silverman, D. (2001). Interpreting Qualitative Data: Methods for Analysing Talk, Text
and Interaction. London: Sage Publications, Ltd.
Silverman, D. (2005). Doing Qualitative Research: A Practical Handbook. London:
Sage Publications Ltd.
Soh, C., S. Kien Sia and J. Tay-Yap (2000). Cultural fits and misfits: Is ERP a universal
solution? Communications of the ACM 43(4): 47-51.
Soh, C. and S. K. Sia (2004). An Institutional Perspective on Sources of ERP PackageOrganisation Misalignments. The Journal of Strategic Information Systems
13(4): 375-397.
Somers, T. M. and K. G. Nelson (2004). A Taxonomy of Players and Activities across
the ERP Project Life Cycle. Information & Management 41(3): 257-278.
Stake, R. E. (2005). Qualitative Case Studies. In N. K. Denzin and Y. S. Lincoln (Eds.),
The Sage Handbook of Qualitative Research. Thousand Oaks: Sage Publications
Inc, 443-466.
Stylianou, A. C. and R. L. Kumar (2000). An integrative framework for IS quality management. Communications of the ACM 43(9): 99-104.
Sumner, M. (2003). Risk Factors in Enterprise-wide/ERP Projects. In P. B. Seddon, L.
Willcocks and G. Shanks (Eds.), Second-Wave Enterprise Resource Planning
Systems: Implementing for Effectiveness. Cambridge: Cambridge University
Press, 157-179.
137
Sumner, M. (2009). How Alignment Strategies influence ERP Project Success. Enterprise Information Systems 3(4): 425 - 448.
Svejvig, P. (2008). The Maiden Voyage of Enterprise Systems and Micro Foundational
Institutionalism. IADIS International Conference Information Systems 2008,
Carvoeiro, Portugal.
Svejvig, P. and A. Carugati (2009). Practices for Deinstitutionalization of an Enterprise
System: A Case Study. Proceedings of Organizations and Society in Information
Systems (OASIS) 2009 Workshop. U. Gal. Phoenix, Arizona, USA, Sprouts:
Working Papers on Information Systems, 9(56). http://sprouts.aisnet.org/9-56:
11-13.
Svejvig, P. and A. Carugati (forthcoming). On the Deinstitutionalization and Institutionalization Practices of Enterprise Systems Re-implementation. European Group
for Organizational Studies (EGOS) 2010, June 28th to 30th, Lisbon, Portugal.
Svejvig, P. and T. B. Jensen (2009). Enterprise System Adaptation: A Combination of
Institutional Structures and Sensemaking Processes. AMCIS 2009 Proceedings.
Paper 746.
Svejvig, P. and T. B. Jensen (forthcoming). What Can We Learn from a Well-Adapted
Enterprise System? Academy of Management Annual Meeting, August 6-10,
2010, Montréal, Canada.
Svejvig, P. and J. Pries-Heje (2009). Enterprise Information Systems Outsourcing: A
Look behind the Curtain. Professional Journal of the Scientific and Educational
Forum on Business Information Systems (4): 42-50.
Swanson, E. B. and N. C. Ramiller (1997). The Organizing Vision in Information Systems Innovation. Organization Science 8(5): 458-474.
Sæbø, J. I., S. Molla, A. Asalefew and S. Sahay (2008). Interplay of Institutional Logics
and Implications for Deinstitutionalization: Case Study of HMIS Implementation in Tajikistan. Sprouts: Working Papers on Information Systems, 8(11).
http://sprouts.aisnet.org/8-11 University of Oslo, Norway.
Tang, M., C. Soh, S.-K. Sia and W. Boh (2000). A Contingency Analysis of PostBureaucratic Controls in IT-Related Change. ICIS 2000 Proceedings. Paper 50.
Taylor, C. (1976). Hermeneutics and politics. In P. Connerton (Ed.), Critical Sociology,
Selected Readings. Harmondsworth: Penguin Books, 153-193.
Thornton, P. H. and W. Ocasio (2008). Institutional Logics. In R. Greenwood, C. Oliver, K. Sahlin and R. Suddaby (Eds.), The SAGE Handbook of Organizational Institutionalism. London: Sage Publications, 99-129.
Tolbert, P. S. and L. G. Zucker (1996). The Institutionalization of Institutional Theory.
In S. R. Clegg, C. Hardy and W. R. Nord (Eds.), Handbook of Organization
Studies. London: Sage Publications, 169-184.
138
Travers, M. (2001). Qualitative Research Through Case Studies. London: Sage Publications Ltd.
Tsakumis, G. T., D. R. Campbell Sr. and T. S. Doupnik (2009). IFRS: Beyond the Standards. Journal of Accountancy 207(2): 34-39.
Vitharana, P. and R. Dharwadkar (2007). Information Systems Outsourcing: Linking
Transaction Cost and Institutional Theories. Communications of the Association
for Information Systems 20(Article 23): 346-370.
Volkoff, O. (1999). Using the Structurational Model of Technology to Analyze an ERP
Implementation. AMCIS 1999 Proceedings. Paper 84.
Wailgum, T. (2009a). ERP Investments Still Top the List for Corporate IT Spending.
CIO magazine. Retrieved 9th December 2009, from
http://www.cio.com/article/507663.
Wailgum, T. (2009b). The Future of ERP. CIO magazine. Retrieved 20th November,
2009, from http://www.cio.com/article/print/508022.
Wailgum, T. (2009c). The Future of ERP, Part II. CIO magazine. Retrieved 10th December, 2009, from http://www.cio.com/article/508023.
Walsham, G. (1993). Interpreting Information Systems in Organizations. Chichester:
John Wiley & Sons.
Walsham, G. (2002). Interpretive Case Studies in IS Research: Nature and Method. In
M. D. Myers and D. Avison (Eds.), Qualitative Research in Information Systems
- A Reader. London: Sage Publications Ltd., 101-113.
Walsham, G. (2005). Learning about being critical. Information Systems Journal 15(2):
111-117.
Walsham, G. (2006). Doing Interpretive Research. European Journal of Information
Systems 15(3): 320-330.
Ward, C. J. (2006). ERP: Integrating and Extending the Enterprise. Public Manager
35(1): 30-33.
Weber, K. and M. A. Glynn (2006). Making Sense with Institutions: Context, Thought
and Action in Karl Weick's Theory. Organization Studies 27(11): 1639–1660.
Weerakkody, V., Y. K. Dwivedi and Z. Irani (2009). The Diffusion and Use of Institutional Theory: A Cross-disciplinary Longitudinal Literature Survey. Journal of
Information Technology 24(4): 354-368.
Wei, H.-L., E. T. G. Wang and P.-H. Ju (2005). Understanding Misalignment and Cascading Change of ERP Implementation: A Stage View of Process Analysis. European Journal of Information Systems 14(4): 324-334.
139
Weick, K. E. (1995). What Theory is Not, Theorizing Is. Administrative Science Quarterly 40(3): 385-390.
Wenneberg, S. B. (2002). Socialkonstruktivisme. Positioner, Problemer og Perspektiver. Frederiksberg: Samfundslitteratur.
Williamson, O. E. (1981). The Economics of Organization: The Transaction Cost Approach. American Journal of Sociology 87(3): 548-577.
Willis, T. H. and A. H. Willis-Brown (2002). Extending the Value of ERP. Industrial
Management & Data Systems 102(1): 35-38.
Wink, D. J. (2006). Connections Between Pedagogical and Epistemological Constructivism: Questions for Teaching and Research in Chemistry. Foundations of
Chemistry 8(2): 111-151.
Xu, L. and S. Brinkkemper (2007). Concepts of Product Software. European Journal of
Information Systems 16(5): 531-541.
Yin, R. K. (2003). Case Study Research: Design and Methods. Thousand Oaks, California: Sage Publications.
Young, R. A. and A. Collin (2004). Introduction: Constructivism and Social Constructionism in the Career Field. Journal of Vocational Behavior 64(3): 373-388.
Zucker, L. G. (1977). The Role of Institutionalization in Cultural Persistence. American
Sociological Review 42(5): 726-744.
Zucker, L. G. (1991). The Role of Institutionalization in Cultural Persistence. In W. W.
Powell and P. DiMaggio (Eds.), The New Institutionalism in Organizational
Analysis. Chicago: University of Chicago Press, 83-107.
140
PAPER 1
SVEJVIG, P. “USING INSTITUTIONAL THEORY IN ENTERPRISE SYSTEMS RESEARCH - DEVELOPING A CONCEPTUAL MODEL FROM A LITERATURE REVIEW”
Abstract.
This paper sets out to examine the use of institutional theory as a conceptually
rich lens to study social issues of enterprise systems (ES) research. More
precisely, the purpose is to categorize current ES research using institutional
theory to develop a conceptual model that advances ES research. Key institutional
features are presented such as isomorphism, rationalized myths, and bridging
macro and micro structures, and institutional logics and their implications for ES
research are discussed. Through a literature review of 180 articles, of which 18
papers are selected, we build a conceptual model that advocates multi-level and
multi-theory approaches and applies newer institutional aspects such as
institutional logics. The findings show that institutional theory in ES research is in
its infancy and adopts mainly traditional institutional aspects like isomorphism,
with the organization as the level of analysis, and in several cases it is
complemented by structuration theory and other theories.
141
1
INTRODUCTION
Much research on enterprise systems (ES) addresses implementation and use as well as
alignment between organization and ES, but it is often homogeneous and monolithic,
which largely simplifies the complex social settings of modern enterprises (Berente
2009; Boudreau and Robey 2005; Lamb and Kling 2003). The perception of ES has
been dominated by a techno-rational and managerial understanding focusing on economic efficiency leading to improved financial performance, where social considerations are downplayed or even overlooked (Dillard and Yuthas 2006), and these undersocialized understandings may be problematic for ES implementation and use. The implementation of ES is often complex due to enterprise-wide integration and data standardization, adoption of “best practice” business models with re-engineering of business
processes, compressed schedules, and finally the participation of a large number of
stakeholders (Soh et al. 2000: 47). The consequences of the under-socialized understandings are that implementation and integration problems are ignored or at best oversimplified, and instrumental solutions are considered superior and sufficient (Dillard
and Yuthas 2006), which can result in failure-prone ES implementations and/or reduced
value of ES implementations (Davenport 1998) due to users’ resistance (Grabski et al.
2003), lack of social commitment (Sumner 2003), misalignment between the ES and
organization (Sia and Soh 2007), and others.
However, institutional theory can be used to address these issues with its ability to “develop a more structural and systemic understanding for how technologies [enterprise
systems] are embedded in complex interdependent social, economic, and political networks, and how they are consequently shaped by such broader institutional influences”
(Orlikowski and Barley 2001: 154), and with its ability to deal with the logics that ES
imposes on organizations (Gosain 2004). Despite the advantages hinted at by Orlikowski and Barley (2001), IS researchers rarely adopt an institutional perspective (Berente
2009; Orlikowski and Barley 2001; Weerakkody et al. 2009), and when they do it is a
narrow use that does not exploit the potential of institutional theory (Currie 2009). The
state of theory presents a gap related to “institutional theory in IS research”, especially
articulated by Orlikowski and Barley (2001), which encouraged us to take a closer look
at research on ES using institutional theory, because it offers a conceptually rich lens for
142
studying the implementation and use of ES in complex social settings (adapted from
Currie 2009). The research questions are thus: (1) how has institutional theory been
used in ES research and (2) what requirements and elements must a conceptual model
address to advance the use of institutional theory in ES research. The contribution of
this paper lies in theorizing about ES using institutional theory, and this is a response to
Weerakkody et al. (2009: 362), who state in a very recent paper “… that very few conceptual/theoretical studies are published for advancing the use of [institutional theory]
in IS research”.
ES research is related to IS research (Kraemmergaard and Schlichter 2009; Pollock and
Williams 2009), and we define ES as large-scale organizational systems, built around
packaged enterprise systems software, enabling an organization to automate and
integrate a comprehensive part of its business processes, to share common data and
practices, and to produce and access information in real time. The most important class
of ES is enterprise resource planning systems (ERP systems) with other classes such as
customer relationship management systems (CRM systems) and supply-chain
management systems (SCM systems) (Seddon et al. 2003). ES target private
organizations, but also public organizations like hospitals (Sia and Soh 2007) and
municipalities (Caccia and Steccolini 2006). Packaged ES software is generic “semifinished products” from vendors like SAP and Oracle delivered to user organizations,
which tailor the products to their own needs (Brehm and Markus 2000; Seddon et al.
2003). Davenport (1998: 122) expresses the consequences of ES in the much cited line:
“An enterprise system imposes its own logic on a company’s strategy, culture and
organization”, and it is unlikely that there will be a perfect fit between the ES and the
organization, which may cause (severe) misalignment problems (Seddon et al. 2003).
The paper is organized as follows. In the following section, the research methodology is
presented with a focused and detailed literature review. We then explain institutional
theory, focusing on four central institutional concepts and their implications for ES research. Next we analyze 18 selected papers with respect to how institutional theory has
been used in ES research. We then develop a conceptual model to advance ES research
using institutional theory. The paper concludes with a discussion of implications for
theory and practice.
143
RESEARCH METHODOLOGY
In order to answer the research questions, we identified two essential “building blocks”
for the research process consisting of (1) a focused literature review of institutional
theory and (2) a detailed literature review of ES research using institutional theory, as
shown in figure 1 below (inspired by Jones and Karsten 2008):
Figure 1: Research process
The two literature reviews are used to describe the key features of institutional theory
relevant to ES research, to deduce current institutional themes of ES research, and finally to use key features and current themes to build a conceptual model.
The focused literature review of institutional theory is based upon two seminal books
about institutional theory: The New Institutionalism in Organizational Analysis (Powell
and DiMaggio 1991) and the three editions of Institutions and Organizations (Scott
1995; 2001; 2008b). These seminal books are complemented by further references (articles, books) through an analysis of bibliographies in these books by “going backward”
and “going forward” (Webster and Watson 2002). This is not a comprehensive literature
review about institutional theory, but instead a focused selection of institutional theory
in general, related to ES research.
144
The research process for the detailed literature review of ES research using institutional
theory is presented in the following subsections.
Defining the scope of the review
The literature review started with a broadly based improvised literature search (Gray
2004) in order to define the scope of the review. The outcome of this literature search
showed that searching specific journals was a cumbersome process, and that the scoping
of journals should even be very broad in order to embrace papers with ES research using institutional theory. Instead three scholarly databases with search engines were selected: “ProQuest LLC”, “Business Source Complete (EBSCO)”, and “Science Direct
by Elsevier”. This approach also ensured that the search was beyond the IS discipline,
which is highly relevant as ES research is an interdisciplinary field used in other disciplines like organizational theory, operations research, accounting, computer science etc.
(Kraemmergaard and Schlichter 2009; Pollock and Williams 2009).
Searching scholarly databases and selecting papers
A focused search (Gray 2004) was performed using the search word string “Institutional
Theory AND Enterprise System”. The term enterprise system (ES) is the super class
with ERP systems as the most important sub-class (Seddon et al. 2003), so we decided
also to include “ERP” as a search word, in order to embrace papers using this term instead of ES and capture more papers. The result of the search with the two keyword
strings is shown in table 1 below:
Institutional Theory in Enterprise Systems Research*
Search words
ProQuest
EBSCO
Science
Direct
Total
“Institutional Theory” AND “Enterprise System”
4
39
23
66
“Institutional Theory” AND “ERP”
8
55
52
115
Selected papers (doubles removed)
3
5
10
18
*) The result list from the keyword searches has not been checked for doubles caused by more than one
keyword match, but doubles are removed from the selected list of papers.
Table 1: Literature review of ES research using institutional theory
145
A full-text search was performed in all 3 databases in November 2008. Abstracts were
read for all 181 entries and full papers were in some instances skimmed to support the
selection process by searching for the keywords in the papers. Both institutional theory
and enterprise system (ERP) should be used as the main theoretical/empirical perspective possibly juxtaposed with other theoretical perspectives in order for the paper to be
selected. Editorial notes, personal profiles, bibliographies, book reviews etc. were excluded. As specified in table 1, 18 papers were selected.
Classification of the selected papers
An author-centric matrix (Webster and Watson 2002) was used to classify the selected
papers, and the complete list of selected papers is presented in the appendix.
KEY FEATURES OF INSTITUTIONAL THEORY
In discussing institutional theory in ES research it should be emphasized at the outset
that it is a general theory spanning economics, political science, and sociology (Scott
2008b) rather than a theory specific to enterprise systems or information systems. Our
focus in this paper will be on organizational institutionalism (Greenwood et al. 2008b)
used in organization theory and sociology.
Institutional theory attempts to describe the deeper and more resilient aspects of how
institutions are created, maintained, changed, and dissolved (Scott 2004; 2008b), and
deals with the pervasive influence of institutions on human behavior including the
processes by which structures, e.g. rules, routines, and norms, guide social behavior.
Institutions are multifaceted, durable, resilient social structures, made up of symbolic
elements, social activities, and material resources (Currie 2009; Scott 2001: 48-50). Examples of institutions are human rights, societies, families, handshakes, and belief systems like Buddhism. North (1990: 4-5) presents an important, although simplified, distinction between organizations and institutions using a game analogy: institutions are
the rules of the game and organizations are the players. We talk about institutionalization when actions are repeated and given shared meanings by actors (Berger and Luckmann 1966; Scott 2008b), whereby the institution becomes stable and durable (Currie
2009).
146
We will continue with an examination of four key features of institutional theory, which
seems to be important in order to understand and interpret ES research using institutional theory. The four key features are isomorphism, rationalized myths, bridging macro
and micro structures, and institutional logics.
INSTITUTIONAL AND COMPETITIVE PRESSURES LEADING TO ISOMORPHISM
A new approach to institutional analysis was introduced in the 1970s with a focus on
culture and cognition, where taken-for-granted rules led to isomorphism in the formal
structures of the organization, and organizations had to conform to society for legitimacy (Meyer and Rowan 1977; Zucker 1977). DiMaggio and Powell (1983) “moved” the
focus on isomorphism from the society level to the organizational field level with coercive, normative, and cognitive institutional pressures leading to isomorphism, which is
nowadays part of many institutional analyses. Isomorphism means “a constraining
process that forces one unit in a population to resemble other units that face the same set
of environmental conditions” (DiMaggio and Powell 1983: 149) or simply expressed as
structural similarity. Liang et al. (2007) argue that cognitive, coercive, and normative
institutional pressures impact on the assimilation of enterprise systems, for instance the
normative pressure in an organizational field, where suppliers, customers, consultants,
and professional associations collectively assess and endorse IS innovations (Swanson
and Ramiller 1997), shaping the implementation and assimilation of enterprise systems
by providing institutional norms that guide top managers (Liang et al. 2007).
Isomorphism is an important consequence of both competitive and institutional pressures (Scott 2008b), and one of the challenges using institutional theory is to distinguish
between the two kinds of pressures. Competitive pressures assume a system rationality,
often used in ES research (Dillard and Yuthas 2006), that emphasizes market competition where organizations compete for resources and customers, and are closely related
to the technical environment where products and services are expected to be produced
in an effective and efficient way (Scott and Meyer 1991), but “organizations compete
not just for resources and customers, but for political power and institutional legitimacy,
for social as well as economic fitness” (DiMaggio and Powell 1983: 150). Competitive
and institutional pressures “live side by side” and we shall avoid dichotomous explana-
147
tions, where e.g. social explanations exclude techno-rational explanations (adapted from
Greenwood et al. 2008a: 32), and instead acknowledge that social situations, such as ES
in organizations, consist of interdependent non-rational and rational elements (adapted
from Scott 2008b). It is therefore difficult to distinguish between these explanations
empirically, reinforced by the fact that institutional explanations strive to appear technical in nature (Scott and Meyer 1991) as a disguise. Greenwood et al. (2008a) state that
institutional theory is well suited to being juxtaposed with other theories, for instance
competitive pressures “explained” by transaction cost theory and institutional pressures
explained by institutional theory as presented by Vitharana and Dharwadkar in their
paper about IS outsourcing (2007). This facilitates organizational analyses covering
both rational and non-rational elements. The next section about rationalized myths elaborates on the entangledness.
RATIONALIZED MYTHS
A key theme related to institutional isomorphism is that organizations conform to rationalized myths in order to be “proper” organizations (Boxenbaum and Jonsson 2008).
Institutionalized products, services, techniques, regulatory systems, public opinions,
professional standards, etc. “act” as powerful myths exerting institutional pressures on
organizations in multiple and complex ways. Rationalized myths may develop in organizations, where it is believed that the responses to these multiple pressures are aimed at
organizational efficiency, but they are in reality aimed more at achieving legitimacy for
the organization (Meyer and Rowan 1977). Alvarez (2002) examined the role of myths
in an ERP implementation. The old legacy system was deinstitutionalized by creating a
story of “performance crisis”, and a myth-making process took place “constructing the
new ERP system as an integrated system”, which was aligned with the overall organizational goals of the organization, but the benefit of the integration was not supported by
objectively testable facts. The rationalized myth thus legitimized the ERP implementation, “and the story-making process served to align the technology with ideal organizational values” (Alvarez 2002: 82). The case study by Alvarez also shows the deinstitutionalization process of the old legacy system followed by the reinstitutionalization
process of the new integrated ERP system (Greenwood et al. 2002; Scott 2008b; Tolbert
and Zucker 1999), and that narratives can support the institutionalization process (see
148
also Hedman and Borell 2004), which can be a relevant “technique” in practical ERP
implementations.
MULTIPLE LEVELS IN INSTITUTIONAL THEORY BRIDGING MACRO
AND MICRO STRUCTURES
Institutional and competitive pressures are often exerted by the society and the organizational field at the organization, where the organizational field is defined as “those organizations that, in the aggregate, constitute a recognized area of institutional life: key suppliers, resource and product consumers, regulatory agencies, and other organizations
that produce similar services or products” (DiMaggio and Powell 1983: 148). Scott
(2008b) argues that it is beneficial to look at multiple levels in a given study in order to
enrich the understanding in institutional analysis, and this is exactly one of the powerful
features of institutional theory with its ability to operate at varying levels ranging from
society, organizational field, and organization to individual actor level (Scott 2008b: 8590). What is likewise important is the reciprocal interaction between levels, where macro structures in society are bridged by organizational fields to micro structures in organizations or even “down” to the individual actor level. Institutional creation and diffusion thus happen, where top-down processes allow higher level structures to shape the
structure and action of lower levels, while bottom-up processes shape, reproduce, and
change the context within which they operate (Scott 2008b: 190-195). Scott’s argument
is mirrored by that of Currie (2009), who encourages IS researchers to work with multiple levels and multiple stakeholders as this is the mainstay of institutional theory. A
study of ERP implementation in three hospitals in Singapore by Soh and Sia (2004)
emphasizes the interplay between different levels. The selected ERP package was developed for the European and US markets, where the institutional context at society
level for health care is marked by being either highly subsidized (European market) or
paid for by health-care insurance (US market), and the ERP package was inscribed (Latour 1987) with this logic, which is contrary to the tradition in Singapore, where a complicated co-payment calculation depending on bed-class etc. is widely implemented
with invoices sent to both the patient and the state for a stay in hospital. This is an example of a clash between the Western and Singaporean institutional contexts (macro
level), shaping the implementation and use of the ERP package in the three hospitals
149
(micro level). The next section will further advance how macro and micro levels can be
combined by using institutional logics, which can be understood as a “social mechanism” (Hedström and Swedberg 1996) mediating the top-down and bottom-up processes.
INSTITUTIONAL LOGICS
There has been much focus on isomorphism within institutional theory (Greenwood et
al. 2008a), which is reflected in ES research (cf. appendix), but this focus has changed
nowadays and it is no longer so much on isomorphism, whether in society or within the
organizational field, but more on the effects/processes of different, often conflicting,
institutional logics on individuals and organizations. “Institutional logics shape rational,
mindful behavior, and individual and organizational actors have some hand in shaping
and changing institutional logics” (Thornton and Ocasio 2008: 100). Institutional logics
link institution and action (see also Barley and Tolbert 1997) and provide a bridge between macro-structural perspectives (DiMaggio and Powell 1983; Meyer and Rowan
1977) and micro-process approaches (Zucker 1991). Multiple institutional logics are
“available” for organizations and individuals (Scott 2008b), and the embedded agency
in institutional logics presupposes partial autonomy for individuals and organizations
(Thornton and Ocasio 2008), so actions, decisions, and outcomes are a result of interaction between an individual agency and an institutional structure (Friedland and Alford
1991; Thornton and Ocasio 2008: 103-104). Some IS researchers have addressed institutional logics related to information systems or enterprise systems (Berente et al. 2007;
Currie and Guah 2007; Gosain 2004). Gosain (2004) argues that a mismatch between
the institutional logics in an enterprise system and the incumbent institutional logics in
an organization can lead to institutional misalignment. Varying degrees of mismatch
between institutional logics in enterprise systems and organizations can lead to varying
degrees of institutional misalignment, which again can have problematic consequences
like resistance to the new enterprise system (Gosain 2004). Other researchers discuss
misalignment between enterprise systems and organizations, which is similar to Gosain’s account, although they do not use the “institutional logic” concept directly (Sia
and Soh 2007; Soh and Sia 2004).
The concept of institutional misalignments presented by Gosain can be used to emphasize several aspects of institutional logics. First, Fligstein (2001: 100) criticizes institu150
tional theory for considering organizational actors to be passive recipients or “cultural
dopes”, using readily available scripts provided by the government, professionals, or
other institutional carriers to structure their actions. However, applying institutional
logics counters this critique, where an individual agency plays an important role in selecting and changing institutional logics in the working practices, since “institutional
logic is the way a particular social world works” (Thornton and Ocasio 2008: 101). Users of an enterprise system might adopt the embedded institutional logics in ES and then
change the incumbent organizational institutional logics to fit “the ES logics” so institutional misalignment is reduced, whatever consequences this may have, which nevertheless implies an agency from the organizational actors, who are guided by interest, power, and opportunism. Second, the changes in institutional logics are part of (or are) the
institutional/organizational changes (see also perspectives on institutional change in
Hargrave and Van De Ven 2006) taking place in an organization, for instance by implementing an enterprise system that could be designated a “precipitating technological
jolt” starting a change (Greenwood et al. 2002: 60). We can thus analyze the process
and stages of change using “institutional logics as a method of analysis” (Thornton and
Ocasio 2008: 109-111). Finally, the institutional logics perspective provides an approach to bridging macro and micro perspectives, where e.g. the institutional logics
“built into” enterprise systems from the original US/Western European development
context (macro perspective) are then used for e-procurement by a purchaser in a Singaporean defense organization (micro perspective) (Sia and Soh 2007).
The move away from focusing on institutional pressures, leading to isomorphism, to the
effects of institutional logics seems to be promising not least in enterprise systems research, because it is a way to “open” the enterprise system artifact (see also Orlikowski
and Iacono 2001) and gives it a prominent role.
151
SUMMARY OF THE KEY FEATURES OF INSTITUTIONAL THEORY
AND IMPLICATIONS FOR ES RESEARCH
As shown in the previous sections, institutional theory has a number of attractions to
offer to enterprise systems research, especially as a way to look beyond the technorational perspective so often embraced in enterprise systems research, and it emphasizes
social considerations to complement technical considerations. Below is a table with a
summary of the key features of institutional theory presented in the previous sections
augmented with their possible implications for ES research:
Institutional
and competitive pressures leading
to isomorphism
Key features of institutional theory
Implications for enterprise systems
research
Organizations are facing both competitive
and institutional pressures leading to isomorphism (structural similarity).
Researchers should look beyond rational explanations to institutional explanations with regard to understanding the
management, implementation, and use
of ES.
Institutional pressures could be coercive,
normative, and cognitive.
Social situations consist of interdependent
non-rational and rational elements.
Institutional pressures are shaping the
implementation, use, and management
of ES, for instance decisions to adopt a
specific ES.
Rationalized Rationalized myths related to technology
are technical procedures, accounting, permyths
sonnel selection, or data processing. Such
institutionalized techniques establish an
organization as appropriate, rational, and
modern, quite apart from their possible
efficiency (Meyer and Rowan 1977).
We are surrounded by rationalized
myths in enterprise systems research,
whether it is the ES itself that is a rationalized myth or “best practices” like
BPR, TQM, BPM etc. embedded in ES
or the implementation and use process
of ES.
Rationalized myths can be used as techniques in ES implementations.
Multiple
levels of
analysis
bridging
macro and
micro structures
Institutional theory can be applied at varying levels of analysis ranging from society,
organizational field, and organization to
individual actor level.
ES research can be performed at different levels, for instance at the organizational field level examining the diffusion of specific enterprise systems, or at
the organizational level understanding
Top-down processes allow higher-level
institutional misalignment between
structures to shape the structure and action enterprise systems and organization.
of lower levels, while the bottom-up
processes shape, reproduce, and change the ES research can also take advantage of
context within which they operate.
combining micro and macro perspectives where the institutional macro
context shapes the management, implementation, and use of ES in an organization’s micro practices executed
by actors.
Institutional Institutional logics are a set of material
152
Enterprise systems embed institutional
logics
Key features of institutional theory
Implications for enterprise systems
research
practices and symbolic constructions linking institution and action, and they provide
a bridge between macro-structural perspectives and micro processes.
logics, which are inscribed during development and implementation. The
institutional logics in the ES constrain
the use process (Gosain 2004).
Institutional logic is the way a particular
social world works.
Institutional logics “open” the ES and
thereby counteract the tendency to black
box the IT artifact in ES/IS research.
Institutional logics are a promising
theoretical lens for understanding the
interaction between enterprise systems
and organization both statically (structures) and dynamically (processes).
Table 2: Key features of institutional theory and their implications for ES research
The table above highlights the key features of institutional theory, which offers a distinctive perspective on organizations, enterprise systems, and their interplay, which is
highly relevant to enterprise systems research and enables us to extend beyond the techno-rational and managerial perspective. However, this chapter also illuminates the complexity, ambiguity, and diversity associated with institutional theory, so it is both an
opportunity and a challenge to use institutional theory for instance to provide conceptual clarity (Currie 2009). The following chapter will use the features in table 2 to categorize ES research using institutional theory.
ANALYZING THE USE OF INSTITUTIONAL THEORY IN
ENTERPRISE SYSTEMS RESEARCH
This chapter presents, based on 18 selected papers, the result of the detailed literature
review of ES research using institutional theory. An author-centric matrix (Webster and
Watson 2002) is shown in the appendix, while this chapter condenses and elaborates on
important aspects from the matrix.
The use of institutional theory in ES research is recent as all the selected papers are distributed in the period from 2003 to 2008. Institutional theory applied within organizational theory dates back to the late 1940s (e.g. Selznik 1949), while new institutional
theory was established in the 1970s (e.g. Meyer and Rowan 1977) and has now reached
adulthood as a mature social theory (Scott 2008a). One of the early uses of institutional
153
theory in IS research was Barley’s seminal paper about CT scanners (1986), while the
use in ES research is recent and still in its infancy. However, a consequence of this infancy is that there are many unexplored research avenues available, where we can widen
ES research to embrace complex social situations. An appropriate starting point is to
understand “how institutional theory has been used in ES research”, which is shown in
the table below:
Key features of
institutional
theory1
Institutional and
competitive pressures
Number of Examples of institutional theory used in ES research
papers2
13
The paper addresses how organizations respond to “radio frequency
Identification” (RFID) mandates under uncertainty and with what consequences. Especially varying degrees of uncertainty are discussed as
well as the corresponding rate of isomorphic change (Barratt and Choi
2007).
Institutional pressures play a critical role in the implementation of the
Sarbanes Oxley Act, particularly coercive pressures but also normative
pressures to act as a socially acceptable public company (Braganza and
Desouza 2006).
Institutional pressures are reflected in enterprise systems configurations
that exert control over organizational actors. This is a powerful control
that is often difficult to resist (Gosain 2004: 160-165).
Coercive, normative, and cognitive institutional pressures impact on the
assimilation of enterprise systems by guiding top managers in their
decisions (Liang et al. 2007).
Organizational learning interplays with the institutional environment
(consisting of institutional pressures), where e.g. government authorities
request a successful public organization that has implemented an enterprise system to share its experience with other public organizations
(Phang et al. 2008: 113).
Rationalized myths
4
Innovations in financial and accounting techniques [implemented in ES]
can shape the vision of reality that organizational actors have, spreading
concepts like value for money, accountability, efficiency, and effectiveness, turning them into new shared meanings and values (Caccia and
Steccolini 2006: 155 our emphasis).
“Narratives are used to persuade, convince and make people act and
believe in certain ways” (Hedman and Borell 2004: 286). Narratives or
stories are powerful rationalized myths that can be used in ERP systems
evaluation (i.e. learning and understanding) (Hedman and Borell 2004).
The evaluation and selection of enterprise systems are said to be rational
and deterministic, but ceremonial aspects seem to play an important role
in order to legitimize the organization (Tingling and Parent 2004).
Multiple levels of
analysis
6
The US fixed by law the Sarbanes Oxley Act as a federal law to address
the effectiveness of internal controls of public companies as a response
to company scandals, which impose society requirements on organizations (Braganza and Desouza 2006).
154
Key features of
institutional
theory1
Number of Examples of institutional theory used in ES research
papers2
The ERP package was developed for the European and US markets,
where the institutional context at the society level for health care is
marked by being either highly subsidized (European market) or paid for
by health-care insurance (US market), and the ERP package was inscribed with this logic, which is contrary to the tradition in Singapore,
where a complicated co-payment calculation depending on bed-class
etc. is widely implemented and impacts on the target organization (Soh
and Sia 2004).
B2B e-commerce adoption in Taiwan’s electronics industry is highly
impacted on by government policies and subsidies (Thatcher et al. 2006:
96).
Institutional logics
4
Radical changes in accounting systems are justified by external events,
often by legal requirements (coercive pressure), but they have hardly an
impact on routines (decoupling) especially if the change is not consistent with traditional shared values. This is institutional misalignment
(Caccia and Steccolini 2006).
Package–organization misalignment is due to an institutional context,
which can be divided into an imposed context being country-specific,
industry-specific, or a voluntarily acquired context due to organizational
differences (Sia and Soh 2007; Soh and Sia 2004).
Other features of
institutional theory
7
Institutional entrepreneurship
Enterprise systems, such as IT innovation, are likely to be launched
successfully if institutional entrepreneurs perform legitimation activities
like “spreading definitive success stories from users and vendors”
(Wang and Swanson 2007).
Process of institutionalization
The process of institutionalization for resource planning systems is
investigated through a historical analysis covering the MRP period, the
MRPII period, and the ERP period. The key mechanism to embed these
systems in organizations’ business practices (institutionalize) are role
specialization, the production of discourse, and boundary-spanning
activities (Scarbrough et al. 2008).
Technology as institutions (ES/ERP as institutions)
Technology can assume the properties of an institution (structure) where
regulative, normative, and cultural-cognitive pillars are applicable to the
institutional entity (Cadili and Whitley 2005).
1) The categories are based on table 2 in the previous chapter with an additional category “Other” to capture other uses of institutional theory.
2) See the appendix for further information.
Table 3: Examples of how institutional theory is used in ES research
ES research makes use of different key features of institutional theory although a majority of papers apply the foundational themes of new institutional theory (Powell and DiMaggio 1991) like institutional pressures (DiMaggio and Powell 1983), rationalized
myths, and legitimacy (Meyer and Rowan 1977). Newer theme such as institutional
155
logics (Friedland and Alford 1991; Thornton and Ocasio 2008) has also found its way to
ES research, while topics like institutional entrepreneurs (DiMaggio 1988; Fligstein
2001) and institutional processes (Greenwood et al. 2002; Hargrave and Van De Ven
2006) are only touched by few papers. Organizational institutionalism in general has
advanced considerably from the foundational themes of new institutional theory (see for
instance Greenwood et al. 2008b), and thereby it provides several opportunities for ES
research to advance in using institutional theory, especially by focusing on process approaches instead of variance approaches (Currie 2009).
Applying institutional theory as a theoretical lens is often only one side of the coin,
while the other side of the coin might imply juxtaposing it with other theoretical lenses.
The table below shows which of the selected 18 papers take a single-theory or a multitheory approach, respectively:
Paper number
Theoretical perspective
1
2
3
Organizational change
X
Stakeholder theory
X
Contingency theory
X
Complementary approach
X
Structuration theory
Organizational sensemaking
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18
X
X
X
X
Organizational learning
X
Evaluation and measurement
X
X
Technology assimilation
X
Innovation diffusion
X
Culture (organization and
nationwide)
X
Organizational politics
X
Ontology model of IS
X
X
Decision theory
X
Resource-based view
X
1) Grey bars indicate papers where institutional theory is used as the only explicitly applied theory.
2) See the appendix for further information.
Table 4: Multiple theoretical perspectives used together with institutional theory
156
Table 4 enforces Greenwood et al.’s (2008a: 28) account that “institutional theory has
gained enormously for many years from its combination with, or incorporation of, other
theories” as 12 out of 18 papers are used in a multi-theory approach. Structuration
theory, organizational learning, and culture are the preferred choices to combine with
institutional theory, but the multiplicity of theories in the table emphasizes the many
alternatives for ES researchers to gain from when combining institutional theory with
other theories. Currie (2009) argues that some of the shortcomings of institutional
theory might be overcome by using a multi-theory approach to stimulate empirical investigations to create rich insights. One of the stated shortcomings in institutional theory
is the “lack of agency”, i.e. we are cultural dopes (Fligstein 2001). This is taken up by
Cadili and Whitley (2005) as they use a “structuration theory” to theorize about agency
(and structure in an interdependent duality) and “institutional theory” to theorize about
the wider environmental structural properties (society, organizational field etc.), thereby
combining the two theoretical perspectives.
One of the strong features of institutional theory, as presented in the previous chapter, is
its ability to operate at multiple levels and bridge macro and micro structures. The majority of the papers in the literature review (16 out of 18 cf. appendix) use the organization as the level of analysis while the remaining operate at the organizational field level,
and only 4 papers apply multiple levels bridging micro and macro structures. There
seems to be a potential for ES research to shift the level of analysis to the organizational
field, society, or down to the individual level and furthermore utilize the bridging of
micro and macro structures, for instance by institutional logics. Currie draws the same
result from her examination of institutional theory in IS research in general, and she
points out that “IT-related constructs, such as adoption intention, assimilation and implementation, without considering wider environmental and inter-organizational levels”
are problematic, because important tenets of institutional theory are based on multi-level
and multi-stakeholder analyses (2009: 66).
The analysis in short shows that institutional theory in ES research is in its infancy, and
adopts mainly “traditional” institutional aspects like isomorphism, with the organization
as the level of analysis, and is in several cases complemented by structuration theory
and other theories.
157
The analysis in this chapter together with the key features of institutional theory from
previous chapters will be used to develop a conceptual model.
DEVELOPING A CONCEPTUAL MODEL
This chapter synthesizes the use of institutional theory in ES research from previous
chapters into a coherent conceptual model. The purpose of the model is to provide an
analytical model to advance both theoretical and empirical ES research using institutional theory.
Existing models do not seem to be able to fulfill the purpose specified above or at least
only partially. Scott’s (2008b) seminal analytical framework for organizational analysis
is very comprehensive, but also very general, and does lack the focus on ES research we
are aiming for in this paper. Currie (2008) presents a constructive theoretical research
framework to study IT compliance, which has some of the features fulfilling the purpose above, but one of the shortcomings is the missing specificity of the ES artifact,
which is important, as argued and discussed later in this chapter. Several other models
have been consulted, but they mainly address focused research issues and are not even
intended as more generic models for ES research (e.g. Cadili and Whitley 2005; Gosain
2004; Liang et al. 2007). The existing models have nevertheless served as inspiration
for the conceptual model presented in this chapter.
Four requirements for the conceptual model
The requirements for the model are mainly derived from previous chapters and make up
the following:
1. It must include core features of institutional theory like institutional and competitive isomorphism and rationalized myths, but also newer features such as
institutional logics and institutional processes.
2. It should support multi-level analysis bridging macro and micro structures, and
thereby implicitly address a multi-stakeholder approach, all strong features of
institutional theory.
158
3. It must contain multi-theory elements to gain from juxtaposing institutional
theories with other theories.
The three requirements above are noticeably derived from previous chapters, but there
is “a missing link” best expressed by Orlikowski and Iacono (2001: 121), who strongly
and provocatively argue that the IT artifact in IS research is desperately needed. There
is a tendency in the IS research using social theories to give theoretical and empirical
significance to the context and leave technology unspecified (Orlikowski and Iacono
2001). IS researchers have “difficulties in grasping the inner structure of the technology
artifact” (Czarniawska 2009: 50), and it is important to be specific about the technology
(Monteiro and Hanseth 1996), as there is a great difference between tailor-made software neatly aligned to specific requirements in a specific organization and ES designed
by vendor organizations as “semi-finished products” delivered to customer organizations for configuration and implementation (see also Carton et al. 2008; Davenport
2000). ES itself is furthermore a comprehensive and broad category (Pollock and Williams 2009) with very diverse systems like payables applications, which are highly
transaction-oriented with structural similarity to “a Fordist assembly line” (Czarniawska
2009: 57), and advanced supply-chain planning and optimization applications used for
tactical and strategic decision making (e.g. SAP 2008). All this is summed up in the
fourth and final requirement for the conceptual model:
4. It must incorporate the ES artifact in order to be specific about the technology.
159
The conceptual model is presented in the figure below in the quest to fulfill the four
requirements:
Figure 2: Conceptual model
The model is divided into three frames: “institutional theory”, “phenomenon of study”,
and “juxtaposed theories”. First, the “institutional theory frame” catches many of the
institutional elements considered in this paper, but the shown elements should not be
taken as “fixed and complete” but on the contrary as a menu to choose from, with the
possibility to add other dishes (i.e. an open model). The different elements can be used
at different levels, and there is no direct correspondence between the location of elements and the levels of analysis presented in the “phenomenon of study” frame. Second,
the “juxtaposed theories” frame is a placeholder to visualize the possibility to juxtapose
other theories with institutional theory, where these theories could “attach” at different
levels of analysis in the “phenomenon of study” frame. Finally, the “phenomenon of
study” frame presents the multi-level approach often used in institutional theory (Scott
2008b). The three elements “organization”, “enterprise system”, and “actor” have similarities to Orlikowski’s structural model of technology (1992) and seek to give the ES
artifact a salient role in the organization. The arrows are conceptual links between the
160
elements in the frame, and they are not meant to be causal, but instead descriptive and
exploratory relationships to aid theoretical discussions and empirical analyses (adapted
from Fligstein 2001: 115).
Here are some closing remarks about using the conceptual model: first, the model can
be used as a complete model with all the levels in the “phenomenon of study” frame
etc., but it is also possible to work on a single element in the model, e.g. the enterprise
system, which can produce insights into other elements, i.e. an eclectic approach.
Second, the model is applicable to both process and variance studies (see Newman and
Robey 1992: 251-252). An institution can be perceived as a state/result of an existing
order (e.g. a society, a legal system, or an organization) or as a process of institutionalization, deinstitutionalization, preinstitutionalization, and reinstitutionalization (Greenwood et al. 2002) through either incremental or radical changes (Scott 2008b: 50). The
conceptual model thus supports both process and variance studies. Finally, research
using the model can apply different paradigmatic stances (e.g. Burrell and Morgan
1979) possibly spanning from social realism to social constructivism, embracing quantitative and qualitative research (Scott 2008b: 62-66) (cf. also appendix).
DISCUSSION
The implementation and use of ES in modern enterprises are very complex, and both
researchers and practitioners need to understand the multifaceted connections between
the technical and social domains embedded in wider institutional contexts. Current research on ES is dominated by monolithic and techno-rational understandings where
social considerations are downplayed or even overlooked (Berente 2009; Dillard and
Yuthas 2006), and the wider institutional context plays a limited role (adapted from
Currie 2009). We have developed a conceptual model to theorize about ES using institutional theory to address these issues, as urged by Weerakkody et al. (2009), and this
has several implications for theory and practice.
IMPLICATIONS FOR THEORY
Reviewing the landscape of research on ES using institutional theory shows its infancy
and supports the claim about a lack of institutional theory in IS research (Orlikowski
161
and Barley 2001; Weerakkody et al. 2009). This opens quite a few unexplored research
avenues that this paper has briefly addressed. King et al. (1994: 141) argue that “Institutional factors are ubiquitous and essential components to understand and explore IT
innovations that cross organizational and firm boundaries”, and this is very applicable
for the implementation and use of ES (see also Currie 2009) and underlines the motivation for using institutional theory in ES research. We give an overview of what institutional theory is and how it can be used in ES research. Despite the small numbers of
papers and the claimed infancy, the literature review provides a foundation for researchers to use institutional theory in ES research, and to some extent identifies possible gaps
for future research, although the latter part is fairly coarse and limited.
In terms of theory development, the conceptual model presented in this paper suggests a
rich contextualized lens to study social considerations in ES research, which can be
adapted to specific research issues, to support theoretical and empirical research. One
underutilized example is to apply the actor or organizational field as the level of analysis instead of the organization, which seems to be the “default” level of analysis in ES
research (see also Weerakkody et al. 2009).
Previous conceptual models using institutional theory have either been very general
(e.g. Scott 2008b) and not sufficiently related to ES research or address very specific
research purposes (e.g. Liang et al. 2007). With these problems in mind, we claim to
have developed a model, inspired by previous models, with a reasonable balance between generality and specificity applicable to diverse ES research issues covering multilevels, multi-stakeholders, multi-theory etc.
Another important aspect of the model is the inclusion of the ES software in the model,
which appears very obvious, but is fairly infrequent in ES research using institutional
theory. This might be reasoned by the fact that organizational studies tend to black box
technology (Orlikowski and Barley 2001; Orlikowski and Iacono 2001), and that IS
researchers adopt this “black boxing” when they use institutional theory coming from
organizational studies. The inclusion is important anyway because the inner structure of
the ES software plays a significant role, for instance with its inscribed institutional logics impacting on the alignment process between the organization and ES (Berente 2009;
Gosain 2004).
162
IMPLICATIONS FOR PRACTICE
Currie (2009) argues that institutional theory offers conceptual tools and techniques for
practitioners to understand complex change management scenarios involving information systems. We follow her line of thought and elaborate possible implications for
practice.
The conceptual model can be used as a rich analytical tool to describe and analyze wider social issues, but also has the potential to be used normatively. The wider social issues and institutional structures are inevitably ingredients of complex projectmanagement and change-management scenarios involving ES, but at the same time we
have to retain the focus on technology (ES software), actors, and organization(s), which
are obvious parts of the practical implementation and use of ES. The model thus enforces a more holistic view of the environment in which the ES implementation and use
take place. This could be illustrated with an example: organizations that are implementing ES are strongly recommended to adopt the best practices inscribed in ES and to
avoid customizations (e.g. Hildebrand 2009; Parr and Shanks 2003; Seddon et al. 2003).
This is, however, not always the best approach, because it can lead to severe misalignment problems (Sia and Soh 2007) and lock the organization into an inappropriate structure preventing future optimizations (Lindley et al. 2008). This recommendation has
even so become a strong rationalized myth as the legitimized way of implementing ES.
Here the model can be used to understand the observed phenomenon in a more informed
way and to inspire approaches to deviate knowingly and willingly from this strong rationalized myth arising from the organizational field as an institutional pressure.
Working with ES projects could be challenging, and practitioners often embed a technorational understanding (Dillard and Yuthas 2006) where they search for rational solutions to organizational and technical issues. This mindset could lead to frustration
among project personnel, when they experience problematic unintended situations or
apparent obscure management decisions deviating from “the rational path”. Institutional
theory might in these cases offer complementary understandings and explanations, motivating alternative solutions or “just” reducing the frustrations among project personnel, which in itself can be beneficial for the organization.
163
The conceptual model even has the potential to be stretched to more normative purposes, which are shown with an illustrative example: the institutional logics with a given
ES and the institutional logics in a given department are analyzed during the design
phase of an ES project, and the obtained knowledge is used to reduce the anticipated
misalignment by carrying out identified technical and organizational activities in order
to adapt the ES and department to each other.
CONCLUSION
In this paper we have examined the use of institutional theory in ES research. Institutional theory has been advocated as highly relevant to IS research (Orlikowski and Barley 2001), and promoted as distinctively well-suited to ES research (Berente et al. 2007;
Gosain 2004). Institutional theory offers a conceptually rich lens to study social considerations in ES research, which has often been downplayed or even overlooked in extant
studies. The paper set out to review current ES research using institutional theory and to
develop a conceptual model to provide a foundation for further ES research applying
institutional theory. We identified 18 selected papers, which were analyzed and used to
identify key features of institutional theory and to develop a conceptual model. In the
process we found that institutional theory in ES research is in its infancy and adopts
mainly “traditional” institutional aspects like isomorphism, with the organization as the
level of analysis, and in several cases complemented by structuration theory and other
theoretical lenses. We presented a conceptual model from the analysis of the literature,
which advocates multi-level and multi-theory approaches and applies newer institutional aspects as institutional logics. The model offers a conceptually rich lens for analyzing
the implementation and use of ES in organizations for both researchers and practitioners.
There are some limitations to this paper. Institutional theory is a highly complex and
diverse theory (Currie 2009), so even the selective presentation in this paper has the
danger of being “everything about everything”, while others might claim the opposite,
that important institutional elements are absent. Both positions are valid critiques and it
is a difficult trade-off with constraints, and the response may be that the conceptual
model is flexible so institutional elements may be included or excluded depending on
164
specific research issues. Another limitation is the small number of papers in the detailed
literature review, which warrants our claim of infancy of institutional theory in ES research. It might be possible to increase the number of papers by enhancing the keyword
searches to include “isomorphism”, “rationalized myths”, “institutional”, “CRM”, “enterprise information systems” etc., which would possibly give a more embracing study.
Much ES research takes a life-cycle approach (Esteves and Bohorquez 2007), and this
perspective is lacking in the analysis of the current themes and the conceptual model,
which is a significant drawback calling for future research, e.g. to enhance the conceptual model with life-cycle elements (see also Gosain 2004) or other ways to address this
issue. Finally, the focus in this paper is deliberately on ES research and not IS research
in general, because ES is a special class of information systems, as argued in previous
chapters. However, this is a limitation but also an opportunity for generalization (Seddon and Scheepers 2006) in future studies.
165
APPENDIX
ID
Alphabetic Title
listing of
papers
Journal
Type of Organization
1
Barratt and Mandated RFID and instituChoi (2007) tional responses: Cases of
decentralized business units
Production
and Operations Management
Large defense
RFID technology Empirical,
contractor in the US used in supplypositivistic
chain management
Communications of Association of
Information
Systems
Three global organizations (one financial services and
two professional
services)
2
Braganza
and Desouza (2006)
Implementing section 404 of
the Sarbanes Oxley Act:
Recommendations for
information systems organizations
Application
Category
Research
Approach
Level of Key Features of Institutional Theory
Analysis c Institutional and Competitive Pressures
d Rationalized Myths
e Multiple Levels of Analysis
f Institutional Logics
g Other
Organization
c Institutional pressures within varying degrees of uncertainty
(DiMaggio and Powell 1983; Meyer and Rowan 1977; Scott
1995; Zucker 1987)
Multi-Theory Approach
No
e Department of Defense at the organizational field level impacting on the specific organization
Implementation
Empirical, Organiof Sarbanes
explorative, zation
Oxley Act (SOX) descriptive
(interpretive)
c Institutional pressures to implement SOX (DiMaggio and
Powell 1983)
No
e The US fixed by law the SOX as a federal law to address the
effectiveness of internal controls of public companies (from
society to organization)
g Institutional interventions in IT innovations (King et al. 1994)
3
Brignall
Strategic enterprise manand Ballan- agement systems: New
tine (2004) directions for research
Management
Accounting
Research
Public and private
organizations in
general
Strategic enterConceptual, Organimixed
zation
prise management (SEM)
software “on top”
of operational ES
c Institutional theory is used to gain insight into the internal and • Organizational change with
external contexts in which SEM software is designed, implecontext, content, and process
mented, and used, especially the relative bargaining power
(Pettigrew 1985)
determining what aspects of performance will be measured
• Stakeholder theory (Atkinson et
(DiMaggio and Powell 1983)
al. 1997)
d SEM software constitutes some form of instrumental rational• Contingency theory (Gordon
ism (Meyer and Rowan 1977; Scott 1987)
and Miller 1976)
• Complementarities approach
(Milgrom and Roberts 1995)
4
Caccia and
Steccolini
(2006)
Accounting change in Italian Critical Perslocal governments: What’s
pectives on
beyond managerial fashion? Accounting
Local governments
in Italy
Accounting
system
Empirical,
positivistic
Organization
c Institutional pressures (isomorphism) and institutional rules
(Powell and DiMaggio 1991; Scott 1987; 2001)
d Innovations in financial and accounting techniques [implemented in ES] can shape the vision of reality that organizational
(SAP/R3)
166
ID
Alphabetic Title
listing of
papers
Journal
Type of Organization
Application
Category
Research
Approach
Level of Key Features of Institutional Theory
Analysis c Institutional and Competitive Pressures
d Rationalized Myths
e Multiple Levels of Analysis
f Institutional Logics
g Other
Multi-Theory Approach
actors have, spreading concepts like value for money, accountability, efficiency, and effectiveness, turning them into new shared
meanings and values
f An accounting system is designed as technically perfect and
formally consistent with the most advanced managerial ideas and
fashion (managerialist fashion), but lacks taking the organizational context into consideration
5
Cadili and
Whitley
On the interpretative flexibility of hosted ERP systems
Journal of
Strategic
Information
Systems
Central accounting
department in the
UK of a large
multinational petrol
company
Hosted ERP
systems
(SAP/R3)
Empirical,
interpretive
Enterprise information
systems as objects and
carriers of institutional
forces: The new iron cage
Journal of
Association
for Information Systems
Public and private
organizations in
general
Enterprise information systems
Conceptual, Organipositivistic zation
(2005)
6
Gosain
(2004)
Organization
g Technology (enterprise systems) as institutions, i.e. systems as • Structuration theory (Giddens
an infrastructure have similarities to institutions (Scott 2001;
1984; Orlikowski 1992)
Zucker 1977) and they are “infused with values”
c Institutional pressures (DiMaggio and Powell 1983) are
reflected in the ES configurations that exert control over organizational actors – control that is powerful, sometimes not apparent, and often difficult to resist (Gosain 2004: 160)
Organizational sensemaking
(Weick 1990)
f Institutional logic (DiMaggio 1997) and institutional misalignment
7
Hedman
and Borell
Narratives in ERP systems
evaluation
Journal of
Enterprise
Information
Management
Assimilation of enterprise
systems: The effect of
institutional pressures and
the mediating role of top
management
MIS Quarterly Chinese organizations (77 organizations representing a
wide range of
geographical and
cultural diversity)
(2004)
8
Liang et al.
(2007)
Public and private
organizations in
general
ERP systems
Conceptual, Organiinterpretive zation
d “Narratives are used to persuade, convince and make people
• Theory of action and learning
act and believe in certain ways” (Hedman and Borell 2004: 286).
(Argyris and Schön 1974)
Narratives or stories are powerful rationalized myths that can be
used in ERP systems evaluation (i.e. learning and understanding) • Evaluation and measurements
(Hedman and Borell 2004; Meyer and Rowan 1977)
(Hoebeke 1990)
Enterprise systems
Empirical,
c Mimetic, coercive, and normative institutional pressures
(DiMaggio and Powell 1983) impact on the assimilation of
enterprise systems by guiding top managers in their decisions
positivistic
Organization
• Technology assimilation (Purvis
et al. 2001)
• Innovation diffusion theory
(Jones and Beatty 1998)
167
ID
Alphabetic Title
listing of
papers
Journal
Type of Organization
Application
Category
9
Phang
(2008)
Journal of
Strategic
Information
Systems
National Library
Board in Singapore
Enterprise-wide Empirical,
IS to support HR,
finance, and
positivistic
administrative
functions
Investigating organizational
learning in eGovernment
projects: A multi-theoretic
approach
Research
Approach
Level of Key Features of Institutional Theory
Analysis c Institutional and Competitive Pressures
d Rationalized Myths
e Multiple Levels of Analysis
f Institutional Logics
g Other
Organization
Multi-Theory Approach
c Institutional pressures (isomorphism) (DiMaggio and Powell
1983), and increase legitimacy and survival prospects (Meyer
and Rowan 1977)
• Organizational learning (Argyris and Schön 1978; Levitt and
March 1988)
g Organizations actively seek to shape the institutional environment surrounding organizations (Oliver 1991)
• Organizational culture (Schein
1985)
• Organizational politics (Jasperson et al. 2002; Silva and Backhouse 2003)
10
Reimers
(2003)
Developing sustainable B2B Electronic
e-commerce scenarios in the Markets
Chinese context: A research
proposal
Chinese organizations in general
Interorganizational supplychain management
Conceptual, • Orgapositivistic
nizational
field
c Institutional legitimacy (Meyer and Rowan 1977)
e Organizational field with institutional pressures (DiMaggio
and Powell 1983). Structures of domination and patterns of
coalition emerge from the organizational field, which interorganizational supply chain management must (SCM) reproduce (and
thereby in the respective organizations linked by SCM)
g Rationing transactions from institutional economics (Commons and Rutherford 1990)
11
12
Scarbrough
(2008)
Developing the processual
analysis of institutionalization: The case of resource
planning systems
Sia and Soh An assessment of package(2007)
organization misalignment:
institutional and ontological
Academy of
Management
Proceedings
Public and private
organizations in
general
MRP, MRPII,
and ERP
European
Journal of
Information
Defense industry
and health-care
Enterprise systems
Conceptual
(historical
analysis)
Empirical,
168
• Organ- e Interaction between organizational field and organization
ization g Process of institutionalization (Barley and Tolbert 1997): the
analysis [of MRP, MRPII, and ERP] highlights the way in which
• Orga- RP systems have become institutionalized in business practice
nizathrough a cycle of disembedding knowledge, which links local
tional and field-level actions. The analysis indicates, at some points,
field
that this cycle was complemented and reinforced by the operation of the institutionalizing mechanisms identified (role specializations, production of discourse, and boundary-spanning activities)
• Organ- c Enterprise systems are subject to institutional forces/pressures
ization (Gosain 2004)
f Institutional context is embedded in software (Soh and Sia
• Ontology model of information
system (Wand and Weber 1990)
ID
Alphabetic Title
listing of
papers
structures
13
Soh and Sia An institutional perspective
(2004)
on sources of ERP packageorganization misalignments
Journal
Type of Organization
Systems
industry
Journal of
Strategic
Information
Systems
Three hospitals
implementing the
same package
Application
Category
Research
Approach
Level of Key Features of Institutional Theory
Analysis c Institutional and Competitive Pressures
d Rationalized Myths
e Multiple Levels of Analysis
f Institutional Logics
g Other
positivistic
ERP package
Empirical,
positivistic
Multi-Theory Approach
2004)
• Organ- e The selected ERP package was developed for the European
• Structuration theory (DeSanctis
ization and US markets, where the institutional context at the society
and Poole 1994; Orlikowski
level for health care is marked by being either highly subsidized
1992)
(European market) or paid for by health-care insurance (US
market), and the ERP package was inscribed with this logic. This
is contrary to the tradition in Singapore, where a complicated copayment calculation depending on bed-class etc. is widely implemented with invoices sent to both the patient and the state for
a stay in hospital
f Package–organization misalignments arise when the package
and the implementing organization’s embedded structures differ,
which can lead to “package customization” and/or “organizational adaptation”
14
Thatcher
(2006)
B2B e-commerce adoption
decisions in Taiwan: The
interaction of cultural and
other institutional factors
Electronic
Commerce
Research and
Applications
Electronic and
textile manufacturing industry in
Taiwan
B2B e-commerce Empirical,
positivistic
linked to ERP,
SCM
• Organ- c and e Institutional pressures exerted by industries, governization ments, and national cultures (DiMaggio and Powell 1983).
Institutional factors influencing IT adoption decisions (King et
al. 1994)
Journal of
Strategic
Information
Systems
A large Canadian
financial institution
Email system
Empirical,
interpretive
Organization
c and d The evaluation and selection of enterprise systems are • Decision theory (Simon 1960)
said to be rational and deterministic, but ceremonial aspects seem
to play an important role in order to legitimize the organization
(DiMaggio and Powell 1983; Meyer and Rowan 1977)
A large Spanish
electricity company
Accounting and
financial system
Empirical,
Organization
c Changes in accounting and financial IS (ERP) due to institutional forces (DiMaggio and Powell 1983). Interplay between
institutional and market forces (Oliver 1992), i.e. institutional
and competitive pressures
15
Tingling
and Parent
(2004)
An exploration of enterprise
technology selection and
evaluation
16
Tsamenyi
et al.
Changes in accounting and
Management
financial information system Accounting
Research
in a Spanish electricity
company: A new institutional theory analysis
(2006)
interpretive
g Interplay between institutional forces and intra-organizational
power relations (Oliver 1991)
169
• National cultures (Hofstede
1984)
• Giddens’ articulation of power
(2006)
ID
Alphabetic Title
listing of
papers
17
Wang and
Swanson
(2007)
18
Zhang and
Dhaliwal
(2008)
Journal
Type of Organization
Application
Category
Launching professional
Information
services automation: Institu- and Organization
tional entrepreneurship for
information technology
innovations
Multiple industries
involved in launching professional
services automation
(PSA): IT research
firms and analysts,
IT professional
services organization, consultants,
PSA vendor etc.
Professional
Empirical,
services automa- positivistic
tion (PSA), that
is, ERP to service
organizations
(Burns 2008)
Organizational
field
g Institutional entrepreneurship research (DiMaggio 1988;
Maguire et al. 2004) falls short of explaining the working of the
launch process, which could be explained by “collective actions”
(mobilization and legitimation) (Wang and Swanson 2007: 65) as
an institutionalization process (Zucker 1988) of IT innovations
An investigation of resource-based and institutional theoretic factors in
technology adoption for
operations and supply-chain
management
Chinese firms (101
firms in multiple
industry segments
public or privately
owned)
IT-enabled
operations and
supply-chain
management
Organization
c Coercive, imitative, and normative legitimization processes in
relation to innovation diffusion (DiMaggio and Powell 1983)
International
Journal of
Production
Economics
Research
Approach
Empirical,
positivistic
Level of Key Features of Institutional Theory
Analysis c Institutional and Competitive Pressures
d Rationalized Myths
e Multiple Levels of Analysis
f Institutional Logics
g Other
Table 5: Author-centric categorization of ES research using institutional theory
170
Multi-Theory Approach
• Resource-based view (Barney
1991; Melville et al. 2004)
REFERENCES
Alvarez, R. (2002). The Myth of Integration: A Case Study of an ERP Implementation.
In F. F. H. Nah (Ed.), Enterprise Resource Planning Solutions and Management.
Hershey: Idea Group Inc., 63-88.
Argyris, C. and D. A. Schön (1974). Theory in Practice: Increasing Professional Effectiveness. San Francisco, California: Jossey-Bass Publishers.
Argyris, C. and D. A. Schön (1978). Organizational Learning: A Theory of Action
Perspective. Reading, MA: Addison-Wesley
Atkinson, A. A., J. H. Waterhouse and R. B. Wells (1997). A stakeholder approach to
strategic performance measurement. Sloan Management Review 38(3): 25-37.
Barley, S. R. (1986). Technology as an Occasion for Structuring: Evidence from Observations of CT Scanners and the Social Order of Radiology Departments. Administrative Science Quarterly 31(1): 78-108.
Barley, S. R. and P. S. Tolbert (1997). Institutionalization and Structuration: Studying
the Links between Action and Institution. Organization Studies 18(1): 93-117.
Barney, J. (1991). Firm Resources and Sustained Competitive Advantage. Journal of
Management 17(1): 99-121.
Barratt, M. and T. Choi (2007). Mandated RFID and Institutional Responses: Cases of
Decentralized Business Units. Production & Operations Management 16(5):
569-585.
Berente, N. (2009). Conflicting Institutional Logics and the Loose Coupling of Practice
with NASE's Enterprise Information System. Department of Information Systems, Case Western Reserve University, Cleveland. PhD Thesis.
Berente, N., K. Lyytinen and Y. Yoo (2007). An Institutional Analysis of Pluralistic
Responses to Enterprise System Implementations. ICIS 2007 Proceedings, Paper 96.
Berger, P. L. and T. Luckmann (1966). The Social Construction of Reality. A Treatise in
the Sociology of Knowledge. New York: Doubleday.
Boudreau, M. C. and D. Robey (2005). Enacting Integrated Information Technology: A
Human Agency Perspective. Organization Science 16(1): 3-18.
Boxenbaum, E. and S. Jonsson (2008). Isomorphism, Diffusion and Decoupling. In R.
Greenwood, C. Oliver, K. Sahlin and R. Suddaby (Eds.), The SAGE Handbook
of Organizational Institutionalism. London: Sage Publications, 78-98.
171
Braganza, A. and K. C. Desouza (2006). Implementing Section 404 of the Sarbanes
Oxley Act: Recommendations for Information Systems Organizations. Communications of AIS 2006(18): 464-487.
Brehm, L. and M. L. Markus (2000). The Divided Software Life Cycle of ERP Packages. Proceedings of 1st Global Information Technology Management (GITM)
World Conference, Memphis, Tennessee.
Brignall, S. and J. Ballantine (2004). Strategic Enterprise Management Systems: new
directions for research. Management Accounting Research 15(2): 225-240.
Burns, M. (2008). Enterprise software survey 2008. CA Magazine 141(7): 14-15.
Burrell, G. and G. Morgan (1979). Sociological Paradigms and Organisational Analysis. Elements of the Sociology of Corporate Life. London: Heinemann Educational.
Caccia, L. and I. Steccolini (2006). Accounting change in Italian local governments:
What's beyond managerial fashion? Critical Perspectives on Accounting 17(23): 154-174.
Cadili, S. and E. A. Whitley (2005). On the interpretative flexibility of hosted ERP systems. The Journal of Strategic Information Systems 14(2): 167-195.
Carton, F., F. Adam and D. Sammon (2008). Project Management: A Case Study of a
Successful ERP Implementation. International Journal of Managing Projects in
Business 1(1): 106-124.
Commons, J. R. and M. Rutherford (1990). Institutional economics: its place in political economy. News Brunswick: Transaction Publishers.
Currie, W. (2008). Institutionalization of IT Compliance: A Longitudinal Study. ICIS
2008 Proceedings. Paper 182.
Currie, W. (2009). Contextualising the IT artefact: Towards a Wider Research Agenda
for IS using Institutional Theory. Information Technology & People 22(1): 6377.
Currie, W. L. and M. W. Guah (2007). Conflicting Institutional Logics: A National Programme for IT in the Organisational Field of Healthcare. Journal of Information
Technology 22(3): 235-247.
Czarniawska, B. (2009). How Institutions are Inscribed in Technical Objects and What
it may mean in the case of the Internet. In F. Contini and G. F. Lanzara (Eds.),
ICT and Innovation in the Public Sector - European Studies in the Making of EGovernment. Houndmills, Basingstoke, Hampshire: Palgrave Macmillan, 49-65.
Davenport, T. H. (1998). Putting the enterprise into the enterprise system. Harvard
Business Review 76(4): 121-131.
172
Davenport, T. H. (2000). Mission Critical: Realizing the Promise of Enterprise Systems.
Boston: Harvard Business School Press.
DeSanctis, G. and M. S. Poole (1994). Capturing the Complexity in Advanced Technology Use: Adaptive Structuration Theory. Organization Science 5(2): 121-147.
Dillard, J. F. and K. Yuthas (2006). Enterprise Resource Planning Systems and Communicative Action. Critical Perspectives on Accounting 17(2-3): 202-223.
DiMaggio, P. (1997). Culture and Cognition. Annual Review of Sociology 23: 263-287.
DiMaggio, P. J. (1988). Interest and Agency in Institutional Theory. In L. G. Zucker
(Ed.), Institutional Patterns and Organizations: Culture and Environment.
Cambridge, Massachusetts: Ballinger Publishing Company, 3-21.
DiMaggio, P. J. and W. W. Powell (1983). The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields. American Sociological Review 48(2): 147-160.
Esteves, J. and V. Bohorquez (2007). An Updated ERP Systems Annotated Bibliography: 2001-2005. Communications of the Association for Information Systems
19(Article 18): 386-447.
Fligstein, N. (2001). Social Skill and the Theory of Fields. Sociological Theory 19(2):
105-125.
Friedland, R. and R. R. Alford (1991). Bringing Society Back In: Symbols, Practices,
and Institutional Contradictions. In W. W. Powell and P. DiMaggio (Eds.), The
New Institutionalism in Organizational Analysis. Chicago: University of Chicago Press, 232-263.
Giddens, A. (1984). The Constitution of Society: Outline of the Theory of Structure.
Berkeley: University of California Press.
Giddens, A. (2006). Sociology. Cambridge: Polity Press.
Gordon, L. A. and D. Miller (1976). A Contingency Framework for the Design of Accounting Information Systems. Accounting, Organizations & Society 1(1): 5970.
Gosain, S. (2004). Enterprise Information Systems as Objects and Carriers of Institutional Forces: The New Iron Cage. Journal of the Association for Information
Systems 5(4): 151-182.
Grabski, S. V., S. A. Leech and B. Lu (2003). Enterprise System Implementation Risks
and Controls. In P. B. Seddon, L. Willcocks and G. Shanks (Eds.), Second-Wave
Enterprise Resource Planning Systems: Implementing for Effectiveness. Cambridge: Cambridge University Press, 135-156.
Gray, D. E. (2004). Doing Research in the Real World. London: Sage Publications.
173
Greenwood, R., C. Oliver, K. Sahlin and R. Suddaby (2008a). Introduction. In R.
Greenwood, C. Oliver, K. Sahlin and R. Suddaby (Eds.), The SAGE Handbook
of Organizational Institutionalism. London: Sage Publications, 1-46.
Greenwood, R., C. Oliver, K. Sahlin and R. Suddaby (2008b). The SAGE Handbook of
Organizational Institutionalism. London: Sage Publications.
Greenwood, R., R. Suddaby and C. R. Hinings (2002). Theorizing Change: The Role of
Professional Associations in the Transformation of Institutionalized Fields.
Academy of Management Journal 45(1): 58-80.
Hargrave, T., J. and A. H. Van De Ven (2006). A Collective Action Model of Institutional Innovation. Academy of Management. The Academy of Management Review 31(4): 864-888.
Hedman, J. and A. Borell (2004). Narratives in ERP systems evaluation. Journal of Enterprise Information Management 17(4): 283-290.
Hedström, P. and R. Swedberg (1996). Social Mechanisms: An Introductory Essay. In
P. Hedström and R. Swedberg (Eds.), Social Mechanisms - An Analytical Approach to Social Theory. Cambridge: Cambridge University Press, 1-31.
Hildebrand, C. (2009). The Value of Sticking with Vanilla. Profit Online. Retrieved
25th March, 2009, from
http://www.oracle.com/profit/smb/122808_ziegele_qa.html.
Hoebeke, L. (1990). Measuring in Organisations. Journal of applied systems analysis
17: 115-122.
Hofstede, G. (1984). Culture's Consequences: International Differences in WorkRelated Values. Newbury Park, California: Sage Publications Inc.
Jasperson, J., A. Carte Traci, S. Saunders Carol, S. Butler Brian and et al. (2002). Review: Power and Information Technology Research: A Metatriangulation Review. MIS Quarterly 26(4): 397-459.
Jones, M. C. and R. C. Beatty (1998). Towards the development of measures of perceived benefits and compatibility of EDI: a comparative assessment of competing first order factor models. European Journal of Information Systems 7: 210220.
Jones, M. R. and H. Karsten (2008). Giddens's Structuration Theory and Information
Systems Research. MIS Quarterly 32(1): 127-157.
King, J. L., V. Gurbaxani, K. L. Kraemer, F. W. McFarlan, K. S. Raman and C. S. Yap
(1994). Institutional Factors in Information Technology Innovation. Information
Systems Research 5(2): 139-169.
174
Kraemmergaard, P. and B. R. Schlichter (2009). A Comprehensive Literature Review of
the ERP Research Field. CIM Working Paper No. 1, Vol. 1. Aalborg University,
Aalborg.
Lamb, R. and R. Kling (2003). Reconceptualizing Users as Social Actors in Information
Systems Research. MIS Quarterly 27(2): 197-235.
Latour, B. (1987). Science in action: How to follow scientists and engineers through
society. Cambridge, Massachusetts: Harvard University Press.
Levitt, B. and J. G. March (1988). Organizational Learning. Annual Review of Sociology 14: 319-338.
Liang, H., N. Saraf, H. Qing and X. Yajiong (2007). Assimilation of Enterprise Systems: The Effect of Institutional Pressures and the Mediating Role of Top Management. MIS Quarterly 31(1): 59-87.
Lindley, J. T., S. Topping and L. T. Lindley (2008). The hidden financial costs of ERP
software. Managerial Finance 34(2): 78-90.
Maguire, S., C. Hardy and B. Lawrence Thomas (2004). Institutional Entrepreneurship
in Emerging Fields: HIV/AIDS Treatment Advocacy in Canada. Academy of
Management Journal 47(5): 657-679.
Melville, N., K. Kraemer and V. Gurbaxani (2004). Information Technology and Organizational Performance: An Integrative Model of IT Business Value. MIS Quarterly 28(2): 283-322.
Meyer, J. W. and B. Rowan (1977). Institutionalized Organizations: Formal Structure as
Myth and Ceremony. American Journal of Sociology 83(2): 340-363.
Milgrom, P. and J. Roberts (1995). Complementarities and Fit: Strategy, Structure, and
Organizational Change in Manufacturing. Journal of Accounting & Economics
19(2-3): 179-209.
Monteiro, E. and O. Hanseth (1996). Social shaping of information infrastructure: on
being specific about the technology. Information Technology and Changes in
Organizational Work: 325-343.
Newman, M. and D. Robey (1992). A Social Process Model of User-Analyst Relationships. MIS Quarterly 16(2): 249-267.
North, D. C. (1990). Institutions, Institutional Change and Economic Performance.
Cambridge: Cambridge University Press.
Oliver, C. (1991). Strategic Responses to Institutional Processes. Academy of Management. The Academy of Management Review 16(1): 145-180.
Oliver, C. (1992). The Antecedents of Deinstitutionalization. Organization Studies
13(4): 563-589.
175
Orlikowski, W. J. (1992). The Duality of Technology: Rethinking the Concept of Technology in Organizations. Organization Science 3(3): 398-427.
Orlikowski, W. J. and S. R. Barley (2001). Technology and Institutions: What can Research on Information Technology and Research on Organizations Learn from
each other. MIS Quarterly 25(2): 145-165.
Orlikowski, W. J. and C. S. Iacono (2001). Research Commentary: Desperately Seeking
the 'IT' in IT Research - A Call to Theorizing the IT Artifact. Information Systems Research 12(2): 121-135.
Parr, A. and G. Shanks (2003). Critical Success Factors Revisited: A Model for ERP
Project Implementation. In P. B. Seddon, L. Willcocks and G. Shanks (Eds.),
Second-Wave Enterprise Resource Planning Systems. Cambridge: Cambridge
University Press, 196-219.
Pettigrew, A. M. (1985). Contextualist Research and the Study of Organizational
Change Processes. In E. Mumford, R. Hirschheim, G. Fitzgerald and A. T.
Wood-Harper (Eds.), Research Methods in Information Systems. Amsterdam:
Elsevier Science Publishers B.V., 53-78.
Phang, C. W., A. Kankanhalli and C. Ang (2008). Investigating Organizational Learning in eGovernment Projects: A Multi-theoretic Approach. The Journal of Strategic Information Systems 17(2): 99-123.
Pollock, N. and R. Williams (2009). Software and Organizations: The Biography of the
Enterprise-wide System or How SAP Conquered the World. Milton Park, Abingdon, Oxon: Routledge.
Powell, W. W. and P. DiMaggio (1991). The New Institutionalism in Organizational
Analysis. Chicago: The University of Chicago Press.
Purvis, R. L., V. Sambamurthy and R. W. Zmud (2001). The Assimilation of Knowledge Platforms in Organizations: An Empirical Investigation. Organization
Science 12(2): 117-135.
Reimers, K. (2003). Developing Sustainable B2B E-Commerce Scenarios in the Chinese Context: A Research Proposal. Electronic Markets 13(4): 261 - 270.
SAP (2008). SAP Supply Chain Management - Your Business, Your Future. SAP, Document number 50 076 037 (08/07).
Scarbrough, H., M. Robertson and J. Swan (2008). Developing the Processual Analysis
of Institutionalization: The Case of Resource Planning Systems Innovation.
Academy of Management Proceedings.
Schein, E. H. (1985). Organizational Culture and Leadership. San Francisco: JosseyBass.
176
Scott, W. R. (1987). The Adolescence of Institutional Theory. Administrative Science
Quarterly 32(4): 493-512.
Scott, W. R. (1995). Institutions and Organizations: Theory and Research. Thousands
Oaks: Sage Publications.
Scott, W. R. (2001). Institutions and Organizations. Thousands Oaks: Sage Publications
Scott, W. R. (2004). Institutional Theory: Contributing to a Theoretical Research Program. In K. G. Smith and M. A. Hitt (Eds.), Great Minds in Management: The
Process of Theory Development. Oxford: Oxford University Press, 460-485.
Scott, W. R. (2008a). Approaching Adulthood: The Maturing of Institutional Theory.
Theory and Society 37(5): 427-442.
Scott, W. R. (2008b). Institutions and Organizations: Ideas and Interests. Thousands
Oaks: Sage Publications.
Scott, W. R. and J. W. Meyer (1991). The Organization of Societal Sectors: Propositions and Early Evidence. In W. W. Powell and P. DiMaggio (Eds.), The New
Institutionalism in Organizational Analysis. Chicago: University of Chicago
Press, 108-140.
Seddon, P. B. and R. Scheepers (2006). Other-settings Generalization in IS Research.
ICIS 2006 Proceedings. Paper 70.
Seddon, P. B., L. Willcocks and G. Shanks (2003). Introduction: ERP - The Quiet
Revolution? In P. B. Seddon, L. Willcocks and G. Shanks (Eds.), Second-Wave
Enterprise Resource Planning Systems: Implementing for Effectiveness. Cambridge: Cambridge University Press, 1-19.
Selznik, P. (1949). TVA and the Grass Roots. Berkeley, University of California Press.
Sia, S. K. and C. Soh (2007). An Assessment of Package-Organisation Misalignment:
Institutional and Ontological Structures. European Journal of Information Systems 16(5): 568-583.
Silva, L. and J. Backhouse (2003). The Circuits-of-Power Framework for Studying
Power in Institutionalization of Information Systems. Journal of the Association
for Information Systems 4(6): 294-336.
Simon, H. A. (1960). The New Science of Management Decision. New York: Harper.
Soh, C., S. Kien Sia and J. Tay-Yap (2000). Cultural fits and misfits: Is ERP a universal
solution? Communications of the ACM 43(4): 47-51.
Soh, C. and S. K. Sia (2004). An Institutional Perspective on Sources of ERP PackageOrganisation Misalignments. The Journal of Strategic Information Systems
13(4): 375-397.
177
Sumner, M. (2003). Risk Factors in Enterprise-wide/ERP Projects. In P. B. Seddon, L.
Willcocks and G. Shanks (Eds.), Second-Wave Enterprise Resource Planning
Systems: Implementing for Effectiveness. Cambridge: Cambridge University
Press, 157-179.
Swanson, E. B. and N. C. Ramiller (1997). The Organizing Vision in Information Systems Innovation. Organization Science 8(5): 458-474.
Thatcher, S. M. B., W. Foster and L. Zhu (2006). B2B E-commerce Adoption Decisions
in Taiwan: The Interaction of Cultural and Other Institutional Factors. Electronic Commerce Research and Applications 5(2): 92-104.
Thornton, P. H. and W. Ocasio (2008). Institutional Logics. In R. Greenwood, C. Oliver, K. Sahlin and R. Suddaby (Eds.), The SAGE Handbook of Organizational Institutionalism. London: Sage Publications, 99-129.
Tingling, P. and M. Parent (2004). An Exploration of Enterprise Technology Selection
and Evaluation. The Journal of Strategic Information Systems 13(4): 329-354.
Tolbert, P. S. and L. G. Zucker (1996). The Institutionalization of Institutional Theory.
In S. R. Clegg, C. Hardy and W. R. Nord (Eds.), Handbook of Organization
Studies. London: Sage Publications, 169-184.
Tsamenyi, M., J. Cullen and J. M. G. González (2006). Changes in Accounting and Financial Information System in a Spanish Electricity Company: A New Institutional Theory Analysis. Management Accounting Research 17(4): 409-432.
Vitharana, P. and R. Dharwadkar (2007). Information Systems Outsourcing: Linking
Transaction Cost and Institutional Theories. Communications of the Association
for Information Systems 20(Article 23): 346-370.
Wand, Y. and R. Weber (1990). An Ontological Model of an Information System. IEEE
Transactions on Software Engineering 16(11): 1282-1292.
Wang, P. and E. B. Swanson (2007). Launching Professional Services Automation: Institutional Entrepreneurship for Information Technology Innovations. Information and Organization 17(2): 59-88.
Webster, J. and R. T. Watson (2002). Analyzing the Past to Prepare for the Future:
Writing a Literature Review. MIS Quarterly 26(2): 13-23.
Weerakkody, V., Y. K. Dwivedi and Z. Irani (2009). The Diffusion and Use of Institutional Theory: A Cross-disciplinary Longitudinal Literature Survey. Journal of
Information Technology 24(4): 354-368.
Weick, K. E. (1990). Technology as Equivoque: Sensemaking in New Technologies. In
P. S. Goodman and L. S. Sproull (Eds.), Technology and Organizations. San
Francisco: Jossey-Bass Publishers, 1-44.
178
Zhang, C. and J. Dhaliwal (2009). An Investigation of Resource-based and Institutional
Theoretic Factors in Technology Adoption for Operations and Supply Chain
Management. International Journal of Production Economics 120(1): 252-269.
Zucker, L. G. (1977). The Role of Institutionalization in Cultural Persistence. American
Sociological Review 42(5): 726-744.
Zucker, L. G. (1987). Institutional Theories of Organization. Annual Review of Sociology 13: 443-464.
Zucker, L. G. (1988). Where do Institutional Patterns come from? Organizations as Actors in Social Systems. In L. G. Zucker (Ed.), Institutional Patterns and Organizations: Culture and Environment. Cambridge, MA: Ballinger, 23-49.
Zucker, L. G. (1991). The Role of Institutionalization in Cultural Persistence. In W. W.
Powell and P. DiMaggio (Eds.), The New Institutionalism in Organizational
Analysis. Chicago: University of Chicago Press, 83-107.
179
PAPER 2
SVEJVIG, P. AND PRIES-HEJE, J. “ENTERPRISE SYSTEMS OUTSOURCING “BEHIND THE CURTAIN” A CASE
STUDY SHOWING HOW RATIONAL AND INSTITUTIONAL
EXPLANATIONS CO-EXIST AND COMPLEMENT EACH
OTHER”
Abstract.
Outsourcing is now a feasible means for enterprise systems (ES) cost savings, but
does however increase the complexity of coordination substantially when many
organizations are involved. We set out to study ES outsourcing in a large Scandinavian high-tech organization, SCANDI, a case setting with many interorganizational partners, trying to answer the question: Why does SCANDI engage in these
very complex outsourcing arrangements? To answer this question we have analyzed documents, observed meetings and gathered data from interviews in four
parts of SCANDI. The first data analysis found just the rational frontstage costsaving explanation; but then, with a more careful analysis focusing on institutional factors, other backstage explanations “behind the curtain” were uncovered,
such as management consultants with a “best practice” agenda, people promoting
outsourcing, thereby being promoted themselves, and a belief in outsourcing as a
“silver bullet”: a recipe to success, solving everything.
180
2
INTRODUCTION
The development of Enterprise Systems (ES) has been a major trend within both the
private and public sectors over the past decade. They often trigger major organizational
changes and at the same time, introduce high-risks with potential high rewards (Chae
and Lanzara 2006: 100; Markus 2004). ES can be defined as large-scale organizational
systems, built around packaged ES software, and composed of people, processes, and
information technology (Seddon et al. 2003).
Outsourcing of ES is becoming increasingly important in today’s global business environment in order to gain cost savings (Olson 2007). Gartner valued the application outsourcing market at $76.9 billion in 2009 and forecasted that it would increase to $97.9
billion in 2012 (Young et al. 2008), which is a high growth rate in these turbulent economic times, indicating that an outsourcing trend is in progress.
ES outsourcing is the practice whereby the organization “purchases goods or services
that were previously provided internally” (Lacity and Hirschheim 1993: 74). In this
paper, we interpret this further, defining it as any type of outsourcing involving ES and
Information Technology (IT). This includes application development on top of the ES,
operation of a data centre running ES, business processes enabled by the ES, or the entire IS function (inspired by Dibbern et al. 2004). Thus, outsourcing in this context
means aggregating specific tasks or entire processes and moving them to one or more
outsourcing vendors (McFarlan and DeLacey 2004).
Organizations have claimed that ES/IS outsourcing reduces cost and time, increases
quality and reliability of product and services, improves business performance, and releases organizations to concentrate on core competencies (McFarlan and DeLacey 2004;
Vitharana and Dharwadkar 2007).
The reasons why organizations outsource are dominated by rational explanations related
to bounded rationality and opportunism (Williamson 1981), and cost savings seem to be
a prevailing explanation, which is repeated in the literature (e.g. Dibbern et al. 2004;
Olson 2007).
181
However, though a majority of extant literature finds mainly rational explanations for
engaging in ES/IS outsourcing, our own experience from many companies and from
many outsourcing arrangements made us speculate whether there was more to it. Using
a theatrical metaphor we ask, could it be that there were both “frontstage” explanations
as well as other, different, “backstage” explanations hidden “behind the curtain”?
To address this curiosity, we set out to study ES outsourcing in a large Scandinavian
high-tech organization, SCANDI, which has many inter-organizational partners. Our
research question was: Why does SCANDI engage in very complex ES/IS outsourcing
arrangements?
The paper is organized as follows. First we introduce the Transaction Cost Theory and
the Institutional Theory. Then we explain our research methodology based on the interpretive paradigm. The section that follows outlines the SCANDI case study. The analysis of frontstage and backstage explanations is then presented and is followed by a discussion and implications. Finally, we present some concluding remarks and sum up significant new insights, both from a theoretical point of view and from a practitioner perspective.
RATIONAL AND INSTITUTIONAL EXPLANATIONS
The theories used in this study have evolved over time in reaction to our progressive
understanding of data collected during field work. Rational explanations were easily
identifiable, and institutional explanations resulted from searching for “behind the curtain” explanations while interpreting the field data and IS literature. Two alternative
theories were then adopted for this research: (1) Transaction Cost Theory (TCT), and
(2) Institutional Theory (INT).
Both theories can be applied with an interpretive approach, where they are viewed as
alternative “lenses” for interpreting outsourcing decisions, and not as the objective deterministic truth about outsourcing decisions (adapted from Lacity and Hirschheim
1993). TCT was adopted because it has the widely accepted view that organizational
actors make outsourcing decisions based on the economic rationale used in a majority of
outsourcing literature, embedding an understanding about bounded rationality and op-
182
portuunism (Scottt and Daviss 2007: 53-556). We ado
opted INT too explain ouutsourcing decid
sionss in line witth Vitharanaa and Dharw
wadkar (200
07), becausee the theorettical constru
ucts
in IN
NT are recoggnizable in our empiriccal data. Thee two theoreetical lensess are presen
nted
in thee conceptuaal frameworrk in Figuree 1, which iss used to answer the ressearch quesstion.
Figgure 1: Concep
ptual frameworrk combining rational
r
and in
nstitutional facctors
Figurre 1 shows that both raational and institutional
i
l factors imppact the deccision-making
proceess to adoptt outsourcinng, and thereeby make up
u the determ
minants for outsourcing
g.
Ratioonal and insstitutional faactors are afffecting eacch other (hence the arroows between
n
them
m) where, for instance, rational
r
facttors, which are rooted in
i organizattional efficienccy, elaborate into rationnalized instiitutional my
yths that cannnot be objeectively testted
(Meyyer and Row
wan 1977). The followiing sectionss detail TCT
T and INT.
TRA
ANSACTIO
ON COST
T THEORY
Y
TCT is an organnizational ecconomics thheory linkin
ng economicc theory witth organizattional
realitty, and TCT
T focuses onn the relatioonships betw
ween organiizations (Sccott and Dav
vis
20077; Shafritz et
e al. 2005). TCT suggeests that the only reasonn for an organization to
o
outsoource is cosst savings (L
Lacity and Hirschheim
H
1993). Trannsaction hass been defin
ned
as “w
when a goodd or service is transferrred across a technologiccally separaable interfacce”
(Willliamson 19881: 552). TC
CT is basedd on several dimensionss: a) Costs, b) transactiion
type,, c) threat off opportunissm, and d) uncertainty
u
(Lacity andd Hirschheim
m 1993; Scott
and Davis
D
2007)).
183
The total costs of a transaction consist of the production cost (the cost to produce the
transaction measured in capital, labour and/or materials), and the coordination cost (related to controlling and monitoring the transaction); this is a transaction cost (Williamson 1981: 552). Both kinds of costs incur when transactions are produced internally or
externally, but economic efficiency can be achieved by a comparative analysis of production and transaction costs for different alternatives; this is called a make or buy decision. This seems to be a straightforward rational economic decision, but the challenge is
that both production and transaction costs are often difficult to estimate (Lacity and
Hirschheim 1993) implying bounded rationality, which limits the optimal decision either to outsource or to insource (Vitharana and Dharwadkar 2007).
There are different types of transactions; they vary in frequency and asset specificity.
Frequency could be occasional or frequent. Asset specificity refers to the degree of customization of the transaction (Lacity and Hirschheim 1993; Scott and Davis 2007). An
internally tailor-made system has higher asset specificity than a packaged enterprise
system. The transaction type impacts on the make or buy decisions and the sourcing
strategies (Williamson 1991).
Scott and Davis (2007) state that people are not only boundedly rational, but also sometimes opportunistic, which means that they act on “self-interest seeking with guile”
(Williamson 1981: 554); where people (and hence organizations) are not trustworthy
and honest, they even lie, cheat, and steel (Scott and Davis 2007; Williamson 1981).
The last dimension is uncertainty, which expresses the extent to which it is possible to
map out all future contingencies (Scott and Davis 2007). Transactions with a high degree of certainty are relatively uninteresting, for instance, in situations where it is easy
to substitute the outsourcing vendor; but transactions with a high degree of uncertainty
should attract attention, depending on the asset specificity (Williamson 1979: 253-254).
The combination of high uncertainty and a high degree of asset specificity especially
calls for careful screening and evaluation of vendors. ES outsourcing uncertainties include a vendor’s inability to deliver agreed services, because of their bankruptcies, misaligned cultures, uncontrolled expenditures etc. (Dibbern et al. 2004; Olson 2007;
Vitharana and Dharwadkar 2007). Several ES outsourcings could be placed in a cate-
184
gory with both a high uncertainty and a high degree of asset specificity, which is also
the case for the study presented in this paper.
INSTITUTIONAL THEORY
As we noted above, we found Institutional Theory (INT) to be very useful in coding and
understanding our data. Institutional Theory deals with the pervasive influence of institutions on human behaviour, including the processes by which structures (for example,
rules, routines, and norms) guide social behaviour (Scott 2008).
Rationalized myths are part of the institutional context and they are often disguised as
rational arguments used by organizations to “maximize their legitimacy and increase
their resources and survival capabilities” (Meyer and Rowan 1977: 53). Meyer and
Rowan argue that rationalized myths and taken-for-granted rules lead to “isomorphism”
(structural similarity), where the formal structures of organizations should conform to
society to obtain legitimacy. DiMaggio and Powell (1983) “move” the focus on isomorphism from society level to the organizational field level and introduce the concepts of
coercive, normative, and cognitive institutional pressures. They argue that these pressures lead to isomorphism where organizations live in an iron cage.
Institutional Theory has been applied within IS outsourcing to some extent (Ang and
Cummings 1997; Bridgman and Willmott 2006; Kshetri 2007; Miranda and Kavan
2005; Miranda and Kim 2006; Vitharana and Dharwadkar 2007), for instance the muchpublished Kodak-IBM outsourcing engagements (Loh and Venkatraman 1992), and a
study linking TCT and INT (Vitharana and Dharwadkar 2007). In 1989, IBM took over
the work done by four Kodak data centres and 300 Kodak workers became IBM employees. Kodak hoped to cut about 50% of the costs by this outsourcing engagement
(Loh and Venkatraman 1992: 336). The Kodak-IBM outsourcing was a watershed in IS
outsourcing and created what is called the “Kodak effect”, where many other organizations mimicked the outsourcing practices legitimated by Kodak (Ang and Cummings
1997; Loh and Venkatraman 1992).
Another study is a conceptual paper by Vitharana and Dharwadkar (2007) where they
define a conceptual framework linking TCS, INT, and three stages of institutionalization designated “innovation diffusion”, “stability”, and “deinstitutionalization” (see also
185
Lawrence et al. 2001). The paper takes a macro-view of organizations with the level of
analysis being the organizational field, being exposed by INT and TCT factors. Vitharana and Dharwadkar (2007) describe organizations as mimicking each other in order to
reduce uncertainty in the innovation diffusion phase of the institutionalization, where
more and more organizations adopt an innovation, like a specific outsourcing practice.
Normative forces take over in the stability phase, where the innovation is becoming
well known and widespread by professional associations, consultancy organizations,
and educational systems. Finally, coercive forces, arising from technological changes,
can lead to the deinstitutionalization of outsourcing practices. Explanations using institutional forces have limitations, which could be illustrated like this (Vitharana and
Dharwadkar 2007: 358):
When only 15 per cent of the firms have adopted the practice, the focal firm is
not intensely pressured to mimic others. Hence, in the absence of strong institutional pressures, the firm is likely to employ TCT prescriptions for IT outsourcing governance…. On the other hand, if 80 per cent of the firms have adopted
an IT outsourcing practice, then the focal firm faces greater mimetic pressures.
The nature and extent of mimetic forces will vary within any organizational field, and
be intertwined by rational arguments, which sometimes prevail.
In contrast to the above studies, our research also takes a micro-view on outsourcing
events in a specific organization, SCANDI, in order to understand how intertwined rational and institutional explanations shape the outsourcing decision process, based on
empirical data.
THE INTERDEPENDENCE OF RATIONAL AND INSTITUTIONAL FACTORS
Individuals do not always make rational decisions and neither do organizations. Any
social situation consists of interdependent non-rational and rational elements (Scott
2008: 217-218). ES outsourcing decisions are shaped by both competitive pressures
(rational elements) and institutional pressures (non-rational elements), which are even
more interdependent, as shown on the conceptual framework in Figure 1. Competitive
186
pressures are prevailing in certain situations and institutional pressures are prevailing in
other situations.
Today’s global outsourcing services are characterized by an increasing number of organizations that adopt “multisourcing”. This means moving away from full-scope, fullservice, and long-term outsourcing contracts with a single outsourcing vendor (Nagle
and Maughan 2007), and moving into complex multisourcing arrangements. An empirical study of nine organizations (Jayatilaka 2006) underlines this multisourcing trend,
where determinants for outsourcing arrangements were analysed – both the initial arrangement and the subsequent changes in sourcing. The study showed that both institutional and non-institutional arguments for these outsourcing arrangements were prevailing. The study by Jayatilaka emphasizes the argument that rational and institutional
elements are intertwined and interdependent.
RESEARCH METHODOLOGY
To answer our research question, we studied many relevant documents, gathered data
from interviews in four parts of SCANDI, and we observed numerous meetings; our
aim was to obtain in-depth knowledge about outsourcing arrangements and events in
SCANDI, surrounding the ES, which makes up the scenery for the outsourcing events.
The methodology we adopted was a contextualized, interpretive one, using the technique of case study research (Pettigrew 1990; Walsham 2006). Interpretive research
attempts to understand phenomena through the meanings that people assign to them,
and access to reality is through social constructions, such as language, consciousness,
and shared meanings (Myers and Avison 2002).
We designed our study to be longitudinal starting from the beginning of 2008 to the end
of 2009. Fieldwork was conducted in four parts of SCANDI by one of the authors, to
gain access to ongoing actions, multiplicity among actors, and immediate interpretations. This approach gives a more vivid and dynamic picture than an historical reconstruction. Naturally, however, some events in SCANDI had taken place before January
2008 and they are necessarily historical reconstructions from documents and recollections from interviews.
187
We used a number of data collection methods, as specified in Table 1, below:
Data Collection
Methods
SCANDI
Project Group
Semi-structured 5 interviews
Interviews
Short
unstructured
phone
interviews
3 interviews
SCANDI
Finance
Dept.
SCANDI
Purchasing
Dept.
SCANDI
Supply Chain
Requester
Oracle
7 interviews 10 interviews
5 interviews
3 interviews
8 interviews 9 interviews
9 interviews
1 interview
Focus group
interviews
Participant
observations
18 meetings
1 meeting
2 meetings
Document
analysis
Unpublished documents: plans, reports, minutes and presentations; press releases
from SCANDI; Oracle information (e.g. www.oracle.com) news articles; magazine
reports etc.
Table 1: Data Collection Methods
The aim of the data collection was primarily to understand and interpret, from the interviewees’ perspectives, why SCANDI engages in multiple outsourcings. A number of
participant observations were also conducted, which helped us to understand the relationships between SCANDI and the outsourcing vendors, as well as the many – often
problematic – consequences of outsourcing; we were then able to contrast the explanations for outsourcing given at the executive level. Finally, a comprehensive document
analysis was performed to supplement and complement other sources.
Data analysis followed the interpretive tradition, using hermeneutics (Myers 2009). Interviews were transcribed and notes were taken from the meetings we observed. These
data and documents were then coded and analysed. First, we found the rational cost saving explanations. However, after a subsequent, more careful qualitative data analysis
(Miles and Huberman 1994) we uncovered a number of other “behind the curtain” explanations. Here, we followed a process inspired by Goetz and LeCompte (1981). First,
we scanned and coded the data and documents. This identified several categories of
explanations for outsourcing as well as attributes of both complex and interorganizational partnerships. Second, we compared data and categories to uncover all
relevant elements of the data. During this exercise, we realized that there was more than
rational cost saving explanations in our data. Third, we decided to use the INT as a sen-
188
sitiziing device (Patton
(
20022: 452-462)) as we foun
nd it to be very useful w
while contraasting
rationnal explanaations suppoorted by TCT. Finally, in the fourthh step, we ccreated our conclusiion, in the foorm of tablees showing the relation
nships discoovered. Quallitative dataa
analyysis softwarre (NVivo) (Bazeley 20007) was ussed to suppoort the data managemen
nt
and coding
c
proccess. In Figuure 2, we shhow part of a screen im
mage illustratting codes from
f
the data
d analysiss, which dem
monstrates the institutiional explannations for ooutsourcing
g
from
m the codingg process:
Figurre 2: Codes in NVivo
THE
H SCA
ANDI CASE STUDY
T
This section desscribes the empirical
e
caase study an
nd provides a time-ordeered processs
view
w of the casee study.
CASSE INTRO
ODUCTION
N
SCA
ANDI (a pseuudonym) iss a Scandinaavian compaany with moore than 10,,000 employ
yees.
It bellongs to thee utility induustry field, where
w
it pro
oduces and sells high-teech servicess.
The first
f
companny in SCAN
NDI was esttablished in
n the late 1890s, and thee company today is
i a result off a merger between
b
sevveral compaanies. SCAN
NDI, and itss predecesso
ors,
operaated for manny years as territorial companies
c
in
i a context that was hiighly regulaated,
for example, cusstomers werre only allowed to buy utility servvices from thhem (i.e. a mom
nopooly). This sittuation channged in the 1990s, wheen the Scanddinavian countries deciided
to deeregulate annd liberalizee the utility market.
m
SCA
ANDI now faced higheer competitiion
althoough only feew strong coompetitors exist
e
today.
189
Low level competition has marked SCANDI’s culture. A consultant describes SCANDI
as: “…a super tanker that does not have all the needed engines to react promptly enough
or all the engines are not started simultaneously…or said more directly you cannot lay
off more than a [certain] number of employees and SCANDI has a huge backlog”. This
statement indicates one of the business challenges that SCANDI faces; where a truce
between trade unions and SCANDI limits its manoeuvring. At the same time, the company benefits from its many years of monopoly by having a large market share and
owning a considerable part of the utility infrastructure in the present market, so the shift
from monopoly to competition implies both opportunities and challenges.
SCANDI has been a traditional full-fledged company, but started to outsource activities
including IS from the beginning of the new millennium, and the rate of outsourcing has
increased since then.
ENTERPRISE SYSTEM IMPLEMENTATION AND USE
In 1996, SCANDI decided to implement a leading, global enterprise system in the market. Figure 3 shows a timeline with milestones (M1 to M6) and outsourcing events (O1
to O5).
Year
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
Financial System
(FinSys)
Launch FinSys
(M1)
Upgrade of ES
(M3)
Supply Chain
System (SCS)
Outsourcing
Events
Start SCS project
Launch SCS (M2)
Upgrade of ES
(M3)
Major conversion
of legacy system
(M4)
Outsourcing event O1
Start of RE-ES project (M5)
Outsourcing event O2
Outsourcing event O3
Outsourcing event O4
Outsourcing event O5
2009
Research
Launch of RE-ES project (M6)
Figure 3: Important milestones and outsourcing events
190
Research
Period
The financial system (FinSys) was launched in 1996 (M1), while the supply chain system (SCS) was operational from 2001-2002 (M2). Both FinSys and SCS was upgraded
in 2003 (M3), and SCS had a major extension in 2004 (M4), where a logistic legacy
system was closed down and functionality was transferred to the ES. A reimplementation project (RE-ES project) was started in 2007 (M5), and the project was
launched in January 2009 (M6).
FinSys covers financial management issues, including general ledger, account receivables, and payables etc. and has about 40 heavy users. SCS is used by more than 100
heavy users for purchasing management and inventory management. Another user
group of 3,000+ ad hoc users uses SCS much like an online shopping solution. The ES
is integrated into 40+ other information systems, which mean that the ES is part of a
fairly complex information infrastructure.
The five outsourcing events are detailed in Figure 4 below:
Figure 4: Outsourcing events
Four different functions are shown to the left (of the timeline) in Figure 4: (1) IT development, which is configuring the ES and developing interfaces; (2) IT operation, taking
191
care of operating the ES infrastructure; (3) IT project management, responsible for
managing ES projects; and finally (4) other project tasks.
Four organizations are involved in performing the functions. The bars in Figure 4 indicate (containing vendor names inside) which organization is assigned to which function
at a given period of time; the mixture of bars and textures illuminates the increasing
complexity with many inter-organizational partners, and it is fairly easy to imagine that
the outsourcing events may have seriously hampered the RE-ES project.
THE FIRST PERIOD WITH THE ENTERPRISE SYSTEM (1996-2007)
The financial system was operational in 1996 (M1), and was used for more than 10
years ,with several upgrades since that time. Discussions with finance users indicated
that FinSys is well institutionalized at SCANDI
The first part of SCS was in launched in 2002 (M2) with an upgrade in 2003 (M3) and a
major extension to SCS occurred in spring 2004 (M4), and this release was operational
for several years. The first operational period with the extended SCS from spring 2004
was harsh with severe problems. The SCS seems to be less institutionalized compared
to FinSys, and some users expressed concerns about its user friendliness.
The period from the major extension of SCS in spring 2004 until summer 2007, could
be characterized as a period that mainly operated ES and when there were only a few IT
development activities.
The IT development of the ES was outsourced to Outsourcing Vendor A in summer
2005 (O1) as part of a more comprehensive outsourcing agreement covering many applications; and the ES was suddenly managed by two parties. One of the reasons for this
outsourcing was that “the ERP-area is not business critical for SCANDI and therefore
suitable for outsourcing” (Press Release #2, 2005). More than 100 employees from
SCANDI were relocated to Outsourcing Vendor A. This transfer ensured a quick
knowledge transfer. However, it also caused a challenge because the ties between the
former SCANDI employees and SCANDI was strong, “It has been a challenge…to
work with employees that we have taken over because the ties are so strong between
them and their former colleagues…they accept change requests too easily” (Consultant
192
#1, Outsourcing Vendor A). The transition was even more demanding in relation to the
setting up of the arrangements that facilitated cooperation between SCANDI and Outsourcing Vendor A, and in the establishment of a joint project methodology framework.
Cooperation between the two parties became increasingly institutionalized during 20052007, and was established before entering the RE-ES project in summer 2007 (cf. next
section).
The IT operation of the ES was further outsourced in spring 2007 to Outsourcing Vendor B (O2); and about 150 employees were relocated to Vendor B (Press Release #3,
2007). Outsourcing event O2 was completed three months before starting the RE-ES
project.
The situation in spring 2007 was that three organizations had to work together on development, implementation, and operation of the ES. The cooperation between
SCANDI and Outsourcing Vendor B was not stabilized, let alone the necessary working
practices between the two outsourcing vendors, which were not stabilized either before
starting the RE-ES project.
THE RE-IMPLEMENTATION PROJECT (SUMMER 2007 TO JANUARY
2009)
The ES had evolved into a much more customized standard system since the first implementation in 1996. About 400 customizations were part of the “ES luggage” leading
to difficulties in upgrading the ES, and thereby locking SCANDI into an old ES architecture. This situation was untenable, and something had to be done. Analysis of upgrade options were initiated in January 2007 to mitigate the untenable situation, which
resulted in a decision to go for a re-implementation solution.
Consequently, the RE-ES project was started in summer 2007, covering both FinSys
and SCS. The purpose of the project was primarily to: (1) reduce the number of customizations, (2) modernize the application architecture, (3) optimize business processes,
and (4) use standard functionality. The project was considered a technical reimplementation project with very few benefits for business; indeed, some of the technical changes might have had a negative business impact, and users even changed from
193
their local language to the English language. The launch of RE-ES was delayed several
times, but in January 2009, the system finally went live after an eight month delay.
The RE-ES project organization consisted of a steering committee, project management,
three project groups, and staff functions. SCANDI and Outsourcing Vendor A were
represented in all project groups, including management, while Outsourcing Vendor B
was only involved in the technical infrastructure project group and not at management
level. Outsourcing Vendor B did have a vendor project coordinator assigned to the REES project at the beginning of 2008, but this was late, in relation to the establishment of
smooth co-operational working amongst the three inter-organizational partners.
The project continued with these arrangements, and the spirit between SCANDI and
Outsourcing Vendor A can be characterized as good. The same cannot not be said for
Outsourcing Vendor B; this relationship was expressed as “[they are] not really part of
the project” (Supply Chain Manager) and a consultant said: “the cooperation has been
very problematic and [they] lacks competences related to this project” (Consultant #1).
This was also observed from one of the authors during their participation in project
meetings.
SCANDI took another outsourcing step in spring 2008 (O3) by outsourcing more than
200 employees to Outsourcing Vendor B. This meant that “project management” and
“other project tasks” were transferred to Outsourcing Vendor B (cf. Figure 4), so reducing the number of SCANDI employees further. This made the situation even more unclear, since Outsourcing Vendor B was offering several resources to the RE-ES project,
but was still not represented at steering committee level. This additional outsourcing
activity expanded the outsourcing agreements with Outsourcing Vendor B, and the
SCANDI Chief Information Officer (CIO) stated in a press release (Press Release #4,
2008):
We already have an excellent relationship with [Outsourcing Vendor B]
in several areas and are pleased to expand it further to include management of our legacy application portfolio….The new agreement will enable us to further modernize and consolidate our applications so they deliver the capabilities our business requires.
194
The internal anchoring of the RE-ES project in SCANDI was further changed during
summer 2008, when the RE-ES project became part of a large company-wide programme, which was managed by Outsourcing Vendor C (O4). The consequence for the
RE-ES project was that Outsourcing Vendor C took over the responsibility for project
management, and became responsible for the whole RE-ES project. The transfer of project management responsibilities was achieved in a rush during August 2008.
Four interorganizational partners were now “playing the RE-ES project game”, which
had become a very complex situation in terms of implementation, incorporating possible built-in conflicts and different political interests.
A final, and very controversial change in outsourcing partners was conducted in December 2008 (O5), where the responsibility for IT development was changed from Outsourcing Vendor A to Outsourcing Vendor C, just 1-2 months before the launch of the
RE-ES. This caused many political problems and made launch preparation very risky.
Both the Chief Executive Officer (CEO) and CIO left SCANDI at the end 2008 and an
unsuccessful IT strategy seems to have been a major reason for their resignations, especially for the CIO (Press Release #5, 2008). The many outsourcing events were partly as
a result of this unsuccessful IT strategy, and there was a growing understanding at
SCANDI that the outsourcing strategy had gone too far. The project was nevertheless
launched in January 2009.
THE POST IMPLEMENTATION PERIOD (JANUARY 2009 TO AUTUMN
2009)
The post implementation phase was difficult for the stakeholders involved, but the system was slowly stabilized and discussion with users indicates that they perceived the
system to be “in normal operation” by autumn 2009. Several informal discussions with
stakeholders describe the severe problems after the launch, and they refer to the very
late outsourcing event in December 2008 (O5) as a main cause for the problematic
situation. The overall situation is captured very well by a user of the SCS system:
…there is a wave through SCANDI now, where you can see that you did
something stupid by outsourcing so much. But it will always be broken,
195
because we will never get the same competences back as we had before.
But I hope it will be better!
ANALYSIS: BEHIND THE CURTAIN
The SCANDI case study is the “scenery” for a complex set of outsourcing engagements
and also shows the consequences of these engagements; but it only goes so far in answering the research question, “why does SCANDI engage in such a complex ES/IS
outsourcing landscape?” as shown in Figure 4.
The decisions to carry through the many outsourcing engagements are taken at the executive level. The frontstage rational explanations and institutional backstage explanations for outsourcing engagements are related to strategic issues more than to specific
applications and specific projects, and this will be demonstrated in this section.
This chapter is divided into four sections consisting of economic, technical, and symbolic-interpretive explanations, inspired by Hatch (1997: 211) and we also include a
summary of the issues discussed.
ECONOMIC EXPLANATIONS
Outsourcing is, in many cases, directly linked to attained cost savings, but often it is
indirectly linked to other factors, like higher flexibility, access to specialist expertise
etc. (Olson 2007). All of the outsourcing engagements mentioned in this case study are
related “to simplifying and obtaining effectiveness in business operation” (Press Release #4, 2008), and are proxies for cost savings (Williamson 1981). This opinion is
supported by several interviews, where the answer to the question “why outsource?” is
cost savings, although it is often combined with other explanations, like the myth that
outsourced IT-employees will further their careers and develop their competences better
within an IT company than within SCANDI.
Cost savings are also perceived to lead to a healthy and attractive company for the
shareholders. It is suggested that because the current owners (mainly investment funders), by all accounts, want to sell SCANDI in the coming years, the massive outsourcing strategy is part of preparing SCANDI for this sale. Further, this outsourcing strategy
196
is linked to offering financial incentives for the management team as a way of increasing their bonus-based salaries by fulfilling their personal balanced scorecard.
The rational economic explanations for outsourcing are fairly overt when reading press
releases and interviewing people, yet institutional explanations are more covert. One
supply chain manager stated, “my impression is that the management team at SCANDI
is driven by key economic figures and management consultants' analyses”, and this
opinion is supported by another quotation, “I know that management consultants are
working with executive management at the highest level…and very close to the chief
information officer…positioning what to outsource [based on best practices, benchmarks etc.]” (Consultant #2). This means that the management consultants “bring” to
SCANDI analyses with benchmark numbers for companies. This thereby impacts on
outsourcing decisions, whether they are rationally substantiated or not. This is an example of management consultants “travelling with best practices” (adapted from
Czarniawska and Joerges 1996). They can be seen as professionals, enacting normative
institutional pressures on the organizations they encounter (Scott 2008) leading to isomorphism among organizations (DiMaggio and Powell 1983) based on institutionalized
“templates for organizing” (Miranda and Kim 2006).
Another economical related subject that plays an important role in outsourcing decision
is the truce between SCANDI and the trade unions. Here, it was agreed that SCANDI
could not dismiss more than a certain number of employees per year. This was a coercive, institutional pressure (DiMaggio and Powell 1983) constraining the possible actions that the SCANDI’s executive management could take in order to create a healthy
company. The “best” rational decision (Scott and Davis 2007) might have been to dismiss a number of employees, but ES outsourcing, and outsourcing in general, was then
used as the second best instrument to move SCANDI in a certain strategic direction.
TECHNICAL EXPLANATIONS
Most of the technical explanations are frontstage rational explanations, and they are
already presented in the case study section. They are paraphrased as follows: (1) The
ERP-area is not business critical and therefore, is suitable for outsourcing and furthermore, it is controlled internally, so the outsourcing is expected to be low risk. (2) The
197
new contract will make it possible for former SCANDI employees to (better) use their
competences and develop them further. Finally (3), the new agreement will enable
SCANDI to further modernize and consolidate applications required by business.
The first account is related to understanding that the ERP-area is an easy and obvious
area for outsourcing, while the latter two last could be characterized as “standard reasons for outsourcing”, as mentioned in some papers (McFarlan and DeLacey 2004; Olson 2007). It is difficult to evaluate the three accounts. Are they believed by the executive management team? Or are they myths to fertilize the ES/IS outsourcing process?
This study can neither confirm nor deny the accounts, but at least one of the accounts
(about developing outsourced employees’ competencies) is challenged by a supply
chain manager: “They [SCANDI executive management] tell in the organization that
the [outsourced] people will develop their competencies better in a definite IT company
than in SCANDI…but these people have not got any further knowledge...they just work
for another company”.
When we consider the multisourcing arrangements at SCANDI, we argue that the practice of outsourcing was becoming institutionalized at SCANDI, which meant that it “became part of the furniture” (adapted from Silva and Backhouse 1997) and taken for
granted. This implies that we go from a rational-choice situation to a non-choice situation, or at least there was a narrowing of the scope of choices (Berger and Luckmann
1966). One of the benefits of institutionalized habits (in this case outsourcing) is that it
reduces efforts in decision making and frees energy for other purposes in an organization (Silva and Backhouse 1997). Tangible artefacts of this “outsourcing institutionalization” can be observed by analysing press releases and news articles over a long period
of time (2003-2008 and including more events than presented in this paper). From this
analysis, it can be seen that outsourcing practice is repeated over and over; and, as one
of the interviewees sarcastically commented, “SCANDI is using the same template for
outsourcing press releases”, which indicates a habit with a routine approach to outsourcing.
198
SYMBOLIC-INTERPRETIVE EXPLANATIONS
The culture of using IS/IT at SCANDI is driven by dissimilarity, where each department
is different and has its own requirements and needs to be fulfilled by different information systems. One of the reasons for these many interfaces is that many departments
have their own decentralized financial systems, and management has accepted this “culture of dissimilarity”. This causes a complex technical infrastructure, which is not compatible with the IS strategy that was launched in 2007, to reduce the number of applications and to aim at packaged applications systems with minimal customizations. ES/IS
outsourcing is then used to change the culture. However, the culture is difficult to
change as it is highly institutionalized, as some participants commented: “I think
SCANDI tries to break up the structures by virtue of outsourcing, and that might be
possible in the long run” (Consultant #1); and “Outsourcing could be a good instrument
if you want to change a culture which is so highly institutionalized, and driven by [the
fact] that we are all different” (Consultant #2).
Earlier, we argued that the outsourcing practice at SCANDI had become institutionalized, and that this did not arise out of a vacuum, but was created by real actors over time
and was an historical process (Weber and Glynn 2006). From our analysis, there seems
to be at least two key actors at SCANDI doing this institutional work (Lawrence and
Suddaby 2006). This is what two interviewees said about the two key actors, first about
the Executive Business Manager (EBM):
I think…that it gives prestige to undertake outsourcings…the people
who got promoted at SCANDI, are the people who have carried through
several outsourcings and cut down the number of employees…let me
give an example: [EBM] was responsible for outsourcing of IT development [O1] in 2005, he was promoted to an important line of business,
where he has been head of division for the last couple of years, and
nearly everything has been outsourced in his reign…and he gets a lot of
attention from the executive management.
(Supply Chain Manager)
199
Second, about the Chief Information Officer (CIO):
If I should point towards a central individual who has relatively much to
say…it is [CIO]. He was three years ago a successful head of a purchasing department…and he created an extreme large power base and position, by virtue of good outsourcing deals for SCANDI…and I am quite
sure that the “outsourcing silver bullet” from the purchasing department
also characterizes the governance of the IT department.
(Consultant #2)
So, undertaking outsourcing gives prestige to key actors and creates an extreme power
base, which supports values like power, wealth, and management control (Hatch 1997:
211) and it has encouraged key actors in SCANDI to continue to undertake outsourcing.
It is suggested that the “prestige and promotion factors” associated with outsourcing
have had a contagious influence on other actors, both as an attractive norm (unwritten
rule) and as a script to copy (mimetic institutional mechanism) (Scott 2001; 2008) – in
short, a “silver bullet” or “a recipe for success”.
200
SUMMARY OF EXPLANATIONS
Table 2 summarizes the rational explanations and institutional explanations for carrying
out the outsourcings at SCANDI:
Rational Explanations
“Behind the Curtain” Institutional Explanations
Cost savings in order to elicit a healthy and Management consultants “bring” in
Economic
Explanations attractive company for shareholders – to benchmark numbers from other companies
make SCANDI saleable.
to SCANDI. They “travel in best practices”.
Financial incentive for the management
team to outsource (fulfilling the balanced Agreement with trade union about the
scorecard).
maximum number of dismissals per year.
The ERP-area is not business critical and Outsourcing becomes an institutionalized
Technical
practice i.e. a habitual choice.
Explanations therefore suitable for outsourcing.
The ERP area is in control internally, so
outsourcing does not pose a big risk.
New competencies enabling further improvement of the quality of IS.
Outsourcing enables SCANDI to further
modernize and consolidate its applications
to deliver the capabilities that business
requires.
Outsourcing appears to be a way to force a Outsourcing is seen as prestigious in
SymbolicInterpretive cultural change from a “highly institution- SCANDI. Employees are promoted if they
Explanations alized culture of dissimilarity” to a “more have mastered outsourcing.
agile and streamlined company”.
Outsourcing is a silver bullet – a recipe for
success.
Table 2: Summary of Explanations for ES Outsourcing
DISCUSSION AND IMPLICATIONS
The previous sections have described the theoretical and practical findings in terms of
the SCANDI case. One important implication of these findings is the value of the complementary learning from applying not only a classic rational perspective, but also an
institutional perspective. Both perspectives provide a different, but still useful domain
of learning. The two perspectives have progressed not only in terms of new concepts,
but also by shifting the domain of learning. The old rational model remains useful and
important in developing an understanding of the case. Newer institutional work has not
201
refined or supplanted the older work, but rather augmented and extended it. Thus, our
first contribution to research is the value of multiple perspectives and multiple-theory
analysis in studying sourcing decisions. This, however, it is not a new idea. In fact Lacity and Hirschheim (1993) did something like it, and Baskerville and Pries-Heje (2001)
used it in a study of innovation in a software company.
One can reasonably ask whether the multi-perspective analysis is valuable only to researchers, and it is suggested that this is not the case. If you are a manager in an organization that is contemplating outsourcing, you should always consider; are there other
reasons, which favour outsourcing that are hidden “behind the curtain”? If you are a
manager in SCANDI (not the one making the argument) you could use these findings as
a checklist for identifying institutional explanations. This process would open the curtain, and help you make decisions that are based on your own knowledge and information. Please see Table 2 for an example; right column, from the top: Do you have management consultants helping the company? Are they bringing in benchmark numbers
and best practices? How do they know that these numbers and practices are not based
on rational views only – so the people answering the survey behind the benchmark may
have answered what they expected the interviewer wanted to hear? Expressed another
way; the curtains may have been closed when the benchmark was made. Or (following
the right column in Table 2), has outsourcing just become the easy, habitual choice? Is it
seen as prestigious within the organization? Are people involved in promoting outsourcing? If the answers to the latter questions are “yes”, you may consider whether you
should join the stream of events yourself, thereby following a “garbage-can” view on
decision making (Cohen et al. 1972).
There is another contribution to practice: For example, an interesting distinction between cost savings and cost spending. The rational argument of cost saving as a determinant for multiple outsourcings is very persistent, but does it really create cost savings? If we take a transaction cost perspective, it is fairly obvious that the coordination
cost among the many inter-organizational partners is considerable (also designated coordination complexity). Then add to this the cost of the several transitions, we end up
with a large total of coordination costs, which have to be compensated for by lower
production costs in order to have optimal economic efficiency – and there is no warrant
202
for this claim. The cost for the RE-ES project increased from 1.6 million € to
8 million € during the project course. This is a large increase, even for an IS/IT project
covering both production and coordination costs. Thus, it makes sense to suppose that
the coordination cost would have been lower if there had been fewer outsourcing vendors, or there had been a total in-sourced project. This argument is supported by several
informal discussions with project members, and, as one participant said, the first implementation of SCS in 2001/2002 was a bigger challenge but was undertaken with a
core project group of 15 people, which is less than the number people involved in this
project. This is further warranted by industry analysts, who suggest that multi-sourcing
has high transitioning costs and requires two to three times more oversight than working
with a single partner (Overby 2010).
It is thus relevant to consider both rational and institutional explanations in a given
ES/IS outsourcing case, and not just to “run away” with obvious frontstage explanations
without considering behind the curtain, backstage institutional explanations. This finding could also be used to understand one’s own behaviour in similar situations and especially other peoples’ behaviours.
CONCLUSION
Outsourcing arrangements are often quite complex and it is feasible to ask why companies engage in such complex outsourcing arrangements. We analysed documents, observed meetings, and gathered data from interviews in four parts of SCANDI. Our first
round of analysis identified rational cost saving explanations only. However, after a
second round of analysis, using the Institutional Theory, we found other “behind the
curtain” explanations, such as management consultants with a “best practice” agenda,
people promoting outsourcing so being promoted themselves, and a belief in outsourcing as a “silver bullet” or a recipe for success that solves everything.
Other researchers have used an institutional perspective to obtain insights into outsourcing. In contrast to these (primarily macro) studies, our research also applied a micro
view on outsourcing. This gave us significant new insights; we found institutional explanations, such as trade union agreements and benchmarking across organizations at
the macro level, as well as career-motivated thinking and habits in play at the micro
203
level. Another significant insight from our study is that the actors involved – managers,
consultants, people being outsourced –would potentially benefit from “opening the curtains” and looking behind them for institutional explanations that are hidden by convenient rational platitudes.
204
REFERENCES
Ang, S. and L. Cummings (1997). Strategic Response to Institutional Influences on Information Systems Outsourcing. Organization Science 8(3): 235-256.
Baskerville, R. and J. Pries-Heje (2001). A Multiple-theory Analysis of a Diffusion of
Information Technology Case. Information Systems Journal 11(3): 181-212.
Bazeley, P. (2007). Qualitative Data Analysis with NVivo. London: Sage Publications
Ltd.
Berger, P. L. and T. Luckmann (1966). The Social Construction of Reality. A Treatise in
the Sociology of Knowledge. New York: Doubleday.
Bridgman, T. and H. Willmott (2006). Institutions and Technology: Frameworks for
Understanding Organizational Change-The Case of a Major ICT Outsourcing
Contract. The Journal of Applied Behavioral Science 42(1): 110-126.
Chae, B. and G. F. Lanzara (2006). Self-destructive Dynamics in Large-scale Technochange and Some Ways of Counteracting it. Information Technology & People
19(1): 74-97.
Cohen, M. D., J. G. March and J. P. Olson (1972). A Garbage Can Model of Organizational Choice. Administrative Science Quarterly 17(1): 1-25.
Czarniawska, B. and B. Joerges (1996). Travels of Ideas. In B. Czarniawska and G.
Sevón (Eds.), Translating Organizational Change. Berlin: Walter de Gruyter,
13-48.
Dibbern, J., T. Goles, R. Hirschheim and B. Jayatilaka (2004). Information Systems
Outsourcing: A Survey and Analysis of the Literature. Database for Advances in
Information Systems 35(4): 6-102.
DiMaggio, P. J. and W. W. Powell (1983). The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields. American Sociological Review 48(2): 147-160.
Goetz, J. P. and M. D. LeCompte (1981). Ethnographic Research and the Problem of
Data Reduction. Anthropology & Education Quarterly 12(1): 51-70.
Hatch, M. J. (1997). Organization Theory. Modern, Symbolic, and Postmodern perspectives. Oxford: Oxford University Press.
Jayatilaka, B. (2006). IT Sourcing a Dynamic Phenomena: Forming an Institutional
Theory Perspective. In R. Hirschheim, A. Heinzl and J. Dibbern (Eds.), Information Systems Outsourcing: Enduring Themes, New Perspectives and Global
Challenges. Berlin: Springer-Verlag, 103-134.
205
Kshetri, N. (2007). Institutional Factors affecting Offshore Business Process and Information Technology Outsourcing. Journal of International Management 13(1):
38-56.
Lacity, M. C. and R. A. Hirschheim (1993). Information Systems Outsourcing; Myths,
Metaphors, and Realities. New York, USA: John Wiley & Sons Inc.
Lawrence, T. B. and R. Suddaby (2006). Institutions and Institutional Work. In S.
Clegg, C. Hardy, T. Lawrence and W. R. Nord (Eds.), The Sage Handbook of
Organization Studies. London: Sage Publications Ltd, 215–254.
Lawrence, T. B., M. I. Winn and P. D. Jennings (2001). The Temporal Dynamics of
Institutionalization. Academy of Management Review 26(4): 624-644.
Loh, L. and N. Venkatraman (1992). Diffusion of Information Technology Outsourcing:
Influence Sources and the Kodak Effect. Information Systems Research 3(4):
334-359.
Markus, M. L. (2004). Technochange Management: Using IT to drive Organizational
Change. Journal of Information Technology 19(1): 4-20.
McFarlan, F. W. and B. DeLacey (2004). Outsourcing IT: The Global Landscape in
2004. Boston: Harvard Business School.
Meyer, J. W. and B. Rowan (1977). Institutionalized Organizations: Formal Structure as
Myth and Ceremony. American Journal of Sociology 83(2): 340-363.
Miles, M. B. and A. M. Huberman (1994). Qualitative Data Analysis: An Expanded
Sourcebook. Thousand Oaks: Sage Publications Inc.
Miranda, S. M. and C. B. Kavan (2005). Moments of Governance in IS Outsourcing:
Conceptualizing Effects of Contracts on Value Capture and Creation. Journal of
Information Technology 20(3): 152-169.
Miranda, S. M. and Y.-M. Kim (2006). Professional versus Political Contexts: Institutional Mitigation and the Transaction Cost Heuristic in Information Systems
Outsourcing. MIS Quarterly 30(3): 725-753.
Myers, M. D. (2009). Qualitative Research in Business & Management. London: Sage
Publications.
Myers, M. D. and D. Avison (2002). An Introduction to Qualitative Research in Information Systems. In M. D. Myers and D. Avison (Eds.), Qualitative Research in
Information Systems - A Reader. London: Sage Publications, 3-12.
Nagle, A. and A. Maughan (2007). Making Multisourcing Work. Accountancy
140(1372): 42-44.
Olson, D. L. (2007). Evaluation of ERP Outsourcing. Computers and Operations Research 34(12): 3715-3724.
206
Overby, S. (2010). IT Outsourcing: Don't Get Caught in Multi-Sourcing's Costly Trap.
CIO magazine. Retrieved 20th January, 2010, from
http://www.cio.com/article/512867.
Patton, M. Q. (2002). Qualitative Research & Evaluation Methods. Thousand Oaks:
Sage Publications Inc.
Pettigrew, A. M. (1990). Longitudinal Field Research on Change: Theory and Practice.
Organization Science 1(3): 267-292.
Scott, W. R. (2001). Institutions and Organizations. Thousands Oaks: Sage Publications
Scott, W. R. (2008). Institutions and Organizations: Ideas and Interests. Thousands
Oaks: Sage Publications.
Scott, W. R. and G. F. Davis (2007). Organizations and Organizing: Rational, Natural,
and Open System Perspectives. Upper Sadle River: Pearson Prentice Hall.
Seddon, P. B., L. Willcocks and G. Shanks (2003). Introduction: ERP - The Quiet
Revolution? In P. B. Seddon, L. Willcocks and G. Shanks (Eds.), Second-Wave
Enterprise Resource Planning Systems: Implementing for Effectiveness. Cambridge: Cambridge University Press, 1-19.
Shafritz, J. M., J. S. Ott and Y. S. Jang (2005). Classics of Organization Theory. Belmont, CA: Thomson Wadsworth.
Silva, L. and J. Backhouse (1997). Becoming Part of the Furniture: The Institutionalization of Information Systems. In A. S. Lee, J. Liebenau and J. I. DeGross (Eds.),
Information systems and qualitative research. proceedings of the IFIP TC8 WG
8.2 International Conference on Information Systems and Qualitative Research,
31st May-3rd June 1997. Philadelphia, Pennsylvania, USA: Chapman & Hall,
389-414.
Vitharana, P. and R. Dharwadkar (2007). Information Systems Outsourcing: Linking
Transaction Cost and Institutional Theories. Communications of the Association
for Information Systems 20(Article 23): 346-370.
Walsham, G. (2006). Doing Interpretive Research. European Journal of Information
Systems 15(3): 320-330.
Weber, K. and M. A. Glynn (2006). Making Sense with Institutions: Context, Thought
and Action in Karl Weick's Theory. Organization Studies 27(11): 1639–1660.
Williamson, O. E. (1979). Transaction-Cost Economics: The Governance of Contractual
Relations. The Journal of Law and Economics 22(2): 233-261.
Williamson, O. E. (1981). The Economics of Organization: The Transaction Cost Approach. American Journal of Sociology 87(3): 548-577.
207
Williamson, O. E. (1991). Comparative Economic Organization: The Analysis of Discrete Structural Alternatives. Administrative Science Quarterly 36(2): 269-297.
Young, A., D. S. Anderson, K. F. Brant, R. H. Brown, L. R. Cohen, S. Cournoyer, C. D.
Rold, M. Goldman, H. Huntley, V. K. Liu, R. T. Matlus, W. Maurer, B. Pring,
C. Tornbohm, G. Tramacere, J. Longwood, Ian Marriott, D. Blackmore, T.
Singh, C. Dreyfuss and R. Sood (2008). Gartner on Outsourcing, 2008-2009.
Gartner, Document number G00164206.
208
PAPER 3
SVEJVIG, P., AND JENSEN, T.B. “ENTERPRISE SYSTEM
ADAPTATION: A COMBINATION OF INSTITUTIONAL
STRUCTURES AND SENSEMAKING PROCESSES”
Abstract.
In this paper we set out to investigate how an Enterprise System (ES) adaptation
in a Scandinavian high-tech organization, SCANDI, can be understood using a
combination of institutional and sensemaking theory. Institutional theory is useful
in providing an account for the role that the social and historical structures play in
ES adaptations, and sensemaking can help us investigate how organizational
members make sense of and enact ES in their local context. Based on an analytical framework, where we combine institutional theory and sensemaking theory to
provide rich insights into ES adaptation, we show: 1) how changing institutional
structures provide a shifting context for the way users make sense of and enact
ES, 2) how users’ sensemaking processes of the ES are played out in practice, and
3) how sensemaking reinforces institutional structures.
209
3
INTRODUCTION
Enterprise Systems (ES) have been a major trend in both private (Seddon, Willcocks
and Shanks, 2003) and public sectors (Sia and Soh, 2007) over the past decade. They
have been on the market since the beginning of the nineties (Jacobs and Weston, 2007)
as an organizational solution to the growing trend of globalization, mergers, and acquisitions (Chang, Gold and Kettinger, 2003). ES can be defined as large-scale organizational systems composed of people, processes, and information technology (IT) enabling an organization to: 1) automate and integrate business processes, 2) share common
data and practices, and 3) produce and access information in real time. ES cover various
business areas and encompass a large range of products such as enterprise resource
planning (ERP), customer relationship management (CRM), and supply chain management (SCM).
ES often trigger major organizational changes and at the same time introduce high-risk
with a potential high reward (Markus, 2004). Despite many expectations of high rewards by investing in ES the reality often seems to be fairly different, indicating that the
value of ES has been overestimated (Davenport, 1998).
In order to better understand the mechanisms at play when adapting ES in organizations, we argue that higher priority should be given to social and organizational aspects
of ES adaptation (Barley, 1986; Vaast and Walsham, 2005). This suggestion is based on
the understanding that an ES adaptation is influenced and created by those people in
organizations who are going to use the system (Gosain, 2004). Similarly, institutional
structures form the way users make sense of and enact ES in practice. Consequently we
argue that ES emerge from and are subject to institutional structures that set the agenda
for their implementation and use. In the local practice they are acted out by organizational members and this enactment influences back on the institutional structures. We
use the term “ES adaptation” to imply that organization and ES adapt to each other in a
reciprocal way during use (adapted from Henfridsson, 1999).
In order to examine the structures and processes that are at play in ES adaptations we
draw upon institutional theory and sensemaking theory. Institutional theory is useful in
providing an account for the role of larger social and historical structures of ES adapta210
tions, and sensemaking theory is a central analytical perspective for investigating how
organizational members make sense of and enact ES in their local context. We investigate the potential of combining the two theories with a case study in a large Scandinavian high-tech organization, SCANDI, where an ES was introduced and used by the
employees in a financial department.
Next we present the foundations of institutional and sensemaking theories and outline
an analytical framework for combining the two theories when investigating ES adaptation. This is followed by a methodology section and a presentation of the empirical findings. We analyze and discuss the findings and present implications.
INSTITUTIONAL THEORY AND SENSEMAKING FOR
INVESTIGATING ES ADAPTATION
Both sensemaking theory and organizational institutionalism build on similar philosophical traditions inspired by Berger and Luckmann (1966) and Schutz (1967), which supports a combined view. In this section we first outline the foundations of each perspective and then combine them into an analytical framework.
AN INSTITUTIONAL PERSPECTIVE ON ES ADAPTATION
Institutional theory deals with the pervasive influence of institutions on human behavior
including the processes by which structures as e.g. rules, routines and norms guide social behavior and how they are created and adapted over time. An institution can be perceived as a state/result of existing order (e.g. a society, a legal system, or an organization) or as a process of institutionalization and deinstitutionalization through either incremental or radical changes. We talk about institutionalization when actions are repeated and when shared meanings by actors are formed (Scott, 2008)
Scott (2008) presents two dimensions to describe the institutional perspective. One dimension relates to pillars; i.e. regulative, normative, and cognitive. The regulative pillar
represents repeated actions because of explicit rules or laws, for example standardized
work procedures. The normative pillar implies repeated actions due to values and
norms. The cognitive pillar relies on repeated patterns due to a desire to be or look like
other institutions. The three pillars exert institutional pressures on organizations leading
211
to a process of homogenization captured as isomorphism, i.e. the desire to “resemble
other units that face the same set of environmental conditions” (DiMaggio and Powell,
1991). Rational arguments are sometimes stated, also known as rationalized myths that
are part of the institutional context.
The other dimension reflects how institutions are embedded in various kinds of carriers
like symbolic systems, social structures, routines, ideas and artifacts (Scott, 2008), involving structure and action (Giddens, 1984) Carriers like ideas and artifacts move over
time and place, thus being altered, modified and combined with other ideas and artifacts
(Scott, 2008). ES are artifacts carrying institutional logics (Gosain, 2004), which are
adopted in organizations but also customized to align with local requirements.
In IS adaptation research, institutional theory has been applied as an analytical tool to
investigate for example transformation of an enterprise (Avgerou, 2000), implementation and use of CT scanners (Barley, 1986), and diffusion of mobile services (Knutsen
and Lyytinen, 2008).
A SENSEMAKING PERSPECTIVE ON ES ADAPTATION
Sensemaking theory provides a framework for focusing on the relationship between
cognition and action in organizations. Weick defines sensemaking as the “making of
sense” (Weick, 1995) where making refers to the activity of constructing or creating
something and sense refers to meaning. Meaning is created when extracted cues are
related to a specific frame, also known as enactment. For instance, implementing an ES
in an organization may constitute the stimulus that the organizational members try to
place in a frame (the organizational context) through a process in which they attempt to
relate their understandings of the ES to the organizational structures, procedures, role
responsibilities, tasks, etc.
Sensemaking is ongoing and grounded in the identity construction of the individual
members. The need to make sense is intensified in circumstances when organizational
members face new or unexpected situations and when high degree of ambiguity and
uncertainty is present (Weick, Sutcliffe and Obstfeld, 2005). It is a retrospective development of a plausible story to explain why organizational members act the way they do.
The individual interpretations are negotiated through social interaction, where the cog-
212
nitive processes happen within the individual, but the individual’s meaning construction
is reified through social construction processes (Kjærgaard and Vendelø, 2008).
Orlikowski and Gash (1994) suggest that sensemaking theory is a useful lens for studying how people develop particular assumptions, expectations, and knowledge of technology. Similarly, Bansler and Havn (2004) argue that a sensemaking perspective helps
managers to clarify the values, needs, and priorities of users when implementing IS.
Organizational members develop particular assumptions and expectations about the
technology which then shape their actions towards it. It is when actions are repeated and
when shared meanings by actors are formed that we talk about institutionalization.
AN ANALYTICAL FRAMEWORK FOR INVESTIGATION ES ADAPTATION
The focus in institutional theory is primarily on how institutionalized environments
shape organizations, where actors are agents of institutional forces or rationalized myths
(Fligstein, 2001). Less emphasis is on describing how the institutionalized structures are
constructed in practice in the first place, and how social practices are internalized,
changed, or reproduced through actions. Consequently, sensemaking theory is useful as
it focuses on how and why organizational members act and make sense of IS in their
local context. However, sensemaking theory does not explicitly conceptualize the institutional structures within which these sensemaking processes take place. It only refers
to the notion of frame as a way to explain how sensemaking occurs.
213
This calls for a combination of the two theories (see Jensen, Kjærgaard and Svejvig,
2008). Inspired by Weber and Glynn (2006), we introduce the following analytical
framework:
te
on
stru
ctu
re
ec
vid
Pro
c on
t ex
t
Macro: Institutional structures
for
Re
xt
Actors enact ES in practice
Enterprise system
Micro: Sensemaking processes
Time
Figure 1: Enterprise Systems Adaptation
As illustrated in the figure, the institutional structures (macro level) provide context for
the sensemaking processes and thus the enactment of the ES system in practice (micro
level). These processes are ongoing and restructure the institutional context over time
once they are repeated and shared by the organizational members.
RESEARCH METHODOLOGY
In line with our framework, we studied institutional structures and sensemaking
processes that formed the ES adaptation in SCANDI. The methodology adopted was a
contextualized, interpretive one, building on case study research (Pettigrew, 1990; Walsham, 2006). Interpretive research attempts to understand phenomena through the meanings that people assign to them, and access to reality is through social constructions
such as language, consciousness, and shared meanings (Myers and Avison, 2002).
214
The study was designed as longitudinal, starting January 2008 and is still ongoing. A
combination of data collection techniques has been used in different functional areas of
SCANDI by one of the authors. Participant observation served primarily as a way to
understand the context. The richest source of empirical data stems from semi-structured
interviews with five employees in the financial department. Each interview lasted from
50-70 minutes where the accounts clerks and other finance personnel talked about their
experiences with the ES adaptation. All interviews were taped and transcribed verbatim.
They were analyzed by using institutional theory and sensemaking theory as sensitizing
devices.
The aim of our fieldwork was to describe context, content, and process. Fieldwork is a
valuable approach as it gives a more vivid and dynamic picture compared to a historical
reconstruction with post hoc rationalization. However, no large organization starts out
from scratch which means that some events in SCANDI took place before January
2008, prior to the initiation of our research. Therefore such events are necessarily historical reconstructions from archival documents and recollections from interviews.
THE SCANDI CASE STUDY
This section describes the empirical case study and provides a time-ordered process
view of the ES adaptation.
CASE INTRODUCTION
SCANDI (a pseudonym) is a Scandinavian company with more than 10.000 employees.
It belongs to the utility industry segment where it produces and sells high-tech services.
The first company in SCANDI was established in the late 1890s, and the company today is a result of a merger between several companies.
SCANDI, and its predecessors, operated for many years as territorial companies in a
context which was highly regulated, meaning that customers were only allowed to buy
utility services from them (i.e. monopoly). This situation changed in the 1990s when the
Scandinavian countries decided to deregulate and liberalize the utility market. SCANDI
now faced higher competition although only few strong competitors exist today (i.e.
oligopoly).
215
The low level of competition has marked SCANDI’s culture. A consultant describes
SCANDI as: “…a supertanker that does not have all the needed engines to react
promptly enough or all the engines are not started simultaneously…or said more directly you cannot lay off more than a [certain] number of employees and SCANDI has a
huge backlog”. This statement indicates one of the business challenges that SCANDI
faces where a truce between trade unions and SCANDI limits its maneuvering. A recent
press release from SCANDI’s CEO, however, responds to this by stating that: “[CEO]
declines to…comment whether he expects to keep the earlier agreement between
SCANDI…and the employees about the maximum yearly downsizing of 5-7 percent of
the workforce”. At the same time, the company benefits from its many years of monopoly by having a big market share and owing a considerable part of the utility infrastructure in the present market, so the shift from monopoly to competition implies both
opportunities and challenges.
A STORYLINE OF THE ES ADAPTATION
In the mid 1990s SCANDI decided to implement one of the leading enterprise systems
(ES) in the market. This was done to have an up-to-date and common system in the
merged SCANDI.
The financial system (FinSys) was the first to be launched in 1996. FinSys covered financial management including general ledger, account receivables and payables etc.
with around 40 heavy users. Subsequently, the supply chain system (SCS) was introduced and operational from 2002 and used by more than 100 users for purchasing and
inventory management. Both FinSys and SCS were upgraded in 2003.
FinSys and SCS were technically implemented as two separate enterprise systems based
on the same standard. Figure 2 below shows a timeline of the implementation and use of
FinSys and SCS:
216
Year
Financial System
(FinSys)
1996
Go live of FinSys
Supply Chain System
(SCS)
Research
1997
1998
1999
2000
2001
Go live of SCS
2002
2003
2004
Upgrade of ES
Upgrade of ES
Major conversion of legacy system completed
2005
2006
2007
Re-implementation project started
“One Finance and Supply Chain”
Research started at
SCANDI
2008
2009
Go Live of “One Finance and Supply Chain” system
Figure 2: Timeline of the Implementation and Use of FinSys and SCS
The ES has evolved into a customized standard system since the first implementation in
1996. About 400 customizations were made throughout the years of implementation,
leading to difficulties in upgrading the system and locking SCANDI into an old ES architecture – an untenable situation which had to be changed. Consequently, a reimplementation project (RE-ES project) started in the summer 2007 covering both FinSys and SCS functionality with the motto “one finance and supply chain”. The purpose
of the project was primarily to: 1) reduce the number of customizations, 2) modernize
the application architecture and prepare for future functionality, 3) optimize business
processes, 4) utilize standard functionality, and 5) upgrade to latest version of ES. The
RE-ES project was considered a technical re-implementation project with very few ben-
217
efits for business. The launch of RE-ES was delayed several times, but in January 2009
the system finally went live.
The next section focuses on the adaptation of FinSys in the financial department.
ADAPTING THE FINANCIAL SYSTEM (FINSYS)
The financial department is a central function collecting data from several decentralized
financial functions like approval of vendor invoices before payment. Some decentralized financial functions even run their own accounting system interfacing with FinSys.
In 2006, the number of finance transactions amounted to 54 million.
Most of the accounts clerks have worked ten, twenty or more in SCANDI and have used
FinSys since its launch in 1996. There are users of FinSys both inside and outside financial department fulfilling accounting work procedures. Interviews with the primary
users indicate that FinSys is well institutionalized at SCANDI: “We are used to work
with [FinSys] and this is part of one’s workday… it becomes a habit to use the system”
(accounts clerk #1).
The management of SCANDI has accepted that each department has had its specific
requirements and needs fulfilled by various information systems. Consequently many
interfaces to FinSys exist as many departments (including daughter/sister companies)
have their own decentralized financial systems. This has caused a complex technical
infrastructure with more than 40 interfaces to FinSys, different work practices, and diverse interpretations of the system.
A typical work practice has the following steps: c An accounts clerk receives an email
from a feeding system indicating that a transaction file is ready to import, d the file is
pre-checked and reconciled, e the file is imported 5 to 6 times (one per legal company
in SCANDI), f An error list is created which goes back to the feeding system, g erroneous transactions are corrected in a spreadsheet and then sent to the accounts clerk, and
h the erroneous transactions are corrected in FinSys by the accounts clerk.
This work practice is so embedded in SCANDI that if you ask for ideas for changes or
improvements, the answer is that step e could be optimized to one import instead of 5
to 6 imports (accounts clerk #2). The possibility to have one integrated financial system
218
across SCANDI and avoid many of the interfaces is not considered a solution. This is
probably not because the accounts clerks cannot see this as a possibility; rather, they
perceive the structural setup of FinSys as somehow “fixed” and generally accept it as is.
The users seem satisfied with FinSys and even if they believe that there is room for improvement, the situation is described as: “I can’t really immediately give you some examples [of improvement]…we have had [problems] on creations of multiple supplier
records, where the system of course had to be adapted, and some system changes was
made…but that is changed” (accounts clerk #1). The accounts clerks are furthermore
confident with FinSys and know how to correct errors.
The accounts clerks more or less take the system for granted: “…I am very familiar with
[FinSys]” (accounts clerk #1). A new accounts clerk who was employed in 2007 states
that FinSys is easy to get familiar with and especially the “multiple creations of supplier
records” is a clever feature because you only need to access one screen display and fill
in all information needed, and then press the button for the mass creation. This increases
the productivity considerably. There are, however, many fields to access before starting
a payment job, but this is not discussed as a problem since detailed business procedures
support the process and compensate for the missing guidance in FinSys.
FinSys is not updated with new features as the users are awaiting the RE-ES project.
There are few shortcomings, for instance a “notice of customers” that has to be done
manually and it would increase the performance if it was done automatically by FinSys.
This is to be implemented as a feature in the RE-ES project. Response times are generally good and FinSys has a high availability, “but we have had unfortunate incidents
where the system is down especially at month-end closing of accounts where there is a
mad rush on the system and it can be frustrating” (accounts clerk #2).
ANALYSIS AND DISCUSSION
CHANGING INSTITUTIONAL STRUCTURES: FROM MONOPOLY TO
COMPETITION
As described in the case, SCANDI and its predecessors had for many years been public
companies (or semi public concessionary companies) living in a monopoly. The institu219
tional logic was characterized by security of supply, delivering standard outputs, and
following agreed procedures, operating in a governmental regulated environment, enacting coercive institutional pressures (DiMaggio and Powell, 1991). The employees in
SCANDI had stable jobs with reasonable salaries and good working conditions. According to a former employee the organizational culture was marked by pride and loyalty.
The consequence from an IS perspective was that diversity in applications was accepted. Each department had their own specific homegrown application(s) adapted to
their specific work practices. Individual employees were allowed to make a demand on
dedicated functionality without presenting a strong business case – the 400 customizations in the ES were a result of this practice. This was possible due to the monopoly
status where focus on efficiency was less predominant.
Because of the shift from monopoly to competition in the 1990s and a more competitive
market with few strong competitors, SCANDI now had to operate as a private company
with focus on efficiency and profit. Consequently, the reduction of the workforce was
5% from 2000 to 2007 and increased to 15% from 2007 to 2008. The downsizing thus
increased dramatically the last years and this trend seems to continue, impacting the
financial department. This implies that employees live in uncertainty because of comprehensive and radical organizational changes. It shows how the competitive pressures
are becoming stronger than the institutional pressures from a business context (Scott
and Meyer, 1991).
The competitive pressures have had some advantages seen from an IS architectural
point-of-view, because they have fertilized the ground for highly integrated applications
like enterprise systems without many customizations, which is exactly the goal with the
RE-ES project. Consequently, the ES has become the coercive force (DiMaggio and
Powell, 1991), enforcing employees to adapt to its inscribed institutional logics (Gosain,
2004).
These changing institutional structures “provide the context for” the enactment of ES in
the local practices (see figure 1) as described in the next section.
220
ENACTING THE ES IN THE LOCAL PRACTICE: ACCEPTING THE ES
AS PART OF DAILY WORK PRACTICES
In the case description we outlined how the financial department was subject to the institutional structures in which SCANDI existed. The ES adaptation with a common financial system back in 1996 was considered a necessary precondition in a future competitive environment, but FinSys was first adapted with many customizations and interfaces. The much later RE-ES project was based on a rationalized myth of “an efficient
ES” and was created as a response to the highly competitive market situation. This idea
was “travelling” (Czarniawska and Joerges, 1996) into SCANDI as a way to optimize
organizational work procedures, i.e. reduce the number of customizations, optimize
business processes, modernize the architecture, and use standard functionality. The rationale of an effective ES was already in the employees’ mindsets who considered FinSys as “part of one’s workday”. There seemed to be a strong alignment between the
management’s wish for efficiency and the employees’ satisfaction with the ES.
The case also showed evidence of how the employees made sense of their work practices in relation to the ES and how some of their enacted practices reinforced existing
structures. FinSys constituted the stimulus that the accounts clerks tried to place in a
frame (their organizational context) and they continuously tried to relate their understandings of the ES to their existing work processes, role responsibilities, and organizational structures. This had consequences for their actions. The typical work procedures
with steps c to h were neither questioned nor changed as they made immediate sense
as “fixed” and rational procedures. This meant that the users re-enacted existing routines and procedures and thus reinforced the institutional structures.
The accounts clerks were the primary users and they interpreted, created as well as determined the use of the ES in practice (i.e. enacted), which to some extent is related to
pre-established conventions of use and ways of thinking. They bracketed the flow of
cues about the ES continuously available to them. In this way, they ‘discovered’ the
system and made necessary changes to it. However, these changes were only minor, e.g.
changes to the creation of multiple supplier records.
The clerks’ perceptions of and actions towards the system were also grounded in their
identity construction as they continuously engaged in processes of association with the
221
system in relation to their identities where they defined themselves in relation to the
core of their work and their mission as clerks. They considered the ES adaptation an
explicit way to improve the overall productivity in a competitive environment. Their
perceptions of themselves were to be productive and this influenced the way they acted
and reacted to the ES.
RESTRUCTURING THE CONTEXT: INSTITUTIONAL STRUCTURES
ARE REINFORCED
In the two previous sections we have analyzed how the institutional structures of the
market created the ES as a rationalized myth that travelled to the organizational as well
as individual level in SCANDI. The market mechanisms with an increased level of
competition influenced how the accounts clerks perceived and used the ES (i.e. as a way
to increase productivity and become competitive) and this reinforced the structures of a
competitive market. The ES became the coercive force, enforcing employees to adapt to
its inscribed institutional logics.
The “enactment of ES in practice” is an ongoing process which “restructures the context” in which SCANDI is situated (see figure 1). The restructuring can either result in
new institutional structures or in a reinforcement of existing structures. The employees
will continue to relate their interpretations of the ES to their tasks and responsibilities,
and they will act according to what makes sense. Only time will show whether the institutional structures will change or remain reinforced. As this study is still ongoing, we
wish to follow the development of the ES adaptation in SCANDI with a particular focus
on the development of institutional structures and local sensemaking processes. In this
way we will be able to further develop figure 1.
CONCLUSION AND IMPLICATIONS
The purpose of this study has been to investigate how the ES adaptation in SCANDI’s
financial department can be understood using a combination of institutional theory and
sensemaking. To answer the question we conducted a longitudinal interpretive case
study which is still ongoing. Our analysis showed: 1) how changing institutional structures provide a shifting context for the rationale of implementing ES and for the way
222
users make sense of and enact ES, 2) how users’ sensemaking processes of the ES are
played out in practice, and 3) how the local sensemaking processes reinforce the institutional structure.
The practical implications emphasize that implementation and use of ES have to take
the institutional structures and local sensemaking processes into account as they are
providing the context for and use of the specific implementation. An ES project cannot
be planned only by focusing on the project processes but also by adapting and aligning
with the broader context that the project is part of. This is where a combination of institutional and sensemaking theories proves useful.
223
REFERENCES
Avgerou, C. (2000). IT and Organizational Change: An Institutionalist Perspective. Information Technology & People 13(4): 234-234.
Bansler, J. P. and E. Havn (2004). Technology-Use Mediation: Making Sense of Electronic Communication in an Organizational Context. Scandinavian Journal of
Information Systems 16(57): 57-84.
Barley, S. R. (1986). Technology as an Occasion for Structuring: Evidence from Observations of CT Scanners and the Social Order of Radiology Departments. Administrative Science Quarterly 31(1): 78-108.
Berger, P. L. and T. Luckmann (1966). The Social Construction of Reality. A Treatise in
the Sociology of Knowledge. New York: Doubleday.
Chang, K.-c., A. Gold and W. Kettinger (2003). The Extent of Enterprise System Adoption in Companies: A Multiple Theoretical Perspective. AMCIS 2003 Proceedings. Paper 56.
Czarniawska, B. and B. Joerges (1996). Travels of Ideas. In B. Czarniawska and G.
Sevón (Eds.), Translating Organizational Change. Berlin: Walter de Gruyter,
13-48.
Davenport, T. H. (1998). Putting the enterprise into the enterprise system. Harvard
Business Review 76(4): 121-131.
DiMaggio, P. and W. W. Powell (1991). The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality. In W. W. Powell and P. DiMaggio (Eds.), The
New Institutionalism in Organizational Analysis. Chicago: The University of
Chicago Press, 63-82.
Fligstein, N. (2001). Social Skill and the Theory of Fields. Sociological Theory 19(2):
105-125.
Giddens, A. (1984). The Constitution of Society: Outline of the Theory of Structure.
Berkeley: University of California Press.
Gosain, S. (2004). Enterprise Information Systems as Objects and Carriers of Institutional Forces: The New Iron Cage. Journal of the Association for Information
Systems 5(4): 151-182.
Henfridsson, O. (1999). IT-Adaptation as Sensemaking: Inventing New Meaning for
Technology in Organizations. Department of Informatics, Umeå University,
Umeå. PhD Thesis.
Jacobs, F. R. and F. C. T. Weston, Jr. (2007). Enterprise Resource Planning (ERP) - A
brief History. Journal of Operations Management 25(2): 357-357.
224
Jensen, T. B., A. Kjærgaard and P. Svejvig (2008). Two Perspectives on Information
System Adaptation: Using Institutional Theory with Sensemaking, Informatics
Research Group working paper series I-2008-06, Aarhus School of Business,
Aarhus University.
Kjærgaard, A. L. and M. T. Vendelø (2008). What Makes a Reference Discipline? Investigating the Role of Theory Adaptation. Proceedings of the Academy of Management Annual meeting, Anaheim, CA.
Knutsen, L. A. and K. Lyytinen (2008). Messaging Specifications, Properties and Gratifications as Institutions: How Messaging Institutions shaped Wireless Service
Diffusion in Norway and Japan. Information and Organization 18(2): 101-131.
Markus, M. L. (2004). Technochange Management: Using IT to drive Organizational
Change. Journal of Information Technology 19(1): 4-20.
Myers, M. D. and D. Avison (2002). An Introduction to Qualitative Research in Information Systems. In M. D. Myers and D. Avison (Eds.), Qualitative Research in
Information Systems - A Reader. London: Sage Publications, 3-12.
Orlikowski, W. J. and D. C. Gash (1994). Technological Frames: Making Sense of Information Technology in Organizations. ACM Transactions on Information Systems 12(2): 174-201.
Pettigrew, A. M. (1990). Longitudinal Field Research on Change: Theory and Practice.
Organization Science 1(3): 267-292.
Schutz, A. (1967). Phenomenology of the Social World. Evanston, Illinois: Northwestern University Press.
Scott, W. R. (2008). Institutions and Organizations: Ideas and Interests. Thousands
Oaks: Sage Publications.
Scott, W. R. and J. W. Meyer (1991). The Organization of Societal Sectors: Propositions and Early Evidence. In W. W. Powell and P. DiMaggio (Eds.), The New
Institutionalism in Organizational Analysis. Chicago: University of Chicago
Press, 108-140.
Seddon, P. B., L. Willcocks and G. Shanks (2003). Introduction: ERP - The Quiet
Revolution? In P. B. Seddon, L. Willcocks and G. Shanks (Eds.), Second-Wave
Enterprise Resource Planning Systems: Implementing for Effectiveness. Cambridge: Cambridge University Press, 1-19.
Sia, S. K. and C. Soh (2007). An Assessment of Package-Organisation Misalignment:
Institutional and Ontological Structures. European Journal of Information Systems 16(5): 568-583.
Vaast, E. and G. Walsham (2005). Representations and Actions: The Transformation of
Work Practices with IT use. Information and Organization 15(1): 65-89.
225
Walsham, G. (2006). Doing Interpretive Research. European Journal of Information
Systems 15(3): 320-330.
Weber, K. and M. A. Glynn (2006). Making Sense with Institutions: Context, Thought
and Action in Karl Weick's Theory. Organization Studies 27(11): 1639–1660.
Weick, K. E. (1995). Sensemaking in Organizations. Thousand Oaks: Sage Publications
Inc.
Weick, K. E., K. M. Sutcliffe and D. Obstfeld (2005). Organizing and the Process of
Sensemaking. Organization Science 16(4): 409-421.
226
PAPER 4
SVEJVIG, P., AND JENSEN, T.B. " MAKING SENSE OF
ENTERPRISE SYSTEMS IN INSTITUTIONS: A CASE
STUDY OF A WELL-ADAPTED SYSTEM"
Abstract.
Contrary to previous research that provides numerous accounts of failure prone
enterprise systems (ES) adaptations in organizations, empirical data from an ES
adaptation in a Scandinavian high-tech company, SCANDI, shows how the system was highly integrated, accepted by its users, and well-aligned to the work
processes. It is therefore natural to ask: Why is the enterprise system so welladapted in SCANDI and what can we learn from this case study? Building on
concepts from institutional theory and sensemaking theory, we present three sets
of reasons for the well-adapted system: (1) “a rationalized myth” about an efficient ES that will create effective work practices travels from a national to a local
level; (2) a long transition process from “match to current business processes”
towards “match to standard package”; and (3) the users in practice find it easy to
adjust to the ES and reinforce existing structures. We present the lessons learned
from the study such as: first customize then un-customize, be prepared for a long
term adaptation process, and consider the match between the users and the system. We discuss theoretical and practical implications.
227
4
INTRODUCTION
Since the beginning of the nineties, enterprise systems (ES) have been a major trend in
both the private and public sectors as an organizational solution to the growing tendency
of globalization, mergers, and acquisitions (Chang et al. 2003) and as a way to optimize
and improve business operations (Häkkinen and Hilmola 2008). ES often trigger major
organizational changes and at the same time introduce high risk with a potential high
reward (Chae and Lanzara 2006: 100; Markus 2004). Several companies have gained an
important increase in productivity and speed (Häkkinen and Hilmola 2008), while others have experienced failure prone ES implementations (Grabski et al. 2003; Sumner
2003). Still others have highly overestimated the value of ES (Davenport 1998; Robbins-Gioia 2002) and realized that the benefits did not materialize. ES already in use
might prevent future optimizations in the organization due to the rigidity built into the
systems (Hanseth and Braa 1998; Lindley et al. 2008).
The missing organizational fit (Sumner 2003) or lack of alignment between ES and
business (Grabski et al. 2003) have long been recognized as explanations of some of
these problems (Sia and Soh 2007). Studies have paid attention to the fundamental
problem of misalignment by understanding the critical nature of the ES adaptation
process (Hong and Kim 2002; Lucas et al. 1988; Swan et al. 1999; Wei et al. 2005),
including the sources of misfit and misalignment (Soh and Sia 2004). The term ES
adaptation is here used to imply that organization, human actors and ES adapt to each
other in a reciprocal way during design and use (inspired by Henfridsson 2000; Tyre
and Orlikowski 1994).
Contrary to the existing literature on non-adaptation and misalignment, empirical data
from an ES adaptation in a Scandinavian high-tech organization, SCANDI, indicate that
the system is well-adapted in the financial department. The system is highly integrated,
accepted by its users, and well-aligned to the work processes. This triggers our research
question: Why is the ES so well-adapted in the financial department at SCANDI and
what can we learn from this case study? To address this question and to better understand the underlying mechanisms of an ES adaptation process, we argue that higher
priority should be given to social and organizational aspects (Barley 1986; Currie 2009;
228
Vaast and Walsham 2005). We therefore build on the structural model of technology
introduced by Orlikowski (1992) to investigate the institutional logics and the sensemaking processes at play in order to provide rich insights into the ES adaptation. Concepts from institutional theory are useful in providing an account of the role of larger
social and historical structures of ES adaptations that set the agenda for their implementation and use (Gosain 2004). Concepts from sensemaking theory provide a central analytical perspective for investigating how organizational members make sense of and
enact ES in their local context (Orlikowski and Gash 1994).
This paper is structured as follows. In the next section, we describe the dual structural
model of technology adapted from Orlikowski and present three main concepts from
institutional theory and sensemaking theory that we use as sensitizing devices to understand what is at play in ES adaptations. The research methodology behind the study is
then reported followed by a presentation of the empirical findings. We analyze the findings and list the key lessons learned along with a presentation of the implications for IS
research in general and ES adaptations in particular.
THE DUAL STRUCTURAL MODEL OF TECHNOLOGY
Tyre and Orlikowski (1994) argue that understanding IS adaptation is critical as the
operating efficiency achieved by the system depends heavily on users’ modifications of
the system, and the corresponding adaptation of the physical and organizational context.
We apply the structural model of technology by Orlikowski (1992) where: (1) technology is an outcome of creative human design processes embedding interpretive schemes
(rules), facilities, and norms (design process). Human actors appropriate the technology
by assigning shared meaning to the technology and sustain the technology through ongoing maintenance and modifications (use process); (2) technology facilitates and constrains human action, but only to some extent as human agency implies the possibility
for humans to act otherwise; (3) the interaction between the human actors and the technology is influenced by the institutional context (organization), which both facilitates
and constrains the interaction (like the technology itself); and (4) humans’ interaction
with the technology impacts the institutional properties by either reinforcing (more typically) or transforming these (less frequently). There is a reciprocal interaction between
229
human actors, technology, and the organization that reinforces and transforms all three
elements in a continuous manner. This is illustrated in Figure 1 below where we have
adjusted the figure to fit to an ES implementation process (adapted from Brehm and
Markus 2000; Orlikowski 1992):
Figure 1: The dual structural model of technology for enterprise systems
(The numbers on the arrows refer to the structural model of technology)
Figure 1 shows the four processes outlined by Orlikowski, both in the supplier and customer organization. The design process in the supplier organization results in an ES as a
semi-finished product, which stands out as “a complete, though flexible, ready to implement solution” (Soh and Sia 2004: 376) crossing the border to the customer organization. The system is then configured and customized through another design process
depending on the perceived gap between requirements and the functionality provided by
the semi-finished product. The dual design processes followed by the use process are
repeated as new releases of the system are produced (Brehm and Markus 2000).
The ES adaptation process implies that the institutional properties of the organization,
the human actors, and the enterprise system adapt to each other in a reciprocal way during a dual design and use process. To fully understand what goes on in this adaptation
process, however, we must investigate the institutional structures that condition the
adaptation and we need to zoom in on the local sensemaking practices where the tech-
230
nology becomes enacted by the users. Next we introduce concepts that help us focus on
the interplay between structures and practices.
RATIONALIZED MYTHS, ISOMORPHISM AND LOGICS
Three concepts from institutional theory, rationalized myths, isomorphism, and institutional logics (Scott 2004; 2008) are useful when investigating the institutional properties of the organization that are part of the dual structural model (Jensen et al. 2009;
Svejvig and Jensen 2009). Institutional theory deals with the pervasive influence of institutions on human behavior including the processes by which structures, for example,
rules, routines, and norms, guide social behavior (Scott 2004).
Rationalized myths are rational arguments that are used by organizations to “maximize
their legitimacy and increase their resources and survival capabilities” (Meyer and Rowan 1977: 53). Organizations conform to these myths in order to be a “proper” organization (Boxenbaum and Jonsson 2008). Institutionalized products, services, techniques,
regulatory systems, public opinions, professional standards, etc. serve as powerful
myths exerting institutional pressures on organizations in multiple and complex ways.
Rationalized myths and taken-for-granted rules lead to isomorphism (structural similarity), where the formal structures of organizations need to conform to society to obtain
legitimacy (Meyer and Rowan 1977). DiMaggio and Powell (1983) move the focus on
isomorphism from society level to the organizational field level and introduce the concepts of coercive, normative, and cognitive institutional pressures. They argue that these
pressures lead to isomorphism where organizations live in an iron cage. Liang et al.
(2007) postulate that cognitive, coercive, and normative institutional pressures impact
the assimilation of enterprise systems. For instance, the normative pressures in an organizational field, where suppliers, customers, consultants, and professional associations
collectively assess and endorse IS innovations (Swanson and Ramiller 1997), will shape
the implementation and assimilation of enterprise systems by providing institutional
norms that guide top managers (Liang et al. 2007).
There has been much emphasis placed on isomorphism within institutional theory
(Greenwood et al. 2008), but this focus has more recently progressed to address the effects of different, often conflicting, institutional logics on individuals and organizations.
231
We therefore extend the two core constructs with a third construct, institutional logics
that “…shape rational, mindful behaviour, and individual and organisational actors have
some hand in shaping and changing institutional logics” (Thornton and Ocasio 2008:
100). Institutional logics link institutions and actions, and they provide a bridge between
macro structural perspectives (DiMaggio and Powell 1983) and micro process approaches (Zucker 1977). Multiple institutional logics are available for organizations and
individuals (Scott 2008), and the embedded agency in institutional logics presupposes
partial autonomy for individuals and organizations (Thornton and Ocasio 2008).
Although institutional logics imply an embedded agency, we lack a detailed understanding of how individuals, as well as organizations, choose between the available multiple
logics, often contradictory, and then “edit” the roles and scripts (Weber and Glynn
2006) embedded in institutional logics. This is where sensemaking theory serves as an
appropriate approach. We suggest complementing institutional concepts with sensemaking concepts (Weick 1995) to ascertain the interplay between action and interpretation
at the micro-level where meaning is created, thus guiding further action and interpretation.
BRACKETING, ENACTMENT AND IDENTITY
Sensemaking theory (Weick 1995) is useful when examining social aspects of IS adaptations (Jasperson et al. 2005; Vaast and Walsham 2005). As part of a sensemaking
process, people develop certain assumptions and expectations of the technology that
shapes their subsequent actions with it (Orlikowski and Gash 1994). The sensemaking is
intensified when organizational members face new or unexpected situations, e.g., when
new technology is implemented. The new technology normally imposes a high degree
of ambiguity or uncertainty for those who are going to use it as there is no predetermined way to act (Weick et al. 2005). Weick refers to such a situation as a “shock” that
triggers an intensified period of sensemaking (Anderson 2006). We choose to focus on
three constructs from sensemaking theory that are particularly relevant for studying the
micro-level mechanisms at play when investigating IS adaptations in organizations:
bracketing, enactment, and identity.
232
When members of an organization interact with a technology, they single out items
and/or events in order to connect them and make sense of the technology. This is known
as a bracketing process (Weick 1979) in which the users of a given technology identify
specific cues that signify desired preferences and ends. In the process of bracketing, the
cues are related to a specific set of frames that an individual holds and it is the connection of the cues to existing frames that creates meaning. The frames represent certain
institutional logics and we thus see an obvious link to institutional structures. The
bracketing process is ongoing where the technology is contextualized, managed, and
adapted in the specific context of use.
The output of the bracketing process may be the creation of new structures or the reinforcement of existing ones. In this way, the users of a technology create the reality that
they respond to in a process of enactment (Weick 1995). The meaning that the users
create will guide their future actions and attention in the situations they face. Enactment
relates to the human agency aspect as presented by Orlikowski (1992), where human
actors appropriate the technology by assigning shared meaning to the technology and
sustain the technology through ongoing maintenance and modifications.
The third construct, identity, relates to: “who we think we are as organizational actors
(identity) shapes what we enact and how we interpret” (Weick et al. 2005: 416). Introducing new technology in an organization may constitute the stimulus that the users try
to place in a frame. Through this process, users attempt to relate their interpretations of
the technology to the expectations they have of their roles and responsibilities and thus
their identity. The identity thereby forms the sensemaking, but the sensemaking also
informs the identity by confirming or questioning the existing understandings of identity
(i.e., who we are).
In the analysis we will show how the combination of concepts from institutional theory
and sensemaking theory support one another as part of the structural model to explain
what goes on in an ES adaptation process.
233
RESEARCH METHODOLOGY
SETTING AND RESEARCH APPROACH
We studied institutional structures and sensemaking processes that shaped the design,
implementation and use of the enterprise system to obtain in-depth knowledge about the
adaptation process in SCANDI (a pseudonym). SCANDI is a Scandinavian company
with more than 10,000 employees. It belongs to the utility industry segment where it
produces and sells high-tech services. SCANDI performs primary activities such as logistics, operations, sales and marketing, and services, as well as supporting activities
such as administration, human resource management, procurement, information systems
services etc. The first company in SCANDI was established in the late 1890s, and the
company today is the result of a merger between several companies. We chose to focus
on the finance department in SCANDI as our main focus and unit of analysis since the
enterprise system is perceived to be highly adapted in this department. This appears to
be a deviant case (Creswell 2007) compared to the extant literature on ES adaptations
(e.g. Soh and Sia 2004).
The research approach we adopted was a contextualized, interpretive one (Pettigrew
1990; Walsham 2006), building on a case study to answer our research questions (Yin
2003). Interpretive research attempts to understand phenomena through the meanings
that people assign to them (Myers and Avison 2002), and access to reality is through
social constructs such as language, consciousness and shared meanings (Berger and
Luckmann 1966).
DATA COLLECTION
The study was designed as longitudinal, running from January 2008 to the end of 2009.
The first author gathered empirical data based on a combination of data collection techniques as presented in Table 1:
234
Data Collection
Methods
SCANDI
Project Group
Semi-structured 5 interviews
Interviews
Short
unstructured
phone
interviews
3 interviews
SCANDI
Finance
SCANDI
Supply Chain
Procurement
SCANDI
Supply Chain
Requester
Oracle
7 interviews 10 interviews
5 interviews
3 interviews
8 interviews 9 interviews
9 interviews
1 interview
Focus group
interviews
Participant
observation
18 meetings
1 meeting
2 meetings
Document
analysis
Unpublished documents: plans, reports, minutes and presentations; press releases
from SCANDI; Oracle information (e.g. www.oracle.com) news articles; magazine
reports etc.
Table 1: Summary of data collection methods
We made use of the data sources presented in Table 1, i.e., interviews, participant observations and documents to obtain an overall understanding of the customer organization SCANDI and its context. Furthermore we studied Oracle as a supplier organization
mainly through publicly available documents and a few interviews with current and
former employees. The richest source of empirical data stems from interviews with five
employees in SCANDI’s finance department, and was performed over two periods
(from February 2008 to May 2008 and again from January 2009 to June 2009). The 16
interviews varied in length from 15 to 70 minutes where the accounts clerks and other
finance personnel talked about their experiences with the ES adaptation. The type of
engagements with the interviewee ranged from in-depth semi-structured interviews to
short unstructured phone interviews. The interview guide for the semi-structured interviews focused on initiating and stimulating the discussion about institutional structures
and sensemaking processes (see appendix for excerpt of interview guide). All in-depth
interviews were taped and transcribed verbatim, while the short unstructured phone interviews were transcribed from handwritten notes immediately after the phone interview.
The aim of our fieldwork was to describe context, content, and process (Pettigrew 1990)
in order to understand the ES adaptation process by gaining access to ongoing actions
and immediate multiple interpretations at SCANDI. Such fieldwork is a valuable approach as it gives a more vivid and dynamic picture compared with a historical recon235
struction with post hoc rationalization. However, some events in SCANDI took place
before January 2008 and they are of necessity historical reconstructions from documents
and recollections from interviews.
DATA ANALYSIS
The data analysis followed the interpretive tradition (Walsham 2002; 2006) using hermeneutics (Myers 2009), where concepts from institutional theory and sensemaking
were used as sensitizing devices (Patton 2002: 452-462) to support the coding and analysis process, although our approach can be described as being between grounded theory
and direct application of theory to data (see also Noir and Walsham 2007). We constructed 30 codes based on several readings of the interview transcripts. The codes were
combined with information from other sources (e.g., archival data) in an iterative fashion – an excerpt of codes is presented in Table 2 below:
Why is the ES so well-adapted in the finance department at SCANDI
Sources
References Created on
Tree Node
Working Practice w ell established
6
14
Tree Node
Well-defined w ork procedures roles
1
1
26-10-2009 08:07
Tree Node
Well established support organization
4
5
14-08-2009 08:39
Tree Node
We are in the same boat
3
3
13-08-2009 13:38
Tree Node
Time The first FinSys version w as implemented in 1996
2
3
11-08-2009 10:38
Tree Node
The several delays in launch of RE-ES project have increased the quality
1
1
13-08-2009 13:42
Tree Node
Support employees at vendor know the system very w ell
1
1
13-08-2009 13:31
Tree Node
Started w ith many customizations w hich has decreased since
6
9
11-08-2009 10:44
Tree Node
Routines (it has alw ays been like that)
1
1
26-10-2009 08:07
Tree Node
Revisions and regulations (norms, rules, procedures)
1
1
26-10-2009 08:06
Tree Node
Quality of w ork
1
1
26-10-2009 08:06
Tree Node
Perceived improvements after RE-ES launch
5
8
14-08-2009 09:14
13-08-2009 12:31
Table 2: Excerpt of codes from NVivo
NVivo (Bazeley 2007) was used to support the data management and coding process as
shown in Figure 2. A number of themes emerged from this process and these were investigated by the use of theory. The finance employees, who had been interviewed earlier, were invited to a focus group interview (October 2009) where the themes from the
coding process were presented in order to validate our results and to get further explanations related to the research questions. The draft paper was submitted to Oracle to verify
the factual description of Oracle and their products.
236
THE SCANDI CASE STUDY
In this section, we describe the implementation of the enterprise system in SCANDI.
First, we provide some background information about the company and the course of
the ES implementation. Second, we describe the enterprise system being introduced in
the organization, and third, we describe the finance department where the system was
implemented.
THE STORY OF SCANDI’S IMPLEMENTATION OF THE ORACLE EBUSINESS SUITE
SCANDI, and its predecessors, operated for many years as territorial companies in a
highly regulated context, meaning that customers were only allowed to buy utility services from them. This meant a low level of competition for SCANDI which marked its
internal context of structure and culture (Pettigrew 1987). A consultant described
SCANDI as: “…a supertanker that does not have all the needed engines to react
promptly enough or all the engines are not started simultaneously”.
The monopoly situation changed in the 1990s when the Scandinavian countries decided
to deregulate and liberalize the utility market. SCANDI now faced higher competition
although only from a few strong competitors. The company still benefitted from its
many years of monopoly by having a big market share and owning a considerable part
of the utility infrastructure in the market. However, in order to keep up with the competition and to have an up-to-date and common system in the merged SCANDI, the management decided in 1996 to implement Oracle E-business Suite (OEBS), also known as
Oracle Financials (James and Seibert 1999). The financial system (FinSys) was the first
system to be launched in SCANDI. It covered financial management including general
ledger, fixed assets, receivables and payables etc. with around 40 frequent users. Figure
2 shows the timeline for the implementation and use of FinSys, and the connection to
major OEBS releases from Oracle.
237
Year
Financial System
(FinSys)
Oracle
E-business Suite
Research
Oracle Financials Launched in US
1989
1990
1991
1992
International edition
1993
1994
1995
1996
Launch FinSys
1997
1998
1999
Release 11i launched
2000
2001
2002
2003
Upgrade of FinSys
2004
2005
2006
2007
Start of RE-ES project
Release 12 launched
Research period
2008
2009
Launch of RE-ES project
Figure 2: Timeline of implementation and use of Oracle E-business suite
SCANDI also implemented a Supply Chain System (SCS), which is used by more than
100 users for purchasing, inventory and other logistics functionalities and about 3,000
ad hoc users for purchase order requests. Our main focus in this paper is FinSys and the
finance department, but as SCS is part of the complete ES at SCANDI, it is also relevant to mention it briefly. FinSys was integrated with more than 40 other systems which
meant that a fairly complex technical information infrastructure existed. FinSys was
upgraded in 2003 to OEBS Release 11i.
A high number of customizations were made throughout the years of implementation,
leading to difficulties in upgrading the system and locking SCANDI into an old ES architecture. This was an untenable situation which had to be changed, and in 2007 a reimplementation project (RE-ES project) started, covering both FinSys and SCS functio238
nality. The purpose of the project was to: (1) reduce customizations from 400 to 150;
(2) implement standard Release 12 functionality; (3) optimize standard business
processes; (4) improve use of standard functionality; (5) modernize the application architecture; and (6) reduce IT cost by approximately 40%.
The RE-ES project was considered a technical re-implementation project with very few
benefits for the business. It was a one-to-one implementation that did not allow new
functionality although Release 12 offered a number of possibilities. The implementation
approach was midway between a “complete replacement of a legacy system” and “a
technical upgrade” (Parr and Shanks 2000). In January 2009 the system went live and
has been operational since then.
ORACLE E-BUSINESS SUITE (OEBS) AS A MULTI-EVERYTHING
ENTERPRISE SYSTEM
The first financial modules of OEBS were originally developed for the US market, but
an international version became available in 1993 with multicurrency capabilities and
support for sites outside the US. The early international versions of OEBS were a poor
fit to Scandinavian customers and many thus chose to heavily customize OEBS. This
was necessary even when Oracle ensured country-specific functionalities such as local
language and local accounting principles in OEBS. In SCANDI, 400 customizations
were made, however, the initial understanding of an adjustable software package
changed to an awareness of the problems with customizations. In the public debate, arguments in favor of implementing ES were to “avoid customization” and to “redesign
business processes to support the software” (Sumner 1999: 302).
Release 11i in 2000 was a landmark for Oracle as a highly integrated and web architected suite (Oracle 2007). The chairman and CEO of Oracle Corporation stated: “companies should forgo shopping for best-of-breed applications and standardize business
processes the Oracle way”, and the marketing vice president continued: “Oracle isn’t
telling enterprises to stop customizing entirely. Rather, they should simply standardize
on Oracle for common tasks” (Wagner 2001: 12).
Finally the newest release, R12, was launched in the beginning of 2007 (Songini 2007)
being a multi-everything enterprise system (Kholeif et al. 2008) with more than 100
239
integrated modules targeting most industries as well as public and private organizations.
R12 would ensure that the companies could: (1) manage business systems globally; (2)
drive compliant business processes worldwide; and (3) deploy country-specific capabilities to operate anywhere in the world (Oracle 2008a). The underlying philosophy of
R12 was articulated as a global, highly integrated, and flexible standard solution. There
was no need to customize R12 due to its flexibility, which included country-specific
capabilities. An area such as accounting and finance was moving towards a global standardization because of international accounting standards that promulgated, e.g. “International Financial Reporting Standards” (Tsakumis et al. 2009), which were addressed
in Oracle’s latest release (Oracle 2008b).
The message from Oracle was mirrored in the purpose of the RE-ES project in SCANDI, and could be conceptualized as “un-customize customizations” (Beatty and Williams 2006). A finance super-user from SCANDI described the situation thus: “FinSys
has become more and more like the Oracle standard, and we cannot get a Rolls-Royce
any longer…there has been a culture change as more people accept default options.
Administrative systems like [OEBS] have evolved much over the years and hence there
is no great need for customizations, since the standard system fulfills the requirements…there is perhaps also a mutual adaptation between the system [FinSys] and the
organization – you get used to the system”. Also a former Oracle employee said: “The
vast majority who have bought a standard system for the first time adapt the system to
the company, which means you have a totally twisted system, but then the company
starts one or more re-implementations, where the company increasingly adapts to the
standard system”, and this indicated exactly the chain of events in SCANDI.
The management of SCANDI accepted that each department had its specific requirements and needs fulfilled. Consequently, many interfaces to FinSys existed as many
departments (including daughter/sister companies) had their own decentralized financial
systems. This “fits for me culture” was furthermore mirrored in the IT department
where IT developers were brought up to serve their internal customers with local custom-made or heavily tailored applications. This obviously caused a complex technical
infrastructure with more than 40 interfaces to FinSys and different work practices with
240
FinSys as a very central system: “FinSys is the hub of the universe and we import data
from many sources” (FinSys super-user).
FINSYS IN THE FINANCE DEPARTMENT
The finance department is well established and has the central function of collecting
data from several decentralized financial functions, such as approval of supplier invoices before payment. In 2006, the number of financial transactions amounted to 54 million.
Most of the accounts clerks have worked for 10, 20 or more years in SCANDI and are
thus very familiar with the tasks they have to perform: “[We have] a nice, well-defined
job [where] we know exactly what to do and what deadlines to keep” (accounts clerk
#2). Another employee adds: “We work within a highly predefined setting. I’m in
charge of paying the bills and I’m not allowed to set up one. It is rather rigid” (accounts
clerk #3). The accounts clerks are very focused on maintaining high quality levels
which means that they deliver what is expected from them: “The users expect to find the
data the day after and all data are in the system before the end of the month” (accounts
clerk #2). It is important that the customers are satisfied with the job done in the finance
department. In order to maintain high quality levels, the employees have implemented
control mechanisms: “We play a police role with regard to terms of payment…one of
our colleagues controls the work that is accomplished the day before and another person
controls her work… it is like a peer review” (accounts clerk #1).
Furthermore, the employees in the finance department argue that they are a harmonious
group: “Collaboration is very important, as are a good working spirit, as well as trust”
(accounts clerk #1). They help each other out if any problems are encountered when
using the system: “We are good at supporting each other. Nobody is left on his own…
we’re all in the same boat, so to speak” (accounts clerk #3).
A typical work process has the following steps: c An accounts clerk receives an email
from a feeding system indicating that a transaction file is ready to import; d the file is
pre-checked and reconciled; e the file is imported 5 to 6 times (one per legal company
in SCANDI); f An error list is created which goes back to the feeding system; g erroneous transactions are corrected in a spreadsheet and then sent to the accounts clerk; and
241
h the erroneous transactions are corrected in FinSys. These work procedures are so
embedded in SCANDI that if you ask for ideas for changes or improvements, the answer is that step e could be optimized to one import instead of 5 to 6 imports (accounts
clerk #2). The possibility of having one integrated financial system across SCANDI to
avoid the many interfaces is not considered a solution.
The users mention that FinSys “… suits our needs” (accounts clerk #3) and even if they
believe that there is room for improvement, the situation is described as: “I can’t really
immediately give you some examples [of improvement]…we have had [problems]
creating multiple supplier records, where the system of course had to be adapted, and
some system changes were made…but that’s been changed” (accounts clerk #1). Furthermore, they are confident with FinSys and know how to correct errors. The accounts
clerks more or less take the system for granted: “…I am very familiar with [FinSys]”
(accounts clerk #1). A new accounts clerk who was employed in 2007 states that FinSys
is easy to become familiar with and especially the “multiple creations of supplier
records” is an appropriate feature because the users only need to access one screen display and fill in all the information needed, and then press the button for mass creation.
This increases productivity considerably.
The user perception with FinSys was discussed with three accounts clerks before and
after the RE-ES project. There appears to be a minor decrease in user satisfaction after
Release 12 which is common for ES implementations with a stabilization phase. The
accounts clerks present different reasons for the decrease: (1) the shift from local language to English; (2) data problems where manual checks are needed and with other
data related challenges; (3) unstable system where users are thrown off the system. It is
however remarkable that none of the interviewees refer to the dramatic reduction in
customization as a reason for the decreased user satisfaction. The users do not put forward misalignment or misfit problems with FinSys. Rather they argue how easy FinSys
is to use and state that: “…we’re living in an R12 world now” (accounts clerk #2). The
problems that the users mention are considered as small “hiccups” and they may disappear as time goes by.
242
ANALYSIS OF THE CASE: WHY FINSYS IS SO WELL
ADAPTED
Several mechanisms are at play that may provide reasons for why FinSys is so welladapted in the finance department in SCANDI. By building on the dual structural model
of technology, including institutional logics and sensemaking practices, we now present
these mechanisms.
INSTITUTIONAL PROPERTIES: A RATIONALIZED MYTH ABOUT EFFECTIVE WORK PRACTICES AND AN EFFICIENT ENTERPRISE SYSTEM
One way to understand the adaptation process and to provide possible explanations for
the well-adapted ES in SCANDI’s financial department is to look at the changes that
happened in the institutional properties for SCANDI and Oracle (see Figure 1).
The institutional context was for several years marked by SCANDI’s monopoly situation where it was allowed to build or buy systems that would match current business
processes despite increased costs. The monopoly situation changed in the 1990s when
the utility market was deregulated and liberalized. This reflected a new era for SCANDI
with market-orientation and a pressure to be competitive. However, it took many years
for SCANDI to adapt and accept the best practices inscribed in the ES, as the “fits for
me culture” was a strong institutional force that prevailed more than ten years after liberalization, as embodied by the 400 customizations.
During the 1990s, enterprise systems had become a major international trend as a way
to optimize and improve business operations. Furthermore, they were perceived as valuable in a context of globalization and in a situation where the number of mergers and
acquisitions increased. We thus witnessed the creation of a rationalized myth around ES
that was expressed in the public debate and stimulated by consultants, professional associations, etc. (see also Swanson and Ramiller 1997) In SCANDI, FinSys represented
an artifact worth implementing, based on its potential to make the accounting practices
more effective and streamlined. Considerable investments were made to initiate its implementation in SCANDI. Czarniawska and Joerges (1996) designate this as the “travels
243
of ideas,” meaning that the idea of “an efficient ES” travels from place to place, or, as in
this example, from the discourse on the international and national level into this specific
organization, and further on to the financial department and its employees.
OEBS has been valuable for Oracle with many customers worldwide, and Oracle is
ranked as the number two global ES supplier after SAP (D’Aquila et al. 2009). This
means that Oracle is a very strong ES supplier and OEBS can exert a strong coercive
institutional pressure on organizations like SCANDI (see also Vitharana and Dharwadkar 2007: 348-349). Oracle is also one of the important institutional entrepreneurs (Scott
2008) that diffuse and institutionalize the rationalized myth about “the un-customized
efficient ES” materialized by their marketing communication and the ostensible global,
integrated and flexible OEBS software with no need to customize (see also Wagner et
al. 2006).
Related to the changes in the institutional properties, we now analyze the changes that
happened with respect to the enterprise system.
THE ENTERPRISE SYSTEM: “MATCH TO CURRENT BUSINESS
PROCESSES” VERSUS “MATCH TO STANDARD PACKAGE”
The early international versions of OEBS fitted badly to Scandinavian customers, as the
system was originally designed for the US market and thus reflected the US institutional
context. For SCANDI it was therefore necessary and logical to heavily customize FinSys in order to fit the system to its local institutional properties, such as interfaces to
many decentralized financial systems, institutionalized financial practices, and Scandinavian legislation for accounting standards. The semi-finished ES was adapted to the
customer organization where the institutional logic “match to current business
processes” prevailed.
Customizations are problematic for most organizations, when they wish to upgrade the
standard package to a newer version, because it both increases the cost and the duration
of the upgrading process (Beatty and Williams 2006). This is a challenge for suppliers
of standard packages and Oracle therefore sought to overcome the challenge by: (1)
convincing its customers to “standardize business processes the Oracle way”, i.e., make
them adhere to the institutional logic “match to standard package”, and (2) building
244
highly flexible multi-everything standard packages, thereby reducing the need for customizations.
We see here how SCANDI was influenced by two conflicting institutional logics in the
form of “match to current business processes” versus “match to standard package” (Berente 2009; Currie 2009). The former logic was predominant until January 2009 when
the RE-ES project was completed, and the number of customizations for FinSys was
reduced. However, the deinstitutionalization process (Scott 2008) started long ago when
the RE-ES project was initiated.
It is interesting to note that user satisfaction with FinSys decreased after Release 12,
however, the dramatic decrease of customizations was not presented as a problem by
any of the interviewees from SCANDI. A possible explanation could be that the accounts clerks accepted the institutional logic of “match to standard package”, which is
reflected in the comment “we cannot get a Rolls-Royce any longer” (FinSys superuser). According to Orlikowski (1992), technology facilitates or constrains human action, but only to some extent as human agency may imply the possibility for humans to
act otherwise. The question is then, why did the accounts clerks accept matching their
work practices to the standard package represented by FinSys? A possible reason may
be that the logic of “matching to standard package” was well-aligned with how the accounts clerks perceived themselves (i.e., their identity) as good and loyal employees.
The way they conducted their work tasks was highly standardized and structured and
this was well in accordance with the procedures already integrated in FinSys. The cues
that the employees bracketed about FinSys (i.e., to support accounting practices and
deliver high quality work) appeared to be well-aligned with the overall frame (i.e., the
standards for accounting procedures).
Another explanation might be the globalization of accounting standards which is an
aligning factor between the standard package and the accounting practices at SCANDI,
so the two elements converge on each other (isomorphic coercive pressure from standards both on SCANDI and Oracle). Accounting systems were early to be automated
(Gorry and Morton 1971) and have been “part of the furniture” for several decades in
organizations. The same could be said about the accounting systems and the work practices at SCANDI. FinSys has been operational since 1996, serving as a common ac245
counting system in the then newly merged SCANDI, and replacing a number of older
accounting systems.
HUMAN ACTORS: ACCOUNT CLERKS ENACTING FINSYS IN PRACTICE AND REINFORCING EXISTING STRUCTURES
The finance employees have many years of experience at SCANDI, and they have been
part of the institutionalization of FinSys and associated business processes over the last
13 years. They have “grown up” with FinSys as a kind of secondary socialization
(Berger and Luckmann 1966), and this long-term institutionalization and legitimization
process may be the key reason for the well-adapted system today. The long-term
process means that the employees did not really experience the RE-ES project as a new,
abrupt, and unexpected situation (i.e., what Weick calls a shock) where there was no
predetermined way to act. Consequently, no high degree of uncertainty or ambiguity
was encountered by the accounts clerks as is normally the case when new systems are
implemented in organizations.
The accounts clerks were active in forming/enacting the course and the outcome of the
RE-ES implementation project. They were the primary users of FinSys and they interpreted, created, as well as determined the use of the system in practice. However, since
it was a one-to-one implementation, the employees did not have much to say with regard to the functionality of the system. This did not appear to be a problem since their
perceptions and actions in practice seemed to be well-aligned with the idea behind the
system, i.e., to slowly but surely optimize and standardize work procedures. The procedures in the system were thus to a high extent related to pre-established conventions of
use and ways of thinking.
We may find other explanations of a well-adapted ES in the culture of the finance department. The accounts clerks express that they are a harmonious group, and that they
know what to do, how to do it, and what deadlines to meet. This implies that their job
fits well with their personal working style. Accounting practices are marked by regulative institutional forces (Scott 2008) such as accounting legislation, auditors’ requirements etc., and we interpret the degree of freedom for finance personnel to be more restricted than many other occupations due to these institutional structures. It is however
246
reasonable to assume that the finance employees at SCANDI find these structures appropriate, whether it is auditor approval procedures or operating constraints by FinSys,
because they fit with their personality and identity. In this way, the procedures in FinSys “make sense” to the employees as they fit well with their preferred working style
and their self-perception, and their enacted practices thus reinforce existing structures.
In other words, what they bracket as specific cues (i.e., standard procedures and regulative practices in FinSys) are well-aligned with the overall frame of accounting practices
and the overall institutional structures that are present and have existed for a number of
years in SCANDI. The situation can be interpreted as follows: we have typified actors
(i.e., accounts clerks) performing typified actions (i.e., accounting practices from steps
c to h) that are well-adapted in a typified technology (i.e., FinSys). This isomorphism
has been created over a long time due to regulative, normative and cognitive institutional pressures existing in the finance department and its environment (DiMaggio and
Powell 1983; Scott 2008).
There are approximately 40 frequent users of FinSys, which is quite a small user group
compared to, for example, SCANDI’s supply chain system (SCS) with more than 3,000
users. It is obviously easier to implement ES for a small user group than a large user
group and to support few people after the go-live point. This may also be a contributory
factor to the well-adapted system, and a super-user supports this argument by the following statement just after the launch of Release 12: “Finance users can use the system
now after a week’s time, but that is because there are few users who need to learn it, and
help is close to us”. The finance user group is not only a small group, but also a homogenous group as emphasized by the interviewees.
FinSys is perceived as the “hub of universe” and thus interpreted as a very important
system. Such an understanding has a positive impact on users’ perceptions of themselves as finance employees doing an important job. FinSys is an important system seen
from a rational perspective (a modern company cannot survive without a wellfunctioning financial system), but this shared understanding of a very important system
also has a symbolic meaning (Meyer and Rowan 1977), which reinforces the institutional structures in the financial department and FinSys, supporting the symbiosis of
finance employees and the financial system.
247
SUMMARY OF THE FINDINGS: EXPLANATIONS OF A WELLADAPTED SYSTEM
We have used the dual structural model (see Figure 1) to analyze our empirical data
where concepts from institutional theory and sensemaking theory served as sensitizing
devices for the emerging themes presented in this analysis. In Table 3 below, we summarize the findings from the analysis and the explanations of why FinSys is so welladapted in SCANDI’s finance department:
248
Oracle Corporation
(Supplier organization)
SCANDI
(Customer organization)
Institutional
Properties
- SCANDI has experienced a shift from
- Oracle is today a global ES supplier
monopoly to competition, which means
(second after SAP), which implies a
focus on more effective and streamlined
high number of OEBS customers worldwork practices
wide
- SCANDI accepts the rationalized myth
- ES has become an international trend
about “the un-customized ES” because
and thus exerts a strong coercive
of the changed context
institutional pressure on customer
organizations
- Oracle adapts and reinforces the
rationalized myth about “the efficient
un-customized ES” and uses marketing
communication to diffuse the message
and persuade customers
Enterprise
System
- From a US specific system to a “multieverything” global, highly integrated
(globalization process) and flexible
system with no need of customizations
- OEBS is adhering to global standards
(e.g., accounting standards)
- First a highly customized version
adhering to the institutional logic “match
to current business processes”
- Then after the RE-ES project a much less
customized version coming closer to the
institutional logic of “match to standard
package”
- The cues that the account clerks bracket
about FinSys are well-aligned with the
overall frame of accounting practices
Human
actors
- [Not investigated in this study]
- Finance employees belong to a small and
harmonious group
- Finance employees have high seniority
in SCANDI
- Long adaptation process from 1996 to
2009 – no high degree of ambiguity or
uncertainty is experienced
- Work procedures in FinSys fit well with
personal working style and the
professional identity as account clerks
(typified actors with typified actions)
- The enacted practices reinforce existing
structures
- FinSys is perceived as the “hub of
universe”, i.e., a very important system
Table 3: Summary of findings
The findings in Table 3 show the many and complex explanations of the research question “why the ES at SCANDI is so well adapted”. It is thus obvious to ask “what can we
learn from this case study?” which is discussed next.
249
LESSONS LEARNED
As argued in the introduction, several studies have pointed to misfits and misalignment
between system and the organization as the main reasons for unsuccessful ES adaptations (Sia and Soh 2007; Soh and Sia 2004; Sumner 2003). In our case we witnessed a
reciprocal adaptation between the organization, the actors, and the OEBS after several
upgrade projects and the major re-implementation project. In what follows, we point to
three lessons learned: (1) the process of customizing and then un-customizing the system; (2) the long-term institutionalization and legitimization course of events leading to
secondary socialization; and (3) the good match between the enterprise system and users.
LESSON 1: ES ADAPTATION AS A LONG-TERM INSTITUTIONALIZATION AND LEGITIMIZATION PROCESS LEADING TO SECONDARY
SOCIALIZATION
Institutionalization of well-adapted enterprise systems is valuable for organizations,
which means that the system “becomes part of the furniture” and is taken for granted.
Institutionalization happens with the presence of a reciprocal typification of habitualized
actions, e.g., the typified accounting practice from steps c to h. The habits imply that
we go from a choice situation to a non-choice situation or at least it narrows our scope
of choices (Berger and Luckmann 1966) as “we do as we usually do”. One of the benefits of the institutionalized habits is that it reduces efforts in decision making for managers and employees and releases energy for other purposes in an organization (Silva
and Backhouse 1997).
Institutionalized habits are enacted over a long period (Berger and Luckmann 1966),
and the lessons learned from this case show that the well-adapted ES evolved over a
long period, maybe as much as five to ten years. The system was enacted in practice by
the users and it was in this ongoing process that the users made sense of the system with
respect to their work practices and roles as account clerks. The cues that were bracketed
about the enterprise system led to actions that were well-aligned with existing practices
and that to a high extent reinforced existing institutional structures. In this way, the system and the actors were engaged in a reciprocal cultivation process (Avgerou et al.
250
2004). This is comparable to secondary socialization, which is defined as “the internalization of institutional-based sub-worlds” (Berger and Luckmann 1966: 158) where the
accounts clerks internalized accounting practices and information systems (mainly
OEBS) in the finance department. The socialization process requires the acquisition of
role-specific vocabularies, structuring routine interpretations and conduct in institutional
settings. Secondary socialization is a long-term process, e.g., acquiring a second language takes several years where one continues to retranslate into the original language,
and then it slowly becomes possible to manage without retranslation, when the second
language is sufficiently internalized (Berger and Luckmann 1966). We posit that the
same kind of social processes apply when individuals internalize enterprise systems and
take them for granted.
LESSON 2: FIRST CUSTOMIZE THEN UN-CUSTOMIZE THE SYSTEM
The conventional wisdom for implementing ES is to minimize or even avoid customizations, also known as “sticking to vanilla” in consultancy language. The arguments put
forward for this are many, such as “problems with upgrade”, “increased costs”, and
“hamper growth strategy” (Beatty and Williams 2006; Hildebrand 2009; Wenrich and
Ahmad 2009). The rationalized myth about “the un-customized ES as the most efficient
system” has been built up over a long period of time and is the conventional approach to
implementing ES nowadays (see also Wagner et al. 2006).
As illustrated in the case study, SCANDI’s initial approach was to “match to current
business processes” which resulted in many customizations. This was later perceived as
highly problematic and one of the main drivers for the RE-ES project was thus to uncustomize customizations, where SCANDI reversed its approach to “match to standard
package”. However, a considerable amount of customizations still existed, of which
some are unavoidable in the current information infrastructure (e.g., integration with
other systems).
In hindsight, we may argue that SCANDI should have avoided the many customizations
in the first place and followed the conventional approach for implementing ES. The
question is, however, whether FinSys would then have been so well-adapted today? It
may be that the approach of “first customize then un-customize” is a better approach in
251
some ES implementations. It means that we accept that people are used to some current
business processes that are part of the institutionalized practice, and that it takes time
(several years) to deinstitutionalize these processes and institutionalize new ones. In this
process, the organization develops together with the ES in a reciprocal manner. We
cannot argue for any specific degree of “over-customizations”, but emphasize that the
conventional approach for implementing ES at least could be questioned, and thereby
challenge practitioners (and academia) to reconsider this rationalized myth in order to
have a well-adapted system in the long run. We are fully aware that SCANDI’s shift in
institutional properties is particular with respect to moving from monopoly to competition, but SCANDI has been in competition since the second half of the nineties, so this
shift can only be taken into account as part of the explanation.
LESSONS 3: THE GOOD MATCH BETWEEN ENTERPRISE SYSTEM
AND USERS
The findings from our case indicate that a match between the logics of the enterprise
system and the logics of those people who are going to use the system is vital. This is
not new and has been emphasized in much of the existing literature (e.g. Berente 2009;
Gosain 2004; Sia and Soh 2007). What is important is how we can foresee a good match
between user and system (a priori) and how to cultivate for a good match.
Users are not just users in general terms; they are finance users, logistics users, core
users of a system spending most of their working time using a system, or occasionally
users with infrequent use of a system, and there are many other categories of users – the
finance personnel at SCANDI are core users of the finance system. Use of the information system is either voluntary or mandatory, but mandatory for SCANDI finance personnel as they “must use the system to perform their job functions, there are no alternatives to actual use” (Wang and Butler 2006: 449). Furthermore some user categories
such as finance personnel in private organizations might be more receptive to ES compared to, for example, sales people (Ahearne et al. 2007; Speier and Venkatesh 2002),
nurses and doctors in hospitals (Currie and Guah 2007; Jensen and Aanestad 2007) and
faculty support staff in universities (Wagner and Newell 2004), due to different institutional contexts for organizations (see also Scott and Meyer 1991), identity, roles, and
typifications of users (Berger and Luckmann 1966; Weber and Glynn 2006). We there-
252
fore suggest analyzing the institutional context for a given organization and the characteristics of its users in order to consider match (or lack of match) between users and
system. This “fit-analysis” complements the ordinary analysis and design activities related to ES implementations where requirements are elicited, gap analysis between ES
best practices and current practices is made, and tailoring (configurations and/or customizations) is considered to prepare for a suitable ES implementation. Goodhue and
Thompson (1995) discuss how individuals’ performance is impacted upon by IT which
is related to utilization and task-technology fit. The latter can be deconstructed into task
characteristics, technology characteristics, and individual characteristics, and the individual characteristics have some resemblance to user characteristics presented in this
paper although our concept appears to be a more compounded element involving social
norms, habits, socialization, etc. The deconstruction of task-technology fit and institutional context for organization appear to be relevant diagnostic elements to evaluate an a
priori match and to propose/cultivate forward pointing actions although the deinstitutionalization and institutionalization processes (Oliver 1992; Tolbert and Zucker
1996) should not be underestimated (see e.g. the story by Speier (2002)).
IMPLICATIONS AND CONCLUSION
Earlier studies in the IS field have provided several accounts of failure prone ES adaptations mainly because of misalignment between system, users, and organization. The
empirical data from the ES adaptation in SCANDI show that the system was welladapted in the finance department. Consequently, the purpose of this study was to investigate the reasons behind the well-adapted system in order to cultivate a higher degree of
adaptation in future projects.
To address this topic, we used the dual structural model of technology as a theoretical
framework. We showed how concepts from institutional theory and sensemaking theory
allow us to extend the structural model of technology and thus obtain a deeper and more
reciprocal understanding of the interaction between technology, people, and institutional
properties. Both theories are useful as sensitizing devices to provide an understanding of
the mechanisms at play in an ES adaptation process. Whereas concepts from institutional theory provide explanations of the outcomes of institutional pressures and logics on
253
the ES adaptation process, concepts from sensemaking theory direct our attention to
how the system is enacted in practice and related to the overall frame.
The practical implications emphasize that managers ought to take the institutional structures and local sensemaking processes into account as they provide the context for and
specific use of the enterprise system, and furthermore the complicated interaction between customer and supplier organization embedded in different institutional contexts
that both enable and constrain the adaptation process. An ES adaptation cannot be
planned only by focusing on the project processes but also by adapting and aligning
with the broader context of which the project is a part. By highlighting possible explanations of why FinSys was so well-adapted in the finance department, managers in other companies may learn from this experience. The approach of first customizing and
then un-customizing the system, the acknowledgment of who the users are, and the
long-term institutionalization process constitute important lessons learned in this study
that may benefit managers in other companies who are about to implement or upgrade
their enterprise system.
We have conducted a single case study from an idiographic perspective and we should
thus be cautious about generalization. Managers should have their own context in mind
as the findings may differ across contexts. Aspects which have proved to be main explanation mechanisms in this study may have less importance in another context. Other
explanations may also appear if a similar topic was investigated in another company and
we therefore encourage other researchers to study the processes leading to well-adapted
systems.
254
APPENDIX EXCERPT OF EARLY INTERVIEW GUIDE
1. Personal data
a. Name , sex and estimated age
b. Work experience (number of years)
c. Employed at SCANDI (number of years)
d. Worked with Oracle E-Business Suite (OEBS) (number of years)
e. Background such as education, work experience etc.
2. Work tasks at SCANDI
a. What role do you have with OEBS?
b. What tasks do you carry out?
c. Discuss specific work tasks
3. Organizational performance
a. How effectively does OEBS function in your daily work?
b. How much of your working time directly involves the use of OEBS?
c. How satisfied are you with OEBS in your daily work? (on a scale from 1 to
10 where 10 is best – please comment on this)
d. How user-friendly do you find OEBS? (on a scale from 1 to 10 where 10 is
best – please comment on this)
e. How good is the quality of the information in the system? (on a scale from 1
to 10 where 10 is best – please comment on this)
f. How good is the quality of the system? (on a scale from 1 to 10 where 10 is
best – please comment on this)
g. How good is the service of the system? (on a scale from 1 to 10 where 10 is
best – please comment on this)
h. Description of the process and specification of who is talking to whom
4. Institutional structures and sensemaking processes
a. Imagine that a new employee becomes a member of your group – please answer the following questions: (1) What would the new employee notice? (2)
What should the new employee learn? (3) What would the new employee regard as special?
255
b. Interpretation of process: (1) How is the process interpreted? (2) Could you
have acted differently? (3) Would you have acted differently if you were to
decide?
c. What procedures (rules) do you have to follow when you use ES? (1) To
what extent are those procedures formal or informal? (2) How does the company follow up on the use of the procedures? (3) What happens if you do not
follow the procedures?
d. What attitudes, values, norms characterize your work in the department?
e. Do you share these norms in the entire group? Or are there differences?
f. Does ES influence attitudes, norms and values – and how?
g. How would you describe the culture?
h. How would you describe your work situation? (Metaphors, analogies, stories, anecdotes, plays).
i. How would you describe your work with ES? (Metaphors, analogies, stories,
anecdotes, plays).
5. Expectations to the re-implementation project
a. What are your expectations of the new Release 12?
b. How do you think that the new Release 12 will affect your work? (First three
months and in the long run)
6. Open questions
a. Are there other conditions that you would like to mention in connection with
our talk today?
b. Are there questions or topics that you would have liked me to ask or talk
about?
c. Could it be useful for me to discuss with other persons the topics that we
have already discussed?
256
REFERENCES
Ahearne, M., D. E. Hughes and N. Schillewaert (2007). Why Sales Reps should Welcome Information Technology: Measuring the Impact of CRM-based IT on
Sales Effectiveness. International Journal of Research in Marketing 24(4): 336349.
Anderson, M. H. (2006). How Can We Know What We Think Until We See What We
Said?: A Citation and Citation Context Analysis of Karl Weick's The Social
Psychology of Organizing. Organization Studies 27(11): 1675-1692.
Avgerou, C., C. Ciborra and F. Land (2004). The Social Study of Information and
Communication Technology: Innovation, Actors, and Contexts. Oxford: Oxford
University Press.
Barley, S. R. (1986). Technology as an Occasion for Structuring: Evidence from Observations of CT Scanners and the Social Order of Radiology Departments. Administrative Science Quarterly 31(1): 78-108.
Bazeley, P. (2007). Qualitative Data Analysis with NVivo. London: Sage Publications
Ltd.
Beatty, R. C. and C. D. Williams (2006). ERP II: Best Practices for Successfully Implementing an ERP Upgrade. Communications of the ACM 49(3): 105-110.
Berente, N. (2009). Conflicting Institutional Logics and The Loose Coupling of Practice
with NASE's Enterprise Information System. Department of Information Systems, Case Western Reserve University, Cleveland. PhD Thesis.
Berger, P. L. and T. Luckmann (1966). The Social Construction of Reality. A Treatise in
the Sociology of Knowledge. New York: Doubleday.
Boxenbaum, E. and S. Jonsson (2008). Isomorphism, Diffusion and Decoupling. In R.
Greenwood, C. Oliver, K. Sahlin and R. Suddaby (Eds.), The SAGE Handbook
of Organizational Institutionalism. London: Sage Publications, 78-98.
Brehm, L. and M. L. Markus (2000). The Divided Software Life Cycle of ERP Packages. Proceedings of 1st Global Information Technology Management (GITM)
World Conference, Memphis, Tennessee.
Chae, B. and G. F. Lanzara (2006). Self-destructive Dynamics in Large-scale Technochange and Some Ways of Counteracting it. Information Technology & People
19(1): 74-97.
Chang, K.-c., A. Gold and W. Kettinger (2003). The Extent of Enterprise System Adoption in Companies: A Multiple Theoretical Perspective. AMCIS 2003 Proceedings. Paper 56.
257
Creswell, J. W. (2007). Qualitative Inquiry and Research Design. Thousand Oaks: Sage
Publications Inc.
Currie, W. (2009). Contextualising the IT artefact: Towards a Wider Research Agenda
for IS using Institutional Theory. Information Technology & People 22(1): 6377.
Currie, W. L. and M. W. Guah (2007). Conflicting Institutional Logics: A National Programme for IT in the Organisational Field of Healthcare. Journal of Information
Technology 22(3): 235-247.
Czarniawska, B. and B. Joerges (1996). Travels of Ideas. In B. Czarniawska and G.
Sevón (Eds.), Translating Organizational Change. Berlin: Walter de Gruyter,
13-48.
D’Aquila, M., J. Freyermuth, S. Jacobson, J. Shepherd, N. Tohamy, M. N. Rizza, M.
Burkett, D. Aquino and C. Fletcher (2009). The Global Enterprise Application
Market Sizing Report, 2008–2013 AMR Research, Inc, Document number
AMR-R-20495.
Davenport, T. H. (1998). Putting the enterprise into the enterprise system. Harvard
Business Review 76(4): 121-131.
DiMaggio, P. J. and W. W. Powell (1983). The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields. American Sociological Review 48(2): 147-160.
Goodhue, D. L. and R. L. Thompson (1995). Task-Technology Fit and Individual Performance. MIS Quarterly 19(2): 213-236.
Gorry, G. A. and S. Morton (1971). A Framework for Management Information Systems. Sloan Management Review, Fall 13(1): 55-70.
Gosain, S. (2004). Enterprise Information Systems as Objects and Carriers of Institutional Forces: The New Iron Cage. Journal of the Association for Information
Systems 5(4): 151-182.
Grabski, S. V., S. A. Leech and B. Lu (2003). Enterprise System Implementation Risks
and Controls. In P. B. Seddon, L. Willcocks and G. Shanks (Eds.), Second-Wave
Enterprise Resource Planning Systems: Implementing for Effectiveness. Cambridge: Cambridge University Press, 135-156.
Greenwood, R., C. Oliver, K. Sahlin and R. Suddaby (2008). The SAGE Handbook of
Organizational Institutionalism. London: Sage Publications.
Hanseth, O. and K. Braa (1998). Technology as Traitor: Emergent SAP Infrastructure in
a Global Organization. ICIS 1998 Proceedings. Paper 17.
Henfridsson, O. (2000). Ambiguity in IT Adaptation: Making Sense of First Class in a
Social Work Setting. Information Systems Journal 10(2): 87-104.
258
Hildebrand, C. (2009). The Value of Sticking with Vanilla. Profit Online. Retrieved
25th March, 2009, from
http://www.oracle.com/profit/smb/122808_ziegele_qa.html.
Hong, K.-K. and Y.-G. Kim (2002). The Critical Success Factors for ERP Implementation: An Organizational Fit Perspective. Information & Management 40(1): 2540.
Häkkinen, L. and O.-P. Hilmola (2008). ERP Evaluation during the Shakedown Phase:
Lessons from an After-sales Division. Information Systems Journal 18(1): 73100.
James, D. and G. H. Seibert (1999). Oracle Financials Handbook: Planning and Implementing the Oracle Financial Applications Suite. Berkeley: Oracle Press Edition from Osborne.
Jasperson, J., P. E. Carter and R. W. Zmud (2005). A Comprehensive Conceptualization
Of The Post-Adoptive Behaviors Associated With It-Enabled Work Systems.
MIS Quarterly 29(3): 525-557.
Jensen, T. B., A. Kjaergaard and P. Svejvig (2009). Using Institutional Theory with
Sensemaking Theory: A Case Study of Information System Implementation in
Healthcare. Journal of Information Technology 24(4): 343-353.
Jensen, T. B. and M. Aanestad (2007). How Healthcare Professionals "Make Sense" of
an Electronic Patient Record Adoption. Information Systems Management 24(1):
29-43.
Kholeif, A. O., M. G. Abdel-Kader and M. J. Sherer (2008). Enterprise Resource Planning: Implementation and Management Accounting Change in a Transitional
Country. Basingstoke: Palgrave Macmillan.
Liang, H., N. Saraf, H. Qing and X. Yajiong (2007). Assimilation of Enterprise Systems: The Effect of Institutional Pressures and the Mediating Role of Top Management. MIS Quarterly 31(1): 59-87.
Lindley, J. T., S. Topping and L. T. Lindley (2008). The hidden Financial Costs of ERP
Software. Managerial Finance 34(2): 78-90.
Lucas, H. C., E. J. Walton and M. J. Ginzberg (1988). Implementing Packaged Software. MIS Quarterly 12(4): 537-549.
Markus, M. L. (2004). Technochange Management: Using IT to drive Organizational
Change. Journal of Information Technology 19(1): 4-20.
Meyer, J. W. and B. Rowan (1977). Institutionalized Organizations: Formal Structure as
Myth and Ceremony. American Journal of Sociology 83(2): 340-363.
Myers, M. D. (2009). Qualitative Research in Business & Management. London: Sage
Publications.
259
Myers, M. D. and D. Avison (2002). An Introduction to Qualitative Research in Information Systems. In M. D. Myers and D. Avison (Eds.), Qualitative Research in
Information Systems - A Reader. London: Sage Publications, 3-12.
Noir, C. and G. Walsham (2007). The Great Legitimizer. Information Technology &
People 20(4): 313-333.
Oliver, C. (1992). The Antecedents of Deinstitutionalization. Organization Studies
13(4): 563-589.
Oracle (2007). Oracle Timeline. Profit: The Executive’s Guide to Oracle Applications
12(2): 26-33.
Oracle (2008a). Meet the Challenges of Globalization. Oracle E-Business Suite, Brochure, 2008, Document number C16140-01.
Oracle (2008b). Oracle’s Financial Management Solutions: Preparing U.S. Companies
for the Transition to IFRS.
Orlikowski, W. J. (1992). The Duality of Technology: Rethinking the Concept of Technology in Organizations. Organization Science 3(3): 398-427.
Orlikowski, W. J. and D. C. Gash (1994). Technological Frames: Making Sense of Information Technology in Organizations. ACM Transactions on Information Systems (TOIS) 12(2): 174-207.
Parr, A. N. and G. Shanks (2000). A Taxonomy of ERP Implementation Approaches.
Proceedings of the 33rd Hawaii International Conference on System Sciences,
Hawaii.
Patton, M. Q. (2002). Qualitative Research & Evaluation Methods. Thousand Oaks:
Sage Publications Inc.
Pettigrew, A. M. (1987). Context and Action in the Transformation of the Firm. The
Journal of Management Studies 24(6): 649-671.
Pettigrew, A. M. (1990). Longitudinal Field Research on Change: Theory and Practice.
Organization Science 1(3): 267-292.
Robbins-Gioia. (2002). ERP Survey Results Point to Need For Higher Implementation
Success. Robbins-Gioia Press Release. Retrieved 20th August, 2007, from
http://www.robbinsgioia.com/news_events/012802_erp.aspx.
Scott, W. R. (2004). Institutional Theory: Contributing to a Theoretical Research Program. In K. G. Smith and M. A. Hitt (Eds.), Great Minds in Management: The
Process of Theory Development. Oxford: Oxford University Press, 460-485.
Scott, W. R. (2008). Institutions and Organizations: Ideas and Interests. Thousands
Oaks: Sage Publications.
260
Scott, W. R. and J. W. Meyer (1991). The Organization of Societal Sectors: Propositions and Early Evidence. In W. W. Powell and P. DiMaggio (Eds.), The New
Institutionalism in Organizational Analysis. Chicago: University of Chicago
Press, 108-140.
Sia, S. K. and C. Soh (2007). An Assessment of Package-Organisation Misalignment:
Institutional and Ontological Structures. European Journal of Information Systems 16(5): 568-583.
Silva, L. and J. Backhouse (1997). Becoming Part of the Furniture: The Institutionalization of Information Systems. In A. S. Lee, J. Liebenau and J. I. DeGross (Eds.),
Information systems and qualitative research. proceedings of the IFIP TC8 WG
8.2 International Conference on Information Systems and Qualitative Research,
31st May-3rd June 1997. Philadelphia, Pennsylvania, USA: Chapman & Hall,
389-414.
Soh, C. and S. K. Sia (2004). An Institutional Perspective on Sources of ERP PackageOrganisation Misalignments. The Journal of Strategic Information Systems
13(4): 375-397.
Songini, M. L. (2007). Oracle rolls out E-Business Suite 12. Computerworld. Retrieved
1. April, 2009, from
http://www.computerworld.com/action/article.do?command=viewArticleBasic&
articleId=9010084.
Speier, C. and V. Venkatesh (2002). The hidden Minefields in the Adoption of Sales
Force Automation Technologies. Journal of Marketing 66(3): 98-111.
Sumner, M. (1999). Critical Success Factors in Enterprise Wide Information Management Systems Projects. Proceedings of the 1999 ACM SIGCPR conference on
Computer personnel research, New Orleans, Louisiana, United States.
Sumner, M. (2003). Risk Factors in Enterprise-wide/ERP Projects. In P. B. Seddon, L.
Willcocks and G. Shanks (Eds.), Second-Wave Enterprise Resource Planning
Systems: Implementing for Effectiveness. Cambridge: Cambridge University
Press, 157-179.
Svejvig, P. and T. B. Jensen (2009). Enterprise System Adaptation: A Combination of
Institutional Structures and Sensemaking Processes. AMCIS 2009 Proceedings.
Paper 746.
Swan, J., S. Newell and M. Robertson (1999). The Illusion of 'Best Practice' in Information Systems for Operations Management. European Journal of Information
Systems 8(4): 284-293.
Swanson, E. B. and N. C. Ramiller (1997). The Organizing Vision in Information Systems Innovation. Organization Science 8(5): 458-474.
261
Thornton, P. H. and W. Ocasio (2008). Institutional Logics. In R. Greenwood, C. Oliver, K. Sahlin and R. Suddaby (Eds.), The SAGE Handbook of Organizational Institutionalism. London: Sage Publications, 99-129.
Tolbert, P. S. and L. G. Zucker (1996). The Institutionalization of Institutional Theory.
In S. R. Clegg, C. Hardy and W. R. Nord (Eds.), Handbook of Organization
Studies. London: Sage Publications, 169-184.
Tsakumis, G. T., D. R. Campbell Sr. and T. S. Doupnik (2009). IFRS: Beyond the Standards. Journal of Accountancy 207(2): 34-39.
Tyre, M. J. and W. J. Orlikowski (1994). Windows of Opportunity: Temporal Patterns
of Technological Adaptation in Organizations. Organization Science 5(1): 98118.
Vitharana, P. and R. Dharwadkar (2007). Information Systems Outsourcing: Linking
Transaction Cost and Institutional Theories. Communications of the Association
for Information Systems 20(Article 23): 346-370.
Vaast, E. and G. Walsham (2005). Representations and Actions: The Transformation of
Work Practices with IT use. Information and Organization 15(1): 65-89.
Wagner, E. L. and S. Newell (2004). 'Best' for Whom?: The Tension between 'Best
Practice' ERP Packages and Diverse Epistemic Cultures in a University Context.
The Journal of Strategic Information Systems 13(4): 305-328.
Wagner, E. L., S. V. Scott and R. D. Galliers (2006). The Creation of 'Best Practice'
Software: Myth, Reality and Ethics. Information and Organization 16(3): 251275.
Wagner, M. (2001). Customers Warming To Oracle 11i. Internetweek (887): 12.
Walsham, G. (2002). Interpretive Case Studies in IS Research: Nature and Method. In
M. D. Myers and D. Avison (Eds.), Qualitative Research in Information Systems
- A Reader. London: Sage Publications Ltd., 101-113.
Walsham, G. (2006). Doing Interpretive Research. European Journal of Information
Systems 15(3): 320-330.
Wang, W. and J. E. Butler (2006). System Deep Usage in Post-acceptance Stage: A
Literature Review and a New Research Framework. International Journal of
Business Information Systems 1(4): 439-462.
Weber, K. and M. A. Glynn (2006). Making Sense with Institutions: Context, Thought
and Action in Karl Weick's Theory. Organization Studies 27(11): 1639–1660.
Wei, H.-L., E. T. G. Wang and P.-H. Ju (2005). Understanding Misalignment and Cascading Change of ERP Implementation: A Stage View of Process Analysis. European Journal of Information Systems 14(4): 324-334.
262
Weick, K. E. (1979). The Social Psychology of Organizing. New York: McGraw-Hill.
Weick, K. E. (1995). Sensemaking in Organizations. Thousand Oaks: Sage Publications
Inc.
Weick, K. E., K. M. Sutcliffe and D. Obstfeld (2005). Organizing and the Process of
Sensemaking. Organization Science 16(4): 409-421.
Wenrich, K. I. and N. Ahmad (2009). Lessons Learned During a Decade of ERP Experience: A Case Study. International Journal of Enterprise Information Systems
5(1): 55-73.
Yin, R. K. (2003). Case Study Research: Design and Methods. Thousand Oaks, California: Sage Publications.
Zucker, L. G. (1977). The Role of Institutionalization in Cultural Persistence. American
Sociological Review 42(5): 726-744.
263
PAPER 5
SVEJVIG, P., AND CARUGATI, A. “MAKING NEW SYSTEMS IS BREAKING OLD SYSTEMS – A CASE STUDY
ABOUT PRACTICES FOR DEINSTITUTIONALIZING AN
ENTERPRISE SYSTEM”
Abstract.
This paper investigates the process of implementation of new enterprise systems
(ES) as the process of decommissioning of the old system takes place. Decommissioning is a vastly overlooked and oversimplified process in IS research and
the role of the pressures and practices leading to the abandonment of the practices
supported and embedded in the old system is not well understood. To answer the
research question we use institutional theory as a lens to make sense of the case
study of SCANDI, a large Scandinavian high-tech organization, in the process of
implementing a new ES after 13 years’ use of their first ES. By drawing on institutional theory we seek to understand the pressures and practices related to making a new system and breaking the old system in a multi level analysis. Through
the institutional lens we examine intended and unintended practices deployed by
SCANDI to deinstitutionalize their current system and replace it with a new one.
The analysis shows that deinstitutionalization is inseparable from institutionalization, some practices impact both processes, while other practices only apply to
either deinstitutionalization or institutionalization. Comparing with previous experiences found in the literature we conclude that the knowledge of the pertinence of
the practices to either process is vital in the successful implementation of deeply
embedded and pervasive systems like ES.
264
5
INTRODUCTION
It is widely recognized that implementations of Enterprise Systems (ES) are challenging
endeavors with high risk and potentially high rewards (Chae and Lanzara 2006: 100;
Markus 2004). Several companies have gained an important increase in productivity
and speed (Häkkinen and Hilmola 2008), while others have experienced failure prone
ES implementations due to users’ resistance (Grabski et al. 2003), lack of social commitment (Sumner 2003), misalignment between ES and organization (Sia and Soh
2007), and many other reasons. Still others have highly overestimated the value of ES
(Davenport 1998; Robbins-Gioia 2002) and realized that the benefits did not materialize
(Lindley et al. 2008). The study of the implementation process (e.g. Chen and Yang
2009; Grabski et al. 2003; Somers and Nelson 2004; Sumner 2009; Wenrich and Ahmad 2009), and the succeeding stabilization (e.g. Ross et al. 2003) has therefore become
a classic topic in IS research. The process has been framed as an institutionalization
process (Gosain 2004) where the ES is so taken for granted that it becomes “part of the
furniture” as Silva and Backhouse (1997) metaphorically express it. However, although
we usually understand ES implementation as “making new systems”, it is also about
“breaking old systems” (Alvarez 2001), for instance, when an old legacy system is replaced by a standard enterprise system. This breaking of the old system or decommissioning of existing technology and related organizational practice (Oliver 1992) is vastly overlooked in IS research with very few exceptions (Alvarez 2001; 2002; Mattila et
al. 2009; Nicholson and Sahay 2009; Sæbø et al. 2008). Empirical studies of deinstitutionalization are rare; (Maguire and Hardy 2009) and Nicholson and Sahay (2009) argue
that the IS domain is ripe for continuing research into deinstitutionalization to understand change enabled by IS/IT. The consequence is that the influence of the old ES on
the new as well as the effect of deinstitutionalization practices (if any are used) is largely under-theorized. The relevance of filling this gap, at the current time where many
companies are implementing their second wave of ERP systems, encouraged us to take
a closer look at this phenomenon in the large Scandinavian high-tech organization
SCANDI, which re-implemented its ES over a two-year period, with the following research question: What practices can be enacted to deinstitutionalize an existing enterprise system and fertilize the ground for a new one?
265
We draw upon institutional theory (Greenwood et al. 2008; Powell and DiMaggio 1991;
Scott 2008) to analyze the context, content and process (Pettigrew 1987) of the deinstitutionalization practices related to the re-implementation. “Institutional analysis examines how broad social and historical forces, ranging from explicit laws to implicit
cultural understandings, affect and are affected by the actions of individuals and organizations” (Orlikowski and Barley 2001: 153). Inside institutional theory, we focus on
deinstitutionalization by which the legitimacy of an established or institutionalized practice erodes or discontinues (Oliver 1992: 564). We pursue an understanding of the political, functional and social practices leading to the deinstitutionalization of the old system in a multi level perspective (Currie 2009) ranging from organization field down to
individual actor level.
The paper starts therefore from the assumption that we have to deinstitutionalize the old
system (technology, organizational practice) before we can institutionalize the new system. This dependency was recognized many years ago by Lewin (1947) and is supported by newer research in the IS domain (Alvarez 2001; Avgerou 2000).
The rest of the paper is structured as follows. In the next section, we describe institutional processes with particular focus on deinstitutionalization and the connection to
replacing old systems with new systems. The research methodology behind the study is
then reported followed by a presentation of the case study and empirical findings. We
analyze the findings and categorize them into financial, political, functional and social
pressures for deinstitutionalization. We list the key lessons learned along with a presentation of the implications for IS research and ES implementations in particular.
DEINSTITUTIONALIZATION
In this paper we pay attention to institutional processes associated with the change from
old systems to new systems and the outcome. Change (often radical in this context) can
be theorized as consisting of several institutional processes (stages) starting with precipitating jolts initiating the change followed by deinstitutionalization, preinstitutionalization, theorization, diffusion and (re)institutionalization according to Greenwood et al
(2002). Technological changes, market forces and legislation are categories of precipitating jolts or events initiating a change (Scott 2008) where established practices start to
266
erode or discontinue leading to the next stage of deinstitutionalization (discussed in detail below). Preinstitutionalization is the stage where organizations start to innovate and
look for technically viable solutions. In order for new practices to become widely
adopted, they have to be theorized, which means development and specification of abstract categories, which leads to simplified and distilled properties of new practices.
Successful theorization is followed by diffusion where ideas are transported within organization communities. Full institutionalization is achieved when the ideas become
taken-for-granted (Greenwood et al. 2002) and become part of the furniture (Silva and
Backhouse 1997). Greenwood and colleagues’ stages of institutional change address the
organizational field level, while we are bringing them down to the organizational level,
as comparable institutional stages or processes are played out at this level, and this
transfer fits well with the mainstay of institutional theory as a multi-level theory (Currie
2009).
Deinstitutionalization is defined as “the process by which legitimacy of an established
or institutionalized practice erodes or discontinues…the de-legitimation of an established organizational practice or procedure as a result of organizational challenges to or
the failure of organizations to reproduce previously legitimated or taken-for-granted
organizational actions” (Oliver 1992: 564). Below is the deinstitutionalization framework proposed by Oliver (1992):
Figure 1: Deinstitutionalization framework
Oliver (1992) discusses three key antecedents that contribute to deinstitutionalization:
political, functional and social pressures. Political pressures tend to happen as a consequence of the utility or legitimacy of an institutionalized practice being called seriously
into question. Organizational members may mount a performance crisis to delegitimize
267
an organizational practice, because their interest or belief conflicts with the status quo,
the organization faces increased pressures to adopt an innovative practice, and/or there
is a reduction in the dependency on the institutional constituents that have encouraged
or enforced persistent procedural compliance with their expectations. Functional pressures relate to technical or functional considerations that have a propensity to compromise or raise doubts regarding the instrumental value of an institutionalized practice due
to potential innovative pressures or performance problems. The perceived worth of an
institutional practice is vulnerable to technical re-evaluation and reconsideration. This
can occur when institutional constituents withdraw the reward associated with an institutionalized practice, when social and economic criteria of organizational success begin
to conflict notably with each other, and/or when an organization ascertains an increase
in its technical specificity or goal clarity. The third antecedent, social pressures,
represents unintended deinstitutionalization, where the organization is neither a proactive agent of deinstitutionalization nor centrally intent on rejecting the particular institutional practice. Social pressures comprise disruptions to organizations’ historical continuity, increasing normative fragmentation as a by-product of organizational changes,
and/or disaggregation of collective norms and values due to structural changes in the
organization or environment (Oliver 1992). The framework in Figure 1 above shows
that two opposing forces moderate the dissipation where entropy accelerates the dissipation process as a driving force while the organizational inertia decelerates the process as
a resisting force. The outcome of the deinstitutionalization process is erosion or discontinuity of the institutionalized practice (Oliver 1992; see also Lewin 1947).
A number of studies examine deinstitutionalization processes, such as erosion via replacement in the case of classic French cuisine (Rao et al. 2003), gradual abandonment
by Japanese companies of permanent employment (Ahmadjian and Robinson 2001) and
several other studies (see Scott 2008: 196-200 for an overview). These studies analyze
change processes, especially the driving and constraining forces for deinstitutionalization.
DEINSTITUTIONALIZATION IN IS RESEARCH
Deinstitutionalization studies within IS are rare although a few exceptions exist (Alvarez 2001; 2002; Mattila et al. 2009; Nicholson and Sahay 2009; Sæbø et al. 2008).
268
These studies present, however, contrasting conclusions. Sæbø et al. (2008) present a
case study from Tajikistan (former Soviet Republic) with an existing institutional practice stemming from the Soviet Health System, which had to be deinstitutionalized during the implementation of a new Health Management Information System (HMIS).
However, the outcome was not heartening: “while some seeds for deinstitutionalization
had been planted, adequate political pressure could not be created for deinstitutionalization to take place in practice. This also emphasizes, institutional change processes are
painfully slow” (ibid.: 15). The case showed that the restraining forces (or inertia) were
more durable and bigger than the driving forces although the deinstitutionalization
might happen in future. Alvarez (2002) presents on the contrary a successful story: she
examined the role of myths as deinstitutionalization practice in an ES implementation.
The old legacy system was deinstitutionalized by creating a story of “performance crisis” (mounting performance crisis, cf. above), and a myth-making process took place
“constructing the new ES as an integrated system”, which was aligned with the overall
organizational goals of the organization, but the benefit of the integration was not supported by objectively testable facts. The case study by Alvarez showed the deinstitutionalization process of the old legacy system followed by the re-institutionalization
process of the new integrated ES (Greenwood et al. 2002), and that narratives can support both the deinstitutionalization and institutionalization process (see also Hedman
and Borell 2004). The two studies demonstrate the importance of understanding the
practices that lead to the deinstitutionalization, as they are decisive as to whether the
deinstitutionalization will succeed or not.
We have so far isolated the deinstitutionalization process from the institutionalization
process, but that is a too simplified representation in ES implementations, although relevant from an analytical perspective. Avgerou (2000: 235) argues that an IT innovation
can be conceptualized as “a dual process of institutionalization of IT and deinstitutionalization of established organizational structures and practices”. Alvarez
(2001) adds a third component, “deinstitutionalization of existing IT”, and we will finally add a fourth component, “institutionalization of the new organizational structures and
practices”. This is shown on the conceptual model below:
269
Figure 2: Conceptual model for the dual process of deinstitutionalization and (re-)institutionalization
The dual processes of deinstitutionalization and (re-)institutionalization shown in Figure
2 above indicate that replacing an ES with a new or changed ES also involves new or
changed organizational structures and practices. There might be different approaches to
ES implementations moving from the current to the future situation; one is the piecemeal approach where technology (ES) is changed followed by organizational practices,
path 1 (or vice versa, path 2), and another is the concerted approach (path 3) where both
technology and organizational practices are changed at the same time (Robey et al.
2002). In any case, deinstitutionalization has to happen in both approaches in order to
break the old system and cultivate the sealing of the new system, for instance, misalignment problems might be caused by lack of deinstitutionalization of current organizational practices, so employees continue to work with old routines, thereby preventing
harvesting the benefits of the new ES.
The antecedent pressures to deinstitutionalization (cf. Figure 1) come from the organization and the environment, and are thus both internal and external to the organization.
Maguire and Hardy (2009) extend this notion to pressures outside and inside an organizational field, which is defined as “those organizations that, in the aggregate, constitute
270
a recognized area of institutional life: key suppliers, resource and product consumers,
regulatory agencies, and other organizations that produce similar services or products”
(DiMaggio and Powell 1983: 148). This calls for a multi level perspective on deinstitutionalization to analyze driving and constraining forces at different levels ranging from
society, organizational field, and organization to organizational subsystem (Scott 2008).
The HMIS study from Tajikistan (Sæbø et al. 2008) was an example of societal institutional structures preventing the deinstitutionalization of the current system, while the
other study mentioned above (Alvarez 2002) about creating a performance crisis
showed a driving force mounted by actors in the organization.
RESEARCH METHODOLOGY
SETTING AND RESEARCH APPROACH
To
answer
the
research
question,
we
studied
intended
and
unintended
deinstitutionalization practices in SCANDI (a pseudonym) during a re-implementation
project from 2007 to 2009. SCANDI is a Scandinavian company with more than 10,000
employees. It belongs to the utility industry segment where it produces and sells hightech services. The first company in SCANDI was established in the late 1890s, and the
company today is the result of a merger between several companies.
The research approach we adopted was a contextualized, interpretive one (Pettigrew,
1990; Walsham, 2006), building on a case study (Yin 2003). Interpretive research
attempts to understand phenomena through the meanings that people assign to them
(Myers and Avison 2002), and access to reality is through social constructs such as
language, consciousness and shared meanings (Berger and Luckmann 1966). We chose
a multi-level approach to understand deinstitutionalization pressures arising from
different levels as the organizational field, the SCANDI organization, the reimplementation project and finally the purchasing department in SCANDI representing
the local level. The multi-level approach is relevant for obtaining a rich and holistic
understanding of complex social settings (Currie 2009; Jensen et al. 2009).
271
DATA COLLECTION
The study was designed to be longitudinal from the beginning of 2008 to the end of
2009. The first author gathered empirical data based on a combination of data collection
techniques as presented in Table 1:
Data Collection
Methods
SCANDI
Project Group
Semi-structured 5 interviews
Interviews
Short
unstructured
phone
interviews
3 interviews
SCANDI
Finance
Dept.
SCANDI
Purchasing
Dept.
SCANDI
Supply Chain
Requester
Oracle
7 interviews 10 interviews
5 interviews
3 interviews
8 interviews 9 interviews
9 interviews
1 interview
Focus group
interviews
Participant
Observations
18 meetings
1 meeting
2 meetings
Document
analysis
Unpublished documents: plans, reports, minutes and presentations; press releases
from SCANDI; Oracle information (e.g. www.oracle.com) news articles; magazine
reports etc.
Table 1: Summary of data collection methods
We made use of the data sources presented in Table 1, i.e., interviews, participant
observations and documents to obtain an overall understanding of SCANDI and its
context. An important source of empirical data stems from interviews with employees in
the purchasing department and a few other key stakeholders although other interviews
have also contributed to the overall understanding of the case study. The interviews
varied in length from 15 minutes to two hours where purchasers and other stakeholders
talked about the re-implementation project and their work situation in general. The type
of engagements with the interviewees ranged from in-depth semi-structured interviews
to short unstructured phone interviews. The interview guide for the semi-structured
interviews used the DeLone and McLean IS success model (2003) qualitatively to
initiate and stimulate the discussion on institutional processes. The IS success model
was in addition applied to scoring of subjective user perception of ES in the purchasing
department. All in-depth interviews were taped and transcribed verbatim, while the
short unstructured phone interviews were transcribed from handwritten notes
immediately after the phone interview. Document analysis has also been a vital data
272
collection method ranging from internal emails, newsletters, and formal project
documents to public documents such as newspaper articles and industry reports.
The aim of our fieldwork was to describe context, content, and process (Pettigrew 1990)
in order to understand the deinstitutionalization practices by gaining access to ongoing
actions and immediate multiple interpretations at SCANDI. Such fieldwork is a
valuable approach as it gives a more vivid and dynamic picture compared with a
historical reconstruction with post hoc rationalization. However, of course some events
in SCANDI had taken place before January 2008 and they are of necessity historical
reconstructions from documents and recollections from interviews.
DATA ANALYSIS
Data analysis followed the interpretive tradition (Walsham 2002; 2006) using hermeneutics (Myers 2009) where concepts from institutional theory were used as sensitizing
devices to support the coding and analysis process (Patton 2002: 452-462), although our
approach can be described as being between grounded theory and direct application of
theory to data (see also Noir and Walsham 2007). We constructed 31 codes based on
several readings of the transcriptions and analysis of documents in an iterative fashion.
Major themes emerged from the coding and analysis process, and we have chosen to
categorize the themes into two dimensions adapted from theory (Oliver 1992; Scott
2008). The first dimension is the antecedent categories of deinstitutionalization suggested by Oliver (1992) – political, functional and social pressures – however, we have
added a fourth category of financial pressures, as this category emerged during the data
analysis as being important for classifying the empirical data. The second dimension
related to the varying levels already discussed in theory chapter (organizational field,
organization etc.) (Scott 2008). Qualitative data analysis software (NVivo) (Bazeley
2007) was used to support the data management and coding process, especially the first
part of the data analysis process, and theorizing (Weick 1995) took place outside NVivo. We have used the principles of Klein and Myers (1999) for evaluating interpretive
research using hermeneutics, for example, the principles of contextualization and of
abstraction and generalization.
273
THE SCANDI CASE STUDY
This section describes the empirical case study as a time-ordered process (Miles and
Huberman 1994; Newman and Robey 1992). The chapter starts off with SCANDI in a
historical perspective, and this is followed by the account of the implementation and use
of Oracle E-Business Suite (OEBS). We then describe the re-implementation project
taking place from 2007-2009, and present the local perception of the project in the
purchasing department.
THE SCANDI ORGANIZATION IN A HISTORICAL PERSPECTIVE
SCANDI, and its predecessors, operated for many years as territorial companies in a
context which was highly regulated, meaning that customers were only allowed to buy
utility services from them. This monopoly situation changed in the 1990s when the
Scandinavian countries decided to deregulate and liberalize the utility market. SCANDI
now faced higher competition although only from a few strong competitors.
The low level of competition has marked SCANDI’s internal context of structure and
culture (Pettigrew 1987). A consultant describes SCANDI as: “…a supertanker that
does not have all the necessary engines to react promptly enough or all the engines are
not started simultaneously…or said more directly you cannot lay off more than a
[certain] number of employees and SCANDI has a huge backlog”. This statement
indicates one of the business challenges that SCANDI faces where a truce between trade
unions and SCANDI limits its maneuvering. SCANDI’s CEO, however, responds to
this by stating that: “[he] declines to…comment whether he expects to keep the earlier
agreement between SCANDI…and the employees about the maximum yearly
downsizing of 5-7 percent of the workforce” (newspaper article, February 2009). At the
same time the company benefits from its many years of monopoly by having a big
market share and owning a considerable part of the utility infrastructure in the present
market, so the shift from monopoly to competition implies both opportunities and
challenges.
The current situation for SCANDI is that it is operating in a recessive or stagnated
market because the services they sell have reached saturation point, thus limiting the
274
acquisition of new customers, and in addition some of its older services are declining.
The global economic crisis has also hit SCANDI, and new players are entering the
utility industry, so the market conditions are very tough and highly competitive for
SCANDI. This keeps high pressure on SCANDI’s executive management to optimize
the organization and they employ several methods to maintain a healthy company. First,
they are selling shares in associated companies to focus on the Scandinavian home
market and to reduce their debt (newspaper article, November 2009). Second, business
process outsourcing and IT outsourcing is heavily used, because it makes SCANDI
more flexible and is anticipated to reduce costs – at least in the longer term. Finally,
they have been making employees redundant in order to reduce operating costs, as more
than 2,500 employees have left the company during the last couple of years. All these
initiatives obviously create an uncertain and nervous atmosphere for SCANDI
employees, and they are left with the question “is it me next time?” (purchaser #2).
SCANDI has been a traditional, fully fledged company, performing primary activities
such as logistics, operations, sales and marketing, and services, as well as supporting
activities such as administration, human resource management, procurement,
information systems services etc. (Porter and Millar 1985) SCANDI has used IS/IT in
all parts of their business for many years, both as business support systems and
operations support systems (that is IS/IT systems for controlling and managing the
utility infrastructure).
The IS/IT strategy is marked by the market conditions for SCANDI to reduce overall
costs including IT cost, but also by the fact that SCANDI has a wealth of legacy systems, some of them more than 25 years old. A large-scale and ambitious enterprise architecture program (EA program) was launched at the end of 2007 to settle with the old
jumble of more than 450 IT systems connected in all directions, and replace it with a
well-structured three-layered model in a four-year period. A SCANDI executive states
that “the goal is to reduce development time of new products by two thirds and to increase the overall productivity by between 20 and 30 percent” (newspaper article, October 2007). The IS/IT strategy was turned completely upside down with the EA program.
Instead of building IT systems from scratch, matching to SCANDI’s needs, SCANDI
must now match the standard systems that are already on the market (occasionally re-
275
ferred to as packaged software applications (see also Xu and Brinkkemper 2007)),
which means that the employees must abide by the functionality provided by these standard systems. SCANDI is consolidating their IT systems so many old and small systems
will be scrapped and replaced by fewer standard systems. However this “revolutionary”
EA program was abandoned in August 2008 due to the cost being too high, and was
maybe also “close to mission impossible” (independent IT journalist). The program was
replaced by an “evolutionary” case-by-case investment plan, but still aimed at the same
strategic goals such as using standard systems as the “default choice” instead of custom
built systems.
THE STORY OF SCANDI’S IMPLEMENTATION AND USE OF
ORACLE E-BUSINESS SUITE
SCANDI decided in 1996 to implement the Oracle E-Business Suite (OEBS), also
known as Oracle Financials (James and Seibert 1999). Figure 3 shows the timeline for
the implementation and use of the enterprise system at SCANDI:
Year
Financial System
(FinSys)
1996
Launch FinSys
Supply Chain
System (SCS)
Research
1997
1998
1999
2000
2001
Start SCS project
2002
Launch SCS
2003
2004
Upgrade of FinSys
Upgrade of SCS
Major conversion
of legacy system
2005
2006
2007
Start of RE-ES project
Research period
2008
2009
Launch of RE-ES project
Figure 3: Enterprise System Implementations at SCANDI
276
The financial system (FinSys) was the first part of the enterprise system to be launched
in SCANDI in 1996 based on one of the earliest international versions of OEBS. This
was done to have an up-to-date and common system in the merged SCANDI. FinSys
covered financial management including general ledger, fixed assets, receivables and
payables etc. with around 40 frequent users. Discussions with finance users indicate that
FinSys is well institutionalized at SCANDI as one user explains: “We are used to
working with [FinSys] and this is part of one’s working day… it becomes a habit to use
the system” (accounts clerk).
SCANDI took the decision in 1999 to enhance their use of packaged software to include
the supply chain. According to the IT development manager: “the development in the IT
domain is so fast that we have to buy standard software packages, and then highly tailor
our business processes to the package. We will therefore not prepare a detailed
requirement specification, but instead cooperate with Oracle, where we have confidence
that they will contribute to develop the right solution” (newspaper article, December
1999). Subsequently the Supply Chain System (SCS) was introduced and operational
from 2002 and used by more than 100 users for purchasing, inventory and other
logistics functionalities. Another 3,000 employees used SCS on an ad hoc basis much
like an online shopping cart solution to register purchase order requisitions.
Both FinSys and SCS were upgraded in 2003 to OEBS Release 11i. FinSys and SCS
were technically implemented as two separate enterprise systems based on the same
standard software package OEBS. A major extension to SCS was implemented in spring
2004, and this release was operational for several years. The SCS seems to be less institutionalized as compared to FinSys, and some users express concerns about the user
friendliness of SCS, e.g., “there are double business processes for creation of supplier
records” (purchaser #1), so you have to key in the same information twice, which is
neither user friendly nor efficient.
The period from the major extension of SCS in spring 2004 and until summer 2007
could be characterized as a period mainly of operation with only a few IT development
activities like bug fixing and implementation of minor enhancements.
277
FinSys and SCS were integrated with more than 40 other systems via middleware
integration or simple batch-oriented integration, which meant that they were part of a
fairly complex technical information infrastructure.
THE RE-IMPLEMENTATION OF THE ENTERPRISE SYSTEM (20072009)
The enterprise system (FinSys and SCS) evolved into a customized standard system
since its first implementation in 1996. About 400 customizations were made throughout
the years of implementation, leading to difficulties in upgrading the system even for
small bug-fixing patches. Furthermore Release 11i was not supported by Oracle from
July 2006, which meant that qualified support was not available and bugs could not be
fixed by patches. SCANDI was thus locked into an old ES architecture.
This was an untenable situation which had to be changed. Consequently, a reimplementation project (RE-ES project) started in the summer of 2007 covering both
FinSys and SCS functionality. One important objective of the project was to have “one
finance and supply chain” supporting a strategy of a coherent ES platform merging the
two technical independent ES platforms into one platform with FinSys and SCS
functionality. Another objective was to support SCANDI’s strategic goal of “one
company”, where SCANDI wanted to reduce from eight legal entities into two legal
entities (in any case, designated “one company”).
The purpose of the project was to: (1) reduce the number of customizations from 400 to
about 150; (2) implement standard Release 12 functionality; (3) optimize standard
business processes; (4) improve use of standard functionality; (5) modernize the
application architecture; and (6) reduce IT cost by approximately 40% (this last point is
the financial justification for the project). The project manager explains in a project
newsletter why the RE-ES project was started (excerpt from May 2008):
One Finance and Supply Chain – Re-Implementation Project: One big step towards a
common Scandinavian ERP platform
…
After deployment of FinSys in 1996 and SCS in 2002, there have been so many customizations to
the applications [ES] that they are no longer compliant with IT strategy and are not supported by
Oracle. This has meant that SCANDI has a very large upgrade backlog which inhibits the
development of purchasing, inventory management and finance functions. SCANDI have also a
very limited return of the annual support from Oracle, as they can no longer use Oracle’s standard
278
patches. One example is that SCANDI did not even have the ability to fix bugs in the application
that Oracle addressed long ago. You can compare it with a tree that has branched out in all sorts of
directions. The many specific crossed branches over time simply led to “the SCANDI Oracle
system” moving so far away from the trunk that the structure today is complex and inefficient. The
time for “customization” is over, and FinSys / SCS applications will be re-engineered back to
standard solutions that can be used across SCANDI. With an IT strategy that will create a coherent,
cost efficient and simple IT architecture across SCANDI and a “One Company Strategy” which
consolidates legal entities of SCANDI, the RE-ES project is not just an upgrade of FinSys / SCS,
but a large step towards a common Scandinavian ERP platform.
…
The newsletter gives a good insight into what is communicated about the project and the
re-implementation of the ES. The statement about reducing customizations is a
persistent message repeated both in RE-ES project and the EA program. The RE-ES
project was considered a technical re-implementation project with very few benefits for
the business. The implementation approach of the RE-ES project was midway between
a “complete replacement of a legacy system” and “a technical upgrade” (see also Parr
and Shanks 2000). The data model was heavily reengineered where, for example,
suppliers were moved from their own data tables to the Trading Community
Architecture data table, which complicated the data conversion process (Swanton 2008).
The RE-ES project was swallowed up by the EA program in June 2008, because the
project fits well with the EA program (standard system, reduce customizations etc.).
The RE-ES project was among the very few projects that survived when the EA
program was abandoned – for instance, the flagship project “CRM” was closed down
with few tangible results. A political angle to the course is that the “RE-ES project
became subject to the EA program, as it provides an opportunity for the EA program to
deliver something [tangible business results] and the RE-ES project was an obvious
candidate” (consultant).
The RE-ES project was very turbulent due to several shifts in outsourcing partners and
higher complexity in the project than expected, which caused several delays but in
January 2009 the system finally went live after eight months’ delay.
The post implementation phase has been a struggling phase, and a logistics manager
expresses the situation like this: “We have outsourced to [outsourcing vendor] who does
not know our business, and the transfer process has not been managed very well, and
279
the necessary agreements have not been settled…we are handling a lot of manual transactions between SCS and the Warehouse Management System, and this takes a lot of
time” (logistics manager #1). Anyway, the re-implemented ES has slowly stabilized
since the launch in January 2009 and several interviews with users in the finance, logistics and purchasing departments indicate that they perceive the re-implemented ES to be
“in normal operation” in the autumn of 2009, which means that the system is recognized
as being stabilized (Silva and Backhouse 1997) and routinized (Cooper and Zmud
1990), although it still has some bugs and open issues.
LOCAL PERCEPTION OF THE ENTERPRISE SYSTEM IN THE PURCHASING DEPARTMENT BEFORE AND AFTER THE LAUNCH OF RELEASE 12
The purchasing department is a central group in SCANDI taking care of creating supplier records, preparing purchase orders, managing supplier agreements, and other purchasing tasks. The central purchasing department has existed for about five years, and
the purchasers are a group brought together from different lines of business, as purchasing was formerly decentralized. The purchasers work two by two and thereby ensure
overlap in working practices, which is beneficial in case of illness, vacation and other
kinds of absence. The working situation could be characterized as: “We have not chosen
to be together, and we notice that we were twice as large 4-5 years ago as we are
now…we have lost colleagues each year…due to rationalization…and the workload has
also fallen substantially” (purchaser #1). But the tone of communication is nonetheless
open and supportive. The purchasers perceive the united group as fairly homogenous
and harmonious today, and they express solidarity, as articulated by the statement “we
stand together” (purchaser #2).
The purchasing department has been using the SCS since 2002 and the overall perception of the old system (Release 11i) is that they accept the system although with several
reservations, especially regarding user friendliness and missing decision support. The
latter is compensated for by a “data warehouse system” (DWS), which is highly integrated with OEBS as a look-up system (read only). DWS gives detailed information
about stock status and historical usage statistics, which is not directly available in
OEBS, at least not compiled as “a one screen overview” as in DWS. The interviewed
280
purchasers emphasize the importance of DWS, and claim that OEBS as a stand-alone
system would be very problematic. Another issue with the old system is the lack of
email integration for submitting purchase orders. This is done through an error prone
fax module – “when we say that we have sent a fax, suppliers believe that we are back
in the Stone Age” (purchaser #2). The fax module is expected to be replaced by email
integration in the new system (Release 12).
The expectation regarding the launch of the new system is that it will be a fairly painless process e.g., “my expectation for the [new system] is that it is similar to the system
that I have, only in English” (purchaser #3). Several bugs are expected to be resolved in
the new version, and the transition to the new system is anticipated to be manageable as
the new system is a major enhancement of the old system and not a complete replacement.
Three purchasers were asked to score SCS user perception of OEBS Release 11i and
OEBS Release 12, using the DeLone and McLean success model (2003). The result is
presented below with a relative scale from 1 to 10 expressing the subjective user’s
perception:
Interview period
User perception of Release 11i
(old system)
User perception of Release 12
(new system)
March 2008 – April 2008
September 2009
Purchaser Purchaser Purchaser Purchaser Purchaser Purchaser
#1
#2
#3
#1
#2
#3
User Satisfaction
7,5
4
6,5
8,5
8
6
User friendliness
4,5
5,5
3
7
8
2,5
Information Quality
4,5
6,5
8
6
8
8
System Quality
6
8
7,5
8
8
4
Service Quality
4
4
7
2
5
7,5
5,3
5,6
6,4
6,3
7,4
5,6
Average per person
Average before / after
5,8
6,4
Table 2: Subjective user perception of SCS before and after launch of Release 12
Table 2 shows that there is a smaller increase in average subjective user perception
which is supported by several statements from the purchasers. First, the new system
(Release 12) has greatly improved the search facilities in the system with the possibility
of customizing queries, and the much needed replacement of the old fax module with
281
email integration is working well. Second, the transition from the old to the new system
has been smoother than expected – one user expresses the shift as “much more painless
as I had expected. Fortunately, it is because I have been through system shifts before,
where there have been a lot of unexpected things that have cost blood, sweat and tears.
And of course, there have been some slight glitches here, but not in the same way I
thought” (purchaser #2). The relatively smooth transition might also be connected with
the involvement of users from the purchasing department during the RE-ES project.
Third, local support from the super-user in the purchasing department has ensured
prompt and practical relevant help, which has been important in getting the purchasers
to embrace the system. Finally, the majority of the purchasers (seven out of nine) want
to keep the new Release 12 according to purchaser #2, which indicates that the new release has bedded down well in the purchasing department despite the ups and downs
during the launch process.
An unintended positive effect of the RE-ES project is increased knowledge-sharing in
the purchasing department. The purchasers are allowed to help each other with things
they can, and refer to the process as something they have been through together. This
has strengthened cohesion in the group very much, and substantiates the increase in subjective user perception shown in the table above.
A MULTI LEVEL ANALYSIS OF DEINSTITUTIONALIZATION PRACTICES
In this chapter we set out to answer the research question about intended and unintended
deinstitutionalization practices. The analysis is built around two dimensions, the first
dimension being the categories of deinstitutionalization pressures, i.e., financial, political, functional and social; the second dimension is the varying levels of organizational
field, the organization, the RE-ES project and the purchasing department representing
the local level. The classification of deinstitutionalization pressures into the two dimensions can in some instances be questioned because there is some overlap, for instance, is
“reduction in customizations” a political pressure or a functional pressure? It is possible
282
to use both categories, and we will argue that the breaking of a complex topic into
smaller parts enables us to gain a better understanding of the topic, while the prize can
be that the splitting is artificial and debatable, but still appropriate from an analytical
perspective.
FINANCIAL PRESSURES
Organizational field level: The highly competitive pressure that SCANDI experiences
today started back in the nineties when the Scandinavian countries decided to liberalize
the utility infrastructure. More and more parts of the infrastructure were deregulated
leading to higher and higher competition. The growth opportunities are limited due to
high rates of penetration in most services, and a potential revenue growth should be
driven by increasing diversity and complexity of services that are offered (newspaper
article, April 2009). The utility industry is at the same time under revenue pressure from
the sharp deterioration of the global economy (industry report, February 2009), so this is
a shaky cocktail for SCANDI that impacts all parts of the company. Total service
spending in the utility industry is forecasted to decrease from 2008 to 2009, and will
most likely not be back on the same level until 2011 (as compared to 2008) (ibid.), so
the competitive situation will probably not be better for SCANDI in the coming years.
The market conditions increase “the change readiness” both for business and IT people,
and thereby decreases the potential resistance to change, which appears to make it easier
to break the old system.
Organizational level: SCANDI has decided to reduce IT costs as one means of being
more competitive. The EA program and RE-ES project supports the goal of reducing IT
costs although it might be questioned whether it is actually achievable, but was nevertheless put forward by the management as a deinstitutionalization practice. The limited
value of the yearly support fee to Oracle was perhaps not the most important factor in
deinstitutionalizing the old system, but was at least a contributing factor.
RE-ES project: There are many reasons for doing the RE-ES project, but IT/IS management’s primary entrance is the expected IT cost reduction of 40% related to OEBS,
which is clearly communicated (RE-ES project description, October 2007). Business
management is focused on the platform that Release 12 will create for future functional-
283
ity and understand the RE-ES project to be a necessary evil (logistics manager #2). The
new platform will make it possible to insource the logistics of a high-tech service and
thereby reduce logistics costs.
Below is a summary of the financial pressures leading to deinstitutionalization:
Financial Pressures
Organizational field level: Very tough and highly competitive market conditions
Recessive or stagnated market for SCANDI high-tech services
Organization level:
Reduce IT cost in general
Limited value of yearly license and support cost
RE-ES Project level:
Reduce IT cost by 40% concerning Oracle E-Business Suite
New platform will make it possible to insource logistics of a high tech
service and thereby reduce logistic cost
Sub group / Actor level:
None
Table 3: Financial pressures leading to deinstitutionalization
The financial pressures would not have had the same power if SCANDI were operating
in a less competitive market, and the deinstitutionalization would have been more difficult, at least based on financial arguments.
POLITICAL PRESSURES
Organizational field level: “ERP has become the software that no business can live
without” (Wailgum 2009) and “standardized packaged solutions account for the bulk of
systems used today” (Pollock and Williams 2009: 4) – this global trend filtered into
SCANDI as early as 1996, when they jumped on the ES bandwagon, but at that moment
only in well-defined functional areas such as the finance department. The concept then
penetrated more business areas of SCANDI and was cemented by the EA program and
RE-ES project as the preferred strategy for implementation and use of IS/IT. Organizations are furthermore advised to go for “vanilla” implementations (i.e., configuring
without customizations) by consultants (Hildebrand 2009) and academia (Parr and
Shanks 2003; Seddon et al. 2003) or at least to minimize customizations as much as
possible (Beatty and Williams 2006). The reason is that vanilla ES implementations are
easier, cheaper (Fitz-Gerald and Carroll 2003) and quicker (Willis and Willis-Brown
2002) and furthermore reduce future maintenance and upgrading costs (Hildebrand
284
2009). This compelling advice was adopted by SCANDI in their RE-ES project and EA
program, and the mantra “un-customize customizations” (Beatty and Williams 2006)
was met by the RE-ES project where the number of customizations were reduced from
400 to 150. The reduction of customizations was a strong organizational field pressure
on the SCANDI organization to break the old system and make a re-engineered vanillalike system, although a considerable amount of customizations still existed after reimplementation (e.g., due to integration with 40 other systems).
Organizational level: The political pressures from the organizational field can be read
in the IS/IT strategy with standard systems (packaged software applications) as the preferred choice, and the discourse about the nice three-layered model instead of the old
jumble of 450 IT systems is a clear message from IS/IT management of the direction
SCANDI is taking with their information infrastructure. In the same vein is the statement that “IT systems define the structures and employees shall abide by the system”,
which means use standard functionality and reduce customizations as much as possible.
Although the RE-ES project was not part of the EA program from the very beginning,
the EA program has nonetheless fertilized the ground for the necessary shift from Release 11i to Release 12. The vision about one coherent ES platform and one company is
also a driving force in deinstitutionalizing the old system as initiated by the business
and IT strategy created by management, and based on their interest and belief for the
direction SCANDI should take.
RE-ES Project: The RE-ES project description states “SCANDI wants more simplified
workflow and reduced costs associated with IT. Therefore, SCANDI wishes to consolidate as many small systems as possible into the Oracle E-Business Suite, where they are
implemented as standard functionality” – this statement is nicely aligned with the EA
program and reflects the overall IS/IT strategy. The EA program was very ambitious
and was abandoned due to costs being too high and was probably also close to being
mission impossible, and the RE-ES project course was carried out in stormy weather
with several launch delays, but both initiatives were important in supporting the execution of the IS/IT strategy. Management showed the importance of the RE-ES project
when it was included in the EA program; this was a strong deinstitutionalization pressure, because it clearly signaled that management wanted this RE-ES project to succeed
285
no matter what, and the RE-ES project became an icon of success in spite of its troublesome project course, and saved the EA program with some tangible business results
which arose from it. Even the dramatic reduction of customizations from 400 to about
150 is accepted by business without flinching, and this is really a culture shift from the
old monopoly days where IT systems were built exactly to a customer’s specific needs.
The political pressures are summarized below:
Political Pressures
Organizational field level: Implementation of standard software packages as enterprise system
Vanilla or close-to-vanilla ES implementations as the legitimized approach
Organization level:
Standard systems or packaged software applications are the strategy
New Enterprise Architecture promotes going from “chaotic network of
systems” to a layered model (get rid of the old jumble of 450 IT systems)
IT Systems defines the structures and employees shall abide by the system
Use standard functionality and reduce customizations as much as possible
Vision about one coherent ES platform
Vision about one company
RE-ES Project level:
Simplified business processes
RE-ES project becomes part of overall EA program (icon for success in this
program)
Reduce customizations from 400 to 150
Sub group / Actor level:
None
Table 4: Political pressures leading to deinstitutionalization
The political pressures presented in Table 4 above are mainly intended practices although some of them, such as “vanilla or close-to-vanilla…”, are pressures from the
organizational field adapted by management, but still enacted as an intended practice at
the organizational level.
FUNCTIONAL PRESSURES
Organizational field: Since SCANDI decided to do business with Oracle (i.e., buy licenses and support for OEBS) they also agreed to rely on Oracle’s OEBS strategy for
better or worse. ES vendors are developing new releases of their ES software on a regular basis and support of older releases ends some years after the new release is launched.
SCANDI was using Release 11i.5 and Oracle’s support of this release expired in July
2006 (Oracle 2009: 6), which was obviously a problem for SCANDI. The good news
286
was that Oracle had launched their newest Release 12 in the beginning of 2007 (Songini
2007) being a multi-everything enterprise system (Kholeif et al. 2008) with more than
100 integrated modules targeting most industries as well as public and private organizations. The underlying philosophy of the new release was a global, highly integrated, and
flexible standard solution (Oracle 2008). The upgrade to Release 12 was necessary,
there being no support for Release 11i.5, but this was also a desirable opportunity for
SCANDI to utilize extensive functional enhancements and a viable technical architecture (Songini 2007).
Organization level: There are several functional pressures within the organization:
First, the overall IS/IT strategy to consolidate small IT systems into the ES. Second, the
problems with the current version of OEBS stand to reason as bugs cannot be fixed and
SCANDI is locked into an old ES architecture. Third, SCANDI has customized the old
ES system more and more, as this has been an acceptable approach, but creates many
problems related to the upgrade of OEBS, and is therefore not acceptable in the future.
Finally, the new version of OEBS (Release 12) contains much needed functionality,
which is important for business such as enhanced search functionality for the purchasing department. All these functional pressures are arguments for moving to the new version.
Sub group / actor level: The purchasers are positive or neutral toward the RE-ES
project, and they expect a fairly painless re-implementation of OEBS because it is the
same enterprise system, albeit re-engineered, and they have been through several implementations and upgrades, which mean that they know the launch process very well
with its ups and downs. Furthermore they look forward to the new functionality that the
new system offers (e.g., enhanced search functionalities and email integration functionality replacing error prone fax module). Involvement of the super-user from the purchasing department in the RE-ES project prepares the department well for the shift from
the old to the new system especially because the local super-user is well-liked and participatory in his approach. The pressures are a mixture of deinstitutionalization of the old
system and institutionalization of the new system, but the latter acts as indirect deinstitutionalization.
The functional pressures are summarized below:
287
Functional Pressures
Organizational field level: Release 11 is no longer supported by vendor
The development of ES into a multi-everything system with more and more
functionalities
Organization level:
Consolidate small IT systems into Enterprise System
Few patches can be implemented in Release 11, thereby locking SCANDI
into an old system
SCANDI is customizing more and more, which is problematic for upgrade
of ES
Release 12 is perceived as an enabler for much needed business functionality
e.g., enhanced search facilities to purchasing department
RE-ES Project level:
None
Sub group / Actor level:
Local support in purchasing group is a big help
Painless implementation of Release 12
Purchasers look forward to new system with enhanced search functionalities
and email integration replacing error prone fax module
Table 5: Functional pressures leading to deinstitutionalization
The functional pressures indicate that there is a close relationship between some pressures or practices taking place at the organizational level and the expectations at the
local level. It is similar to the travels of ideas (Czarniawska and Joerges 1996) from the
organizational field to the organization and finally to the purchasing department (ES as
a multi-everything systems → Release 12 as an enabler for much needed business functionality → enhanced search functionality).
SOCIAL PRESSURES
Organization level: The competitive situation for SCANDI has changed the institutional rules and values (Oliver 1992). The institutional logic in the monopoly days was characterized by security of supply, delivering standard outputs, and following agreed procedures, operating in a governmental regulated environment. The employees in SCANDI had stable jobs with reasonable salaries and good working conditions. According to
a former employee the organizational culture was characterized by pride and loyalty.
This has changed dramatically to a private company with the focus on efficiency and
profit (Svejvig and Jensen 2009). The downsizing (outsourcing and/or making redundant) implies that employees live in uncertainty, and has the consequence that employees tone down any complaints about the new system, as this might result in problems for the concerned employee.
288
Sub group / Actor level: The high degree of participation in the re-implementation
project acts as a social pressure to accept the new system, because the users can influence the process and the result, and it is well known that user participation reduces resistance (e.g. Mumford 2003). Most purchasers have jumped on the “new system bandwagon” as seven out of nine purchasers prefer the new system, which means that the
new system has been broadly accepted by its users (institutionalized), and this deinstitutionalizes the old system indirectly (see also Vitharana and Dharwadkar 2007: 358). The
interviewed purchasers state that there has been an increased knowledge sharing and a
higher spirit of solidarity in the purchasing department after Release 12 and this is perceived as a positive effect of the re-implementation project, which also cements the new
system.
Below is a summary of the social pressures leading to deinstitutionalization:
Social Pressures
Organizational field level: None
Organization level:
Dismissing and/or outsourcing of employees is a pressure on current
employees to accept the new system
RE-ES Project level:
None
Sub group / Actor level:
Involvement of users in purchasing department during re-implementation
process
The majority of purchasers (seven out of nine) prefer Release 12, which
indicates the degree of institutionalization of the new system
Increased knowledge sharing in purchasing department after Release 12 and
a higher spirit of solidarity
Table 6: Social pressures leading to deinstitutionalization
The social pressures in the table above are a mixture of unintended and intended pressures, and are therefore important pieces in the puzzle of deinstitutionalization of the old
system and institutionalization of the new system.
DISCUSSION
The multi level analysis shows the diversity and complexity of deinstitutionalization
practices, intended as well as unintended, related to breaking the old system and fertilizing the ground for the new system. We have drawn two essential elements from the
analysis, which will be discussed below.
289
First, we have used the terms pressure and practice indiscriminately or at least not in a
well-defined way. Pressure has quite a few meanings according to Oxford English Dictionary, but the understanding of it as a “metaphorical force” is well aligned with Oliver’s (1992) use of the term. Practice could mean “the action of doing something” (Oxford English Dictionary) and expresses our use of the term in this paper. If we elaborate
on the two terms, then practice implies action by some actors while pressure is something imposed on the organization, project, department etc. We can apply the two terms
more specifically on the deinstitutionalization elements described in the analysis and
establish patterns of pressure-practice chains as exemplified in Figure 4 below:
Figure 4: Pressure-practice chains
Figure 4 above shows three examples of pressure-practice chains where, for example,
the competitive and stagnated market pressure leads to a practice in SCANDI to reduce
IT cost, and this becomes a pressure on specific projects to reduce IT cost. This is reflected in the RE-ES project where the business case calculates a decrease of 40% for
operational IT cost for OEBS. These chains could also be understood as means-ends
chains (goal hierarchies) where factual premises from the observable world combined
with organizational value premises (i.e., preferred or desirable ends) leads to decisions
290
and activities in the organization (Scott and Davis 2007: 53-56). The EA program and
the RE-ES project in SCANDI are examples of such activities based on strategic decisions, and they become important drivers for deinstitutionalization of the old system.
Figure 4 emphasizes furthermore the value of the multi-level analysis as an approach to
analyze complex social phenomena to get a rich conceptual understanding (Currie 2009;
Jensen et al. 2009).
Second, our study confirms the dual processes of deinstitutionalization and the institutionalization as argued by Avgerou (2000) and supported by Alvarez (2001) (cf. Figure
2). The empirical data shows that the old system is completely replaced by the new system, but our data from the field work do not go far enough to conclude to what degree
organizational practices are changed, although specific practices are changed, such as
communication with suppliers via email instead of old fashioned faxes.
Another challenge in our study is to distinguish between deinstitutionalization and institutionalization pressures/practices. Some are mainly deinstitutionalization pressures
such as “Release 11 is not supported by vendor”, which clearly erode the old system as
this will become very problematic in the longer run. Others are mainly institutionalization pressures fertilizing the ground for the new system, e.g., “enhanced business functionality in Release 12” acting as a relevant argument for shifting to the new version
especially because SCANDI needs the new platform in order to develop business. The
final category are practices which imply both deinstitutionalization and institutionalization properties where “reduce IT cost” is an example as this pressure/practice deinstitutionalize the old system (due to its complexity with two separate technical platforms,
many customizations etc.) and institutionalize the new system as the cost reduction is
not possible without the consolidation and un-customization contained in the RE-ES
project (although alternative solutions might have been considered).
There are many pressures/practices described in the analysis section and a consolidation
into a few key practices will give a relevant overview of the dual deinstitutionalization
and institutionalization processes, but also wipe out the details, which we will argue are
needed to understand the complexity in these institutional processes. Below is a figure
modeling the dual processes with the consolidated pressures/practices with the purpose
of giving an overview:
291
New platform
for future
business
New
Reduce
IT cost
Use standard
functionality and
reduce
customizations
Future
Situation
Enhanced
functionality
One coherent ES
platform and
one company
Old
Organizational Structures & Practices
(Re-)institutionalization
Locked into
old ES
architecture
?
Current
Situation
Release 11i
not supported
any longer
Deinstitutionalization
Old
New
Enterprise System
Figure 5: Model of the dual deinstitutionalization and institutionalization process
The model in Figure 5 shows a few important practices leading to the deinstitutionalization of the old system and the institutionalization of the new system as well as practices
influencing both deinstitutionalization and institutionalization. The figure furthermore
indicates the uncertainty about the degree of changed organizational practices (dashedoutline block arrow with question mark).
What can we learn from this case study about deinstitutionalization? First, making new
systems is breaking old systems! The deinstitutionalization process is inseparable from
the institutionalization process, and both processes have to happen in order to implement the new system. Second, the two processes overlap each other, which is contrary
to Greenwood and colleagues’ (2002) linear stages of institutional change although their
model addresses the organizational field level while our model targets the organizational level which might imply differences. However, Hinings (2004) and Greenwood and
colleagues (2004) have in a newer model, “the dynamics of change”, also stated that the
processes of de- and re-institutionalization should pass off in parallel. Third, deinstitutionalization should take place for both the old system and the old structures and practices, and in addition, institutionalization should target both the new system and the new
structures and practices. Fourth, the pressures/practices toward deinstitutionalization
292
might as well influence the institutionalization process. Institutionalization of the new
system / new practices might indirectly act as pressures on deinstitutionalization
A final question, does it make sense to focus on deinstitutionalization when it is inseparable from institutionalization in ES implementations? The answer is yes from an
analytical point, because deinstitutionalization processes are not very well understood
and the study in this paper can shed some light on the mechanisms at play. It is however
different from a practical point of view where deinstitutionalization and institutionalization processes are interwoven, and we have to take that fact into account in practical
implementations of ES.
CONCLUSION
This paper has examined the intended and unintended practices deployed by SCANDI, a
large Scandinavian high-tech organization, to deinstitutionalize their current system and
replace it with a re-implemented enterprise system. Deinstitutionalization is vastly overlooked in IS research and the role of pressures and practices leading to deinstitutionalization is not well understood. We adapted the model of Oliver (1992) to analyze the
pressures and practices related to making a new system and breaking the old system,
presented as a multi level analysis categorized into financial, political, functional and
social pressures/practices. The paper provides a valuable in-depth insight into these
processes where deinstitutionalization is inseparable from institutionalization and the
processes are overlapping; some practices impact on both processes while other practices only apply to either deinstitutionalization or institutionalization.
293
REFERENCES
Ahmadjian, C. L. and P. Robinson (2001). Safety in Numbers: Downsizing and the
Deinstitutionalization of Permanent Employment in Japan. Administrative
Science Quarterly 46(4): 622-654.
Alvarez, R. (2001). "It was a great system": Face-work and the Discursive Construction
of Technology during Information Systems Development. Information Technology & People 14(4): 385-405.
Alvarez, R. (2002). The Myth of Integration: A Case Study of an ERP Implementation.
In F. F. H. Nah (Ed.), Enterprise Resource Planning Solutions and Management.
Hershey: Idea Group Inc., 63-88.
Avgerou, C. (2000). IT and Organizational Change: An Institutionalist Perspective. Information Technology & People 13(4): 234-234.
Bazeley, P. (2007). Qualitative Data Analysis with NVivo. London: Sage Publications
Ltd.
Beatty, R. C. and C. D. Williams (2006). ERP II: Best Practices for Successfully Implementing an ERP Upgrade. Communications of the ACM 49(3): 105-110.
Berger, P. L. and T. Luckmann (1966). The Social Construction of Reality. A Treatise in
the Sociology of Knowledge. New York: Doubleday.
Chae, B. and G. F. Lanzara (2006). Self-destructive Dynamics in Large-scale Technochange and Some Ways of Counteracting it. Information Technology & People
19(1): 74-97.
Chen, L. and S. C. Yang (2009). Managing ERP Implementation Failure: A Project
Management Perspective. IEEE Transactions on Engineering Management
56(1): 157-157.
Cooper, R. B. and R. W. Zmud (1990). Information Technology Implementation Research: A Technological Diffusion Approach. Management Science 36(2): 123139.
Currie, W. (2009). Contextualising the IT artefact: Towards a Wider Research Agenda
for IS using Institutional Theory. Information Technology & People 22(1): 6377.
Czarniawska, B. and B. Joerges (1996). Travels of Ideas. In B. Czarniawska and G.
Sevón (Eds.), Translating Organizational Change. Berlin: Walter de Gruyter,
13-48.
Davenport, T. H. (1998). Putting the Enterprise into the Enterprise System. Harvard
Business Review 76(4): 121-131.
294
DeLone, W. H. and E. R. McLean (2003). The DeLone and McLean Model of Information Systems Success: A Ten-year Update. Journal of Management Information
Systems 19(4): 9-30.
DiMaggio, P. J. and W. W. Powell (1983). The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields. American Sociological Review 48(2): 147-160.
Fitz-Gerald, L. and J. Carroll (2003). The Role of Governance in ERP System Implementation. Australasian Conference on Information Systems, Perth, Western
Australia.
Gosain, S. (2004). Enterprise Information Systems as Objects and Carriers of Institutional Forces: The New Iron Cage. Journal of the Association for Information
Systems 5(4): 151-182.
Grabski, S. V., S. A. Leech and B. Lu (2003). Enterprise System Implementation Risks
and Controls. In P. B. Seddon, L. Willcocks and G. Shanks (Eds.), Second-Wave
Enterprise Resource Planning Systems: Implementing for Effectiveness. Cambridge: Cambridge University Press, 135-156.
Greenwood, R., C. Oliver, K. Sahlin and R. Suddaby (2008). The SAGE Handbook of
Organizational Institutionalism. London: Sage Publications.
Greenwood, R., R. Suddaby and C. R. Hinings (2002). Theorizing Change: The Role of
Professional Associations in the Transformation of Institutionalized Fields.
Academy of Management Journal 45(1): 58-80.
Hedman, J. and A. Borell (2004). Narratives in ERP systems evaluation. Journal of Enterprise Information Management 17(4): 283-290.
Hildebrand, C. (2009). The Value of Sticking with Vanilla. Profit Online. Retrieved
25th March, 2009, from
http://www.oracle.com/profit/smb/122808_ziegele_qa.html.
Hinings, C. R., R. Greenwood, T. Reay and R. Suddaby (2004). Dynamics of Change in
Organizational Fields. In M. S. Poole and A. H. V. d. Ven (Eds.), Handbook of
Organizational Change and Innovation. New York: Oxford University Press
Inc., 304-323.
Häkkinen, L. and O.-P. Hilmola (2008). ERP Evaluation during the Shakedown Phase:
Lessons from an After-sales Division. Information Systems Journal 18(1): 73100.
James, D. and G. H. Seibert (1999). Oracle Financials Handbook: Planning and Implementing the Oracle Financial Applications Suite. Berkeley: Oracle Press Edition from Osborne.
295
Jensen, T. B., A. Kjaergaard and P. Svejvig (2009). Using institutional theory with
sensemaking theory: a case study of information system implementation in
healthcare. Journal of Information Technology 24(4): 343-353.
Kholeif, A. O., M. G. Abdel-Kader and M. J. Sherer (2008). Enterprise Resource Planning: Implementation and Management Accounting Change in a Transitional
Country. Basingstoke: Palgrave Macmillan.
Klein, H. K. and M. D. Myers (1999). A Set of Principles for Conducting and Evaluating Interpretive Field Studies in Information Systems. MIS Quarterly 23(1): 6793.
Lewin, K. (1947). Frontiers in Group Dynamics. In D. Cartwright (Ed.), Field theory in
social science. New York: Harper & Brothers Publishers, 188-237.
Lindley, J. T., S. Topping and L. T. Lindley (2008). The hidden Financial Costs of ERP
Software. Managerial Finance 34(2): 78-90.
Maguire, S. and C. Hardy (2009). Discourse and Deinstitutionalization: The Decline of
DDT. Academy of Management Journal 52(1): 148-178.
Markus, M. L. (2004). Technochange Management: Using IT to drive Organizational
Change. Journal of Information Technology 19(1): 4-20.
Mattila, M., J. Nandhakumar, P. Hallikainen and M. Rossi (2009). Emerging Role of
Enterprise System in Radical Organizational Change. Proceedings of JAIS
Theory Development Workshop, Sprouts: Working Papers on Information Systems, 9(46). http://sprouts.aisnet.org/9-46.
Miles, M. B. and A. M. Huberman (1994). Qualitative Data Analysis: An Expanded
Sourcebook. Thousand Oaks: Sage Publications Inc.
Mumford, E. (2003). Redesigning Human Systems. Hershey, PA: IRM Press.
Myers, M. D. (2009). Qualitative Research in Business & Management. London: Sage
Publications.
Myers, M. D. and D. Avison (2002). An Introduction to Qualitative Research in Information Systems. In M. D. Myers and D. Avison (Eds.), Qualitative Research in
Information Systems - A Reader. London: Sage Publications, 3-12.
Newman, M. and D. Robey (1992). A Social Process Model of User-Analyst Relationships. MIS Quarterly 16(2): 249-267.
Nicholson, B. and S. Sahay (2009). Deinstitutionalization in the Context of Software
Exports Policymaking in Costa Rica. Journal of Information Technology 24(4):
332-342.
Noir, C. and G. Walsham (2007). The Great Legitimizer. Information Technology &
People 20(4): 313-333.
296
Oliver, C. (1992). The Antecedents of Deinstitutionalization. Organization Studies
13(4): 563-589.
Oracle (2008). Meet the Challenges of Globalization. Oracle E-Business Suite, Brochure, 2008, Document number C16140-01.
Oracle (2009). Oracle Information-Driven Support - Oracle Lifetime Support Policy,
Oracle Applications.
Orlikowski, W. J. and S. R. Barley (2001). Technology and Institutions: What can Research on Information Technology and Research on Organizations Learn from
each other. MIS Quarterly 25(2): 145-165.
Parr, A. and G. Shanks (2003). Critical Success Factors Revisited: A Model for ERP
Project Implementation. In P. B. Seddon, L. Willcocks and G. Shanks (Eds.),
Second-Wave Enterprise Resource Planning Systems. Cambridge: Cambridge
University Press, 196-219.
Parr, A. N. and G. Shanks (2000). A Taxonomy of ERP Implementation Approaches.
Proceedings of the 33rd Hawaii International Conference on System Sciences,
Hawaii.
Patton, M. Q. (2002). Qualitative Research & Evaluation Methods. Thousand Oaks:
Sage Publications Inc.
Pettigrew, A. M. (1987). Context and Action in the Transformation of the Firm. The
Journal of Management Studies 24(6): 649-671.
Pettigrew, A. M. (1990). Longitudinal Field Research on Change: Theory and Practice.
Organization Science 1(3): 267-292.
Pollock, N. and R. Williams (2009). Software and Organizations: The Biography of the
Enterprise-wide System or How SAP Conquered the World. Milton Park, Abingdon, Oxon: Routledge.
Porter, M. E. and V. E. Millar (1985). How Information Gives You Competitive Advantage. Harvard Business Review 63(4): 149-161.
Powell, W. W. and P. DiMaggio (1991). The New Institutionalism in Organizational
Analysis. Chicago: The University of Chicago Press.
Rao, H., P. Monin and R. Durand (2003). Institutional Change in Toque Ville: Nouvelle
Cuisine as an Identity Movement in French Gastronomy. American Journal of
Sociology 108(4): 795-843.
Robbins-Gioia. (2002). ERP Survey Results Point to Need For Higher Implementation
Success. Robbins-Gioia Press Release. Retrieved 20th August, 2007, from
http://www.robbinsgioia.com/news_events/012802_erp.aspx.
297
Robey, D., W. Ross Jeanne and M.-C. Boudreau (2002). Learning to Implement Enterprise Systems: An Exploratory Study of the Dialectics of Change. Journal of
Management Information Systems 19(1): 17-46.
Ross, J., M. R. Vitale and L. P. Willcocks (2003). The Continuing ERP Revolution:
Sustainable Lessons, New Modes of Delivery. In P. B. Seddon, L. Willcocks
and G. Shanks (Eds.), Second-Wave Enterprise Resource Planning Systems: Implementing For Effectiveness. Cambridge: Cambridge University Press, 102132.
Scott, W. R. (2008). Institutions and Organizations: Ideas and Interests. Thousands
Oaks: Sage Publications.
Scott, W. R. and G. F. Davis (2007). Organizations and Organizing: Rational, Natural,
and Open System Perspectives. Upper Sadle River: Pearson Prentice Hall.
Seddon, P. B., L. Willcocks and G. Shanks (2003). Introduction: ERP - The Quiet
Revolution? In P. B. Seddon, L. Willcocks and G. Shanks (Eds.), Second-Wave
Enterprise Resource Planning Systems: Implementing for Effectiveness. Cambridge: Cambridge University Press, 1-19.
Sia, S. K. and C. Soh (2007). An Assessment of Package-Organisation Misalignment:
Institutional and Ontological Structures. European Journal of Information Systems 16(5): 568-583.
Silva, L. and J. Backhouse (1997). Becoming Part of the Furniture: The Institutionalization of Information Systems. In A. S. Lee, J. Liebenau and J. I. DeGross (Eds.),
Information systems and qualitative research. proceedings of the IFIP TC8 WG
8.2 International Conference on Information Systems and Qualitative Research,
31st May-3rd June 1997. Philadelphia, Pennsylvania, USA: Chapman & Hall,
389-414.
Somers, T. M. and K. G. Nelson (2004). A Taxonomy of Players and Activities across
the ERP Project Life Cycle. Information & Management 41(3): 257-278.
Songini, M. L. (2007). Oracle rolls out E-Business Suite 12. Computerworld. Retrieved
1. April, 2009, from
http://www.computerworld.com/action/article.do?command=viewArticleBasic&
articleId=9010084.
Sumner, M. (2003). Risk Factors in Enterprise-wide/ERP Projects. In P. B. Seddon, L.
Willcocks and G. Shanks (Eds.), Second-Wave Enterprise Resource Planning
Systems: Implementing for Effectiveness. Cambridge: Cambridge University
Press, 157-179.
Sumner, M. (2009). How Alignment Strategies influence ERP Project Success. Enterprise Information Systems 3(4): 425 - 448.
298
Svejvig, P. and T. B. Jensen (2009). Enterprise System Adaptation: A Combination of
Institutional Structures and Sensemaking Processes. AMCIS 2009 Proceedings.
Paper 746.
Swanton, B. (2008). Oracle EBS R12: New and Re-implementations, but Few Technical
Upgrades. AMR Research, Inc.
Sæbø, J. I., S. Molla, A. Asalefew and S. Sahay (2008). Interplay of Institutional Logics
and Implications for Deinstitutionalization: Case Study of HMIS Implementation in Tajikistan. Sprouts: Working Papers on Information Systems, 8(11).
http://sprouts.aisnet.org/8-11 University of Oslo, Norway.
Vitharana, P. and R. Dharwadkar (2007). Information Systems Outsourcing: Linking
Transaction Cost and Institutional Theories. Communications of the Association
for Information Systems 20(Article 23): 346-370.
Wailgum, T. (2009). The Future of ERP. CIO magazine. Retrieved 20th November,
2009, from http://www.cio.com/article/print/508022.
Walsham, G. (2002). Interpretive Case Studies in IS Research: Nature and Method. In
M. D. Myers and D. Avison (Eds.), Qualitative Research in Information Systems
- A Reader. London: Sage Publications Ltd., 101-113.
Walsham, G. (2006). Doing Interpretive Research. European Journal of Information
Systems 15(3): 320-330.
Weick, K. E. (1995). What Theory is Not, Theorizing Is. Administrative Science Quarterly 40(3): 385-390.
Wenrich, K. I. and N. Ahmad (2009). Lessons Learned During a Decade of ERP Experience: A Case Study. International Journal of Enterprise Information Systems
5(1): 55-73.
Willis, T. H. and A. H. Willis-Brown (2002). Extending the Value of ERP. Industrial
Management & Data Systems 102(1): 35-38.
Xu, L. and S. Brinkkemper (2007). Concepts of Product Software. European Journal of
Information Systems 16(5): 531-541.
Yin, R. K. (2003). Case Study Research: Design and Methods. Thousand Oaks, California: Sage Publications.
299