Roadmap to CPFR: The Case Studies

Transcription

Roadmap to CPFR: The Case Studies
Need to Add Spine: Upon final page count find out from Brown & Kroger what this thickness is
S t a n d a r d s
C o m m e r c e
I n t e r i n d u s t r y
V o l u n t a r y
© 1 9 9 9
Roadmap to CPFR:
The Case Studies
FPO
i
Endorsements
CPFR represents a natural evolution of ECR, deeper trading partner collaboration and
improved supply chain practices. Its impact will be significant because, for the first time,
it will bring demand and supply planning together in one process and drive the entire
supply chain from POS data.
Ralph Drayer, Procter & Gamble
Historically, you can see where industry activities such as EDI, Quick Response, and ECR
have focused on reduced costs in one segment of the supply chain and, at times, increased
costs in other segments. The real power behind CPFR is the global view of the supply chain
and creating win-win solutions for both the retailer and the supplier to reduce total supply
chain costs. This is where we can impact the level of performance at the shelf for our guests,
which is our ultimate goal for CPFR.
Colleen Wickering, Meijer
Great progress has been made in the development and implementation of supply chain
management and category management, yet the full potential of each has not been realized.
CPFR brings together these powerful processes and trading partners, producing increased
sales and improved effectiveness and efficiency far beyond the results that can be expected
using conventional means. The question isn’t why CPFR–but why not.
Joe Andraski, OMI International
I believe that CPFR is the single largest opportunity to move inventory management
forward in the next five years. We plan to implement collaborative relationships with well
over 100 suppliers in the next 12 months. We believe that CPFR is the driver for moving
into the next era of buyer-seller relationships.
Randy Mott, Wal-Mart
© 19 9 9 Vo l u n t a r y I n t e r i n d u s t r y C o m m e r c e S t a n d a r d s A s s o c i a t i o n
I
Synchronizing the supply chain from raw materials to the consumer offers the greatest
opportunity to improve profitability and provide greater value to the consumer. It is quickly
becoming the standard for competition. CPFR provides the process to link business plans
and ensure synchronization.
Specific benefits we are currently seeing and expect to see in the future include higher
in-stock levels at retail, increased turns for supply chain participants, and lower supply
chain costs. One of the key drivers is a better understanding of consumer purchases.
Point-of-sale information provides the ability to track current purchases and develops more
i Endorsements
accurate sales forecasts. We are continuing to expand our use of POS to provide critical
information for managing our supply chain.CPFR is a new process and we are committed
to working with our customers to make full use of business planning tools to better serve
the consumer.
Mark Jamison, Kimberly-Clark
II
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
ii
Benchmarking Partners
Robert Willey
Best Buy
Al Brady
ConAgra Frozen Foods
Deborah Spratt
Eastman Kodak Co.
Jim Sykes
E-Millenium
Ram Viswanathan
Ernst & Young
Ken Woeste
Ernst & Young
Lisa Calkins
Federated Dept Stores
Margaret Irvin
Fieldcrest Cannon/ Pillowtex
Tony Casas
H.E. Butt
Kevin McLaughlin
Hewlett-Packard
George Perry
Hewlett-Packard
Stan Elbaum
i2 Technologies
Mark Fera
i2 Technologies
Bill Green
i2 Technologies
Susan McNally
i2 Technologies
Acknowledgment
Jim Uchneat
Bruce Stenman
Intelink
Larry Roth
Kimberly-Clark
Pilot
Dan Myers
K-mart Corp
Pilot
Dan VanHammond
K-mart Corp
Pilot
Dave Hutchings
Kraft Foods, Inc
Jeff Stiely
Kurt Salmon Associates
Chuck Murphy
Logility
Andrew White
Logility
Brian Dunch
Manugistics
Chris Verheuvel
Manugistics
Beth Lovett
Manugistics
Colleen Wickering
Meijer
Gerry Cantwell
Nabisco Food
Pilot
Stanley Smith
Nabisco Food
Pilot
Dane Workman
Nabisco Food
Pilot
Joseph Andraski
OMI- Americold
Chairman
Noreen Murphy-Andrews
Pillsbury Company, The
Ken Kappner
Pricewaterhouse Coopers
Ralph Drayer
Procter & Gamble
Pilot
Scott Williams
Procter & Gamble
Pilot
Maria Viglionese
QRS Corp
Amy Eagan
Ralston Purina
John Carlee
Sara Lee Corp
Pilot
Lynne Hammer
Sara Lee Corp
Pilot
Barry Westbrooks
Service Merchandise
Kathy Frank
Sure Fit, Inc
Matt Johnson
Syncra Software
Chris Sellers
Syncra Software
Rich Sherman
Syncra Software
Andrew Love
Uniform Code Council
Jean Schenck
Uniform Code Council
Barry Bernstein
VF Services, Inc
Pat Garvey
VF Services, Inc
Randall Mott
Wal*Mart Stores, Inc.
Pilot
David Ferrell
Pilot
Bobbie Aldridge
Pilot
III
ii Acknowledgments
IV
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
iii
Title Page
i.
Endorsements ............................................................I
ii.
Acknowledgements ..................................................III
iii. Table of Contents ......................................................V
1.
Executive Summary ..................................................1
2.
Introduction to CPFR ................................................3
3.
Roadmap to CPFR......................................................5
Step 1: Evaluate Your Current State ........................6
Step 2: Define Scope and Objectives ....................11
Step 3: Prepare for Collaboration ..........................16
Step 4: Execute: Performing the Pilot....................19
Step 5: Assess Performance and
Identify Next Steps..................................................22
4.0 CPFR Pilot Project Overview ..................................29
4.1 Nabisco and Wegmans Pilot ..................................33
4.2 Kimberly-Clark and Kmart Pilot ............................45
4.3 Wal*Mart and Sara Lee Pilot..................................49
4.4 Procter & Gamble Pilot ..........................................57
4.5 Hewlett-Packard Pilot ..............................................69
5.0 CPFR Technical Specifications ................................79
Specification Approach ..........................................81
Data Flows ..............................................................83
Data Format Standards ..........................................93
© 19 9 9 Vo l u n t a r y I n t e r i n d u s t r y C o m m e r c e S t a n d a r d s
V
Ta b l e o f C o n t e n t s
0.
Transport/Network Protocol Guidelines ................95
Welcome!
Security Considerations..........................................97
Application Architecture Considerations ..............99
Centralized Server ................................................101
Emerging Technologies ........................................103
6.0 Conclusion ............................................................105
iii Contents
Appendices
A
CPFR: Capability Assessment ..............................107
B
Metric Matrix ........................................................115
C
Data Model ............................................................117
D
Data Dictionary......................................................121
Entity Definitions ..................................................121
Attribute Definitions ............................................124
Common Attribute Values ....................................129
E
Message Mapping ................................................139
EDI Transaction Maps ..........................................139
SIL Message Maps................................................144
F
VI
CPFR Technology Questionnaire ........................155
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Executive Summary
1.0
Executive Summary
To the CEO:
This message is for you. Your leadership will make CPFR a mainstream business
process. Consumers will benefit, and companies will win together.
Consumers, category management, supply chain management–all reveal the
importance of a holistic, cooperative, and synchronous approach to the business
environment. Where there is little collaboration within companies or between
trading partners, opportunities are lost.
Your leadership is crucial, despite the limited time you’ll have to learn how
CPFR can help achieve the company’s goals:
• Reducing working capital, so funds can be invested in marketing
and new product development.
• Reducing fixed capital and infrastructure expense.
• Reducing operating expense, as cost as a percent of sales is used
to measure performance.
• Growing sales each year, consistent with the expectations
of the stockholders and Wall Street.
You may ask if trading partners can achieve these objectives without resorting
to major restructuring? Without major acquisitions? Without right sizing–or
downsizing? The answer is found in today’s consolidations, and the restructuring
plans announced daily by companies who previously enjoyed market dominance.
Is CPFR a magic bullet? No. But it is an unconventional–yet piloted and
proven–way of achieving those four crucial objectives of reducing working and
fixed capital, reducing operating expense, and growing sales.
Several leading retailers and manufacturers have embraced CPFR and engaged
in pilots. They have shared their results and the learning that they have gained
in order to interest others in the business community. While there is much press,
dialogue, and activity about the growth of CPFR, it will not bear fruit if you do not
sponsor it and make it a company priority.
1
Your concerns about innovative programs can be mitigated because the
investment in systems is minimal; the Internet exists; communication standards
are in place; human resources implications are few; continuous replenishment
provides significant information; trading partners come together in a fashion
not experienced in the past; and the initial investment in developing and testing
the model has already been made by reputable companies.
So become engaged with one of your trading partners to determine if CPFR
can make a difference for you, your trading partner and the consumer–
1.0 Executive Summary
and to your shareholders.
Sincerely,
2
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Introduction to CPFR
The VICS CPFR Roadmap provides a comprehensive and detailed plan for
companies to begin a CPFR relationship. It was prepared by skilled professionals
2.0
who drew from their personal experience in describing the precise steps for
the VICS CPFR Guidelines published in 1998, the novice company can develop and
implement a CPFR business process with up to five partners. This roadmap gives
significant detail and takes little for granted as it goes through each step.
It closely follows the definition found in Webster's Dictionary: “a detailed plan
to guide progress toward a goal.”
The goal of CPFR is to change the relationship paradigm and create significantly
more accurate information that can drive the value chain to greater sales and
profits. There's no doubt that the roadmap will meet and perhaps exceed
your expectations.
Read On.
3
Introduction to CPFR
moving forward with CPFR. When combined with the models that appeared in
2.0 Introduction to CPFR
4
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
CPFR Roadmap
You will probably do several collaborative planning, forecasting, and
replenishment (CPFR) pilots with different objectives–early implementations
emphasizing organization, later ones exploring collaboration. Companies may
want to jump into CPFR pilots by following only parts of the map, depending
3.0
Roadmap to CPFR
on their starting point. The learning curve is steep, and change is an effort–
but knowledge gained is significant. Piloting adapts to a company’s scale,
and the VICS CPFR Roadmap takes you there.
Introduction
You’ve read the VICS CPFR Guidelines. You have circulated it among colleagues
and executives. You’ve attended conferences and have identified potential
partners–and skeptics. So what steps do you take to implement CPFR and
demonstrate how it can increase sales and reduce costs?
Many organizations would like to implement collaborative planning, forecasting,
and replenishment (CPFR). Piloting can help organizations understand the
benefits of CPFR, document changes required for the further expansion, and
develop strategies for overcoming obstacles. This section is designed to help
companies answer the question, “How do I get started?”
The VICS CPFR Guidelines identified the processes that comprise CPFR, from
the creation of a Front-End Agreement through Business Planning, Forecasting,
and Replenishment. This VICS CPFR Roadmap is not meant to replace that work.
Instead, it supplements it with basic steps that help you get started with CPFR,
test it with a trading partner, and develop rollout plans.
This VICS CPFR Roadmap is a path from awareness of CPFR opportunities to
making implementation plans. Based on the CPFR process model in the
VICS CPFR Guidelines, it aligns you and your trading partner to common
objectives, and guides you together through forecast exchange, exception
management, and the review of performance results. Both partners gain
5
a clear understanding of the potential for rolling out CPFR practices throughout
their organizations.
The VICS CPFR Roadmap is divided into five steps:
• Step 1: Evaluate Your Current State
• Step 2: Define Scope and Objectives
• Step 3: Prepare for Collaboration
• Step 4: Execute
• Step 5: Assess Results and Identify Improvements
The VICS CPFR Roadmap provides ready-to-use templates that can be customized
for different partnerships. The Roadmap also provides a checklist for each step to
ensure all critical items have been completed.
Step
3.0 Roadmap to CPFR
1
Evaluate Your Current State
CPFR begins long before piloting, with an assessment of your company’s needs,
values, culture, strategies, trading partner relationships, and track record in
implementing best practices. This step looks for areas where change is needed
to implement CPFR successfully. Only after this step is done will your company
be prepared to articulate a meaningful vision for CPFR. In addition, the senior
leadership of your company must not only understand the concept of CPFR,
but also openly offer their support.
Put CPFR into your company’s terms in order to gain commitment
CPFR can provide both demand and supply benefits. Mapping these potential
benefits to your company’s priorities will assist in cross-functional buy-in.
CPFR Benefits: Demand
Enhanced Relationship
Implicitly, CPFR strengthens an existing relationship and substantially accelerates
the growth of a new one. Buyer and seller work hand-in-hand from inception
through fruition on business plan, base, and promotional forecasts. Continual
CPFR meetings strengthen this relationship.
6
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Greater Sales
The close collaboration needed for CPFR implementation drives the planning
for an improved business plan between buyer and seller. The strategic business
advantage directly translates to increased category sales.
Category Management
3.0
for targeted SKUs to ensure adequate days of supply, and proper exposure to the
consumer. This scrutiny will result in improved shelf positioning and facings
through sound category management.
Improved Product Offering
Before CPFR implementation, the buyer and seller collaborate on a mutual
product scheme that includes SKU evaluation and additional product opportunities.
CPFR Benefits: Supply
Improved Order Forecast Accuracy
CPFR enables a time-phased order forecast that provides additional information,
greater lead time for production planning, and improved forecast accuracy vs.
either stand-alone VMI/CRP or other industry tools.
Inventory Reductions
CPFR helps reduce forecast uncertainty and process inefficiencies. How much
inventory does your company hold to “cover up” for forecasting errors or a
trading partner’s inability to have the product available in a timely manner?
With CPFR, product can be produced to actual order instead of storing inventory
based on forecast.
Improved Technology ROI
Through the CPFR process, technology investments for internal integration can
be enabled with higher quality forecast information. Your company will benefit
by driving internal processes with common, high-quality data.
7
Roadmap to CPFR
Before beginning CPFR, both parties inspect shelf positioning and exposure
Improved Overall ROI
As other processes improve, the return on investment from CPFR can be
substantial.
Increased Customer Satisfaction
With fewer out-of-stocks resulting from better planning information, higher store
service levels will prevail, offering greater consumer satisfaction.
Where does your company stand today?
3.0 Roadmap to CPFR
Consider these questions to learn if your company is ready for CPFR:
Does your company have a culture that values cooperation and communication
between its departments and with its trading partners?
For a CPFR partnership to succeed, there must be a commitment to work with
and share information between functional areas and between trading partners.
Many companies recognize room for improvement here, but both recognizing
it and committing to make progress is an overwhelming indicator of the future
success of both CPFR and other business partnerships. Information sharing
brings strength that technology alone cannot provide.
Has your company implemented other industry best practices?
Companies that have adopted industry best practices, such as standard product
identification (UPC or EAN-13), Electronic Data Interchange (EDI), and the
basic principles of Quick Response or Efficient Consumer Response already
understand both their implications and benefits. Companies that have designed
a VMI or CRP group have also gained valuable learning applicable to CPFR.
Is using information technology to solve business challenges a company priority?
CPFR has been piloted with relatively little technology assistance. However,
once a pilot proves a business case, implementing CPFR on a broader scale may
become more effective when technology is applied. CPFR sponsorship by senior
management becomes particularly important at this point, as CPFR may be
viewed as competition for IT resources. Given proven results, however, it will
become evident that CPFR complements IT priorities.
8
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
The Internet offers an opportunity to improve both internal and external
business communications. Whether using the Internet simply as a vehicle to
expedite shared information or to use software, CPFR is a prime example of how
Internet technology can be used to enhance business-to-business relationships.
Develop Your Company’s CPFR Vision
3.0
Roadmap to CPFR
Before implementation, your company needs a CPFR vision statement. CPFR is
a business process that gives continuity to the strategies and tactics of category
management, integrating your company goals throughout the supply chain.
A well thought out vision answers these questions:
• What are your company’s objectives for CPFR?
• What areas of the organization will be impacted, and how?
• How will success be measured?
• What will be the scope of the project?
• Which product lines will be included?
• How many stores and DC’s will be included?
• How many trading partners will you bring on board in order
to have a significant impact?
• What is the level of technical sophistication of your company
and your trading partners?
• What is the long-term market position of your trading partners?
• What corporate stance will you develop for press releases?
• How will you document the vision?
Are Your Trading Partners Ready for CPFR?
Consider these factors when reviewing potential partners for a CPFR relationship:
Can your trading partner relationships be characterized as open and trusting?
A trusting and open relationship enhances the benefits of CPFR through the
sharing of sensitive information.
9
Do you and your trading partner have complementary strengths and weaknesses?
Think of what each partner can bring to the CPFR initiative. For example,
one may be strong technologically, providing the majority of the technology
infrastructure. The other may be strong in market and consumer knowledge,
and through the CPFR relationship be able to provide this strength. Be cautious
not to assume that your largest trading partner is the most likely candidate.
Does your trading partner have the appropriate commitment and resources required
to make CPFR successful?
3.0 Roadmap to CPFR
Without a strong commitment to CPFR, its potential will not be realized.
Does your trading partner have experience with CPFR with another partner?
Build on learning. Even though development time may be greatly reduced by
leveraging the learning already gained through your partner’s previous efforts,
the template for each new partnership will probably be different.
Can your trading partner quantify the potential internal and external benefits?
CPFR requires a paradigm shift from an adversarial stance to a win-win
relationship, as well as the working together by departments that may have never
communicated directly or effectively in the past.
Develop a Business Case to take to your Partners
The Capability Assessment, in Appendix A, helps answer these questions, and
helps the partners recognize their current state and where they need to change
to implement a CPFR partnership successfully.
Pilot experience has already been brought to bear:
• The most productive deployment method was to have the seller team
members reach consensus on the scores separately from the buyer team
members.
• Once each company reached its consensus on the individual scores, the two
companies reviewed each CPFR assessment element together to agree on the
10
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
score. Primary differences were documented along with the rationale for the
score in each area. The additional scoring rationale provided critical input
to improve the process for each element.
• When averaged and viewed in the overall context of the scoring grid, the raw
scores identified the relative strength of each of the four CPFR process areas.
• The overall information enabled the partners to set priorities for which CPFR
• At the detailed level, the individual scores and comments provided the
information to identify specific actions needed to improve the score on the
particular element.
This assessment process is a method to identify areas that need immediate
attention or that could be sequenced later in CPFR process improvement.
The key to this assessment process is understanding two critical aspects of CPFR:
1. All four process areas identified in the assessment must be maximized
to fully realize the benefits of CPFR.
2. Partnering companies have the flexibility to decide the priority
Step
in which key CPFR processes will be worked on and improved.
2
Define Scope and Objectives
After you create a CPFR Vision, you are ready to begin piloting. Step 2 requires:
1. Gaining commitment from your trading partner.
2. Assigning team members and establishing their roles.
3. Selecting products and locations that will be included in the process.
4. Deciding which part(s) of the nine-step CPFR process to test.
5. Establishing key performance metrics to measure the initiative’s success.
This is truly the implementation of the CPFR Front-End Agreement as designed
in the VICS CPFR Guidelines.
11
Roadmap to CPFR
processes to improve.
3.0
Preparations
The project begins by engaging at least one agreeable trading partner. To keep
the pilot manageable, no more than five trading partners should be involved.
Template Team Members, Roles, and Responsibilities
• Each trading partner identifies a management-level sponsor. The sponsors
select team members, ensure commitment to the project, and review results.
• The management sponsors for each company identify a team of three
or more individuals who can participate in the initiative at least 10 hours
3.0 Roadmap to CPFR
each week. One team member from each side is assigned to each of these
three areas:
Team Members, Roles and Responsibilities
Role
Responsibilities
Typical
Position Buyer
Typical
Position Seller
Sales
Collaboration
The sales collaboration team
is responsible for establishing
sales forecasts, promotion plans,
collecting and reporting sales
results. The team is also
responsible for recommending
and implementing changes to
the replenishment system.
Category ,
Manager
Buyer,
Replenishment
Analyst
Sales
Representative
(Account
Relationship
Owner)
Replenishment
The replenishment team
determines the order forecast,
and collects actual order and
inventory information.
Inventory
Analyst
(Re-) Buyer
Customer Service
Manager,
Forecast
Analyst, Order
Management
Analyst
Collaboration
Technology
The collaboration technology
team sets up the collaboration
environment, monitors
technology effectiveness, and
evaluates technical rollout
requirements.
IT Coordinator,
Project Manager,
Systems
Manager
IT Coordinator,
Project Manager,
Systems Manager
Table 1
12
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
• The management sponsors may also include an external project facilitator
(systems integrator, software vendor, or network services provider)
whose prior experience can help guide the project team.
• The management sponsors select a team captain for daily management
of the project.
3.0
Roadmap to CPFR
Project Kickoff
Once the project sponsors identify team members and roles, the project is
formally initiated with a kickoff meeting.
Template Kickoff Meeting
All team members, including sponsors and affected department liaisons attend
a one-day kickoff meeting.
Agenda
1. Introduce the team.
2. Create the pilot project overview, including performance metrics.
3. Begin joint scope-setting (see template).
4. Begin joint objective-setting (see template).
5. Set future meeting dates, times, and locations.
A CPFR pilot project is designed to give trading partners a sense of the real
impact of CPFR on their organizations by participating in a rapid test. Though it
is limited in scope, a pilot project includes the complete cycle of CPFR planning,
collaboration, and review. It should also investigate information technology that
can automate the process, in preparation for wide-scale future collaboration.
Give all team members the VICS CPFR Guidelines and the VICS CPFR Roadmap. If a
software package is used, distribute the user documentation. IT team members
should also receive software package administration and integration documentation.
(Detailed training sessions on CPFR practices and associated software are held in
the second phase of the project.)
13
During the scope- and objective-setting sections of the meeting, discuss adjustments
to the basic pilot implementation approach that the team would like to consider,
and document them carefully.
Schedule a one-day CPFR/collaboration technology training session for all team
members (except sponsors) within two weeks of the kickoff meeting.
All team members attend recap meetings at the conclusion of project steps 2, 3,
and 4. These meetings should be in-person, though they may also be conducted
3.0 Roadmap to CPFR
as conference calls.
Finalizing Scope and Objectives
If additional work is required after the kickoff meeting to finalize scope and
objectives, the sales and replenishment collaboration members work informally
within their organizations to come to a conclusion. Templates offer guidelines
for a pilot project.
Template Project Scope-Setting
• Limit the initial phase pilot by using only one product category (between
10 to 50 SKUs) and possibly by limiting the number of distribution centers
involved. If store-level forecasting is used, the project should involve fewer
than 50 stores.
• Determine which CPFR processes will be piloted. Starting with the
Front-End Agreement, decide how far into the nine-step process to proceed.
For example, some companies test CPFR through the Sales Forecasting
Collaboration before proceeding on to Order Forecasting Collaboration.
• Determine which trading partner will take responsibility for the creation
of the sales forecast, the order forecast, and generation of the actual order.
Since this is probably already established in your relationship, discuss the
reasons as they relate to the strengths, systems, and resources of each
company. Before making this decision, discover which trading partner
has the best data and technology or software that could be used to increase
14
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
forecast quality. This is critical, since a high-quality forecast on the front end
greatly reduces the amount future exception communication.
• Identify a process for determining both sales and in-stock inventory positions
for each SKUs/location(s) that will be included in the project.
(This information can be collected and entered from other systems,
3.0
or manually, if necessary.) Capture baseline data in order to determine
Roadmap to CPFR
if you have met your objectives.
• Identify where the best forecast data resides and how it can be used
for the project. Use SKUs where historical data is available.
Template Objective-Setting
Set performance metrics. The project should focus on increasing forecast and
planning accuracy, reducing both safety stock requirements and out-of-stock
conditions. The team sets targets for these measures relative to performance
in the same period a year earlier (or some other comparable period).
Here is a sample set of objectives:
• 15% increase in forecast accuracy (from 60% to 75%)
• 10% reduction in supply chain inventory (from 2000 cases to 1800 cases)
• 3-5% increase in retail in-stock position (from 90% in-stock to 94%)
When setting objectives, consider how the data will be collected. One common
example is total supply chain inventory. Often, manufacturers do not tally
inventory by customer, so assessing total supply chain inventory may be difficult.
Don’t be discouraged. Work with what you have. In this example, a manufacturer
may be able to provide good estimates on inventory by analyzing how its total
inventory for each item is affected.
15
Step 2 Closure Checklist
Before beginning Step 3, the team captain verifies the status and reports it
to the project sponsors:
✔ All team members have been assigned, and time has been allocated
for team member participation throughout the project.
✔ Products and locations for collaboration have been identified,
and associated personnel have been notified.
✔ Metrics such as forecast accuracy, inventory reduction,
3.0 Roadmap to CPFR
and out-of-stock targets have been established.
✔ Sources of forecast data have been identified.
✔ CPFR training session and future project team meetings
Step
have been scheduled.
3
Prepare for Collaboration
In Step 3, the project team studies the details of the CPFR business process, and
identifies the technology and additional resources required to support it. Sales
and replenishment team members develop ground rules for managing exceptions
and changes. Collaboration technology team members install and configure the
information systems (purchased, developed, or simple spreadsheets and e-mails)
used to support collaboration between partner pilot teams. At the end of this
step, collaboration is ready to begin.
Process and Technology
Step 3 starts with a training session for the team, so team members can begin
working with CPFR processes and technology.
Template Collaboration Training Session
All team members except sponsors participate in day-long collaboration training.
• The first session reviews the CPFR process model and Front-End Agreement,
concluding with a case study.
16
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
• The second session trains team members how to use the selected technology
to review exceptions, revise forecasts, and monitor scorecard measures.
• Training concludes with an on-line simulation of a trading scenario,
using the selected technology to interact.
• Collaboration Technology Team members take an additional day of training
3.0
on technology administration and integration.
Roadmap to CPFR
Configuration
Once the sales and replenishment collaboration team members are trained,
they begin to prepare their forecasts and exception thresholds. Meanwhile,
collaboration technology team members set up collaboration technology and
integration interfaces.
Template Sales Forecasting Preparation
• The sales collaboration team creates an initial sales forecast for weeks 9-14
of the project. The number of weeks in the forecast can be extended to
align with the next fiscal period.
• Create a sales forecast at the SKU level in normal buckets (such as weekly),
aggregated to the buyer distribution center level or other appropriate
customer ship point (for example, by store, if Direct Store Delivery is being used).
• Specify promotional forecasts (promotions with effective dates, projected
prices and volume impacts) independently of the base demand forecast
values. Seasonal uplift may be combined with base demand for simplicity.
The accuracy of promotional forecast can be greatly enhanced through the
use of high-quality lift tables that incorporate historical demand data and
related causal information.
• Set exception thresholds…for example thresholds could be at 5% change
in base demand, 15% change in promotional demand.
• Share the sales forecast among trading partners. Exceptions are triggered
based upon changes to this forecast, regardless of who makes them.
17
Template Order Forecasting Preparation
• Give replenishment team members early access to promotional plans
to develop the order forecast.
• Set the order forecast to the level of detail that you are going to replenish
(such as SKU level, weekly, by customer distribution center).
• Either the seller or buyer creates an initial order forecast covering the weeks
of sales planned in the sales forecast and incorporating the seller’s order
requirements (such as minimum order quantities). For example, if the
3.0 Roadmap to CPFR
purchase-order-to-store replenishment cycle time is one week, and the sales
forecast is for weeks 9-14 of the project, then the order forecast should be
for weeks 8-13. Additional weeks can be added to support inventory builds
and other logistics to support the sales plan.
• Set exception thresholds. For example thresholds could be at 5% change
in base demand, 15% change in promotional demand.
• Share the order forecast among trading partners. Exceptions are triggered
based upon changes to this forecast, regardless of who makes them.
Template Collaboration Technology Configuration
Each company sets aside server systems for the project or can subscribe to a CPFR
service provider.
• IT staff or software vendor personnel install collaboration software on-site,
or activate the service provider.
• Each company provides extracts of product and location data to be used in
the project. These products and locations remain fixed through the project
life cycle, to simplify master data maintenance and integration interfaces
for the pilot.
• IT personnel (or business partners) load the product and location data
into the collaboration software.
• Each company provides a format for forecast data to be loaded.
IT, consulting, or software vendor personnel develop scripts to load
and update this data.
18
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
To complete Step 3, team members should meet to review their readiness for
on-line collaboration. By the meeting date, the team should have agreed upon
an initial forecast for the collaboration period, and the collaboration software
should be up and running with the required products, locations, and initial
forecast values entered.
3.0
All project team members, including sponsors, participate in a pre-pilot meeting.
Agenda
1. Review initial sales forecasts, including planned promotions.
2. Review initial order forecasts.
3. Discuss logistics issues (such as carrier arrangements) that could affect the plans.
4. Review the status of collaboration technology; demonstrate technology with project
products and locations configured.
Before beginning Step 4, the team captain verifies the status and reports it to the
project sponsors:
✔ All team members have been trained.
✔ Initial sales and order forecasts for all product/location combinations
have been agreed upon.
Step
Collaboration technology is ready to begin.
4
Execute: Performing the Pilot
In Step 4, the sales and replenishment collaboration teams begin to exchange
forecasts with each other, modifying them to respond to changing conditions.
The collaboration technology team gains experience managing the environment,
and prepares for rollout to a large number of locations and projects after the
pilot is complete.
19
Roadmap to CPFR
Template Pre-Pilot Meeting
The Collaboration Process
Sales and order forecast collaboration follow specific ground rules to secure
benefits.
Template Forecast Collaboration
• Determine the rules for changing forecasts.
• The partners exchange new or revised forecasts at least weekly.
• The buyer enters the latest sales results for each product weekly.
3.0 Roadmap to CPFR
• A partner who disagrees with a proposed forecasted value counter-proposes
a change, along with comments. Counter-proposals are made within five
business days. If a partner does not receive a counter-proposal within that
time, the proposal is accepted.
• The order forecast owner is the final arbiter of forecast values. After rounds
of three changes, only the order forecast owner can change a forecast.
Technology Rollout Planning
Consideration of technology rollout begins now, although details of the rollout
steps themselves are part of Step 5.
The collaboration technology team uses the pilot experience to understand
the IT requirements for a larger-scale rollout of forecast collaboration practices,
including electronic commerce, security, and application integration issues,
as well as staffing, training, backup/recovery, and support procedures. A clear
technology rollout plan produced during this phase of the project can greatly
accelerate future results.
Template Collaboration Technology Rollout Planning
Each organization develops its own plans for technology rollout.
• Assess prospective trading partners’ electronic commerce capabilities.
How many will use EDI? How many will have their own CPFR server?
How many will be client users only?
20
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
• Determine which other application data feeds will be required to maintain
up-to-date product, location, and partner master data. Evaluate the level of
effort required to integrate these sources with the collaboration software.
• Determine which applications require or provide forecast information for
the range of products and locations to be covered in a general rollout.
3.0
Evaluate the level of effort required to integrate these sources to the
• Configure the corporate firewall and web servers to allow secure access
by CPFR trading partners.
• Identify the number of core users (planners), occasional users (management,
warehouse personnel, sales people), support personnel, and administrators.
• Estimate training requirements for the proposed user base.
• Develop a technology rollout timeline, including procurement, installation,
integration, training, and mobilization of support resources.
Collaboration Review
After about four weeks of collaboration, the project team meets to discuss
progress, problems, and changes. The team then takes an additional two weeks
of collaboration to determine whether adjustments have an impact.
Template Collaboration Review Meeting
All project team members, including sponsors, participate in the collaboration
review.
Agenda
1. Discuss major exceptions encountered in the first four weeks of collaboration.
2. Evaluate the effectiveness of response to changes.
3. Propose process enhancements.
4. Adjust objectives and collaboration ground rules.
21
Roadmap to CPFR
collaboration software for import/export.
✔ The project team has participated in at least six weeks of on-line sales
and order forecast collaboration.
✔ The initial results and participant comments have been summarized
and reviewed by both trading partners.
✔ The project team has identified future technology and software
modifications that will enhance the process.
✔ The project team has held a collaboration review meeting, applying ideas
from the first four weeks to improving the process in the final two weeks.
3.0 Roadmap to CPFR
Step
✔ Results have been reviewed with the Project Sponsor.
5
Assess Performance
and Identify Next Steps
In Step 5, the team and its management review its progress, report results to their
respective organizations, and make preparations for broader CPFR rollout.
Partnership Review
Every six to 12 weeks of collaboration, the business team reviews actual results
against the target metrics. The team also considers the business process impact
of their partnership.
Template Collaboration Scorecard
Evaluate the actual results against metrics that were agreed upon in the
Joint Planning Meeting and against data available from previous periods.
• Compare weekly SKU/store level sales forecast accuracy with the
year-ago period, and note changes.
• Compare weekly SKU/DC-level order forecast accuracy with results
for the year-ago period, and note changes.
22
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
• Compare total in-stock inventory at distribution centers and stores per
product with that for the year-ago period. Adjust for changes in the number
of products and stores, and note whether the balance of inventory has shifted
up or down the supply chain.
• Compare the number of store-level out-of-stock events with those for the
3.0
year-ago period. Calculate relative to the in-stock percentage per SKU.
• Add any other measures that seem significant after the collaboration period.
Has there been significant sales growth in the category? Were there special
external conditions (weather, new competitors, product changeovers,
promotional merchandise, store openings/closings, staff changes) that
affected results? How can these factors be accounted for in the future?
For a sample copy of the pilot team scorecard, see Appendix F.
Template Business Process Review
• Review cases in which a business process or forecast changed dramatically.
What caused the change? Did the CPFR process reduce problems associated
with the change? Were there specific benefits over past ways of doing
business?
• Based upon the pattern of changes that occurred, what future opportunities
are revealed?
• Can purchase orders be eliminated, and collaborative forecasts be used
to trigger replenishment (actual orders) instead?
• Were any organizational problems identified?
Template Collaboration Technology Review
• Assess the contribution of CPFR technology to achieving project objectives.
• Highlight new opportunities the technology has uncovered, as well as any
issues or concerns.
• Present the collaboration technology rollout plan to other team members.
Incorporate their feedback.
23
Roadmap to CPFR
• Identify the number of exceptions and their root cause.
Determining Next Steps
Before making a final report, team sponsors discuss what next steps they will
recommend to their respective organizations about continued collaboration.
Future initiatives between the implementation partners are determined jointly.
Team members from each organization also consider the next steps to take
relative to their own trading partner networks after the initial phase reviews are
completed. What additional partners should be included? Which product lines
3.0 Roadmap to CPFR
should be added? Training requirements, organizational changes, project costs,
and rollout time frames should all be estimated.
After a successful pilot, there are a number of ways to expand CPFR, and
numerous areas you would want to test and learn about:
Expanding CPFR
Add
Trading
Partners
Automate
the Process
Successful
CPFR
Implementation
Increase
Level of
Detail
Expand to
Other CPFR
Processes
Add
SKU’s
Integrate
the Results
Figure 1
24
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
1. Expand to other CPFR processes
If the CPFR implementation focused on joint business planning and sales
forecasting, expand to order forecasting collaboration
2. Add SKUs
If the CPFR implementations initially focused on a limited set of items,
3.0
Roadmap to CPFR
increase the gains by expanding to other product categories
3. Increase the level of detail
If the implementation focused on warehouse-level information,
better results can be achieved by moving to store-level information.
4. Automate the process
Although few of the companies that have initiated CPFR have had to add human
resources to complete their implementation, automating the collaboration
process will produce increased gains. The vision of CPFR is one of managing
forecasts by exception, which can best be achieved through an automated
process–especially when the number of products and trading partners increases.
5. Add trading partners
The benefits gained from collaborating with one or a small number of trading
partners can be extended to more trading partners, depending on the relationship
with, culture, and capabilities of each partner. Even before a “critical mass” is
achieved, there are benefits to each relationship.
6. Integrate the results
The benefits of CPFR are fully realized only when the outputs of the collaborative
processes are integrated with the internal processes at both companies. For a
seller or supplier, this means using the collaborative forecast in the production
planning, capacity planning, and materials requirements planning processes,
as well as financial planning processes. For the buyer or retailer, this means
integrating the collaborative forecast into buying, merchandising, replenishment,
and financial planning processes.
25
Reporting Results
The project team should publish a joint project report that allows others in each
partner organization to understand the impact and potential for collaboration.
The initial CPFR experience contributes greatly to a successful general rollout.
Template Project Report
A project report has six major sections:
1. Executive Summary
3.0 Roadmap to CPFR
2. Project Scope, Objectives, and Resources
3. Business Process Results
4. Collaboration Technology Results
5. Next Steps
6. Organization-Specific Plans
• Executive Summary provides a high-level overview of the project.
• Project scope, Objectives, and Resources discusses the original conditions and
changes that were made as a result of the initial CPFR experience.
• Business Process Results reports results against key metrics.
• Collaboration Technology Results describes how information technology
supported the project’s objectives.
• Next Steps suggests future collaborations between the partners,
including the business and technology rollout plans.
• Organization-Specific Plans are private to each organization’s version
of the report. They can include proposals to expand collaboration
to other trading partners, projected costs, organizational impact,
and preliminary project plans.
The team captain creates a summary presentation to supplement the
project report.
26
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Project Completion Checklist
✔ The project team held a collaboration review meeting.
✔ The project sponsors agreed upon next steps, and developed
organization-specific deployment plans for additional collaboration initiatives.
✔ The project report for the current review was written and distributed to
3.0
others within the organization.
to the recommendations for expansion.
27
Roadmap to CPFR
✔ Senior Management has been informed of the results and has contributed
3.0 Roadmap to CPFR
28
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
CPFR Pilot Project Overview
As you read the pilot project reports, you will find several significant themes–
significant because their results can move organizations past barriers toward
expanded implementations, and onward to improved corporate performance.
The pilot projects have several common features
Planning
4.0
best practices within the supply chain. In particular, setting common goals for
organizations builds on and extends other ECR successes, such as category
management and CRP. It pulls them together into a cohesive plan, supports
better execution of the plan, and invites improved planning in the next business
cycle. The improved planning drives sales gains through to the consumer and
lowers costs throughout the supply chain.
Steps
All the pilots addressed the first four steps in the business model. Even where
they did not start with step one (the front-end agreement), they concluded that
it was needed for best results.
Process
Across different buying organizations and a growing number of sellers from
various industries, a strong common business process was established, and the
process model was verified. All of the pilots began without collaborative software,
but expansion plans and rollouts are tied to the implementation of collaborative
software. Some of it is now in place.
Trust
The trading partner relationship improved. Greater trust resulted from better
attention to common business goals. Stronger communication was established
within each organization as each business function’s part in the process was
better understood.
29
CPFR Pilots Project Overview
The CPFR concept and pilot efforts are consumer driven without losing focus on
E n t e r i n g
Case Studies
Benefit
Each pilot organization and partnership received different levels of benefit.
Anecdotal evidence shows large improvements in service or retail sales levels,
but the question remains: How many of those benefits are available to all the
partnerships and products in the marketplace? There is no exact answer. The
key is to focus on the pilot companies’ actions. Have they backed away from
the concept or are they moving forward? On this point, the evidence is clear.
Pilot companies have seen enough business benefit (increased sales, lowered costs
and improved in-stock levels) to move forward with additional customers, product
4.0 CPFR Pilot
Project Overview
lines, and integration efforts.
Resources
Few additional resources were dedicated to the CPFR pilots. Most pilots simply
adjusted the priorities of those already assigned to the business relationship.
All pilots emphasized that significant upper management commitment is essential
to ensure resources for the new processes. Scalability issues require that one or
both partners invest technology that enables expansion without staffing changes.
Data
Each pilot used existing EDI transactions or Internet-based information
exchange. Additional systems will permit collaboration on a larger scale,
demanding human attention only for exceptions or discrepancies between
partners and between the planned activities and actual performance.
Expansion
Expansion of CPFR improves Vendor Managed Inventory, Co-Managed Inventory,
and traditional replenishment. The information shared through collaboration
can strengthen these replenishment processes. Pilot participants expect to
expand collaboration across additional product lines, additional channels of
distribution, and additional partners. This demands automated tools for
scalability and integration of forecast information into production scheduling,
product deployment, and replenishment systems.
30
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
There are also three significant differences between the pilots
Variety
First, there is great variety in product types and seller production strategies.
This should encourage organizations to see CPFR’s appropriateness for themselves.
The concept and process are strong enough to bridge diverse industries.
Technology
Next, several technological approaches were used. Even the simplest efforts
4.0
technologically advanced pilots have developed into a mainstream business
process that addresses partnership business goals and implementation issues.
Planning
Finally, all the pilot projects were successful no matter where they began in the
business process or in the supply chain. This should encourage those who are
not prepared to address store-level replenishment yet, or those who want to
develop a common business plan between partners. As one committee participant
stated while talking about CPFR pilots, “Start small, start early, start making it a
mainstream process, but most importantly, start now!”
Final Observations
Partnerships
Critical mass is not needed before significant benefits are realized. The collaboration
effort can help achieve the mutual goals shared by one buyer and one seller.
But to benefit the entire supply chain–from consumers all the way back to the
raw material suppliers–a broad spectrum of CPFR partnerships must be in place.
Forecasting
The CPFR model calls for the comparison of two forecasts (one from each partner)
to generate exceptions. However, the process is still valuable when one forecast
is compared to actual sales or the current forecast is compared to the previous
forecast. Either way, a collaboration can improve forecast accuracy and ensure
that it agrees with the business plan.
31
CPFR Pilots Project Overview
produced results that encouraged participants to move forward. The more
E n t e r i n g
Case Studies
Common Goals
The most powerful concepts seen in pilots are the attention to a single business
plan and the ability to focus only on exceptions. The front-end agreement is a
powerful tool to move organizations toward their common goals. The risks of
avoiding CPFR are the same risks as avoiding a business plan. Partners may be
moving in different directions without knowing it.
Unexpected Benefits
The CPFR process unexpectedly enabled other ECR initiatives. In several
4.0 CPFR Pilot
Project Overview
instances, CPFR moved other business initiatives to meet goals. Simple situations
such as new product rollouts and shelf set changes could be more accurately
managed between partners and within each organization.
Quantifying Benefits
A great deal of effort was given to quantifying benefits derived from CPFR.
This was difficult because many benefits are achieved by avoiding costs and lost
sales. Also, because of proprietary concerns, some facts could not be released
by the organizations involved.
CPFR has been aptly named. As you read the pilot reports, you will see that
collaboration has significantly improved results in each key area:
Planning–business planning between buyer and seller;
Forecasting– for input to production and deployment planning purposes;
Replenishment–from raw material to the consumer.
32
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Nabisco Inc. and Wegmans Food Markets
Introduction
Nabisco is a major international manufacturer of biscuits, snacks, and premium
grocery products, including such well-known U.S. brands as Oreo, SnackWell's,
and Chips Ahoy!; Ritz crackers; A.1. steak sauces; Grey Poupon mustards;
LifeSavers confections; and Planters nuts and snacks. International products
4.1
include Christie, Peek Freans, Terrabusi cookies and crackers; Yemina pastas;
Nabisco and Wegmans Pilot
Royal dessert mixes; Fleischmann's yeast; and several Nabisco global brands–
Oreo, Ritz and Chips Ahoy!. Nabisco markets products in the United States,
Canada and more than 85 other countries around the world.
Wegmans Food Markets, Inc. is a 58-store supermarket chain in New York
and Pennsylvania, with its first store in New Jersey in 1999. The family-owned
company, founded in 1916, is recognized as an industry leader and innovator.
Executive Summary
Category management and supply chain management have been proven to
provide a competitive advantage to firms that can successfully implement them.
Trading partners can gain an even greater advantage by linking these activities
through the CPFR process. CPFR provides the opportunity to link the output
of business plans that were jointly developed between trading partners into the
supply-chain process. The business plans and forecasts are monitored and kept
current by both trading partners. This is accomplished by the creation of a
two-way interactive communication process that enables the transfer of
promotional plans and forecasts among manufacturers and distributors.
These activities can help grow sales and profits between participating partners.
Nabisco and Wegmans engaged in a CPFR pilot to validate the VICS business
model. The pilot was limited to 22 Planters nut items. The pilot was conducted
without increasing resources in the area of headcount or technology. For the
first six months, transfer of information was accomplished using spreadsheets
and e-mail.
33
Actual results from the CPFR pilot from July 1998 through January 1999 include
an increase in category sales by 13% vs. 8% decline for other retailers in the
market (dollars, units and pounds all showed similar results). Sales increases
for the Planters brand was especially dramatic at 53%, as measured by IRI for
30 weeks ending January 17, 1999. The majority of the increases in retail sales
can be attributed to jointly developed business plans that leveraged enhanced
category management strategy and increased category focus.
These results were achieved with minimal stress on the supply chain due to CPFR.
4.1 Nabisco and Wegmans
Pilot Overview
On the operations side, service level to stores increased from 93% to 97%, and
days of inventory declined 2.5 days (18%). These positive results have led
both Nabisco and Wegmans to decide to extend the timeline for this pilot and
to expand its scope to include Milk-Bone pet snack products. In addition,
commercially available collaboration software will be tested as potential
technology solutions. Both companies are also establishing pilots with other
trading partners.
CPFR Processes Addressed
Front-End Agreement, Joint Business Planning, Sales Forecasting, Order
Forecasting, Order Generation, Delivery Execution, and Technology.
Pilot Objectives
The primary goal of the pilot was to test the CPFR concept and related processes.
Both Nabisco and Wegmans wanted to validate the model as prescribed by VICS
to see if CPFR was a proposition that could be expanded to other businesses and
trading partners.
Once it was decided to move forward with the pilot, other more quantifiable
objectives were developed and agreed upon. They included sales growth,
reduced inventory levels, improved forecast accuracy, reduction in spoilage,
34
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
and reduction or elimination of other supply chain inefficiencies. We expected
to meet these objectives by using existing resources; neither company would add
headcount or systems.
Scope
In retrospect, the pilot can be viewed as being split into two phases. This was not
actually planned; it just evolved that way.
4.1
items stocked by Wegmans; bag and canister snack items were excluded. All 58
Wegmans retail outlets participated from the pilot’s inception. Nabisco shipped
all the items from one Nabisco distribution center (DC) to one Wegmans DC.
All planning and forecasting activities were for shipments in the second half,
July 6 through December 31, 1998; however, the planning period was quarterly.
This meant the joint business planning process had to be done twice during the
initial phase of the pilot. The sales forecast was developed simultaneously with
the business plan. The forecasts and the business plan were constructed at the
item level for 13 weeks.
The second phase of the pilot was expanded to include the shipping period for
the first half of 1999. The same 22 Planters nut items from phase one were
included, as well as 20 Milk-Bone pet snack items. The development of the joint
business plans for Planters and Milk-Bone was completed by the first week in
November. All other preparation and planning was completed so the second
phase of the pilot was considered “live” for the first shipping day of 1999.
The formal end date to this stage of the pilot is the end of the second half of 1999.
It is very likely that CPFR will become the standard business process for Wegmans
and Nabisco.
35
The first phase was limited to 22 Planters nut items, representing all can and jar
Technology Used
Phase one of the pilot used limited technology. The pilot was managed with the
852 EDI transaction set, Excel spreadsheets, and e-mail. Because Wegmans was
an established VMI account with Nabisco, the 852 EDI transaction set was already
in place. The Nabisco customer service agent manually performed the exception
process and developed many of the scorecard metrics. (The exception process
is detailed in step three of the Methodology section.) Excel spreadsheets were
developed and exchanged via e-mail to communicate forecasts, to plan changes
Pilot Overview
and exception items, and to measure actual results.
Manugistics of Rockville, Maryland, provided their Networks collaboration
software for the pilot; its use began in January during phase two. This type of
product will be necessary to help manage the CPFR process when moved into a
production mode with multiple trading partners and a broad base of products.
Also, collaboration software will allow users to import and export data into their
supply chain systems.
The application, which works with Microsoft IE 4.0 or Netscape Navigator,
consists of a Java client residing on the local workstation and an Oracle database
and the Java application residing on a remote server. Access to the application
is made through the local Java client which then connects through the Internet
to the remote server, now located at Manugistics in Maryland. The data entered
into the client is then transmitted to the server where both partners are able to
view, compare and manipulate the data. Nabisco, Wegmans and Manugistics have
worked together to identify functional enhancements that are expected to be
included in later releases.
Metrics and Results
Much time and energy was expended to determine the performance metrics,
method for calculation, data sources, and the person responsible for reporting
the results. The following are the key measures being reviewed:
36
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
• Case-fill to Wegmans DC
• Case-fill from Wegmans DC to retail
• Inventory turns at Wegmans DC
• Forecast accuracy
• Number of forecast changes inside and outside the freeze period
• Sales growth for the category, Planters Brand, and private label
• Category and Planters Brand profits
4.1
POS systems within Wegmans and Nabisco, except for the sales data, which was
measured from both POS and IRI.
There has been dramatic improvement in results against the key metrics. Retail
sales as measured by Wegmans POS data and IRI have shown clear increases on a
dollar and unit sales basis. The increases can be attributed to improved category
plans and improved execution of those plans. The sales increases were supported
with less inventory and an improvement in forecasting. This resulted in a 15%
increase in overall profit dollars; however, their profit margin declined by one
percentage point. Also, days of supply and forecast error have been reduced.
The key results are as follows:
Planters Sales at Wegmans
$800,000
$600,000
$400,000
$200,000
1997
1998
1997
1998
$0
Q3
Q4
Source – Point of Sale Data
Figure 2
37
The data sources for the metrics came mainly from internal supply-chain and
Wegman’s Total Snack Nut Category Sales
$1,500,000
$1,000,000
$500,000
1997
1998
1997
1998
$0
Q3
Q4
Source – Point of Sale Data
Pilot Overview
Figure 3
Service Level to Stores = (1-(Cuts/(Demand-Cuts)))*100
98.0%
96.0%
94.0%
92.0%
90.0%
Pre-Pilot
Q3 1998
Q4 1998
Source – EDI 852 Data
Figure 4
Days of Supply = (On Hand Inventory/Weekly Movement) *7
15
10
5
0
Pre-Pilot
Q3 1998
Q4 1998
Source – EDI 852 Data
Figure 5
38
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
45
1998
1998 Average
1997
1997 Average
40
35
25
4.1
20
15
10
5
0
27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
Week
Figure 6
100
95
90
85
1998
1998 Average
1997
1997 Average
80
75
27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
Week
Figure 7
39
Days of Supply
30
Resources Involved
This pilot was conducted without any additional staffing at either Nabisco or
Wegmans; however, like any new initiative, it required a reallocation of priorities
and time for the people involved. Both companies assembled a cross-functional
team that included members from Sales, Category Management, Supply Chain
Management, IS, and Customer Service. In addition, both companies had
executive sponsorship that ensured proper attention and commitment.
Pilot Overview
Project Challenges
There were several challenges to initiating this pilot effectively:
• Participants had to gain a basic understanding of CPFR.
• Process maps developed by VICS had to be translated into the current workflow, or new ones had to be created at each company.
• Major obstacles due to systems limitations had to be overcome.
• Item-level forecast had to be manually developed, since neither Wegmans
nor Nabisco forecasting systems could adequately perform this task.
• Collaboration software was not available until phase two of the pilot.
Methodology
The pilot can be defined in four phases.
1. Training and Education
2. Preparing the Joint Business Plan
3. Sales and Order Forecast Generation
4. Execution of Shipments
Step 1 Training and Education
Participants needed training and education to understand CPFR. The Wegmans/
Nabisco pilot was one of the first in the industry, and the concept was new to both
companies. Syncra Software sponsored Benchmarking Partners, a supply-chain
40
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
consulting company from Cambridge, Massachusetts, to help facilitate education.
They also assisted in defining the overall scope of the pilot and to ensure that
the process maps published by VICS were understood and used. This step
was extremely important to maintain the integrity of the process. The frontend agreement helped establish the benchmarks that defined the roles,
responsibilities, and timelines.
4.1
Step 2 Preparing the Joint Business Plan
shipping the first order. The first need was to select the categories and products
to participate in the pilot. These criteria helped lead to the decision to select
Planters and subsequently Milk-Bone:
• Identify categories susceptible to major competitive erosion.
• Select strategic categories to defend.
• Understand competitive trade marketing strategies.
• Review viability of current category trade marketing plans.
• Consider alternative trade marketing category solutions.
The 22 Planters nut items were selected. Both companies assembled their
internal marketing plans for the period, reviewed historical shipments, IRI data,
the revised category management strategy and anything else the team thought
could help in the planning process. An aggressive merchandising plan was put
together and agreed upon for the third quarter. Subsequent quarterly plans
were developed 13 to 15 weeks in advance of each shipping period.
Step 3 Sales and Order Forecast Generation
The next step was to develop an item level-forecast–the most difficult step, since
neither company’s forecasting systems were designed to do this. The rolling
13-week sales forecasts were developed manually by the Nabisco Sales Manager
and the Wegmans Category Manager. The sales forecast was based on estimates of
the aggregated quantity the retail stores were going to order from their
41
This phase included all the preparation and planning that took place prior to
supplying Wegmans DC. The sales forecast had two components, the base
forecast and the promotional forecast. The base forecast was primarily generated
from the historical 852 movement data captured in the Nabisco CRP system.
In addition, any additional volume that could be attributed to historical and
projected growth of Wegmans was added.
The promotion forecast was developed along with the category and merchandising
plan. The creation of the promotional forecast proved difficult. The sales
forecast converted to the order forecast when it rolled into the freeze period.
Pilot Overview
The freeze period was originally set at five weeks but was later reduced to three
weeks. A five-week freeze period would allow Nabisco to incorporate the order
forecast into upstream supply chain processes if enough trading partners adapted
the CPFR processes. The five-week freeze period was reduced to three weeks
because the sales growth was so dramatic it was difficult to keep up with demand.
This fact turned the freeze period into more of a slush period because changes
were made to actual orders within the freeze period. Initially the unexpected sales
growth was thought to be an anomaly; however, the trend never subsided.
Spreadsheets were used to manage the forecasting process and the collaboration
exceptions. The Nabisco customer service agent was responsible for flagging
items that fell outside the agreed-upon exception criteria. The Nabisco sales
manager and the Wegmans buyer resolved these exception items by phone
or e-mail.
Step 4 Execution of Shipments
The Nabisco customer service representative (CSR) managed the order process.
The CSR played a key role in tactical execution of this pilot. She received and
monitored the forecast and the product availability data contained in the 852 EDI
transaction. The shipments to Wegmans were based on the order forecasts.
Tentative booking orders based on the sales forecast were created by the CSR
for weeks one to 10. These booking orders actually turned into firm orders when
they rolled into week 11. The booking orders were adjusted along the way as
42
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
actual demand fluctuated from the forecast or as the business planned changed;
the changes had to be considered significant for the adjustments to occur.
Collaboration software will automatically advise the users of significant variations
in forecast or demand.
Summary of Pilot Effectiveness
CPFR is about jointly developing and monitoring business plans and forecasts
4.1
between trading partners. This has led to a richer planning process and an
improved understanding of the participants’ marketing plans and category
dynamics for Wegmans and Nabisco.
Overall, the pilot was deemed a success by all participants. It validated the
business process model that was published by VICS. The major goals and
objectives of growing sales while reducing cost were realized. It was a learning
experience that enhanced the understanding of each company’s business
operations and objectives.
The effort proved that collaboratively monitoring and adjusting business plans
leads to improved supply chain performance and profitability.
Trading Partner Relationship Changes
The relationship between Wegmans and Nabisco was already strong at various
levels of the organizations; this relationship was one of the reasons that the pilot
was undertaken. Both companies have worked closely together and have embraced
the entrepreneurial spirit to help grow sales and improve processes over the years.
The pilot only strengthened the partnership.
Model/Guidelines Functional?
The initial sales forecasts were developed jointly between Nabisco and Wegmans
when the business plan was developed. This enabled collaboration to take place
up front. However, forecast revisions and plan updates may have triggered
exceptions as illustrated in the process model. The model shows two independent
43
forecasts being developed and compared. Only items falling outside the exception
criteria go through the collaboration process.
Unexpected Benefits and Key Learning
All new initiatives come with some surprises. The most obvious unexpected
benefit was the dramatic retail sales growth of Planters and the category. CPFR
did not bring this alone but truly enabled enhanced planning and execution of
the merchandising plan. Another benefit–not totally unexpected–was obtaining
Pilot Overview
an improved understanding of the other trading partner’s business processes.
Rollout Plans
Nabisco and Wegmans are in the process of expanding their pilot programs to
learn more. From the Wegmans perspective, they want to test the CPFR process
with larger, higher-volume categories. Nabisco wants to test it with other accounts
that create their own orders. Also, Nabisco wants to develop an internal process
that can easily integrate the collaborated item-level forecast data into its upstream
supply-chain processes. Additional pilots will help determine the expandability
and scalability of the CPFR process. Both companies are committed to
continuous improvement and believe that CPFR will enable that on-going goal.
44
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Wal*Mart and Sara Lee Branded Apparel
Executive Summary
Wal*Mart believes that our vendor partners are among our keys to success, and
that exchanging information with them creates a unique synergy that allows both
companies to be successful.
CPFR is an extension of what we have done for several years and we believe that
order forecasts, will continue to remove cost from the supply chain and increase
overall profitability. By exchanging information such as forecast and replenishment data with our vendor partners, we ensure that we have the right item at the
right time in the right place, resulting in increased customer satisfaction.
Our focus in the pilot was to validate the industry's best-practice document,
and to ensure that all steps required for collaboration between buyer and seller
were clearly documented and executable. Our pilot identified minor technical
changes that were made before the initial publication, and further proved that
collaborating on forecasts with our partners and focusing on exceptions was the
next evolution in reducing costs and increasing profit throughout the supply
chain.
Sara Lee Branded Apparel believes that long-term business relationships depend
on gratifying the consumer. Participation in this CPFR pilot represents Sara Lee’s
endorsement of synergistic, logical, and measurable efforts in this progression
of business processes.
We addressed three phases of CPFR:
• Creating the Sales Forecast
• Identifying Exceptions to the Sales Forecast
• Collaborating and Resolving Exceptions to the Sales Forecast
49
Wa l * M a rt a n d S a ra L e e P i l o t
standardizing the industry’s exchanges of pertinent information, such as sales and
4.3
We followed both the business and technical specifications outlined in the
industry model, validating all steps. We also informally walked through the
Joint Business Process to ensure that key data inputs had been included in the
technical specifications and that all primary business needs were addressed in
the models.
Pilot Objectives
Our primary objective was to validate the industry model documented in
4.3 Wal* Mart and Sara Lee
Pilot Overview
the VICS CPFR Guidelines, and to identify missing or inaccurate steps in the
business case and in the technical specifications (such as ERD). To validate
the specifications in the industry model, part of our pilot created a CPFR site
to address the above processes, which was successfully completed.
Our secondary objective was to address practical application of these principles
and processes during the execution phase of the pilot.
Scope (Execution Phase)
Twenty-three branded women’s underwear items were selected for the pilot.
Five of the items were new introductions to the line and distributed to smaller
stores; the remaining items had either full chain distribution (to approximately
2,400 stores) or were distributed to all but the smaller stores.
Collaboration began in July 1998 and is still actively taking place. The collaboration
focused more on identifying exceptions and resolving the exceptions than on
creating a sales forecast (sales forecasts were already being created, and the
creation of the sales forecasts did not change with the initiation of CPFR).
Changes in the way the sales forecasts are updated continue to evolve as
information is exchanged and exceptions are identified and resolved.
50
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Positions involved in the detailed collaboration discussions included POS
Replenishment Manager, Director of Sales, Sales Analyst, Forecast Manager, and
Manager of Sales Systems and Logistics. Allocation of pilot functions is shown
in the chart below. Duties often overlapped. Results of those discussions led to
adjustments of associates’ placement within each organization.
Execution Allocation
4.3
Sales
Replenishment
Forecasting
Coordination
Analysis
0%
25%
50%
Wal*Mart
Sara Lee
Figure 8
Technology Used
Our collaboration site was developed within Wal*Mart’s existing Internet-based
vendor communications system. We validated the industry specifications and
provided our feedback for the guidelines publication.
The VICS-EDI 830-transaction set (a subset of the ANSI X-12 standard) was
the data vehicle used to transmit the sales forecast between companies. Using
existing standards expedited setup of collaboration and eliminated development
time in managing multiple data exchange mechanisms.
Metrics
We used these metrics for the pilot:
• In-stock
• Weeks on hand at store level
• Forecast accuracy
• Lost sales
51
After 24 weeks of implementation, we realized a 2% improvement in retail store
in-stock, a reduction of 14% in store-level inventory compared to a 32% increase
in sales, and an increase of 17% in retail turns on the pilot items.
Execution Allocation
GMR0II
Weeks Supply
Pilot Overview
Retail Turns
Instock %
Retail Inventory
Sales
(20%)
(10%)
0%
10%
20%
30%
40%
Figure 9
Resources Involved
People from many areas in both organizations participated in the pilot, including
executive sponsors in Information Systems and in Sales/Replenishment from
both companies. Included in the initial pilot were individuals from Information
Systems (multiple application development teams), Forecasting/Replenishment,
Logistics, Marketing, Supply Chain and Sales.
No incremental staff was hired for the pilot.
In addition to the collaborative site, Wal*Mart’s Retail Link Decision Support
System was used for drill-down analysis.
52
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Project Challenges
Since the Front-End Agreement was not formalized, the initial stages of the
pilot determined both the metrics and the criteria for exception reporting (an
iterative process). Ideally, the Front-End Agreement would be part of the initial
merchandising phases of goods deployment; however, as this pilot revealed,
effective collaboration can occur at any stage of a product life cycle.
Although historically seen as barriers, inter- and intra-company system integration
(store- and distribution-level), production, capacity, and over-the-counter forecasts
exist either independently or at a less-than-optimum dependence and at varying
levels of detail. Until attention is given to integrating these forecasts–first at
manufacturer and retail level and then between the trading partners–realizing
substantial savings will be slow.
Varying levels of commitment and understanding exist throughout the organizations.
It became an inspiration during the collaborative process on the basis of its
definition: to work together, especially in a joint intellectual effort. As the process rolls
out and others learn from it, professional and business relationships will have
new standards.
Methodology
Review of the VICS Collaborative Planning Forecasting and Replenishment Voluntary
Guidelines established a framework for pilot execution. We began by agreeing
on metrics and targets.
We established a time frame for discussions. A telephone conference call was
held weekly for the first eight weeks; discussions continued every other week
thereafter, and we are currently on a monthly schedule.
Practical considerations clarified responsibilities as the pilot progressed.
Sara Lee provided the core analyses as the basis of discussion. After agreement
was reached, changes in tactics for either party (such as forecast) were executed.
53
plays an increasingly crucial role in collaboration. Promotion, replenishment
4.3
Areas of discussion included in-stock position, POS, forecast-to-forecast
comparisons, forecast accuracy (bias and absolute), promotional activity, and
product availability issues.
Here are depictions of some of the analyses:
1400
100%
95%
1200
90%
Pilot Overview
1000
85%
800
80%
600
75%
70%
400
65%
1997 POS
1997 In-Stock
1998 POS
1998 In-Stock
200
60%
0
55%
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51
Week
Figure 10
1200
SL FCST
WM FCST
1000
800
600
400
200
0
1
3
5
7
9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51
Week
Figure 11
54
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Discussions about product availability have always been difficult in a manufacturer/
distributor relationship. Through this process, difficulties were handled in a
professional and practical manner, creating winning results for both partners.
Commentaries on events that affected execution were summarized for planning
and to avoid their recurrence.
Product Availability Among Manufacturer/Distributor
4.3
WM Week
Items Affected
Type
Comment
7/3/98
9822
All
Service
Product availability
9/11/98
9832
Item A, Item B
Service
Change in specs and delay
in assembly; service resumed
on all other SKUs
10/1/98
9835, 36
All
Promotion
TAB - Same tab in 9735, 9736
10/16/98
9837
Item D Item I
Service
Should resume service
WM Week 9840; all other
SKUs back in service
Size X
Misc.
Cannibalization by similar
product
See Comment
Misc.
To increase % in-stock,
all but Item K (service)
and Items A-J (solid in-stock
position) will have an extra
1 week purchased to
jump-start in-stock
Rollback
X-day rollback Wal*Mart
will change profile by x%;
will change by y% on
y week of rollback
Promotion
Tab, all items
10/16/98
9838
Table 2
The pilot has been effective, and it met the original project scope of defining and
proving the guidelines as well as designing and developing systems for execution.
As with any large project, it was difficult to define the scope and work within it.
From the early stages of the pilot, one of the obstacles has been the need for
continual growth of the processes and systems.
The execution of the pilot was effective as evidenced by the improvement of
bottom-line results.
55
Date
There was increased communication between partners and within each enterprise.
Multiple operating areas built relationships through this cross-organizational
pilot, and now communicate their strategic efforts. Internal collaboration has
also improved between functional areas of both companies.
A real benefit not within the scope of this project was the redefining of business
relationships and commitment to ‘win/win’ actions.
Pilot Overview
The CPFR model and guidelines were validated and the basis for the execution
of this pilot.
Rollout Plans
CPFR will roll out to all applicable partners and products. As a continuing part
of the rollout of CPFR, development will continue to address scalability needs
for all types of manufacturers and to continue to automate the existing process.
As with any new process or system, on-going analysis will ensure the partners
remain up to standard and provide automated solutions.
56
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Procter & Gamble, Meijer, Target,
Wal*Mart, Sainsbury, and Tesco
Introduction
The following summarizes the objectives, methodology, measurement and
learning from the CPFR pilots undertaken with Procter & Gamble. Due to the
sensitive nature of some of the systems work underway, only certain customers
4.4
These pilots all contributed to establishing a standardized start-up process for
implementing CPFR. The partnerships provided input into the creation of the
CPFR Capability Assessment, which establishes the baseline evaluation of the
partnership’s four core CPFR processes in 18 key elements that are scored.
Completing the CPFR Capability Assessment allows a quick understanding of the
strengths and weaknesses of the partnership and actions needed to improve
the process. See Appendix A.
The pilots demonstrated that CPFR is not simply another form of category
management. Deployed as prescribed in these pilots, CPFR becomes the key
essential process to begin optimizing the supply chain. The focus of these pilots
is not simply to sell more product to the retailer’s distribution center or depot;
it was on selling more product to the consumer by concentrating on delivering
product efficiently and reliably to the retail shelf using retail point-of-sale data.
Read on, and gain an understanding of CPFR from these state-of-the-art retail
distributors in the U.S. and the U.K.
Procter & Gamble Worldwide
Procter & Gamble has operations in more than 140 countries with over 110,000
employees, and with worldwide net sales greater than $37.2 billion (1997/98).
Its 300 brands reach nearly five billion consumers. Procter & Gamble’s seven
Global Business Units include baby care, beauty care, fabric & home care,
feminine protection, food & beverage, health care, and tissues & towel.
57
Procter & Gamble Pilot
are specifically associated with measurements and learning.
• Collaborative Processes
• Integrated Planning and Forecasting Processes
• Replenishment Processes
• Supply Chain Management Processes
SUPPLIER
MANUFACTURER
4.4 Procter & Gamble
Pilot Overview
CUSTOMER
DISTRIBUTION
CENTER
POINT OF SALE
(POS)
Figure 12
58
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Objectives
Procter & Gamble’s CPFR focus is to build on the current success of the
Continuous Replenishment Program (CRP). CRP has delivered greater than
99% service levels, and has reduced customer distribution center inventories
by as much as 50% in customers representing over 40% of our U.S. and European
businesses.
While these accomplishments are substantial, they fall dramatically short of
4.4
consumer. Conservative estimates based on published studies show 8% to 10%
out-of-stocks still exist in retail stores, along with excessive inventory costs
throughout the supply chain. All become additional costs to the consumer.
P&G is deploying CPFR to enable creation and integration of consumer demand
data. This will trigger product flow from our manufacturing plants to our
customers’ DCs, from the customers’ DCs to their retail store shelves, and
ultimately from the store shelves into consumer homes.
The primary objective of these pilots is 100% product availability on the store
shelf, while simultaneously reducing inventory requirements in the retail stores,
customer distribution centers, and P&G plants. Eventually, P&G expects to
produce and ship in response to a consumer demand signal. These pilots will
test and validate methods that can help achieve this.
Our primary CPFR output concentrates on improving inventory and reducing
out-of-stocks. Traditional supply chain management decreased one but forced
the other to increase; trade-offs were made to deal with the lesser evil.
CPFR recognizes that the main causes of these two evils are identical:
1. Ineffective trust-based collaboration.
2. Ineffective planning using visibility of POS consumer demand.
3. Ineffective forecasting.
4. Ineffective product replenishment in response to demand fluctuations.
59
delivering the ECR promise of products available at the lowest cost to the
Each of the four causes builds on the preceding cause. Inventory or out-of-stocks
result when a supply source does not respond to fluctuations in product demand.
CPFR should make it possible to manage a continuous flow of product with no
delays. Achieving this is the essence of ECR.
Pilot Overview
Figure 13
$
Figure 14
60
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Methodology
The ongoing objective of the CPFR pilots is to test and validate the design
requirements and the changes needed to create a responsive, reliable, and
cost-efficient system that links manufacturing plants to customer DCs to retail
store shelves using POS information.
The key is understanding that CPFR is not a technology. It is a process. To test
and deploy new processes, our CPFR pilot partners agreed to three core activities:
2. Assess the current CPFR capability.
3. Create a joint action plan to address improvement opportunities.
Failure to follow the three-step process in order could lead the project down a
path toward unsatisfactory results. It was critical to include this process in the
Front-End Agreement, obtaining top-level consensus for the pilot’s strategies,
measures, and processes.
Below is a brief review of the three components of our methodology. The results
and learning from these components are detailed in the next section.
Supply Chain Lead-Time Mapping of Product and Data Flow
Together, team members from both companies traced product movement and
the signals that triggered it. All of the processes were mapped, and the time lag
between processes and triggers was measured from the point that a package was
scanned at retail to the point new product was replenished on the shelf.
CPFR Capability Assessment
Once the partners understood the supply chain process, there was no easy way
to translate the supply chain improvement opportunity into CPFR action. The
CPFR Capability Assessment was developed and used in some pilots to assign a
numeric value to each of the CPFR key processes. The actual scoring verified
the understanding from the supply chain mapping, and directed the creation
of a CPFR process improvement plan. Much of the information in this assessment
was compiled from previous ECR Scorecard documentation.
61
1. Document and map the current supply chain processes for product and data flow.
4.4
Joint Action Plans and Testing
This step combined the first two steps into a test plan. It was documented
and approved by the team sponsors, and the process improvement testing and
documenting began. Historical POS data was collected on the test category
(limited number of SKUs) and the POS data was continuously analyzed using
actual orders and shipments.
Metrics
Pilot Overview
The pilots measure aspects of nine elements:
1. Forecast Accuracy vs. Actual Orders
2. DC Service Level and Inventory
3. Retail In-Stock Service Level and Inventory
4. Manufacturer Order Fill Rate vs. Original Order
5. Manufacturer Order Fill Rate vs. ASN
6. Delivery Punctuality
7. Transportation Efficiency (Utilization)
8. Shipment Variability By SKU
9. Profitability / Cost Reduction
Summary of Learning and Results
There are three categories of learning and results: General, United States,
and Europe.
General Learning
This section reviews what was learned from all of the CPFR pilots.
All CPFR pilots recognized the need for a partnership founded both on trust
and on the ability and willingness to share information on processes and
systems. A joint learning process would lead to understanding how to improve
difficult-to-improve business results. It would not be a quick action to increase sales.
62
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
The companies involved in the P&G pilots are all competitors within their
respective marketplaces, yet have agreed to associate their involvement in the
CPFR pilot with P&G. This demonstrates the importance that each company
has placed on the value of CPFR to its future success. Once the processes are
understood and institutionalized, a critical mass of partnership involvement
between manufacturers and retail distributors will be essential.
All CPFR pilots recognize CPFR as a process, not a technical solution. To be
4.4
or standardized. Once this is accomplished, the process needs to become
repeatable and scalable so technology can deliver broad-scale capability. If an
existing process is delivering average results, technology will enable broad-scale mediocrity.
At the time of this report, each pilot was in a limited test using limited technology.
For this reason, most of the sharing of learning and results focuses on general
process changes. Specific references to internal proprietary opportunities,
processes, and systems have been avoided to protect the confidentiality and
to maintain the trust that made these CPFR pilots possible.
Lead Time Mapping & Capability Assessment Learning
All partnerships agreed on the need to reduce retail out-of-stocks, while
simultaneously managing the inventory levels required to remain in stock.
Lead-time process mapping offered the fastest understanding of the supply chain
processes and their results.
The value of this analysis was directly proportional to the detail of the
documentation, which highlighted where product movement was delayed and
pinpointed its causes. Non-value-added activity is anything the consumer would
be unwilling to pay for, and the process provided insight into the non-value-added
activity in customer and manufacturer systems and processes. The activity was
charted; before/after time analysis identified the loss for each delay.
At that point, both partners had a clear picture of their supply chain. Objective
decision-making about process changes could begin.
63
successful, the existing process must become either simplified, streamlined,
The results of the process mapping indicated a lead time range of 8.5 to 14.0 days
to replenish product. Variability in the supply chains were due to these key factors:
• Production Frequency
• Delivery Frequency
• Order Frequency
• Sequencing of communication processes (batch mode or overnight)
• Order Multiples
• Case, Unit Load, and Truck Load quantity requirements
Pilot Overview
• Product Mix requirements
• Delivery Lead Time
• Delivery Frequency
• DC Product Flow Through To Retail Store
• Product Flow From Back Room To Shelf
In the very early stages of most P&G CPFR pilots, it was feasible to remove at least
one day from the entire replenishment cycle. In the case of the 8.5-day cycle,
this represented nearly 12% improvement. In other instances, at least 20%
replenishment cycle time improvement was identified.
The CPFR Capability Assessment was a new concept introduced at the time of
this report, and it was not completed by all of the pilot partner companies.
The method found most productive was to have the manufacturer team members
reach consensus on the scores separately from the retail customer team. Once
each company reached its consensus, the two companies reviewed each question
together to agree on the score. Primary differences were documented along
with the rationale for each score.
64
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
This assessment process identified areas that needed immediate attention or that
could be sequenced later in the CPFR process improvement. Two important
concepts were critical to understanding:
1. All four activity areas must be maximized to fully realize CPFR benefits.
2. Partnering companies must have the flexibility to decide process priority.
U.S. Learning
U.S. customers generally scored highest in the area of Replenishment Processes
in the CPFR Capability Assessment, and view this area as a strength. It was
decided to focus initially in this area or in combination with Collaborative
Processes. Each customer used active views of POS data for their logistical
replenishment processes.
Promotional systems and processes for these retail distributors are completely
separate from daily replenishment (turn) systems and processes. Normal daily
turn demand volume is stable and predictable, so the complexity for both the
retail customer and the manufacturer is to determine the demand for promotions.
Numerous unique promotion systems depend on the type of merchandising.
Production scheduling does not look at the demand streams separately. A case
of product rolling off a production line does not care if it is for a promotion or
regular turn–it is heading for a customer’s retail store. Yet in order to determine
the total demand volume to be produced, the turn and multiple promotion
demand streams must be aggregated into a single production schedule. Therein
lies an opportunity to understand and simplify processes.
Most retail distributors handle promotion product differently from the turn
volume in their distribution systems. This creates unique opportunities to
understand the impact of dual processes on the measures of inventory and
out-of-stocks. As these pilots progress, the CPFR results will be documented.
4.4
65
Pilot Overview
Figure 15
Europe Learning
European pilots are leading a global effort by focusing on Promotion
Management as key to CPFR process improvement. Tesco and Sainsbury in the
UK have actively employed Internet-based data exchange technology to assist
robust planning processes in managing promotion creation through evaluation.
Tesco has aligned with General Electric Information Services (GEIS) and Syncra
while Sainsbury is in partnership with EQOS. These partnerships are working on
four promotion processes:
1. Promotion Proposal/Planning
2. Joint Forecasting
3. Execution/Demand Control (using daily POS sales/inventory data)
4. Promotion Evaluation
Each UK CPFR pilot seeks to improve inventory and out-of-stocks by engaging
in the Collaborative Processes and the Integrated Planning & Joint Forecasting
Processes described in the CPFR Capability Assessment. These Internet-based
collaborative systems combine and manage the customer’s internal demand-data
66
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
streams and enable on-line interactivity between the retail distributor and the
manufacturer. This gives immediate access to data as the forecasts become actual
sales activity during a promotional event.
The impact of this visibility on business processes and responsiveness was felt on
the EQOS system, where a promotion was being monitored. The system reported
a reduction in the in-store availability from 99% to 88% with sufficient lead time
to respond. The on-line view of promotional status saved three to four days of
4.4
On-line functionality enables efficient on-line input to the promotion proposal
and joint planning process. As the jointly planned event evolves, the system
captures the latest changes, eliminating confusion or miscommunication.
Finally, since the system monitors promotional sales progress, the results can be
saved and referenced for future event planning.
Forecast accuracy improvement of 20% has been recorded from this on-line
promotion management process. Work is underway within P&G’s internal
demand-planning processes to use the functionality now available from these
customers.
In Germany, early CPFR pilots focusing on Integrated Planning & Forecasting
Processes have dramatically improved forecast accuracy results with two different
retail customers. These pilots experienced 40% forecast accuracy “hit” improvement within the +/-15% range.
Technology Used
EDI Usage
1. EDI 830 information received:
– Order & sales forecasts
– Aggregated at DC & corporate levels
– Containing promotions with turn volume or just turn volume
67
out-of-stocks.
2. EDI 852: for DC withdrawal information to proportion DC forecasts for the
data aggregated to a corporate level
Data File Sharing
1. POS data is downloaded from extranet sites to use with sales planning
and execution
2. Diskettes of POS data compiled weekly
Software/Internet Applications
Pilot Overview
1. Excel to manage the data analysis
2. EQOS Collaborator
3. GEIS (TIE)
4. Retail Link (Wal*Mart)
Resources Involved
P&G
1. Overall corporate champion/sponsorship team
2. CPFR Project Managers: Business and Technical
3. Customer Business Development Team:
Sales Account Executive, Logistics, Systems and Retail Operations Managers.
4. Data Analysts: Two analysts for formatting and evaluating the
830 order forecasts and the POS data received separately.
5. Category Demand Planner
Customers
1. Company Champion/Sponsor
2. Overall Project Leader
3. EDI Business and Technical Managers
4. Demand Forecasting Managers
5. Category Manager/Purchasing Manager
6. Warehouse & Retail Operations Managers
68
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Hewlett-Packard Pilot
Executive Summary
Hewlett-Packard’s hard copy products (ink jet printers and laserjets) are
characterized by a short product life measured in months and sometimes in
weeks. Product which remains in the channel after a SKU becomes obsolete is
returned to HP for disposal. Accurately gauging future demand for production
4.5
minimizes lost sales opportunities as spikes in demand occur, and minimizes
the amount of obsolete product returned.
CPFR provides a mechanism for Hewlett-Packard’s trading partners to receive
information on overall channel inventories from HP, and for them to provide
HP with information on future sales activity that will be significantly different
from past activity.
The Front-End Agreements and joint business planning followed the normal
HP conventions already in place with its Tier One distributors. The CPFR
approach was introduced to all distributors at HP planning sessions. This pilot
incorporated i2 planning tools to create a forecast based on the sales and
inventory numbers reported by the distributors to HP via EDI. The CPFR web
tool was created by a consultant so the distributors could validate the sales and
inventory numbers used by HP and review the i2 forecasts. As the i2 forecasts
were based on sales history, the distributors would make modifications based on
planned promotional or other activities affecting future demand for a product.
The HP forecast, after review by the distributors, became the agreed ship quantity
for the given planning week for each product. These forecasts became “orders”
and reflected actual shipment quantities.
69
Pilot Objectives
• Develop an efficient and secure process for near-real-time collaboration
on sales, inventory, planned promotions, and planned production between
HP and its primary distributors.
• Reduce overall channel inventory while ensuring adequate stock levels and
inventory distribution in order to support base and promotional sales at
individual distributors, while minimizing returns at product end-of-life.
• Tie together SAP for fulfillment, with i2 for production planning and
with CPFR for distributor feedback on the output of SAP and i2.
4.5 Hewlett-Packard
Pilot Overview
Scope
The initial scope was limited to 63 actively managed SKUs of product with
irregular demand, short product life cycles, and, as a category, a high level of
returns. The initial pilot was limited to aggregate demand for each Tier One
distributor and not taken down to the distributor’s individual distribution centers.
Production scheduling for all of the SKUs was managed using i2 and order
fulfillment was performed using SAP. The supply for the SKUs was managed
by a single HP planner. The demand forecasting was performed by a dedicated
inventory manager/forecaster at each distributor.
Technology Used
Mainstream technology was used with a custom database and web GUI design
and development, including Informix Online Dynamic Server, HP Virtual Vault,
Java, and JavaScript. The database, now in its fourth generation, was developed
to follow the CPFR model while providing for integration of legacy, SAP, i2, and
EDI data (sales, inventory, forecast).
EDI
EDI 846 and EDI 867 were used for sales and inventory information.
70
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Database
Informix Online Dynamic Server running on HP-UX with full Informix database
capabilities was employed, including triggers, alerts, and stored procedures.
Informix alerts mechanism, for example, was used with a Java program to
automatically e-mail forecast changes (EDI or manual entry) and comments to
people based on the product line or distributor involved. The database could
grow dynamically as new SKUs, new distributors, and new CPFR codes were
created, and it could store variable length text (forecast comments) and images.
Confidentiality of data being accessed over the Internet was essential. Security
was patterned after systems created for the banking industry. The Web server was
running under HP’s Virtual Vault, a Trusted OS version of HP-UX behind Cisco
filtering routers (restricting access to both “outside” and “inside” data partitions).
Informix role and view mechanisms were used to restrict user access to database
SSL implementation to ensure the server’s identity and encrypt traffic on Internet
segments.
Internet
Users had secure, encrypted access via the Internet to a web server behind the HP
firewall. The Web server used customized Java programs to provide connectivity
to the CPFR database. Data could be safely retrieved, analyzed, modified (for
some fields such as forecast changes, with reason code and comments), and
sent back to the database.
Software
The development team used JavaScript to provide a full-function browser client,
able to perform spreadsheet-like functions while off line. User screens provided
dynamic list boxes, so as new product lines, products, distributors were added,
the boxes automatically incorporated them. As items were added, new rows were
automatically added to the screen. Based on user preferences, the time periods
and other filters (such as weeks of supply value) could be applied to zero in on
critical data.
71
Security
4.5
Manual data entry provided by the browser user interface was configurable
to restrict updates by data value, based on whether the time period was past,
current, or future. For instance, a user may be restricted from changing sales
manually for a future week, but allowed to do it for a past week; changes to
forecast were allowed for future weeks, but not permitted for current and
past weeks. The same mechanism could be used to restrict changes to EDI sales,
inventory, and forecast inputs.
4.5 Hewlett-Packard
Pilot Overview
Metrics
1. HP forecast accuracy vs. distributor forecast accuracy,
tracked at the distributor SKU level by week.
2. DC service level.
3. Promotional sales as a percent of total sales by product line.
4. Returns as a percent of units shipped.
Numbers cannot be released to the public, but the program has already had
sufficient success to be slated for expansion to two additional product groups,
representing over $15 billion in annual sales.
Resources Involved
The management of Hewlett-Packard’s Channel Logistics and Fulfillment
organization provided strong support for the development work on the CPFR
pilot. Support for the i2 integration and rollout to the Tier One distributors
was championed by both the product planning and channel marketing groups
within Hewlett-Packard. Without the strong support of the channel marketing
organization, the CPFR pilot would not have been successful.
No staffing was added to support the pilot. The CPFR technology developed
enabled HP’s existing planners to work more productively. Many hours were
saved from prior paper- and spreadsheet-based processes, which were both
error-prone and labor-intensive.
72
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Project Challenges
The two primary resource issues were IT support and data cleanup. Traditional
IT silo support does not have the range of skills required for the development
and deployment of web technology. A web application’s operation requires
the successful operation and optimization of servers, database instances, the
corporate network with its hubs, routers, firewalls, proxy servers, Java virtual
machines, JDBC or ODBC connectivity, etc. If corporate IT compartmentalizes
performance or troubleshoot problems.
The CPFR implementation and support at HP required the creation of a team
of individuals with diverse skills and a desire to learn new skills. The usual
sharp line between development and production support is less clear with web
applications, which are under constant development as technology changes
and user expectations and requirements grow.
Use of outside contractors to perform the design and development enabled
the latest technology to be used. Internal developers are more inclined to use
familiar technology, which does not necessarily produce the best product or one
that can continue to grow as web technology and user requirements change.
The CPFR database functioned as a planning data warehouse. Data from
transaction systems such as SAP always had integrity issues to be resolved before
it could be loaded into the CPFR database. Ongoing ownership of the integrity
of corporate data was critical. With data coming into the database from a wide
variety of internal and external sources, data validation was paramount to its safe
and profitable use. It was also the single greatest challenge in scaling up the pilot
to include additional trading partners, as additional SAP customer and item data
must be cleaned and imported into CPFR.
An essential ingredient of CPFR is a sophisticated process for generating and
manipulating forecast data. This was not initially in place at Hewlett-Packard or
at many of its distributors. Implementation of i2 provided the core data for
73
4.5
the management of each of these functions, it becomes very difficult to optimize
collaboration. As SAP was still being implemented at HP, there was a scarcity
of internal information technology resources to assist with i2 implementation
and CPFR development. Consultants from i2 were critical to filling the gap and
providing the necessary expertise for the pilot’s success.
Methodology
On a daily basis, EDI sales and inventory data was received from the distributors
and loaded into i2 and the CPFR database. On the weekend, the i2 planning
4.5 Hewlett-Packard
Pilot Overview
engine used this information to generate a forecast for the succeeding weeks.
On Monday and Tuesday, the Hewlett-Packard planners reviewed the forecast
and determined if there was sufficient existing channel inventory and production
capability to meet the demand forecast. The forecast was fed to CPFR and a
suggested ship quantity, by distributor by week, was fed into SAP. The shipment
quantities were for a specific production planning horizon; the i2-generated
shipment quantities, after collaboration, became the order quantities.
From Wednesday through Friday, the distributors were able to review the forecasts
and send to HP any adjustments for a planning period for a given SKU. Along
with the new value they selected a reason code and entered comments. The new
value, reason code, and comments were automatically loaded into the CPFR
database.
Each night, a program ran on the CPFR database and all current forecast
changes–with the original value, new value, reason code, and comments–were
sent as e-mails to the product planners. All changes made by a distributor during
the day were aggregated into a single e-mail message. The following morning the
planner was able to review the messages and determine whether the suggested
changes needed action, or whether the changes could be met by the current
overall production plan.
Changes for SKUs without an adequate supply in the channel were responded
to on a case-by-case basis, with confirmation of the new shipment quantities
74
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
communicated back to the distributors. Adjustments for SKUs with greater than
a preset week’s supply of inventory in the channel as a whole were automatically
incorporated into the production and shipment schedules for the coming weeks.
Changes by the distributors initially were required by end of business on Friday.
The processing by the i2 planning engine had been shifted to accommodate
changes made on the weekend.
4.5
The pilot was extremely effective in several key areas:
• The business process and benefits were demonstrated to a wide audience within HP.
• The required infrastructure changes were identified and have been put into place.
• The required technology to make this an efficient and secure process was fully
developed with four iterations, and now provides a model for future development
of business applications tied both to internal systems and to the Internet.
The CPFR implementation at Hewlett-Packard was developed in a manner that
would support both external forecasts generated by distributors as well as internal
forecasts generated by HP’s product lines. In some instances, the distributor
was better able to generate a forecast, and in others, a particular product line
or division had the better system.
It was critical to the acceptance of CPFR across all of HP’s business units to have
a technical implementation that could support a wide variety of business process
designs, irrespective of who generated the forecast and regardless of whether the
forecast was for sales or orders.
It was recognized that despite other business operation differences between HP’s
divisions and marketing channels, the data and collaboration involved was much
the same. CPFR became a common ground for merging diverse data sources that
could be shared both internally across organizational boundaries and externally
with HP’s distributors and logistics providers.
75
Implementation of a CPFR application requires full support by both the
manufacturer and the reseller throughout their organizations. This includes
support for the manpower and tools, database development, EDI or other data
transfer mechanisms, data cleanup, the Internet, and all the associated hardware
and software.
In particular, mission-critical web/Internet applications cut across these traditional
boundaries, and the traditional IS shop with its islands of expertise and
organizational silos may have great difficulty supporting and managing them.
4.5 Hewlett-Packard
Pilot Overview
With CPFR and business partner collaboration via the Internet, the IS support
group must adapt to be able to provide support for external, non-company
personnel. The IS group must also ensure that traditional processes, policies and
procedures designed for internal users on internal systems still provide the same
functionality and secure access to selected external users. A key issue is how to
monitor access and remove an external user’s access when they have ceased to
work at one distributor and are now working for another distributor and a
competitor of their former employer. The issues are technical, administrative,
and legal.
The principal change in the relationship was an increased trust that the supplier/
manufacturer–HP in this instance–would deliver the goods as committed
and on time. By providing distributors with a deeper look into HP’s supply
capabilities and overall channel inventories, HP engendered greater trust that its
actions were well-considered and appropriate. The pilot also put more emphasis
on ensuring that the data being exchanged was accurate and complete. This has
changed the way HP processes EDI inventory data received from the distributors
to reflect more accurately the product available for sale.
The overall collaboration process, with web tools, database alerts, and Internet
access, has become much more efficient. As a result, the HP planners, distributors’
buyers, and inventory managers are able to focus more on improving business
76
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
operations, spending less time on the mechanics of communicating,
and sharing supply and demand data, promotional, and related information.
The CPFR process model works as long as the underlying database retains a
sufficient level of abstraction to accommodate many forecast sources and locations.
With HP’s printer and PC business, there are products built by third-party
4.5
assemblers, using third-party components, stored in third-party warehouses,
distributor, who having sold the product to the end-user, reports the sale to HP.
With this complexity in the enterprise supply chain model, rigid coding for
traditional manufacturer/distributor relationships can increasingly fail to capture
exceptions. This was overcome by developing a fairly abstract data model that
could accommodate an unlimited number of locations, location types, and
location relationships.
Unexpected Benefits
Individual product lines realized that, although their business models had
significant variations, they could use the CPFR model for auto-replenishment,
co-managed inventory, or planning collaboration. This was a breakthrough at HP,
where each product line operates as an autonomous entity with its own business
and manufacturing processes, its own set of suppliers and distributors, and its own
marketing and information systems organizations.
The commercial and retail channels in particular have been managed by different
organizations within HP, and different fulfillment groups have been charged with
processing orders. As HP moves to improve and optimize its supply chains among
its various enterprises, it becomes crucial that the planning processes share a
common database.
A secondary benefit has been a lessening of the development of departmental
planning systems (often using spreadsheets or PC databases) and the resulting
77
for shipment to a distributor ... who may in turn “sell” the product to another
fragmentation and islands of information. Problems in gaining access to this
information have hindered the efforts of operational and logistical support
groups who need access to enterprise-wide forecast demand data.
Rollout Plans
The CPFR program has been expanded to include a co-managed inventory
initiative with 30 retail distributors of HP product. In addition, a buy plan or
order forecast pilot is being developed for the retail channel. Three program
4.5 Hewlett-Packard
Pilot Overview
managers are in place to move each of the initiatives forward both within the
Hewlett-Packard organization and with HP’s distributors.
In addition to expanding the scope of CPFR in terms of the sales channels
and distributors involved, additional product lines within Hewlett-Packard are
expected to use this technology to provide CPFR capabilities and optimize their
inventory utilization. This has been a multiple-step process, as the product lines
first come off the legacy systems and onto SAP, then planning solution is selected
(Red Pepper, Manugistics, and i2 are all in use by various groups within HP), and
finally a plan to integrate with the HP CPFR applications is developed.
78
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
CPFR Technical Specification
Electronic commerce is at a crossroads. For many buyers and sellers worldwide,
Electronic Data Interchange (EDI) has become the backbone for computerized
business-to-business communication. Meanwhile, the explosion of the Internet
has brought universal access and a host of new technologies. The danger is that
the benefits of electronic commerce standards such as EDI will be swept away by
the excitement of the Internet, leading to a future world of proprietary platforms,
incompatible APIs, and complex, unique collaborative trading practices among
buyers and sellers.
5
inter- and intra-enterprise collaboration through a common, pragmatic approach.
Rather than create a new standard, it leverages the legacy of existing standards
in broad use through the retailing and consumer manufacturing industries to
develop guidelines for collaborative business processes (e.g., EDI and U.P.C).
In this chapter, we explain the principles behind the CPFR technology
specification, describe the data format standards, outline the transport/network
protocol guidelines and security conditions, and present some application
architecture considerations.
Principles
We envision CPFR as a platform- and vendor-independent environment where
multiple parties can interoperate. Partners of different sizes and technical levels
can collaborate through accessible technologies, including Internet and the Web,
private Value Added Networks (VANs), dial-up, or other transport mechanisms.
This communication is supported by formal standards, which evolve through an
open process.
CPFR technological applications can take many forms. There are, however,
several requirements to which they must adhere:
79
C P F R Te c h n i c a l S p e c i f i c a t i o n
CPFR intends to ensure that the industry actually captures the benefits of
• Standards: The system must use existing standards wherever possible.
Where de jure standards have not been established, the committee has
selected de facto standards that have an open process, that are managed by a
non-profit organization, and are supported by multiple technology vendors.
If these criteria have not been met in an area, the committee has declined
to make a recommendation.
• Scaleability: The system must be able to scale to large implementations in
terms of number of products, trading partners, collaborative relationships,
users, and collaboration interactions.
5.0 CPFR Technical
Specification
• Security: Data security is a major issue in a collaborative environment.
For obvious reasons, sensitive information should be accessible only to those
with permission to view it. CPFR technology solutions must ensure data is
secure when exchanged via public networks, enabling robust collaboration
without revealing data to competitors.
• Open Design: Solutions that require a single vendor’s application are
not acceptable in collaborative relationships that have no locus of control.
There is no “master of the supply chain”; each trading partner must
independently consider all of its buyer/seller relationships. It is unlikely that
all of them would choose the same implementation approach. The technical
specification must be robust enough to support the entire business process,
yet simple enough to allow existing applications to connect with minimal
adaptation. By using open and published standards, new trading partners
can come online quickly and the systems can evolve. In addition, an open
solution must be based on mature technologies, because the rapid pace of
development and market acceptance can take evolving technologies in
diverging directions, including extinction.
• Manageability: A collaborative solution must be easily maintainable by all
parties. Custom solutions and annual software updates drive up costs,
introduce incompatibilities, and cause downtime. Again, with no locus
of control, a solution must be robust from the start, must not require
inordinate support services, and must be able to survive over time. CPFR
business practices should offer no technical or economic barriers for trading
partners, large or small.
80
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
• Resiliency: The technical solution must be resilient to failure, not only
in software, but also in the communications infrastructure. Redundancy
provides greater reliability and increased capacity. In the event of hardware
or power failure, the system should be fault-tolerant.
• Collaboration: Collaboration is more than messaging. With no central
planning body, a CPFR solution must support threaded peer-to-peer
communications among trading partners. It must also facilitate one-to-many
communications among participants. The solution will almost certainly
involve a combination of both human and machine input and output.
There should be automated access to data from decision support software,
by human operators.
Specification Approach
To derive the technical specification presented here, the CPFR technical
subcommittee took the following approach. First, it analyzed the CPFR business
process model and produced a set of data flow diagrams. From the data flow
diagrams, the team then developed a logical data model and a data dictionary
that illustrate the consolidated set of data elements and the relationships among
them. These were then compared with the existing data format standards
(ASC X12 and SIL) to identify an appropriate mapping and any gaps. The technical
subcommittee concluded its work by creating guidelines and considerations for
selection of transport, security, and application architecture mechanisms.
Technically, CPFR specifications, recommendations, and discussions of technical
implementation criteria fall into four areas:
• Data Format Standards: Data formats for messages to be exchanged among
CPFR trading partners, selected from the ANSI ASC X12 Electronic Data
Interchange (EDI) standard, and the Standard Interchange Language (SIL)
standard. CPFR data requirements will be included in the UCS and VICS
EDI implementation guidelines, which are subsets of ASC X12, and the
SIL guidelines.
81
execution systems, and so forth, as well as a facility for exception management
5
• Transport/Network Protocol Guidelines: Criteria for selecting the data
transport (e.g., FTP) and underlying network protocol (e.g., TCP/IP)
specifications for transmitting messages between CPFR trading partners.
• Security Considerations: Techniques for authentication, encryption,
non-repudiation, and origin of CPFR messages that implementations
should take into account.
• Application/Middleware: Alternatives for the location, coordination, and
management of the data processing elements (servers, agents, and other
5.0 CPFR Technical
Specification
components) that make up a CPFR implementation.
All CPFR implementations must use the data formats described in this
specification for message interchange. The selection of data transport,
security scheme, and middleware is beyond the scope of the CPFR standard,
however, and is subject to implementers’ agreements. The guidelines and
selection criteria provided here should help trading partners agree on which
approach to use.
Conceptual Model
The CPFR conceptual model has been divided into a set of process flow models,
data flows, a logical data model, and a data dictionary. The CPFR process flow
models are described in the “Future Process State” chapter. This section
describes the other models.
Data Flows
Each CPFR trading partner interaction produces a data flow, which is translated
to standard message formats and data transport requests in online implementations.
Table 3 summarizes each of the CPFR data flows.
82
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Data Flow Summary
Process Step
Data Consumed
Data Produced
Develop Front End Agreement
(None; manual process)
(None; manual process)
Create Joint Business Plan
Buyer’s Corporate Strategy
Seller’s Corporate Strategy
Joint Business Plan
Create Sales Forecast
Joint Business Plan
POS Data
Event
Sales Forecast Revisions
Sales Forecast
Identify Sales Forecast
Exceptions
Sales Forecast
Exception Criteria
Metrics
Events
Identified Exception
Items
Collaborate on Sales
Forecast Exceptions
Buyer’s Secondary Data for Exception
Items
Identified Exception Items
Seller’s Secondary Data for Exception
Items
Sales Forecast Item
Revisions
Create Order Forecast
Order Forecast Revisions
POS Data
Current Inventory on Hand
Inventory Strategy/Seasonal Info
Sales Forecast
Events
Product Historical Demand and Shipments
Product Availability Data
Item Management Profile Data
Order Forecast
Identify Order
Forecast Exceptions
Order Forecast
Exception Criteria and Values
Events
Identified Order
Exception Items
Collaborate on Order Forecast
Exceptions
Buyer’s Secondary Data for Exception
Items
Identified Exception Items
Seller’s Secondary Data for Exception
Items
Order Forecast
Revisions
Generate Order
Order Forecast
Item Management Profile
Order
5
Table 3
The following pages illustrate the high level processes and the data flows associated
with these processes. These data flows are independent of the architectural
implementation of CPFR. In other words, there could be different architectural
implementations of this model, such as peer-to-peer, hub-and-spoke, and third
party. For example, the CPFR data could be managed in a domain shared
(or potentially duplicated) by the buyer and the seller, or it may exist within one
of the two partner domains, or it could even exist within a third-party service.
The important point about the illustration is the agreement on the types and
formats of the data being shared and the nature of the data flows.
83
Develop Front End Agreement
The Develop Front End Agreement data flow illustrates the process of coming to
agreement on joint business goals for the trading partner relationship.
Develop Front End Agreement Data Flow
5.0 CPFR Technical
Specification
Buyer
Seller
Buyer’s
Business Goals
Seller
Business Goals
Develop
Front End
Agreement
Note: When an individual
diagram includes multiple
process steps, the steps
are numbered using the
Front End
Agreement
step numbers in the
CPFR
Policies
CPFR IDEF0 diagrams.
Legend
External
Agent
Process
Datastore
Data Flow
Figure 16
84
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Create Joint Business Plan
The Create Joint Business Plan data flow covers the exchange of strategy, objective,
and goal setting information between CPFR trading partners at the beginning
of a planning period.
Create Joint Business Plan Data Flow
Buyer
Seller
5
Business Goals,
Corporate
Strategies
Business Goals,
Corporate
Strategies
Create Joint
Category
Strategies/
Tactics
Buyer’s
Business
Plans
Seller’s
Business
Plans
Joint Category
Strategies/Tactics
Create Joint
Business
Plan
Item
Management
Profile
Joint Business Plans
CPFR
Policies
Figure 17
85
Create Sales Forecast
The Create Sales Forecast data flow describes the exchange of initial sales forecasts
for a planning period, based on agreed-upon goals, events, sales results, and
prior revisions.
Create Sales Forecast Data Flow
Historical
Sales
Data
Analysis
Buyer
5.0 CPFR Technical
Specification
Historical
Sales
Data
Analysis
Buyer
Event Dates
Seller
Seller Event
Dates
Create Sales
Forecast
POS Data
Sales
Forecast
CPFR
Policies
Joint
Business
Plan
Exception
Resolution
Data
CPFR
Transaction
Data
Figure 18
86
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Identify Sales Forecast Exceptions
The Identify Sales Forecast Exceptions data flow documents the creation and
exchange of exception items related to forecast performance that result from
criteria established in the joint business plan.
Identity Sales Forecast Exceptions Data Flow
Buyer
Seller
5
Forecast
Changes/
Update
Forecast
Changes/
Updates
Supply
Constraints
Identify
Sales
Forecast
Exceptions
Sales
Forecast
CPFR
Policies
Exception
Criteria
Exception Items
CPFR
Transaction
Data
Figure 19
87
Resolve Sales Forecast Exceptions
The Resolve Sales Forecast Exceptions data flow captures the exchange of sales
forecast item revisions created to resolve forecast exceptions.
Resolve Sales Forecast Exceptions Data Flow
5.0 CPFR Technical
Specification
Buyer
Seller
Undocumented
Knowledge/Data
Undocumented
Knowledge/Data
Buyer
Exception
Resolution
Seller
Exception
Resolution
Resolve
Sales
Forecast
Exception
Unresolved
Sales Forecast
Exception
Unresolved
Sales Forecast
Exception
Research
Sales
Exception
Research
Sales
Exception
Resolved
Exception
Items/Change
Sales Forecast
Buyer’s
Event
Dates
Seller’s
Event
Dates
Decision
Support
Data
Decision
Support
Data
Buyer
Knowledge
Base
Seller
Knowledge
Base
Exception
Items
Exception
Items
CPFR
Transaction
Data
Figure 20
88
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Create Order Forecast
The Create Order Forecast data flow describes the information exchanged in an
initial order forecast for products within a planning period.
Create Order Forecast Data Flow
Capacity Limitations
Buyer
POS Data
Seller
Historical
Demand &
Shipment
Data
5
Forecast
Impact
Events
Create
Order
Forecast
Current Inventory
Item
Management
Data
Sales Forecast
Exception
Resolution
Data
CPFR
Policies
Order Forecast
CPFR
Transaction
Data
Figure 21
89
Order/
Shipment
Data
Inventory
Strategy
Identify Order Forecast Exceptions
The Identify Order Forecast Exceptions data flow illustrates the information
exchanged when an order forecast triggers exceptions based upon joint business
plan criteria.
Identity Order Forecast Exceptions Data Flow
Buyer
Seller
5.0 CPFR Technical
Specification
Changes/
Update
Changes/
Updates
Ability to
Supply
Identify
Order
Forecast
Exceptions
Order
Forecast
CPFR
Policies
Static
Exception
Criteria
Exception Items
CPFR
Transaction
Data
Figure 22
90
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Resolve Order Forecast Exceptions
The Collaborate on Order Forecast Exceptions data flow captures the information
exchanged when CPFR trading partners create revisions to order forecasts in
order to resolve an exception condition.
Resolve Sales Forecast Exceptions Data Flow
Buyer
Seller
5
Undocumented
Knowledge/Data
Buyer
Exception
Resolution
Seller
Exception
Resolution
Resolve
Order
Forecast
Exception
Unresolved
Order Forecast
Exception
Research
Order
Exception
Undocumented
Knowledge/Data
Unresolved
Order Forecast
Exception
Research
Order
Exception
Resolved
Exception
Items/Change
Order Forecast
Buyer’s
Event
Dates
Seller’s
Event
Dates
Decision
Support
Data
Decision
Support
Data
Buyer
Knowledge
Base
Seller
Knowledge
Base
Exception
Items
Exception
Items
CPFR
Transaction
Data
Figure 23
91
Generate Order
The Generate Order data flow documents the transmission of a firm order for
products, based upon an order forecast and an item management profile.
Generate Order Data Flow
5.0 CPFR Technical
Specification
Buyer
Seller
Order
Order
Generate
Order
Item
Management
Data
Order
Forecast
CPFR
Policies
CPFR
Transaction
Data
Order
Figure 24
Logical Data Model
The consolidated set of data elements required to produce the CPFR data flows
and the logical relationships among them are included in the CPFR logical
data model. The data model is in relational form (using the IDEF 1X relational
modeling standard) to aid in the construction of Standard Interchange Language
(SIL) data format standard requests (discussed in the Data Format Standards
section of this chapter). Appendix C presents the entity-relationship diagram
for the model.
92
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
This model is provided as a reference; CPFR implementations are not required or
expected to include a physical database with this schema. Not all fields will be
used by each trading partner.
Data Dictionary
The CPFR data dictionary explains the meaning and intent of each logical data
element referenced in the CPFR data model. A common data dictionary guides
the mapping of CPFR data elements to data format standards such as ANSI ASC
X12 EDI and Standard Interchange Language (SIL). Appendix D contains a
5
Data Format Standards
Selection of Standards
Data format standards specify the order, types, and size of data to be exchanged
in files or messages. They do not specify how the data is transmitted or secured.
Standardizing the data format of CPFR messages ensures that implementations
can interpret communications received from CPFR trading partners. Each CPFR
message is specified in one of two data format standards: ANSI ASC X12 EDI
or Standard Interchange Language (SIL).
Use of ANSI ASC X12 Transaction Sets for CPFR Messages
CPFR uses ANSI ASC X12 transaction sets to exchange messages. To support the
Create Sales Forecast data flow, the Planning Schedule with Release Capability [830]
transaction set is used to transmit the sales forecast among trading partners.
To support the Actual Item Performance History Request message, the Item
Information Request [893] transaction set is used to request information
for specific products, locations, and time periods. The Product Activity Data [852]
transaction set supports the Actual Item Performance History Response message
and is used to transfer product activity or metrics among trading partners.
This data can be compared with the forecasts exchanged to determine forecast
accuracy. The Purchase Order [850] or the Grocery Product Purchase Order
[875] transaction set transfers the order during the Generate Order data flow
93
complete listing of CPFR data dictionary terms.
scenario. The Price/Sales Catalog [832] or the Item Maintenance [888]
transaction set is used to update product information to trading partners.
The Promotion Announcement [889] transaction set is used to inform CPFR
trading partners of upcoming promotional events. Appendix E maps CPFR
requirements to these transaction sets. Other EDI transaction sets provide
equivalent data, which could be used to implement CPFR. However, this
specification does not provide any mapping of them. (See the VICS EDI
and UCS implementation guidelines for maps.)
5.0 CPFR Technical
Specification
EDI standards as they currently exist do not provide transactions for some of the
collaboration scenarios represented in CPFR. For example, the establishment
of joint product goals for out-of-stock and inventory levels, exception conditions,
and a rationale for revisions made are all outside the scope of EDI. The CPFR
specification provides a mapping to Standard Interchange Language (SIL) for
these message types.
Use of SIL for CPFR Messages
The Standard Interchange Language (SIL) standard, maintained by the Uniform
Code Council (UCC), is a data interchange language based on ANSI Structured
Query Language (SQL) syntax. A separately maintained data dictionary defines
the standard table and field names that may be exchanged in SIL messages. SIL
benefits include the familiarity of SQL syntax, the self-describing nature of its
messages, and the ability to both update and query remote systems.
The SIL standard is not nearly as widely known or used as EDI. Since its introduction
in 1990, SIL has been used to respond to dynamic data interchange challenges
among grocers and grocery distributors (for example, distributing price changes
among disparate point-of-sale systems or querying stock positions).
The CPFR committee has chosen to use the SIL approach during the pilot phase
to document messages that are not within the scope of EDI today. SIL’s data
dictionary committee uses a streamlined approach to approve enhancements to
the standard very rapidly, providing a convenient forum for rapid standardization.
SIL is the only forum that the committee is aware of that would allow CPFR to
94
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
achieve standards body coverage for its entire specification, without requiring
the CPFR committee to maintain the standard as it evolves. When the
appropriate message types are added to EDI standards they will be considered
for use in CPFR.
The Extensible Markup Language (XML) is another data formatting standard
that the CPFR committee is considering for future versions of the CPFR guidelines.
Use of XML is contingent upon the presence of a standards organization that can
maintain the XML Document Type Definition (DTD) that the CPFR committee
would produce. To date, no appropriate organization is in place, but progress
5
Appendix E specifies the mapping of CPFR message requirements to existing
SIL data dictionary elements and identifies required elements for future SIL
extensions.
Transport/Network Protocol Guidelines
CPFR implementations must agree on the transport and network protocol to be
used. TCP/IP has become the de facto standard protocol for public networks;
private value-added networks also widely deploy IP. Other networks are technically
capable of supporting CPFR, including standard protocol stacks such as ISO OSI.
Since most contemporary computing architectures can support multiple protocols
simultaneously, and virtually all support TCP/IP, network selection is not likely to
be a major implementation concern.
The appropriate physical medium for transporting data depends on the
architecture selected and the level of service desired. CPFR-formatted messages
may be exchanged over the selected network in files, as data streams, or as blocks
of data delivered through a messaging system.
HTTP/S
The Hypertext Transport Protocol (HTTP) and its secure variant (HTTP/S) are
the transport specifications used by Web browsers to transmit Hypertext Markup
Language (HTML), Java applets, images, and other content to and from servers.
95
in this area appears imminent.
HTTP was not originally designed to transport highly structured data or to manage
complex interactions of clients and servers. However, its pivotal role in the
growth of the Internet has propelled its evolution. A number of EDI software
companies now offer Internet-enabled EDI, which allows EDI messages to be
exchanged over HTTP.
The advantage of HTTP is its wide acceptance. Most organizations allow
HTTP messages through Internet security firewalls. HTTP can also be used
synchronously, enabling interactive response.
5.0 CPFR Technical
Specification
FTP
The File Transfer Protocol (FTP) is a widely supported means of transferring text
and binary files among heterogeneous systems. Properly authorized users from
other organizations may transfer files to or from an FTP server and successfully
interpret their contents. Files can be easily organized and archived, and failed
transmissions can be easily reinitiated without data loss. It is more difficult to
support interactive implementations with a file-based approach, however.
Typically file exchange occurs as a background activity.
SMTP and MIME
E-mail systems and their related standards are another option. Simple Mail
Transfer Protocol (SMTP) is the de facto Internet standard for e-mail transmission.
While many e-mail systems support SMTP today, not all of them are interoperable.
Multi-purpose Internet Mail Extension (MIME) provides structured data types
for SMTP that could be used to format CPFR messages. Efforts are underway to
provide a MIME extension for CPFR.
Other Approaches
Sockets
Sockets are a data streaming protocol. More flexible and primitive than other
approaches, sockets require tight agreement among implementations concerning
the control flow, recovery, segmenting, and other strategies to be used. Sockets
are available on all computing platforms. They can provide high performance;
however, more programming is involved.
96
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Message-oriented Middleware
There is a relatively mature market of message-oriented middleware that provides
data transmission, security, assured delivery, multicasting, broadcasting,
publish/subscribe messaging, and other advanced options. Standards for
messaging products are extremely primitive. For this reason, implementations
using message-oriented middleware usually have to agree on a vendor to use
in common.
OMG CORBA IIOP
The OMG CORBA specification provides a protocol called IIOP (Internet
Inter-ORB Protocol) for making remote object requests and collecting results.
5
interactions. As a result, IIOP requires implementations to take the extra step
of mapping CPFR messages to objects. Future revisions of the CPFR specification
may introduce a distributed object model that could be used more effectively
with this approach.
Security Considerations
Selection of a specific security standard is outside the scope of the CPFR
technical specification. However, every implementation will need to consider
how communication will be secured, especially when messages are exchanged
over public networks. This section provides background information on security
issues that may be useful for those organizations contemplating CPFR initiatives.
Key attributes that should be accessed in a system that provides and controls
system access include authentication, authorization, integrity, confidentiality,
auditing, and non-repudiation.
• Authentication: The security system needs to authenticate remote users,
systems, and applets or other downloaded code. This means ensuring that
users connecting to the system are who or what they are supposed to be,
and ensuring that the remote systems to which users are connected are
the ones they meant to address and not impostors. There also needs to be
a means to ensure what is being downloaded is the correct code and that a
substitution has not taken place. Some standards have been established in
this area. For example, there is certificate code that supports X.509v3 and
provides an online Certificate Authority server to provide user authentication.
97
CPFR is currently defined in terms of data to be transmitted, rather than object
• Authorization: Assuming a user is “authentic,” the next area a security system
needs to address is the authorization to a system or application and the level
of access granted. This includes application privileges for field or record
level access, write permissions, and the ability to assign rights to other users.
With numerous users working at many different locations, an effective and
efficient process to administer users at a group level needs to be incorporated
into the security system. Ideally, each group would have an administrator
who would grant lesser levels of access and authorization to the other
members. By restricting the system access to a user and restricting the
5.0 CPFR Technical
Specification
scope of that access (database views vs. direct table access, for instance),
an administrator can minimize the potential damage from a system breach.
Most controls already in place rely on individual applications for containment.
Additional measures must be instituted to provide containment for a specific
system or application server(s).
• Integrity: A digital signature attests to the contents of a message as well as
to the identity of the signer. As long as a secure hash function is used, there
is no way to take someone’s signature from one document and attach it to
another or to alter a signed message in any way. The slightest change in a
signed document will cause the digital signature verification process to fail.
Thus, public-key authentication allows people to check the integrity of
signed documents. If a signature verification fails, however, it will generally
be difficult to determine whether there was an attempted forgery or simply
a transmission error.
• Confidentiality: At a system level, confidentiality means that all attributes of
a connection, including time, frequency, data, queries, and so on, remain
confidential. For CPFR, the primary focus is on the data being accessed,
received, and sent, regardless of the means used. At present, collaboration
is largely by phone and fax. The security with this approach is almost
nonexistent. The amount of data transmitted electronically is very small
or sent over secure private networks. As real-time access to data or
transactions becomes a business requirement and the medium becomes the
Internet, the requirements for confidentiality are best met through some
form of encryption.
98
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
• Auditing: The security system must provide for the auditing of access and
potential breaches. The audit trail provides a mechanism for assessing the
extent of possible damage and the nature of the attack, and for assisting in
the development of protection against future attacks. The system can also
provide a means for alerting administrators to attempted and actual breaches
of the existing security systems.
• Non-repudiation: Non-repudiation is key to ensuring that outside parties
involved in initiating transactions within the applications are tied to those
transactions. Authentication at the appropriate level of granularity ensures
a transaction. This is critical to generating business transactions using these
systems in place of traditional paper-based processes.
Application Architecture Considerations
The remainder of this chapter explains technical scenarios under which CPFR
could be deployed. Every network of CPFR partners will need to agree on the
location, coordination, and management of data-processing elements in their
implementation. Depending on each case scenario, the resources available,
and the need for expediency, the architecture may follow one of the application
architectures described in this document.
Implementation Scenarios
CPFR implementations can range from real-time, Internet-based applications
to file-based exchanges in batch over a VAN. E-mail and FTP (file transfer
protocols) are also viable. Data could be managed by each trading partner
or by an independent, third-party service bureau. Selection of the transport
mechanism depends on the level of service required, the availability of the
technology to each trading partner, and the ease of deployment. In all cases,
the message content exchanged over these transports should conform to the
CPFR selection of data format standards.
99
that users of the system cannot deny knowledge of, or responsibility for,
5
Some aspects of collaboration are time-critical. The level of sensitivity to response
times varies dramatically, depending on the products and trading partners
involved. The CPFR technical guidelines do not include formal specification
of any level-of-service requirements other than to identify which data may be
time-critical in the data flow model. Implementations must negotiate level-ofservice agreements and select transports, and design their applications to achieve
appropriate response times in each communication context.
There are many alternatives for distributing data and information processing
5.0 CPFR Technical
Specification
among collaborating trading partners. Each of these collaboration environments
presents significant tradeoffs. Examples of each are likely to prevail in different
industries, depending on the rate of market acceptance of CPFR.
Coexistence
Coexistence mirrors most electronic commerce schemes today. Each enterprise
operates its own autonomous network of applications and exchanges business
documents with its trading partners as a background activity, processing the
documents nightly or weekly. Figure 25 illustrates this type of trading partner
network. This style is peer-to-peer, but uncoordinated.
Trading Partners Networks Coexistence
Buyer
Seller
Seller
Seller
Figure 25
100
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
The advantage of these networks is that they are mature, widely deployed, and
frequently supported by the ANSI ASC X12 EDI message types required by
the CPFR specification. However, because of the focus on batch execution,
an environment based solely on coexistence is not well suited to time-critical
collaborative exchanges.
Centralized Server
In supply chains that have a few large and many small players, one company
can develop and manage a collaboration environment on behalf of its trading
partners. Because a significant part of any of the smaller companies’ production
develop or maintain the system, smaller companies are usually motivated to
adopt their partner’s technology. Figure 26 shows a centralized environment,
in which the server is managed by a buyer and clients are operated by sellers.
Centralized Trading Partner Network
Buyer
Client
Seller
Figure 26
Distributed Servers
A system built on a network of distributed servers can locally manage the data
relevant to any individual company and restrict distribution of sensitive information
to trading partners that need it. Distributed server networks are highly scaleable
because the number of systems grows with the number of trading partners.
Figure 27 illustrates a distributed server architecture.
101
flows through the larger trading partner, and the burden is not on them to
5
Distributed server networks are more complex than centralized or loosely coupled
systems, so they require more sophisticated engineering and management.
The distributed server approach also requires each trading partner to install
and maintain its own system, which could be an impediment to the widespread
adoption of CPFR. Data synchronization is the most challenging implementation
issue, whether performed via database replication, distributed transactions,
or other means.
Distributed Server Architecture
5.0 CPFR Technical
Specification
Seller
Buyer
Figure 27
Summary of Systems Architecture Options for CPFR
Architecture
Description
Strengths
Weaknesses
Coexistence
Uncoordinated batch
Minimal investment
Reaction time may
data transfer among
required
be too slow for true
trading partners
collaboration
Centralized
All data for a company
Server
and its trading partner
network maintained
its own server(s)
– Centralized
Partners lack visibility
management
– Ease of entry
for partners
control of data across
other buyer/seller in
relationships
Distributed
Peer-to-peer network
Each partner controls
Higher investment and
Servers
of servers installed
and has visibility
IT capability required
at each trading partner
of all its own data
Table 4
102
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Emerging Technologies
Over the past 10 years, a growing proportion of application programs has been
written using object-oriented techniques. Objects combine the data and behavior
of a business entity–such as a forecast or an order–into a single unit. Distributed
object environments allow individual objects to be accessed remotely as though
they were local.
Sellers and buyers have not deployed systems based on distributed objects in
any significant capacity to date. Competing specifications, relatively immature
products, and a lack of object-oriented programming skills on corporate IT staffs
so it merits considering how to evolve CPFR to a distributed object specification
in the future.
Distributed object approaches have a number of advantages over traditional
application protocol specification techniques. Special variants of a specification
can be created without modifying the base system through a process called
sub-typing. Object-based protocols are not restricted to data flows; actions that
are allowed to be performed on business entities can be explicitly specified
through object methods. Finally, data transport and formatting concerns are
simplified because distributed object specifications take care of these details.
The Object Management Group’s (OMG) Common Object Request Broker
Architecture (CORBA) is a public specification for distributed objects that has
been used by other standards groups. It has been available since 1991. Many
technology companies offer implementations of CORBA that can be used to
construct distributed object environments. A popular object-oriented programming language, Java, provides its own distributed object technique called Remote
Method Invocation (RMI), though it supports CORBA as well. Finally, Microsoft
has a distributed object specification called DCOM, which it submitted to The
Open Group to be maintained as though it were a standard.
103
have limited adoption. Many of these limitations are likely to disappear soon,
5
Conclusion
The CPFR technical specification is a work in progress. The CPFR technical
subcommittee has selected the combination of standards and conventions that
best meets the needs of supply-chain trading partners, based on the technologies
available. Technologies in this domain continue to evolve rapidly. To validate the
CPFR business process and continue to develop the technical recommendations
for CPFR, the subcommittee encourages members to conduct reference pilots.
Through the piloting process, the technologies outlined in this document will
5.0 CPFR Technical
Specification
be tested in the context of real business situations and under field conditions.
The committee then hopes to publish revisions to this document that will
enhance these guidelines and standards selections.
104
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Conclusion
Collaborative Planning, Forecasting, and Replenishment seems to be common
business sense. And it is–setting up front-end agreements for business
partnerships, sharing expectations and goals, and taking those expectations
into order fulfillment.
But it goes beyond a few simple ideas. Partners must also trust that they are
working in each other’s best interest. This sort of trust does not always exist in
business relationships; it’s a cultural change that must take place before CPFR
can be successful.
The pilots have proven that CPFR can deliver expected improvements such as
6.0
partner companies by reducing fixed assets and working capital.
But the most important benefit is that it serves to satisfy consumer demand.
In today’s hectic world, the consumer is time-deprived...unable to shop around,
yet demanding convenience and service. CPFR helps to fulfill that demand while
building customer loyalty to both brand and retailer.
This roadmap provides the direction to follow with CPFR, but it is not the end
of the road. It is still a challenge for partnerships to scale the CPFR process
to handle hundreds of items and distribution points without straining their
resources. Many pilots relied on manual techniques to uncover problems,
but getting to full production requires automation. The technical specifications
have been developed to help make that a reality.
Companies currently active with CPFR are progressing in the area of scalability.
Most expect to use the Internet to handle both the volume of information CPFR
requires as well as the timeliness it demands. They are strongly investigating an
industry standard for XML to help them in this pursuit. They are also looking at
CPFR beyond the original concept of manufacturer and retailer relationships,
forward to suppliers and transportation carriers.
These forerunners stand to benefit the most from CPFR. Rather than hunting
for the right path, it will give them a lead along the competitive road.
105
Conclusion
reduced inventory, higher in-stocks, and increased sales. It also benefits the
6.0 Conclusion
106
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Appendix A: CPFR Capability Assessment
Purpose
The purpose of the CPFR Capability Assessment is to...
• Provide a framework for understanding any gap between a company’s
existing practices and scalable CPFR best practices.
• Provide support for a preliminary business case that objectively
quantifies the process strengths and opportunities that exist,
and how the opportunities can be improved.
• Identify the starting point for change, and enable realistic
expectations for a CPFR best-practices development program.
• Provide an industry-standard process to determine if trading partners
A
are ready to engage in CPFR relationships:
– A company’s profile against an aggregated industry average.
– Other companies’ abilities to align/contribute to your specific focus
areas for improvement and development.
The grid on the next page provides an overview of the progression of the
activities through each CPFR process area. The grid illustrates the increasing
difficulty and benefit of progressing vertically through the processes of
Collaboration to Integrated Planning and Forecasting to Replenishment and
finally to Supply Chain Management. There are core competencies identified
that move each of these four processes through the Basic to Developing to
Advanced levels.
The ultimate goal of CPFR is Supply Chain Optimization, which can be achieved
only when advanced levels are achieved for every process area. CPFR is not only
about collaboration or computer-assisted retail replenishment. CPFR is the sum
total of all identified processes of Collaboration, Planning, Forecasting and
Replenishment. This assessment enables partners to understand the total
107
Appendix A CPFR Capability Assessment
– A company against itself, using before-and-after analysis.
journey, and to provide a way to flexibly map the work needed to meet specific
business goals.
The pages following the grid break out the elemental qualifications into
18 individually scored criteria. Each page represents one of the four process
areas of Collaboration, Integrated Planning and Forecasting, Replenishment
and Supply Chain Management. The scoring levels and many definitions were
taken directly from the Global ECR Scorecard in development at the time this
assessment was created. Some definitions were refined to better reflect the
Appendix A
CPFR-focused effort and results. Each criterion is scored based on the ability
to meet all criteria at a certain level. If all criteria are not met for a given level,
then the next lowest score is assigned (provided all criteria at that level are met).
The scoring grid on the last page allows input of the score for each element
within each process area. There is space to include comments and opportunities
for each score given.
Recommended Assessment Process
In actual CPFR pilot studies, the deployment method found most productive was
to have the selling team members reach consensus on the scores separately
from the buying team members. Once each company reached its consensus
on the individual scores, the two companies went through each CPFR assessment
element together to agree on the score.
Primary differences were documented along with the rationale for the score in
each area. The additional scoring rationale provided critical input to the actions
identified to improve the process for each element. When averaged and viewed
in the overall context of the scoring grid, the raw scores helped identify the
relative strength of the partners in each of the four CPFR process areas. At the
highest level, this information enabled the partners to set priorities on the CPFR
processes to be improved. At the detailed level, individual scores and comments
identified specific actions needed to improve the score on the particular element.
108
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
This assessment process identifies areas that need immediate attention or that
may be sequenced later in the CPFR process improvement. The key to this
process is understanding two critical aspects of CPFR:
1. All four process areas identified in the assessment must be
maximized to fully realize the benefits of CPFR.
2. Partnering companies have the flexibility to decide the priority
in which key CPFR processes will be worked and improved.
CPFR Stages of Progression
Process Area
Basic
Developing
Advanced
D
Supply Chain
No Supply Chain
Internal
Supply Chain
Focus/Plan
Enterprise
Optimization
Management
A
Optimization
Pre-DC
DC
Computer Assisted
Limited/ No Retail
Replenishment
Retail Ordering
Processes
Visibility
Focus
Flow-Through
B
Manual
Standardized
Integrated
Non-Standard
Demand Data
Planning &
Forecasting Planning
Creation & Input
Integrated
Forecasting
Planning,
Processes
Forecasting &
Collaboration
A
Limited
Standardized &
Collaborative
One-Way
Integrated
Processes
Communication
Collaboration
Table 5
109
C
Replenishment
A. Collaborative Processes
BASIC
DEVELOPING
ADVANCED
Criteria
Definition
0 Little to
Nothing
Done
1 Plans
Developed
Nothing Done
2 Partially
Implemented
3 Implemented
but Results
Incomplete
4 Fully
Implemented/
Effective
5 Advanced
Concepts
Joint Business
Planning
Process
The partnership
process to create
mutual plans
for promotion
and new item
events which
are measured
against commonly
understood
corporate
strategies.
Strategies exist on
a corporate level
only. There is no
sharing of annual
plans or evaluation
of current activities
against strategic
direction.
Annual plans &
strategies exist and
are shared, but no
specific plans are
created jointly with
the information
provided.
An overall joint
partnership strategy
is defined. Joint
plans are created on
limited items only.
Annual plans are
in place and
quarterly/6-month
plans are linked to
specific category
strategies.
Promotions and new
item events are jointly
planned and joint
sales forecasts
are created and
measured for accuracy.
Results of the events
contribute to the
success of the
annual plan. Reward
structures perpetuate
the JBP process.
Business-to-business
electronic connections
exist to update and
evaluate progress of
promotions against
annual business
plans. Results
are captured and
communicated
instantly based on
exception criteria
and alert functionality. Results are
archived for future
use.
Promotion/New
Product Initiative
Process
The process to
translate the
promotional
plan into
efficient/effective
implementation.
No consistent
process exists,
with frequent
changes and little
consideration of
lead times.
Systems are being
developed to formalize
the promotion
execution program
and there are some
efforts to coordinate
promotion execution.
There are some
efforts to implement
an effective
promotion process
with inconsistent
results for forecast
accuracy, in-stocks
and inventories.
A promotion
process has been
implemented and
is usually followed.
Out-of-stocks
and inventories
have decreased
consistently with
tested promotional
events.
A highly effective
promotion process
is routinely used.
Forecasts are
accurate and clearly
communicated
throughout the supply
chain. Inventories and
in-stocks are within
acceptable range.
A collaborative
promotion process
driven by a storebased POS data
forecast optimizes
the total supply chain
delivering minimal
disruptions within
the supply chain.
Results
Assessment
The degree to
which promotions
and new item
initiatives are
evaluated
jointly against
a common set
of objectives.
There are minimal
efforts to evaluate
promotion results or
promotion costs.
There is an
understanding of
the need to measure
promotions and
their costs, and some
basic level analysis.
Promotions are
evaluated based
on traditional costs
and volume basis.
Reviews with sellers
are conducted.
A broader set of
results and process
measures are
evaluated related
to category
objectives. Reviews
with sellers include
consumer data.
Measures are
evaluated jointly
with partners and
related to the JBP
category/corporate
objectives.
Promotions are
evaluated against
a common set of
measures. Results
are used to influence
future promotion
programs, co-marketing,
frequent shopper
programs, etc.
Table 6
110
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
B. Integrated Planning & Forecasting Processes
BASIC
DEVELOPING
ADVANCED
Definition
0 Little to
Nothing
Done
1 Plans
Developed
Nothing Done
2 Partially
Implemented
3 Implemented
but Results
Incomplete
4 Fully
Implemented/
Effective
5 Advanced
Concepts
Information
Technology
Utilization
The degree to
which information
technology has
been developed
to support
category
management
work and
measurement
systems.
There are basic
transactional
systems and limited
access to data.
Limited internal
customer POS data
is available and a
plan is in place to
utilize data from
outside the
organization, such
as from third-party
data providers.
An integrated
decision support
system is being
developed to
provide consumer,
customer, and
market data. Some
pilots are being
executed.
Integrated decision
support systems
exist, providing
data electronically
between trading
partners. Rulesbased models are
used that allow
for common
interpretation
of data between
trading partners.
Integrated decisionsupport systems
are used to provide
information to all
trading partners
as a basis for joint
evaluation of results,
facilitation of work
and execution of
plans.
Advanced technologies
(such as the Internet)
provide information
seamlessly between
trading partners to
enable collaborative
planning, evaluation,
and execution. Full
access to and use of
individual trading
partner information
(e.g., POS data and
shopper research)
drives joint business
planning and measures.
Demand Signal
Development
The degree
to which the
demand signal
matches true
consumer
demand.
Demand signals
are not used. Orders
are based on buying
cycles and deals.
Inventory information
is shared and used
to guide the demand
signal.
The demand signal is
based on warehouse
withdrawals and
inventories.
Forecasts of changes
in the demand
(promotions) are
applied manually.
Some consumer
forecast information
is incorporated
automatically into
the demand signal
based on historical
data.
Visibility of consumer
forecasts, warehouse
withdrawals and POS
data is used to adjust
the short-term
demand signal.
A collaborative,
integrated and
automated
communication
process exists, and
actual POS consumer
data is used to generate
the short-term
demand signal and
long-term forecast.
Weekly sales
forecasts are
created jointly with
limited internal
communication to
drive processes.
Joint weekly
sales and order
forecasts are
manually
communicated
internally and feed
production and
replenishment
processes with
limited or varying
degrees of
accuracy.
Internal
Integration Of
Demand
Forecasts
Score Twice:
1. Seller
Integration to
Demand
Planning
2. Customer
Integration:
Buyer
Replenishment
The degree to
which accurate
forecasts are
integrated
into existing
production and
replenishment
processes.
No forecasts are
either created or
communicated
across partnerships.
Sales event
forecasts are created
independently
between partners
and not shared.
Table 7
111
Joint sales and
order forecasts are
created and internally
integrated into
production and
replenishment
planning systems.
Accuracy
measurements are an
consistently within
acceptable range.
Joint sales and order
forecasts in daily and
weekly increments
are captured and
communicated, with
exception-management
processes in place to
respond to changes
as they occur.
Forecasts become
“frozen” at a point
that eliminates the
need for purchase
order creation.
A
Criteria
C. Replenishment Processes
BASIC
DEVELOPING
ADVANCED
Criteria
Definition
0 Little to
Nothing
Done
1 Plans
Developed
Nothing Done
2 Partially
Implemented
3 Implemented
but Results
Incomplete
4 Fully
Implemented/
Effective
5 Advanced
Concepts
Delivery
Reliability
The extent to
which the reliability
of deliveries is
measured to
ensure that orders
are delivered
on time, in full,
and in the right
quantity and
quality.
There are no
measures in place,
and no plans to
measure.
Measures are
identified and
responsibilities
are assigned.
There is regular
monitoring of
delivery performance.
Major problems are
regularly analyzed.
There is no close
cooperation
between trading
partners to address
problems.
There is close
cooperation
between trading
partners to ensure
consistent delivery
performance.
Delivery performance
is regularly
measured and
analyzed.
Achieved delivery
standards are
recognized as being
world-class, that is,
greater than 95%
of all orders are
delivered on time
in full.
Delivery results
are consistently maintained by solid, integrated work processes.
Performance is being
measured at orderline level and Line Fill
Rate is maintained
at greater than 99.5%.
A formal emergency
order process is in
place.
Efficient
Receiving
The extent to
which efficient
receiving practices
are used at the
warehouse.
There is no use of
technology (e.g., bar
codes, ASNs, unit
loads) or scheduling
practices. There are
high unloading
times.
Plans are being
developed to improve
the receiving process
using technology
(such as bar codes
and unit loads) and
improve scheduling.
Alternatives (such
as drop and hook)
to maximize dock
resources utilization
are being considered.
Use of technology
(such as bar codes),
handling of unit
loads via third-party
pallet pools and dock
scheduling systems
are being tested.
Drop and hook
capability exists and
is used to maximize
carrier turn-around
and dock resources
use.
Bar codes (such as
UCC 128), unit loads
using third-party
pallet pools, and
dock scheduling
are being used
to receive and
ship product.
Unloading times,
dock utilization and
carrier turn around
have been improved
significantly.
There is joint planning
of shipments using
mutually efficient
loading configurations
& receiving techniques.
There is full use of
advanced techniques
(such as bar codes,
ASNs, and unit
loads/pallets) throughout the DC to receive,
sort, and ship product.
An understanding
of the full costs/
benefits of working
with efficient
sellers is reflected
in warehouse
handling and is used
to influence corporate
decisions.
Retail Process
Reliability &
Compliance
to Standard
The extent to
which standards
are in place to
manage data
integrity, promotional ordering,
performance
compliance,
shelving and
display schematics,
and replenishment
from backroom.
No standards are
in place to manage
processes uniformly
throughout all
stores.
Standards exist but
are not fully enforced
or measured.
Standards are set
and targets are
in place to begin
measuring.
Alignment to change
in direction of
growth.
Standards are set
and targets are fully
aligned by all being
measured. Tests
are underway to
improve key lead
sites (stores)
as examples of
successful
deployment.
100% compliance
is achieved for all
identified key retail
process standards.
Predictability of
promotional sell–
through is not
impacted by issues
of retail compliance.
Rigorous and regular
reviews of the standards
occur to accommodate
specific market or
store sales results.
Store inventory and
in-stock data drives
changes and input into
standards for shelf–
turn and promotional
replenishment.
Automated
Store Ordering
The extent of
automation
within the store
stock and order
systems.
There is a wholly
manual system to
track inventory and
re-order product.
There is partial use
of handheld devices
to record shelf
inventory needs,
combined with
manual inventory
to calculate stock
re-order.
Hand-held devices
are routinely used
to record shelf
inventory needs,
combined with
manual inventory
measures to
calculate stock
re-order.
Scanning and
perpetual inventory
systems replace
manual counting
for store re-stocking.
Adjustments are
anticipated and
applied to order
calculation.
Scanning and
perpetual inventory
systems are used.
Re-order cycles
can be adjusted
based on category
requirements.
Store-level systems
have been developed
to use POS data to
drive store-level
forecasting, which
drives the supply
chain.
Replenishment
Process
The extent to
which the process
used to replenish
stock is driven
by true consumer
demand and
managed all the
way to the retail
shelf.
Changes in inventory
are primarily driven
by deals and price
fluctuation.
Performance
measures are not
identified.
There is an agreement
between trading
partners to begin
developing a system
to drive replenishment
based on some measure of demand.
Some performance
measures have been
identified.
Piloting some form
of demand-driven
replenishment
(RMI or VMI –
CRP) and utilizing
EDI standards is
being done. An
agreed-to set of
performance
measures is being
tracked.
The demand-driven
replenishment system
is being rolled
out with some
improvements in
inventories and
performance
measures. Focus
is on retail results
as well as
DC/warehouse.
A fully implemented
demand-driven
replenishment
system resulting in
high service levels
at DC and retail
stores is in place.
Out-of-stock and
inventory levels are
consistently within
acceptable levels.
The replenishment
system is automated
to use true consumer
POS data. Store shelf
replenishment is the
primary focus.
Promotions and
new items are fully
integrated into the
replenishment
process.
Product Flow
The degree that
physical product
flow techniques
(such as crossdock, flowthrough, DSD,
customer pick-up,
third-party
consolidator)
have been
implemented
within the
distribution
network.
No strategic choices
have been made to
manage product
flow. Shelves are not
designed to support
efficient operation.
There is no labor
cost data.
Strategic plans have
been developed to
implement product
flow techniques.
Benefits are
understood at
conceptual level.
There are fundamental
back-room-to-shelf
processes in place.
Simple pilot
evaluation of the
different product
flow alternatives
have been completed
with only a few
partners. Plans are
in place to improve.
Clear criteria on
when/how to use
the different product
flows has been
established.
Implications of
cross-dock or flowthrough systems
are understood and
broader application
is occurring with
strategic partners
primarily for
promotional surges.
Costs/benefits of
the different product
flow alternatives
are understood
and reflected in
distribution costs.
DC layout and
distribution
capabilities
support systemic
implementation.
Cross-docking, flowthrough is standard.
Process for suitable
SKUs.
Decisions on how
to flow product are
based on retail needs
to support category
initiatives and drive
consumer value.
All products suited
for cross-dock or
flow-through are
handled this way.
Partnership programs
are in place to
evaluate supply
chain optimization,
and partnerships
delivering optimization
are valued.
Score Twice:
1. Seller to
Buyer.
2. DC to retail
Store Shelf
Score Twice:
1. D.C. flow
2. Retail Flow
to Store
Shelf
Table 8
112
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
D. SUPPLY CHAIN MANAGEMENT
BASIC
DEVELOPING
ADVANCED
Definition
0 Little to
Nothing
Done
1 Plans
Developed
Nothing Done
2 Partially
Implemented
3 Implemented
but Results
Incomplete
4 Fully
Implemented/
Effective
5 Advanced
Concepts
Partnering
and Trust
Relationships
The synergy
of strategic
relationships
between buyers/
sellers including
alignment to a set
of measures.
There is no strategic
relationship, only a
traditional buyer/
seller relationship.
The need to work
closer together and
to establish common
measures is
recognized.
A commitment is
made to work together,
piloting programs on
“event-focused” plans.
Internal and external
measures are .
reviewed.
Commitment is
demonstrated, with
widespread efforts to
focus on joint efforts
between trading
partners, including
a common set
of measures.
There is routine joint
multi-functional
efforts focused on
the total system
using a standard
set of joint measures.
Corporate
organizations are
aligned with key
trading partners to
maximize mutual
regional/local
objectives aligned
to corporate strategies.
Business
Process
Re-Engineering
The ability to
jointly understand,
communicate and
document existing
CPFR related
business processes
ultimately leading
to mutually
beneficial process
change.
There is no focus
on improvement.
Business results are
maximized within
current system and
process constraints.
Re-engineering plans
are in place for
internal process
enhancements.
There is no external
consulting or
understanding of
impact.
Internal re-engineering
is being deployed.
There are no measures
of external partner
impact. There is initial
influence on other
departments within
the enterprise to
maximize impact.
The need for external
partner involvement
to achieve desired
results is recognized.
Internal optimization
of business process
results is achieved
across multiple
departments within
the enterprise.
Re-engineering plans
are in place resulting
from joint business
process mapping with
external enterprises.
Tests are underway
to establish best
practices for internal
and external
partnerships.
Systemic improvement
involving policy
changes are underway
to improve the costs
or efficiencies across
enterprises. Measurements
to evaluate the
effectiveness of the
supply chain are in
place. Ongoing re-inventing
process is scheduled.
Operating
Strategy:
– Service Level
– Inventory
Definition of an
operating strategy
that maximizes
service level, and
balances total
supply chain
capabilities, cost,
and inventory
levels.
Inventory levels are
managed based on
experience, with no
communication with
suppliers on results.
No targets are set.
There is no link to
service level (product
availability).
Inventory levels are
managed based on
experience with no
communication with
seller on results.
Targets are set (no
relationship exists
with supply chain
capability and desired
service levels).
Inventory targets are
statistically set based
on product segmentation
(ABC items). They are
linked to the desired
service level. Internal
improvements are
being made to lower
inventory costs.
There are shared
targets, statistically set
based at SKU level,
demand variability at
DC, optimum service
levels among stores,
and total supply
chain capability and
cost. Safety, cycle,
and anticipation
components (and
accountability) are
recognized.
There are frequent joint
reviews of service
levels and targets,
based on SKU level
supply and demand
variability at the store
level; systemic joint
efforts are underway
to maximize service
level, lower inventory
and total supply
chain cost.
Service and inventory
targets are totally
aligned with category
management roles at
the SKU level. Efficient
assortment, promotions
and new product
introduction plans are
properly integrated
into the inventory
strategy and plan.
Measurement/
Reward
Measurements &
reward systems
consistent with
supply chain
efficiency objectives,
goals, strategies
that optimize
the entire supply
chain without
suboptimization of
individual enterprises.
Measurement is
internally focused.
Managers are
measured and
rewarded on
traditional measures
(gross margin, sales,
efficiency, etc.).
Measurement is
internally focused.
Plans are developed
to enhance measures,
along with a
beginning look at
broader measures
on an ad hoc basis.
Measurement is
internally focused.
Broader measures
are being developed.
There is routine use
of some limited
broader measures
(such as market
share).
Measurement is
externally focused.
Managers are
partially measured/
rewarded on a
“balanced” scorecard
(consumer, customer,
and market
measures).
Measurement is
externally focused.
Managers are
measured and
rewarded on their
ability to meet category
objectives that are
driven by a balanced
scorecard. Supply
chain effectiveness
measures are in place.
Measurement is
externally focused.
Managers are
measured and rewarded
based on their ability
to meet corporate and
category objectives.
Supply chain
efficiencies have
been achieved and
documented.
Activity
Costing
Process
The use of
activity costing
techniques to
analyze business
processes and
results between
trading partners.
No activity costs
are collected.
Gross margin is
used between
trading partners.
Plans are in place to
begin using activity
costing concepts.
Joint planning is
still based on
traditional measures.
Activity costs are
measured on part
of the company’s
cost structure and
are used for a
limited amount of
joint planning.
Activity costs are
beginning to influence
joint planning.
All business activity
costs are measured
on an ad hoc basis
and are used for
management planning
as the need arises.
An activity cost
modeling approach
is occasionally used
with consumer,
customer, and
manufacturer data
as the basis for
joint planning.
All business costs are
reported on an activity
basis. Activity costs are
measured and updated
as necessary. Activitybased costing information
is reported as part of
standard management
accounts. ABC modeling
is routinely used with
consumer, customer,
and manufacturer
data for planning.
All information systems
have incorporated
activity measurement
techniques as the
standard. All joint
planning leverages
the use of activity
costs.
On-Shelf
Product
Availability to
Consumers
The extent to
which the on-shelf
availability of
product is
measured and
managed in the
retail stores.
On-shelf availability
is not measured
and there are no
plans to measure it.
Plans are developed
to measure on-shelf
availability, but are
not implemented yet.
is measured from
time to time with
ad hoc actions to
improve it.
is routinely measured.
Performance is
analyzed to identify
root causes, with
actions taken to
systemically improve.
On-shelf availability is
tracked continuously.
Actions are taken
immediately to correct
out-of-stocks. Products
are never missing from
the shelf if available in
back room of store.
Back room & shelf
on-hand quantity are
distinguished and
managed.
Product availability
results are consistently
maintained by solid,
integrated work
processes. An early
warning system is in
place to prevent
out-of-stocks.
Table 9
113
A
Criteria
CPFR Capability Assessment Scoring Grid
A Collaborative Processes
Score
Comments/Opportunity Identified
Score
Score
Score
Joint Business Plan Process
Promotion/ New Item Initiative
Results Assessment Process
Total Score for Section
Average Score for Section
B Integrated Planning & Forecasting
Information Technology Utilization
Appendix A
Demand Signal Development
Internal Integration Of Demand Forecasts (Seller)
Internal Integration Of Demand Forecasts (Buyer)
C Replenishment Processes
Delivery Reliability Process (Manufacturer to Customer)
Delivery Reliability Process (DC to Retail Store)
Efficient Receiving Process
Retail Process Reliability and Compliance
Automated Store Ordering
Replenishment Processes
Product Flow (DC/Warehouse)
Product Flow (Retail Backroom to Shelf)
D Supple Chain Management
Partnering and Trust Relationships
Business Process Re-Engineering
Operating Strategy: Service Level and Inventory
Measurement/ Reward
Activity Costing Process
On-Shelf Availability To Consumers
Total Assessment Score
Average
Table 10
114
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Appendix B: CPFR Sample Metric Matrix Worksheet
Objective
The objective of this worksheet is to facilitate the discussion on performance
metrics for CPFR. It is intended to help trading partners collaborate on the
various metrics that will be tracked over the course of collaboration.
General Information
Selling__________________________ Company ______________________ Rep ______________________
Buying__________________________ Company ______________________ Rep ______________________
Completed Date ________________________________________________
Metric Analysis Matrix
Questions
Frequency
or Interval
MAD/Mean
Daily
Time
Horizon
or Period
14 Days
Granularity
Variance
Threshold
Relative
Impact
SKU
8%
High
1)
2)
3)
4)
5)
6)
Legend
Specific Formula
Agree upon and enter the specific mathematical formula for the performance metric.
Time – Frequency
"Every metric should have an element of time within the calculation. Define the Frequency
(daily, weekly, etc) or the interval (snapshot value every x days)."
Time – Horizon
Some metrics (such as forecast accuracy) require that you set an agreed upon horizon
or period for which the measurement is applied.
Granularity
Each metric needs to have a specific level of detail defined to ensure that the two parties are
measuring the same thing. For example item level accuracy verses category level accuracy.
Variance Threshold
"Metrics, like forecasts can vary between parties, and they should be monitored for variances
between the numbers reported by the two parties."
Relative Impact
It is important to agree upon the relative importance of each of the metrics. This may differ
between the two parties and should be noted.
Benchmark– Initial Value
Once the calculation has been defined each party should calculate an initial position and
record it to track progress against.
Benchmark – Target Value "Even through the process may do better than initial targets, it is important to have an
agreed upon target value for the metric."
Data Provider
"Identify who will provide the data for the metric, the majority of metrics will be provided
by both parties and compared, however some metrics may have only one party capable of
providing the information."
115
Benchmark
Initial
Target
Value
Value
71.20%
75%
Data
Provider
Both
A p p e n d i x B Metrics Matrix
Example
Forecast Accuracy
Specific
Formula
B
Appendix B
Metrics Matrix
116
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Appendix C: Data Model
This Data Model includes the set of data elements required to produce the CPFR
data flows and the logical relationships among them. The data model is in
relational form, using the Crow’s Foot ERD notation. This model is provided
as a reference. CPFR implementations are not required or expected to include
a physical database with this schema. Not all fields will be used by each trading
partner. Each site that implements CPFR will want to refer to this Data Model
and determine how to best implement it in its environment.
Notations:
• An entity is represented by a rectangular box. Each entity has a name,
which is a noun or noun phrase.
• An entity has one or more attributes whose values uniquely identify one
entity from another. Each attribute of an entity is represented by listing it in
key. Foreign Key (FK) indicates that the attribute is the key in another entity.
• A connection relationship is represented by the presence of a Foreign Key
attribute in a child relation.
The following minor corrections have been made since the publication
of the VICS CPFR Guidelines in 1998:
• All mentions of a “Unit of Measure” were consolidated, and a Unit of Measure
entity was created.
• New entity Forecast Type was created to portray whether a forecast
(item level) was total, promotional, seasonal, or fill/replenishment.
• New attribute Forecast Revision Comment was added to Forecast Revision
Item entity to reflect any additional comments made about the reason
for a forecast revision.
• The relationships of Forecast Purpose were changed for modeling correctness
so that Forecast Purpose directly links to Forecast Header, and migrates to
several other entities.
• Changed attribute; Exception Value Comparison renamed Exception Value
Comparison Code.
• Changed attribute; Planned vs. Actual renamed Planned or Actual Code.
117
A p p e n d i x C Data Model
the box. Elements above the line indicate attributes which are part of the
C
Unit of Measure
UOM Code
Operational Exception Criteria
UOM Description
Seller ID (FK)
Buyer ID (FK)
Standard Product ID (FK)
Criteria Type Code (FK)
Effective Date
Buyer Location ID (FK)
Seller Location ID (FK)
UOM Code (FK)
Exception Value
Collaboration Priority Code (FK)
Exception Value Comparison Code
Expiration Date
Operational Result
Time Period Begin Date
Seller ID (FK)
Buyer ID (FK)
UOM Code (FK)
Time Period End Date
Metric Value
Appendix C
Data Model
Product Activity
Standard Product Type
Seller ID (FK)
Buyer ID (FK)
Item Quantity
Standard Product Type Code
Standard Product Type Description
Buyer ID
ID Type Name (FK)
Buyer Name
Buyer Location
Seller ID (FK)
Buyer ID (FK)
Standard Product ID
Standard Product Type Code (FK)
Effective Date
Seller ID (FK)
Buyer ID (FK)
Expiration Date
Order Multiple Quantity
Minimum Order Quantity
Lead Time Days
Order Interval Days
Rounding Rules Description
Transportation Strategy Description
Safety Rules Description
Frozen Period Days
Seller ID (FK)
Buyer ID (FK)
Event Type Code (FK)
Event Begin Date (FK)
Event Item Impact
Business Type Code
Event Impact Type Code (FK)
Seller ID (FK)
Buyer ID (FK)
Planned or Actual Code
UOM Code (FK)
Impact Value
Impact Effect Code
Seller Location
Seller Location ID
Seller ID (FK)
ID Type Code (FK)
Business Type Code (FK)
Data Source
Criteria Type
Criteria Type Code
Event Begin Date
Event Source Code (FK)
Event End Date
Event Description
Event Create Date
Event Item
Business Type
ID Type Description
Collaborative Priority Description
Event
Event Impact Type Description
ID Type Code
Collaborative Priority Code
Forecast Type Description
Data Source Description
Event Impact Code
Business Type Description
Collaboration Priority
Forecast Type Code
Data Source Code
Event Impact Type
ID Type
Forecast Type
Event Item
Event Type Code
Event Category Code (FK)
Event Type Description
Criteria Type Description
Forecast Item
Seller ID (FK)
Buyer ID (FK)
Forecast Generation Date (FK)
Forecast Purpose Code (FK)
Data Source Code (FK)
Forecast Type Code (FK)
Item Quantity
Participating Locations Quantity
Forecast Change Restriction Indicator
C
Forecast Revision Item
Time Period Begin Date (FK)
Seller ID (FK)
Buyer ID (FK)
Revision Generation Date (FK)
Forecast Revision Comment
Revised Item Quantity
Adjustment Reason Code (FK)
Event Category
Event Category Code
Event Category Description
Forecast Item Exception
Forecast Header
Seller ID (FK)
Buyer ID (FK)
Forecast Generation Date
Forecast Sent Date
Forecast Start Date
Forecast End Date
Forecast Revision Header
Seller ID (FK)
Buyer ID (FK)
Revision Generation Date
Revision Sent Date
Forecast Change Status Code
Seller ID (FK)
Buyer ID (FK)
Effective Date
Adjustment Reason
Adjust Reason Code
Adjust Reason Description
Forecast Purpose
Seller
Seller ID
ID Type Code (FK)
Seller Name
Forecast Purpose Code
Forecast Comparison Exception
Forecast Purpose Description
Seller ID (FK)
Buyer ID (FK)
Comparison Forecast Generation Date (FK)
Comparison Data Source Code (FK)
Effective Date
Operational Result Exception
Seller ID (FK)
Buyer ID (FK)
Effective Date
Figure 28
118
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
119
A p p e n d i x C Data Model
Buyer Location ID
Buyer ID (FK)
ID Type Code (FK)
Business Type Code (FK)
Seller ID (FK)
Buyer ID (FK)
Effective Date
Time Delta Days Quantity
UOM Code (FK)
Exception Value
Collaboration Priority Code (FK)
Exception Value Comparison Code
Expiration Date
Collaboration Item
Safety Stock
Buyer
Forecast Item Exception Criteria
Appendix C
Data Model
120
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Appendix E: Message Mapping
EDI Transaction Maps
This section documents how data required by CPFR data flows maps to existing
ANSI ASC X12 EDI transaction sets. Where there are gaps, members of the
technical subcommittee are actively working with the appropriate standards
bodies to update the EDI specifications to ensure complete coverage in the near
future.
Planning Schedule with Release Capability [830]
The 830 transaction set corresponds to two CPFR messages: the sales/order
forecast and the sales/order forecast item revision. As of Version 004030
of the ASC X12 standard, some CPFR-specific extensions of the 830 have been
on Version 004010.
See Page 140 for Message Mapping Charts.
139
E
A p p e n d i x E Message Mapping
approved. These are shown in italics. All other mapping information is based
Sales/Order Forecast Message Mapping
Entity
Attribute
Segment
ID
Reference
Designator
Data
Element
Data
Type
Min/ Code(s)
Max
Forecast
Header
Seller ID
N1
01
98
ID
2/2
Seller
Location ID
N1
04
67
AN
2/80
Buyer ID
N1
01
98
ID
2/2
Buyer
Location ID
N1
04
67
AN
2/80
Forecast
Generation
Date
BFR
08
373
DT
8/8
Forecast
Sent Date
ISA
09
I08
DT
6/6
Forecast
Start Date
BFR
06
373
DT
8/8
Forecast
End Date
BFR
07
373
DT
8/8
Forecast
Type Code
BFR
04
675
ID
2/2
BB=Sales
DL=Order
Forecast
Source
Indicator
ISA
06
I06
AN
15/15
ISA05 will
contain a
qualifier
identifying the
type of ID
Standard
Product ID
LIN
03
234
AN
1/48
LIN02 will
contain code
UK or UP
Quantity
Type Code
FST
11
673
ID
2/2
PH=Promotional
R1=(Fill)
Replenishment
SZ=Seasonal
Time Period
Begin Date
FST
04
373
DT
8/8
Time Period
End Date
FST
05
373
DT
8/8
Buyer
Location ID
SDQ
03, 05, 07,
09, 11, 13,
15, 17, 19,
21
67
AN
2/80
Item Quantity
SDQ
04, 06, 08, 10,
12, 14, 16, 18,
20, 22
380
R
1/15
Event ID
REF
02
127
AN
Number of
Participating
Locations
QTY
02
380
R
1/15
Forecast
Frozen Flag
FST
02
680
ID
1/1
Appendix E
Message Mapping
Forecast
Item
N103 will
contain a
qualifier
identifying the
type of ID
N103 will
contain a
qualifier
identifying the
type of ID
SDQ02 will
contain a
qualifier
identifying the
type of ID
REF01 will
contain code
EVI = Event
Identification
QTY01 will
contain
code AS =
Participant Total
Table 31
140
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Sales/Order Forecast Item Revision Message Mapping
Entity
Attribute
Segment
ID
Reference
Designator
Data
Element
Data
Type
Min/ Code(s)
Max
Forecast
Item
Revision
Seller ID
N1
01
98
ID
2/2
Seller
Location ID
N1
04
67
AN
2/80
Buyer ID
N1
01
98
ID
2/2
Buyer
Location ID
N1
04
67
AN
2/80
N103 will
contain a
qualifier
identifying the
type of ID
Revision
Generation
Date
BFR
09
373
DT
8/8
BRF01 will
contain code
04=Change
Revision
Source
Indicator
ISA
06
I06
AN
15/15
ISA05 will
contain a
qualifier
identifying the
type of ID
Standard
Product ID
LIN
03
234
AN
1/48
LIN02 will
contain code
UK or UP
Quantity
Type Code
FST
11
673
ID
2/2
PH=Promotional
R1=(Fill)
Replenishment
SZ=Seasonal
Time Period
Begin Date
FST
04
373
DT
8/8
Time Period
End Date
FST
05
373
DT
8/8
Forecast
Generation
Date
BFR
08
373
DT
8/8
Forecast
Source
Indicator
N1
01
98
ID
2/2
EBI=Source
Forecast
Type Code
BFR
04
675
ID
2/2
BB=Sales
DL=Order
Quantity
Type Code
FST
11
673
ID
2/2
PH=Promotional
R1= (Fill)
Replenishment
SZ=Seasonal
Adjustment
Reason Code
FST
12
426
ID
2/2
Adjustment
Description
(Comment)
FST
13
352
AN
1/80
Buyer
Location ID
SDQ
03, 05, 07,
09, 11, 13,
15, 17, 19,
21
67
AN
2/80
Item Quantity
SDQ
04, 06, 08,
10, 12, 14,
16, 18, 20,
22
380
R
1/15
SDQ02 will
contain a
qualifier
identifying the
type of ID
Table 32
141
E
Forecast
Revision
N103 will
contain a
qualifier
identifying the
type of ID
Item Information Request [893]
The 893 transaction set can be used to request that a trading partner send
activity data for a product (using the Product Activity Data [852] transaction set).
It corresponds to the CPFR Actual Item Performance History Request message.
Actual Item Performance History Request
Entity
Appendix E
Message Mapping
New
Entity
Attribute
Segment
ID
Reference
Designator
Data
Element
Data
Type
Min/ Code(s)
Max
Purpose
Code
B2A
01
353
ID
2/2
Standard
Product ID
G39
03
234
AN
1/48
G3902 will
contain code
UK or UP
Seller ID
N1
01
98
ID
2/2
FR
Seller
Location ID
N1
04
67
AN
2/80
N103 will
contain a
qualifier
identifying the
type of ID
Buyer ID
N1
01
98
ID
2/2
TO
Buyer
Location ID
N1
04
67
AN
2/80
N103 will
contain a
qualifier
identifying the
type of ID
Activity Period
Begin Date
DTM
02
373
DT
8/8
DTM01 will
contain code
196=Start
Activity Period
End Date
DTM
02
373
DT
8/8
DTM01 will
contain code
197=End
“Bucket Size”
RCR
01
681
ID
1/1
Quantity Type
Code/Criteria
Code
RCR
02
859
ID
2/2
Table 33
142
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Product Activity Data [852]
The 852 Product Activity Data transaction set corresponds to the CPFR Actual
Item Performance History Response message.
Actual Item Performance History Response
Entity
Attribute
Segment
ID
Reference
Designator
Data
Element
Data
Type
Min/ Code(s)
Max
Product
Activity
Standard
Product ID
LIN
03
234
AN
1/48
LIN02 will .
contain code
UK or UP
Seller ID
N1
01
98
ID
2/2
MF
Seller
Location ID
N1
04
67
AN
2/80
N103 will
contain a
qualifier
identifying the
type of ID
Quantity Type
Qualifier
ZA
01
859
ID
2/2
Activity Period
Begin Date
XP
ZA
02
05
373
373
DT
DT
8/8
8/8
ZA04 will
contain 090
=Report Start
Activity Period
End Date
XP
ZA
03
05
373
373
DT
DT
8/8
8/8
ZA04 will
contain 091
=Report Start
Buyer
Location ID
SDQ
03, 05, 07,
09, 11, 13,
15, 17, 19,
21
67
AN
2/80
SDQ02 will
contain a
qualifier
identifying the
type of ID
Item Quantity
SDQ
04, 06, 08,
10, 12, 14,
16, 18, 20,
22
380
R
1/15
Standard
Product ID
LIN
03
234
AN
1/48
LIN02 will
contain code
UK or UP
Seller ID
N1
01
98
ID
2/2
MF
Seller
Location ID
N1
04
67
AN
2/80
N103 will
contain a
qualifier
identifying the
type of ID
Criteria Code
ZA
01
859
ID
2/2
Activity Period
Begin Date
ZA
05
373
DT
8/8
ZA04 will
contain 090
=Report Start
Activity Period
End Date
ZA
05
373
DT
8/8
ZA04 will
contain 091
=Report End
Buyer
Location ID
SDQ
03, 05, 07,
09, 11, 13,
15, 17, 19,
21
67
AN
2/80
SDQ02 will
contain a
qualifier
identifying the
type of ID
Metric Value
SDQ
04, 06, 08,
10, 12, 14,
16, 18, 20,
22
380
R
1/15
Metric Value
UOM
SDQ
01
355
ID
2/2
Table 34
143
Item
Metric
E
Purchase Order [850 or 875]
Companies use either the 850 or 875 transaction set to place a purchase order
for products that have been forecasted (i.e., VICS EDI 850 or UCS 875). It is
expected that they will use existing conventions for the data exchange. There is
no CPFR-specific mapping for purchase orders.
Purchase Order Acknowledgement [855]
Companies can use the 855 transaction for two distinct purposes. First is to
acknowledge a buyer’s purchase order submitted via a Purchase Order
Appendix E
Message Mapping
Transaction [850]. This acknowledgement can contain no changes, nor can it
initiate changes or cancellations in response to the buyer’s purchase order.
The acknowledgement can be used to initiate a seller–created order to the buyer.
There is no CPFR-specific mapping for acknowledgements.
Price/Sales Catalog [832] or Item Maintenance [888]
Companies can use the 832 or 888 transaction set to send updated product
information to trading partners (i.e., VICS EDI 832 or UCS 888). It is expected
that they will use existing conventions for the data exchange. There is no
CPFR-specific mapping for item maintenance transactions.
Promotional Announcement [889]
Companies can use the 889 transaction set to inform CPFR partners of upcoming
promotional events. Promotional events are a subset of the event types defined
by CPFR. There is no CPFR-specific mapping for events.
SIL Message Maps
This appendix documents how data required by CPFR data flows
maps to existing Standard Interchange Language (SIL) data elements
and tables.
Since the publication of the 1998 CPFR Guidelines, new data elements and SIL tables were submitted to and approved by the SIL
Standards Maintenance Committee. An updated mapping follows:
144
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Item Management Profile
CPFR Data
SIL Table
Model Attribute
SIL
SIL
Column ID Data Type
SIL
Category
Size
Item
Management
Profile
Standard
Product ID (K)
ITEMMANAGE
F01
GPC
NUMBER
14
Seller
Location ID (K)
ITEMMANAGE
F213
CHAR
NUMBER
13
Buyer
Location ID (K)
ITEMMANAGE
F414
CHAR
NUMBER
13
Effective Date (K)
ITEMMANAGE
F415
DATE
DATE
7
Expiration Date
ITEMMANAGE
F416
DATE
DATE
7
Minimum
Order Quantity
ITEMMANAGE
F417
NUM
QUANTITY
5.0
Order
Multiple Quantity
ITEMMANAGE
F418
NUM
QUANTITY
5.0
Lead Time Days
ITEMMANAGE
F419
NUM
QUANTITY
4.1
Order Interval
Days
ITEMMANAGE
F420
NUM
QUANTITY
3.1
Rounding Rules
Description
ITEMMANAGE
F421
CHAR
DESC
80
Transportation
Strategy
Description
ITEMMANAGE
F422
CHAR
DESC
80
Safety Stock
Rules Description
ITEMMANAGE
F423
CHAR
DESC
80
Frozen Period
Days
ITEMMANAGE
F424
NUM
QUANTITY
4.1
Table 35
Sales/Order / Forecast Header
CPFR Data
Model Entity
CPFR Data
SIL Table
Model Attribute
SIL
SIL
Column ID Data Type
SIL
Category
Size
Forecast
Header
Seller ID (K)
FCSTHEADER
F213
CHAR
NUMBER
13
Buyer ID (K)
FCSTHEADER
F414
CHAR
NUMBER
13
Forecast
FCSTHEADER
Generation Date (K)
F425
DATE
DATE
7
Forecast Purpose
Code (K)
FCSTHEADER
F426
NUM
CODE
1
Data Source
Code (K)
FCSTHEADER
F427
NUM
CODE
1
Forecast Sent Date
Extracted
from
transaction
DATE
DATE
7
Forecast Start Date FCSTHEADER
F428
DATE
DATE
7
Forecast End Date
F429
DATE
DATE
7
FCSTHEADER
Table 36
145
E
CPFR Data
Model Entity
Sales/Order/ Forecast Item Detail
CPFR Data
Model Entity
CPFR Data
SIL Table
Model Attribute
SIL
SIL
Column ID Data Type
SIL
Category
Size
Forecast Item
Seller
Location ID (K)
FCSTITEM
F213
CHAR
NUMBER
13
Buyer
Location ID (K)
FCSTITEM
F414
CHAR
NUMBER
13
Standard
Product ID (K)
FCSTITEM
F01
GPC
NUMBER
14
Time Period
Begin Date (K)
FCSTITEM
F428
DATE
DATE
7
Forecast
Generation
Date (K)
FCSTITEM
F425
DATE
DATE
7
Forecast Purpose
Code (K)
FCSTITEM
F426
NUM
CODE
1
Forecast Type
Code
FCSTITEM
F432
NUM
CODE
1
Data Source
Code (K)
FCSTITEM
F427
NUM
CODE
1
Time Period
End Date
FCSTITEM
F429
DATE
DATE
7
Appendix E
Message Mapping
Item Quantity
FCSTITEM
F433
NUM
QUANTITY
10.3
Event Type Code
FCSTITEM
F434
CHAR
CODE
4.0
Event Begin Date
FCSTITEM
F435
DATE
DATE
7
Participating
Locations
Quantity
FCSTITEM
F436
NUM
QUANTITY
6.0
Forecast Change
Restriction
Indicator
FCSTITEM
F437
FLAG
FLAG
1
Table 37
146
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Actual Item Performance History
CPFR Data
Model Entity
CPFR Data
SIL Table
Model Attribute
SIL
SIL
Column ID Data Type
SIL
Category
Size
Product
Activity
Standard
Product ID (K)
PERFORMHIST
F01
GPC
NUMBER
14
Seller
Location ID (K)
PERFORMHIST
F213
CHAR
NUMBER
13
Buyer
Location ID (K)
PERFORMHIST
F414
CHAR
NUMBER
13
Time Period
Begin Date (K)
PERFORMHIST
F439
DATE
DATE
7
Time Period
End Date
PERFORMHIST
F440
DATE
DATE
7
Criteria Type
Code (K)
PERFORMHIST
F441
NUM
CODE
4
Operational
Results
PERFORMHIST
F441
NUM
QUANTITY
10.3
UOM Code
PERFORMHIST
F23
NUM
CODE
4
Standard
Product ID (K)
PERFORMHIST
F01
GPC
NUMBER
14
Seller
Location ID (K)
PERFORMHIST
F213
CHAR
NUMBER
13
Buyer
Location ID (K)
PERFORMHIST
F414
CHAR
NUMBER
13
Time Period
Begin Date (K)
PERFORMHIST
F439
DATE
DATE
7
Time Period
End Date
PERFORMHIST
F440
DATE
DATE
7
Criteria Type
Code (K)
PERFORMHIST
F441
NUM
CODE
4
Metric Value
PERFORMHIST
F442
NUM
QUANTITY
10.3
UOM Code
PERFORMHIST
F23
NUM
CODE
4
Table 38
147
E
Item Quantity
Sales/Order Forecast Exception Item
CPFR Data
Model Entity
CPFR Data
SIL Table
Model Attribute
SIL
SIL
Column ID Data Type
SIL
Category
Size
Forecast Item
Exception
Standard
Product ID (K)
FCSTITEMEX
F01
GPC
NUMBER
14
Seller
Location ID (K)
FCSTITEMEX
F213
CHAR
NUMBER
13
Buyer
Location ID (K)
FCSTITEMEX
F414
CHAR
NUMBER
13
Time Period
Begin Date (K)
FCSTITEMEX
F439
DATE
DATE
7
Criteria Type
Code (K)
FCSTITEMEX
F441
NUM
CODE
4
Forecast
Generation
Date (K)
FCSTITEMEX
F425
DATE
DATE
7
Data Source
Code (K)
FCSTITEMEX
F427
NUM
CODE
1
Forecast
Purpose
Code (K)
FCSTITEMEX
F426
NUM
CODE
1
Effective
Date (K)
FCSTITEMEX
F445
DATE
DATE
7
Forecast Type
Code (K)
FCSTITEMEX
F432
NUM
CODE
1
Appendix E
Message Mapping
Table 39
148
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Sales/Order Forecast Comparison Exception
CPFR Data
SIL Table
Model Attribute
SIL
SIL
Column ID Data Type
SIL
Category
Size
Forecast
Comparison
Exception
Criteria Type
Code (K)
FCSTCOMPEX
F441
NUM
CODE
4
Seller
Location ID (K)
FCSTCOMPEX
F213
CHAR
NUMBER
13
Buyer
Location ID (K)
FCSTCOMPEX
F414
CHAR
NUMBER
13
Standard
Product ID (K)
FCSTCOMPEX
F01
GPC
NUMBER
14
Time Period
Begin Date (K)
FCSTCOMPEX
F428
DATE
DATE
7
Forecast
Purpose
Code (K)
FCSTCOMPEX
F426
NUM
CODE
1
Forecast
Generation
Date (K)
FCSTCOMPEX
F425
DATE
DATE
7
Comparison
Forecast
Generation
Date (K)
FCSTCOMPEX
F443
DATE
DATE
7
Data Source
Code
FCSTCOMPEX
F427
NUM
CODE
1
Comparison
Data Source
Code (K)
FCSTCOMPEX
F444
NUM
CODE
1
Effective Date (K)
FCSTCOMPEX
F445
DATE
DATE
7
Forecast Type
Code (K)
FCSTCOMPEX
F432
NUM
CODE
1
Table 40
149
E
CPFR Data
Model Entity
Operational Exception Item
CPFR Data
Model Entity
CPFR Data
SIL Table
Model Attribute
SIL
SIL
Column ID Data Type
SIL
Category
Size
Operational
Result
Exception
Standard
Product ID (K)
FCSTOPEREX
F01
GPC
NUMBER
14
Time Period
Begin Date (K)
FCSTOPEREX
F439
DATE
DATE
7
Criteria Type
Code (K)
FCSTOPEREX
F441
NUM
CODE
4
Seller
Location ID (K)
FCSTOPEREX
F213
CHAR
NUMBER
13
Buyer
Location ID (K)
FCSTOPEREX
F414
CHAR
NUMBER
13
Effective
Date (K)
FCSTOPEREX
F445
DATE
DATE
7
Appendix E
Message Mapping
Table 41
Operational Exception Criteria
CPFR Data
Model Entity
CPFR Data
SIL Table
Model Attribute
SIL
SIL
Column ID Data Type
SIL
Category
Size
Operational
Exception
Criteria
Standard
Product ID (K)
OPEREXCCRT
F01
GPC
NUMBER
14
Criteria Type
Code (K)
OPEREXCCRT
F441
NUM
CODE
4
Effective Date (K)
OPEREXCCRT
F445
DATE
DATE
7
Buyer
Location ID (K)
OPEREXCCRT
F414
CHAR
NUMBER
13
Seller
Location ID (K)
OPEREXCCRT
F213
CHAR
NUMBER
13
Exception Value
OPEREXCCRT
F447
NUM
NUMBER
10.3
UOM Code
OPEREXCCRT
F23
NUM
CODE
4
Collaboration
Priority Code
OPEREXCCRT
F448
NUM
CODE
3
Exception Value
Comparison Code
OPEREXCCRT
F449
NUM
CODE
4
Expiration Date
OPEREXCCRT
F450
DATE
DATE
7
Table 42
150
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Sales/Order Forecast Item Exception Criteria
CPFR Data
SIL Table
Model Attribute
SIL
SIL
Column ID Data Type
SIL
Category
Size
Forecast Item
Exception
Criteria
Standard
Product ID (K)
FCSTITEXCRT
F01
GPC
NUMBER
14
Criteria Type
Code (K)
FCSTITEXCRT
F441
NUM
CODE
4
Forecast Purpose
Code (K)
FCSTITEXCRT
F426
NUM
CODE
1
Effective Date (K)
FCSTITEXCRT
F445
DATE
DATE
7
Time Delta
Days Quantity (K)
FCSTITEXCRT
F446
NUM
NUMBER
5
Buyer
Location ID (K)
FCSTITEXCRT
F414
CHAR
NUMBER
13
Seller
Location ID (K)
FCSTITEXCRT
F213
CHAR
NUMBER
13
Forecast Type
Code (K)
FCSTITEXCRT
F432
NUM
CODE
1
Exception Value
FCSTITEXCRT
F447
NUM
NUMBER
10.3
UOM Code
FCSTITEXCRT
F23
NUM
CODE
4
Collaboration
Priority Code
FCSTITEXCRT
F448
NUM
CODE
1
Exception Value
Comparison Code
FCSTITEXCRT
F449
NUM
CODE
4
Expiration Date
FCSTITEXCRT
F450
DATE
DATE
7
Table 43
151
E
CPFR Data
Model Entity
Sales/Order Forecast Event
CPFR Data
Model Entity
CPFR Data
SIL Table
Model Attribute
SIL
SIL
Column ID Data Type
SIL
Category
Size
Event
Event Type
Code (K)
FCSTEVENT
F434
NUM
CODE
4
Event Begin
Date (K)
FCSTEVENT
F435
DATE
DATE
7
Data Source
Code
FCSTEVENT
F427
NUM
CODE
1
Event End Date
FCSTEVENT
F451
DATE
DATE
7
Event
Description
FCSTEVENT
F452
CHAR
DESC
30
Event Create
Date
FCSTEVENT
F453
DATE
DATE
7
Appendix E
Message Mapping
Table 44
Sales/Order Forecast Event Item
CPFR Data
Model Entity
CPFR Data
SIL Table
Model Attribute
SIL
SIL
Column ID Data Type
SIL
Category
Size
Event Item
Seller
Location ID
FCSTEVENTITM
F213
CHAR
NUMBER
13
Buyer
Location ID
FCSTEVENTITM
F414
CHAR
NUMBER
13
Standard
Product ID
FCSTEVENTITM
F01
GPC
NUMBER
14
Event Type Code
FCSTEVENTITM
F434
NUM
CODE
4
Event Begin Date
FCSTEVENTITM
F435
DATE
DATE
7
Table 45
152
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Sales/Order Forecast Event Item Impact
CPFR Data
Model Entity
CPFR Data
SIL Table
Model Attribute
SIL
SIL
Column ID Data Type
SIL
Category
Size
Event Item
Impact
Seller
Location ID (K)
EVNTITEMIMPT
F213
CHAR
NUMBER
13
Buyer
Location ID (K)
EVNTITEMIMPT
F414
CHAR
NUMBER
13
Standard
Product ID (K)
EVNTITEMIMPT
F01
GPC
NUMBER
14
Event Impact
Type Code (K)
EVNTITEMIMPT
F454
NUM
CODE
Event Type
Code (K)
EVNTITEMIMPT
F434
NUM
CODE
4
Event Begin
Date (K)
EVNTITEMIMPT
F435
DATE
DATE
7
Planned or
Actual Code (K)
EVNTITEMIMPT
F455
NUM
CODE
UOM Code
EVNTITEMIMPT
F23
NUM
CODE
4
Impact Value
EVNTITEMIMPT
F456
NUM
NUMBER
10.3
Impact Effect Code EVNTITEMIMPT
F457
NUM
CODE
1
Sales/Order Forecast Revision Header
CPFR Data
Model Entity
CPFR Data
SIL Table
Model Attribute
SIL
SIL
Column ID Data Type
SIL
Category
Size
Forecast
Revision
Header
Seller
Location ID (K)
FCSTHEADER
F213
CHAR
NUMBER
13
Buyer
Location ID (K)
FCSTHEADER
F414
CHAR
NUMBER
13
Revision
Generation Date
FCSTHEADER
F425
DATE
DATE
7
Forecast Purpose
Code (K)
FCSTHEADER
F426
NUM
CODE
1
Data Source
Code (K)
FCSTHEADER
F427
NUM
CODE
1
Forecast Change
Status Code
FCSTHEADER
F430
NUM
CODE
1
Revision Sent
Date
FCSTHEADER
Extracted
from
transaction
DATE
DATE
7
Table 47
153
Table 46
E
Sales/Order Forecast Revision Item
CPFR Data
Model Entity
CPFR Data
SIL Table
Model Attribute
SIL
SIL
Column ID Data Type
SIL
Category
Size
Forecast
Revision Item
Seller
Location ID (K)
FCSTITEM
F213
CHAR
NUMBER
13
Buyer
Location ID (K)
FCSTITEM
F414
CHAR
NUMBER
13
Time Period
Begin Date (K)
FCSTITEM
F428
DATE
DATE
7
Standard
Product ID (K)
FCSTITEM
F01
GPC
NUMBER
14
Forecast
Generation
Date (K)
FCSTITEM
F425
DATE
DATE
7
Forecast Purpose
Code (K)
FCSTITEM
F426
NUM
CODE
1
Data Source
Code (K)
FCSTITEM
F427
NUM
CODE
1
Revision
Generation
Date (K)
FCSTITEM
F425
DATE
DATE
7
Criteria Type
Code (K)
FCSTITEM
F441
NUM
CODE
4
Forecast Type
Code (K)
FCSTITEM
F432
NUM
CODE
1
Participating
Locations
Quantity
FCSTITEM
TBD
???
???
???
Revised Item
Quantity
FCSTITEM
F433
NUM
QUANTITY
10.3
Adjustment
Reason Code
FCSTITEM
F438
NUM
CODE
4
Appendix E
Message Mapping
Table 48
154
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Appendix F: CPFR Technology Questionnaire
Both parties in any CPFR technology initiative must come to agreement on the
types, format, source, and content of the data that they exchange, as well as
the security scheme and CPFR process alternatives to be used. The following
questionnaire identifies many of the issues that need to be resolved in order
to create a viable front-end agreement.
For each item, note the approach that the buyer and seller intend to use. Where
there are areas of difference, identify a usable common alternative. For example,
the buyer in one relationship may use SKU to identify products, while the seller
uses UPC. The seller may be able to map SKUs to UPCs, allowing products in
their CPFR collaboration to be identified by SKU.
F
A p p e n d i x F C P F R Te c h n o l o g y Q u e s t i o n n a i r e
CPFR Process
Item
Buyer
1
Will sales forecasts be collaborated on?
(if not, skip questions 6-27)
2
Will order forecasts be collaborated on?
(if not, skip questions 18-28)
3
What is the normal forecast revision interval?
(weekly, monthly, quarterly, other)
4
Are full or net-change forecasts sent?
5
How frequently are exceptions processed?
(weekly, daily, continuous, other)
6
Who creates the initial sales forecast?
7
Who will “store” the sales forecast?
(buyer, seller, or both)
8
Who enters events?
9
Who will generate sales forecast-related
exceptions?
Seller
10 Who will make the initial sales forecast revision?
11 Who will make the secondary sales forecast
revision?
12 How many revisions will be supported
and audited?
13 Who is the final arbiter of sales forecast revisions?
14 What horizon period of revision changes is
maintained on-line?
15 Are sales forecasts "frozen" within some time
window? (If so, within what window?)
Table 49
155
CPFR Process
?
(Continued from page 155)
Item
Buyer
Seller
16 Will sales forecasts automatically freeze
at this window?
17 Whose sales forecast will be the preference
for differences remaining at this horizon?
18 Who will create the order forecast?
19 Who will “store” the order forecast?
20 Who will generate order forecast-related exceptions?
21 Who will make the initial revision of the
order forecast?
Appendix F
CPFR Technology
Questionnaire
22 Who will make the secondary revision of the
order forecast?
23 How many revisions of the order forecast will
be supported and audited?
24 Who will be the final arbiter of order forecast
revisions?
25 What horizon period of order forecast revisions
is maintained on-line?
26 Are order forecasts "frozen" within some time
window? (If so, within what window?)
27 Will order forecasts automatically be released
at this window?
Resolution: Choice of CPFR process alternatives
Table 49 (continued)
Data Source (name the application, database, or provider)
Item
Buyer
1
What is the source for product information?
2
What is the source for location information?
3
What is the source for partner information?
4
What is the source for events information?
5
What is the source for sales forecasts?
6
What is the source for order forecasts?
7
What is the source for sales results?
8
What is the source for orders?
9
What is the source for in-stock inventory
information?
Seller
10 What is the source for item profiles?
Resolution: Sources for CPFR data
Table 50
156
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s
Data Attributes
Item
Buyer
1
What bucket size is used for sales forecasts?
(daily, weekly, monthly, other)
2
How many buckets in a sales forecast?
(12, 13, 52, other)
3
What is the sales forecast bucket boundary?
(day of week, fiscal period, other)
4
At what level are sales forecasts shared?
(store, DC, region, account, other)
5
How many location-pairs are sales forecasted
in the partnership?
6
How many products are sales forecasted?
7
Are component sales forecasts generated?
(base, promotional, seasonal)
8
What bucket size is used for order forecasts?
9
What is the order forecast bucket boundary?
Seller
F
A p p e n d i x F C P F R Te c h n o l o g y Q u e s t i o n n a i r e
10 How many buckets in an order forecast?
(112, 13, 52, other)
11 At what level are order forecasts shared?
(store, DC, region, account, brand/category, other)
12 How many location-pairs are order forecasted
in the partnership?
13 How many products are order forecasted?
14 Are component order forecasts generated?
(fill, promotional, seasonal)
15 What bucket size is used for results data?
16 What is the result bucket boundary?
17 At what level are results shared?
18 At what level is in-stock inventory reported?
19 Can the partner’s unit of measure be matched
for the products collaborated on?
Resolution: Detailed attributes of data to be shared
Table 51
157
Data Format
?
Appendix F
CPFR Technology
Questionnaire
Item
Buyer
1
Will data be exchanged between companies,
or managed at one site? (If data will be managed
at one site, skip to the next section, “Identification”)
2
Are X 12 830 transaction sets supported?
(If not, when will they be?)
3
Are X 12 852 transaction sets supported?
4
Are X 12 850/875 transaction sets supported?
5
Is the X 12 893 transaction set being used
to request a forecast?
6
Are VICS EDI guidelines for formatting
messages used?
7
Which revision of EDI standards and
guidelines is used?
8
Is SIL supported?
9
Are EDIFACT/GEDI supported?
Seller
10 Is XML supported? If so, what data dictionary
will be used?
Resolution: Message formats to be used to exchange CPFR data
Table 52
Identification
Item
Buyer
1
What identifiers are used for companies?
(DUNS, vendor/customer number)
2
What identifiers are used for products?
(GTIN, UPC, vendor part number, retail SKU/PLU)
3
What identifiers are used for locations?
(DUNS+4, GLN, other [specify])
Seller
Resolution: Common product, location, and company identification to be used
Table 53
Security
Item
Buyer
1
Will end-users be connecting to the CPFR system
over a public network?
2
Will data be exchanged between CPFR systems
over a public network?
3
How will users be authenticated?
4
How will data be encrypted?
5
Are firewalls in place?
6
How will access be made through a firewall
(if applicable)?
Seller
Resolution: CPFR security scheme to be used
Table 54
158
R o a d m a p
t o
C P F R :
T h e
C a s e
S t u d i e s

Roadmap to CPFR: The Case Studies

Transcription

Similar documents

Registration Form - NC-CCIM

LIPA Front-End Transition Cost Estimate

HPC and numerical modeling tools at the Chilean Weather

Guidance For Development Of An Emergency Fallout Shelter

Jonathan cainer www.CAiNER.COM

Implementing Successful Large Scale CPFR Programs

06_NextGen File

WeatherBELL Analytics: Energy

Bonneville Power Administration - Northwest Hydroelectric Association

Diapositive 1