People Data Activation

People Data Activation
The Premise & Promise
The consumer web has ushered in a new normal. Portable computing, always-on communications, and
ubiquitous network connectivity have transformed our expectations for just about everything, radically
reshaping how we conduct our private and public lives. Whether it comes to selling, buying, socializing,
healing, learning, inventing, building, negotiating, or politicking, we want access to information and services
now. We want it personalized. And we want it on whichever screen we happen to have in our pockets or bags.
The shift began in the early 2000s with the advent of portals offering news and information tailored to
each person’s specific interests, and has evolved over the last decade into a broader set of technologies
that deliver more relevant ads, more effective commerce, and more customized content. Digital media’s
nomenclature is instructive: it refers to technology for turning people data into more relevant web experiences
simply as ‘data management.’ There are many types of data to be sure, but in digital media, people data
reigns supreme.
The trend is likely to reverberate beyond digital media. When we refer to data, we’ll increasingly mean
information about people: who they are, their wants, desires, needs, and expectations. The new basis for
competition will focus on capturing and harnessing data in order to anticipate and deliver what people want,
any way they want it, on any screen, in any context, at any moment. Call it the ‘AnyX’ initiative.
Is Big Data the answer? While its early manifestations have been promising, Big Data is still mostly piping and
processing in search of a purpose. Big Data’s early application area, Software as a Service (SaaS), leveraged cloud
technologies for lowest-common-denominator enterprise processes, such as storage and sharing (DropBox,
Box), collecting and analyzing (Tableau, Splunk), and sales automation (Salesforce.com). While all of these are
unquestionably useful, the market cries out for Big Data applications that focus on people, not processes.
Enterprise software, meanwhile, struggles to keep pace with the implications of the consumer web. It treats
people as a secondary phenomenon, throwing off occasionally useful scraps of data — an address attached to
a purchase, or an online profile of someone who clicked or converted. These systems creak and crumble with
the volume, velocity, and variety of the data that’s generated by literally trillions of real-time web experiences.
Most fatally, legacy software puts organizations, not people, at the center. Indeed, the nomenclature of the
consumer web and enterprise CRM barely veil their fundamental orientation: we, people, are users, at best
customers, faceless minions conscripted into the service of consumption.
A new era is upon us, one that begs businesses to re-consider their online and offline approaches to people.
It’s time to re-frame business-to-people transactions as interactions through a relentless focus on people data.
It is not sufficient for companies to say they are ‘customer-centric;’ the imperative is to bring people data to life
by first changing internal enterprise systems and processes, then delivering experiences that are authentic
and meaningful. We call this People Data Activation (PDA).
2
©2015, All Rights Reserved
Krux Digital, Inc.
www.krux.com
People Data Activation will usher enterprise software out of its process-centric legacy and give it the power
to finally confront the implications of the people web. SaaS was a partial revolution — the application of new
technologies to solve old problems centered mostly on cost reduction and process efficiency for finance
and IT organizations. PDA, in contrast, delivers vertical applications that drive top-line performance for P&L
holders. It achieves this by giving any organization the ability to:
⊲⊲ Protect, capture, control, organize, analyze, and connect people data from any source
⊲⊲ Use people data to create cooler, faster, smarter, and safer web experiences
⊲⊲ Engage with people in individualized ways across any screen, through any channel
⊲⊲ Reach existing customers, new prospects, and emerging audiences
Legacy Enterprise Software
People Data Activation
Key focus
Processes
People
Business Impact
Cost Reduction & Process efficiency
for finance & IT
Effectiveness / Revenue Generation
Technology Method
Horizontal enabling infrastructure
for storing, collecting, and
analyzing data and automating
common enterprise processes
Vertical, purpose-built applications
that drive engagement with people
(users, audiences, consumers,
prospects, customers)
The People Revolution
Regardless of the digital transaction, people want what they’ve always enjoyed about non-digital experiences
at a neighborhood coffee shop: convenience, access, and personalization. We like to be recognized and, even
better, to have our needs anticipated.
Consider, for example, that you’re a bank. Your customer, Sally, wants to see her savings and investments
products in an elegantly designed store catered to her needs and fueled by her past banking habits. She
wants to conduct an online application process that’s paperless and takes 30 seconds to approve and
complete. And she wants to do all of it on her tablet or smartphone from the comfort of her living room.
The consumer web raises the bar on access, recognition, and support in very important ways. People now
want control over the process: to define the location and timing of the transaction, and to have hyper-efficient,
machine-driven experiences that still feel human, personal, and catered to our needs. To respond, companies
need to build and sustain always-on, intelligent awareness of individual customers — instead of treating
people as general market segments, demographic groups, and sectors.
3
©2015, All Rights Reserved
Krux Digital, Inc.
www.krux.com
The AnyX Model of Success
A number of Internet companies recognized the AnyX initiative early and moved quickly to conquer it. They
include familiar names such as Google, Facebook, and Apple, as well as sector-specific players, such as
Netflix and LinkedIn. For the enterprise, however, one company defines the gold standard for PDA: Amazon.
Web scale, delivery efficiency, and pricing power were certainly keys to Amazon’s success. The strategic
breakthrough, however, was simpler, more durable, and ultimately more powerful: with every click and every
page view, Amazon created a personal connection with the consumer. Its early decision to create and share
wish lists and consolidate user reviews presaged the social media revolution to come. Its recommendation
engine deftly replaced the guy behind the counter at the neighborhood coffee shop, enabling Amazon to
tap into a wealth of user and transaction data to inform purchase decisions and discover otherwise unseen
customer demand.
AN
ORG
PEOPLE ENGINE
CO L L
PLE PARTICIPATE
IZE & ANALYZ
PEO
E
EP
ENGAG EOPLE
E C T & P R OT EC T
If consumers feel as though Amazon knows them, it’s because the company does. And it uses that
understanding to cross-sell and up-sell, inspiring their customers to spend a little extra money while on the
site. More importantly, in an era where issues of privacy and data usage loom large, customers implicitly trust
Amazon. With every return visit, they express that trust by sharing more about their wants and interests, which
in turn provides Amazon’s massive people engine with the data it needs to deliver an even smarter personal
touch. Thus, the virtuous circle is complete.
4
©2015, All Rights Reserved
Krux Digital, Inc.
www.krux.com
The Six Capabilities of People Data
Activation
For companies that grew up on the web, delivering on AnyX wasn’t an option — it was a survival condition. But
technology moves inexorably forward. Virtuosity in the consumer web is no longer the domain of an exclusive
few. A decade later, all the components and capabilities are available to any enterprise that seeks to follow the
path Amazon has blazed. PDA is, today, an opportunity available to any business.
PDA best practices are organized into a central discipline, supported by a set of six capabilities:
5
Completeness
Governance
Learning
Recall instantly 100% of
people data, 100% of the
time across all sources
Protect enterprise
people data, control
where it goes
Learn continuously, apply
insight, adapt to new
experiences
Synthesis
Identity
Actionability
Unify and analyze data
across screens, silos,
and systems
Recognize individuals
across browsers, devices,
and screens
Put data intelligence
to work and drive
measurable ROI
©2015, All Rights Reserved
Krux Digital, Inc.
www.krux.com
Completeness
People explore, purchase, socialize, and change their minds at web speed. To meet this
challenge, savvy marketers want to launch new targeting tactics across screens and
contexts with unlimited agility. When they activate a selling or marketing tactic, they need to
know, in seconds, how many users or customers it will reach. Waiting a day for an estimate
or ‘go-live’ is no longer an option.
PDA’s first capability, completeness, allows businesses to keep pace with web velocity. Completeness is about
the total view, recall, and capture of 100% of people data across all systems and silos at any time.
Complete architectures deliver total recall and extreme anticipatory awareness. They provide an everexpanding reservoir of data accessible at any time to fuel new hypotheses, build new theories of customers, or
infer new results. With a complete system, all data is fully available for retrieval and activation at any time, even
if you didn’t define, specify, or refer to it in the past.
Incomplete systems collect and report only what you requested ahead of time. Any data that isn’t explicitly
defined, usually via formal segment definition, is entirely forgotten. Like the amnesiac in the movie Memento
who can remember critical facts only by tattooing them on his arm, incomplete systems are amnesic; they
recall only the facts and numbers that have been explicitly burnt into their schemas. So if you want to
understand how consumers engaged with beauty offers targeting moms last Mother’s Day, you’re out of luck.
Unless the segment was explicitly defined and maintained, there’s no prospect of retrieving it and putting it to
work to support this year’s Mother’s Day promotion. It’s simply gone.
Complete systems provide total recall; your data is always fully available for retrieval and activation at any
moment, even if you didn’t define it, point to it, or refer to it in the past. Completeness entails:
⊲⊲ Capturing 100% of your data, 100% of the time, across 100% of your customers and marketing touch
points, as well as the systems that support them
⊲⊲ Supporting real-time manipulation and computation of derived results from the data
⊲⊲ Synthesizing data from web-based JavaScript, mobile SDKs, direct-log file ingestion, and offline CRM
systems using flexible cloud-based extract, transform and load (ETL)
⊲⊲ Enabling just-in-time, on-demand transport and application of data to the screens and experiences where
it can create the most value
6
©2015, All Rights Reserved
Krux Digital, Inc.
www.krux.com
Governance
Governance is about data ownership, privacy, and data protection. With the ability to capture
nearly limitless data about individual consumers, PDA requires built-in mechanisms for
managing choice and privacy, tools that enable the enterprise to serve as a trusted steward
of people data. Detection systems monitor data collection from third-party services (e.g.,
measurement, analytics, social, ad servers), supported by data usage contracts that enforce
tighter controls and shield the enterprise from rogue data collection. Opt-in services are
woven into the fabric of web systems, allowing people concerned about privacy to decide what companies
know about them and how such information is collected and used. Tag management systems provide a secure
gateway for controlled access to people data on the web, as well as a policy-managed framework for throttling
that data into the hands of trusted partners. Taken together, all of these functions enable the enterprise to
assert full ownership over its data, while demonstrating good governance to its customers and partners.
But data governance is only as strong as the weakest link in the chain that connects an enterprise with its
partners. Accordingly, companies increasingly seek data management partners with no conflict of interests.
For example, does the enterprise own its people data and control it exclusively, or is that data used to fuel a
partner’s business model? More to the point, is the enterprise’s data helping a competitor, masquerading as
a partner, and reaping incremental revenue from advertising, content, or commerce operations? Enterprises
seeking to protect their advantage and minimize opportunity costs will naturally pursue data-management
partners that operate solely as technology-focused facilitators — not content, network, or media principals.
Learning
Iterative real-time learning and granular, precise responses to evolving audience needs are
at the essence of plasticity.
Learning unfolds along three key vectors:
⊲⊲ Any datum: As consumers engage and connect across screens, every piece of data has the potential to
contribute to a more complete picture of who they are and how they’re motivated. Clicks, conversions,
purchases, redemptions, downloads, registrations, hovers, mouse-overs, eye movements, location – all of
these actions are valuable grist for PDA’s mill. PDA captures, learns from, and turns this data into action
through machine-driven methods.
⊲⊲ Any use: Data in ‘raw materials’ form must be easily transformed into ‘finished goods.’ PDA makes people
data portable across uses, channels and systems via structured APIs. Although advertising defined its
initial domain and proving ground, PDA now powers multiple use cases, such as fueling more dynamic
7
©2015, All Rights Reserved
Krux Digital, Inc.
www.krux.com
content, more effective marketing and selling, and more intelligent collaboration.
⊲⊲ Any segment: Prior limits on capturing and processing people data severely limited the computing power
of legacy enterprise software. True, CRM worked well with large chunks of data, but it quickly hit a barrier
when analyzing more detailed combinations of interests, propensities, behaviors, or time. PDA overcomes
that barrier, allowing us to understand the needs of this consumer at a granular level, to overlay everything
that is known about him, and to quickly combine it with everything your partners and vendors know.
Through flexible authoring and discovery of actionable segments (e.g., people between the ages of 24
and 35 on the Eastern seaboard who are socially active electronics enthusiasts and have engaged at least
twice with my electronics content in the last week), PDA allows marketers to design and deploy dynamic
marketing strategies that iteratively improve over time, while measurably improving engagement and
profit.
MOBILE
INDUSTRIAL
GRADE .JS
ACTIONABLE
RELEVANT
OFFLINE
DATA
COLLECTION
& SYNTHESIS
DESKTOP
PORTABLE
PRIVACY PROOF
OCC/CT
OTT/CT
ALGORITHM
REAL-TIME
APPLIANCE
FINISHED GOODS
In any analytical processes, one question leads to insights, which lead to a new set of more astute, more relevant
questions. With its ability to comprehensively capture and make sense of any scrap of data, PDA’s learning lets
marketers learn iteratively from experience, in real time. While completeness frees companies from the drudgery
of pre-thinking all the possibilities, learning allows them to learn and evolve faster than their competitors.
Synthesis
In the enterprise model of the 1990s and early 2000s, data integration projects required
dozens or even hundreds of systems integrators – a huge time and money suck for
enterprises.
By harnessing the power of Big Data and the ubiquity of HTTP, PDA transforms the previously
ugly, unending chore of cross-silo, cross-screen integration into a cloud-enabled process that is flexible,
seamless, and fast. This capability is synthesis, which is about unifying data and analyzing insights across
8
©2015, All Rights Reserved
Krux Digital, Inc.
www.krux.com
screens, silos, and systems.
With synthesis, datasets in catacombs of legacy systems can be streamed within hours to a central cloud and
meshed with people data from any web experience (social, mobile, display, set-top TV, online, gaming). Cloudbased online analytical processing (OLAP) technologies, such as Hive, provide seamless, SQL-like access to
data on-the-fly. Machine-driven methods, free of manual data entry and point-to-point extraction, allow the
transformation of both unstructured and structured data into shared, more actionable formats.
Identity
Where PDA truly ushers legacy enterprise software into the age of the consumer web is in its
ability to manage identity across screens, silos, and systems. For example, Sally starts her
online banking session on a tablet while in her living room in the evening, continues the
session on a handheld during her train ride to work the next morning, and finishes the
session later in the afternoon from a desktop. By providing the data fabric that ties Sally’s
identity across screens and across time (without sacrificing privacy and choice), PDA allows the enterprise to
support Sally at every step of her journey, as well as maintain an always-on, intelligent awareness of her future
needs.
Actionability
PDA’s sixth capability is actionability, which is about turning intelligence into action, data into
dollars. Actionability allows companies to move beyond seeing and knowing to connecting
and doing. It binds data to a unique person in the right context, channel and time, and
delivers an informed interaction to serve his or her individual needs. Ultimately, actionability
transforms people data into lifelong engagement. Three important functions are at the heart
of actionability:
⊲⊲ Inquiry: Like Sally, billions of people around the world are living multi-faceted digital lives, conscientiously
or accidentally using a variety of connected devices, including desktops, tablets, phones, sensors,
connected TVs, set-top boxes and more. Their intentions and needs, however, are uniquely their own.
With a complete, synthesized and real-time view of Sally’s data-driven interactions, businesses can now
ask new questions, such as: How does she make decisions? Who influences her, and whom does she
influence? How often should we reach out to her? What works best — a timely coupon on her mobile
phone when she’s in store, or an informative blog that supports her online purchase at home? What will
she need later? Where might she go next? The data provides new insights and creates game-changing
opportunities for marketers to envision disruptive products and services that transform the end user’s
experience.
9
©2015, All Rights Reserved
Krux Digital, Inc.
www.krux.com
⊲⊲ Control: Intelligent, real-time personalization cannot be achieved at scale by stitching together disparate
purpose-built tools. It requires execution functions architected from the ground up to anticipate Sally’s
needs and coordinate the next interaction with her across outbound systems in real time. All actions are
measured and evaluated, and the controls optimized in real time. All waste is eliminated, and the cycle of
deep engagement continues unfettered.
⊲⊲ Revision: Real-time revision follows inquiry and control. Ubiquitous connectivity, responsive web
technologies, and programmatic decisions combine to deliver globally correlated and uniquely
personalized experiences across all channels and systems. Knowing a person’s preferences ensures her
time is not wasted on irrelevant messages or recommendations. Knowing how often she prefers to be
contacted means you won’t bombard her with disruptive ads at untimely moments.
The successful AnyX enterprise powered by PDA identifies and engages Sally and her cohorts with messages
that resonate, educate, listen and delight, rather than irritate and clutter. The enterprise becomes a trusted,
intelligent confidant and collaborator in Sally’s choices, whether she is browsing, purchasing, referring,
socializing, or consuming content. Each interaction increases the data reservoir and gives the enterprise
another opportunity to dazzle Sally in her next interaction with a brand. The AnyX enterprise anticipates,
predicts, and fundamentally re-orients itself away from the reactive opportunism of ‘customer-relationship
management ’ to a mutually beneficial ‘life-time engagement.’
PDA is the Future of Enterprise Software
Legacy enterprise software was architected for specific processes, such as billing and ordering, pipeline
management, inventory tracking, and balancing the books. Businesses will continue to rely on such systems
of record to support the objectives for which they were designed: minimizing costs and maximizing efficiency.
When it comes to maximizing revenue and relationships with people, however, PDA delivers a system
of engagement that continuously makes sense of the flood of data from different screens and sources,
giving people what they want, any way they want it, on any screen, in any context, at any moment. With an
unswerving focus on people and revenue – as opposed to processes and efficiency – PDA redefines how
businesses engage with existing customers, prospects, and emerging audiences.
The Future of PDA
We envision a world where the flow of people data becomes a new tier that sits atop the TCP/IP network stack,
energized by common standards and open protocols. We anticipate a re-conception of business-to-consumer
commerce built on Big Data and anchored not on horizontal, efficiency-oriented infrastructure, but on
vertical applications that measurably improve top-line revenue performance for the enterprise. We invite our
partners, collaborators, competitors, industry experts, academics, and privacy advocates to join us in building
tomorrow’s technology today. Working together, we have the opportunity to abandon business as usual and
truly deliver on the promise of People Data Activation.
10
©2015, All Rights Reserved
Krux Digital, Inc.
www.krux.com
References
The Cluetrain Manifesto, by Rick Levine, Chrisopher Locke, David Weinberger, and Jake McKee
Mining of Massive Datasets, by Anand Rajaraman and Jeffrey David Ullman.
The Elements of Statistical Learning: Data Mining, Inference, and Prediction, by Trevor Hastie, Robert
Tibshirani, and Jerome Friedman
http://hadoop.apache.org
11
©2015, All Rights Reserved
Krux Digital, Inc.
www.krux.com
KRUX.COM
04142015