This white paper—produced in conjunction with the Interactive Advertising Bureau— will explore four data-driven use cases (audience optimization, channel optimization, advertising yield management and targeted media buying) that collectively represent the foundation of how many are now seeking to leverage the potential of “big” marketing data. In addition to that analysis, it will demonstrate that capitalizing on this opportunity will require:  Rules-driven integration of disparate data sets: The collection, analysis and segmentation of digital data demands the aggregation and anonymization of virtually all data, challenging marketers’ fundamental ability to draw distinct insights from consumers’ cross-channel interactions  Improved operating infrastructures: Though substantial process and data structure challenges also exist, a substantial barrier now inhibiting wider marketing data optimization resides within the marketing organization— characterized by rigid “silos” and the paucity of data-savvy marketing operations, IT and sales talent  A strong network of data-centric technology and service partners: The fastest and most efficient data aggregation, analysis and throughput solutions require a strong ecosystem of partners who understand and can integrate seamlessly with core data assets and supporting technologies  Marketing data governance: While organizations have long employed policy experts to advise on the regulatory ramifications of data utilization, many are coming to see marketing data governance—defining the “rules of the road” for assigning distinct data sources to different promotional tasks—as equally important

1. 3
From Information to Audiences:
The Emerging Marketing Data Use Cases
A Winterberry Group White Paper
January 2012

2. © 2012 Winterberry Group LLC.
Acknowledgements
This white paper would not be possible without the significant contributions of more
than 175 advertising and marketing thought leaders—representing virtually all corners
of the commercial data and technology ecosystem. In particular, Winterberry Group is
grateful to our research partner, the Interactive Advertising Bureau, as well as the
following sponsors for their generous support of this important research initiative:
Presenting Sponsors:
Supporting Sponsors:
To all those whose insights, time and other contributions helped in the development
of this white paper, we thank you.
Notice
This report contains brief, selected information pertaining to the commercial
marketing data industry and has been prepared by Winterberry Group LLC with the
support of Interactive Advertising Bureau. It does not purport to be all-inclusive or to
contain all of the information that a prospective investor or lender may require.
Projections and opinions in this report have been prepared based on information
provided by third parties. Neither Winterberry Group, the Interactive Advertising
Bureau nor their respective sponsors make any representations or assurances that
this information is complete or completely accurate, as it relies on self-reported data
from industry leaders—including advertisers, marketing service providers, technology
developers and agencies. Neither Winterberry Group, the Interactive Advertising
Bureau nor any of their officers, employees, representatives or controlling persons
make any representation as to the accuracy or completeness of this report or any of
its contents, nor shall any of the foregoing have any liability resulting from the use of
the information contained herein or otherwise supplied.
2

3. © 2012 Winterberry Group LLC.
Executive Summary
No matter what analogy you prefer, one truth is undeniably clear: Technology has
fundamentally advanced the creation of what many call “big data.” Consider:
 From the dawn of time through 2003, according to Google’s executive
chairman, Eric Schmidt, human civilization generated approximately 5 exabytes
of aggregate information. In 2009, that much data—captured in the equivalent
of 25 quadrillion tweets—was generated every two days
 Globally, businesses created 1.8 zettabytes of data in 2011, according to IDC.
That output—enough to fill 57.5 billion 32-gigabyte Apple iPads—is growing
approximately 62 percent annually (on a compounded basis)
 In July 2011, Facebook’s 750 million worldwide users uploaded approximately
100 terabytes of data every day to the social media platform. Extrapolated
against a full year, that’s enough data to manage the U.S. Library of Congress’
entire print collection—3,600 times over.
The world’s Twitter feeds, iPads and libraries may not stand a chance against this
onslaught of information. But to the world’s marketers, the proliferation of data has
given rise to what may prove to be the most substantial commercial opportunity since
the emergence of the World Wide Web: the ability to better understand consumers,
seamlessly match “right-time” offers to their needs and optimize the management of
profitable, long-term customer relationships.
The ongoing Not surprisingly, many are working feverishly to capitalize on the new potential of
marketing data, especially with respect to the torrent of highly insightful (but highly
convergence of unstructured) information being generated online. The ongoing convergence of new
new data data sources, targeting technologies and advertising delivery platforms is likewise
sources, shifting their focus—from the management of raw information to the optimization of
granular consumer audiences across discrete advertising channels, product categories
targeting
and geographies.
technologies and
advertising The demands of real-time, rules-driven, audience-centered marketing represent a full-
delivery on paradigm shift in how marketing is done. But with the opportunity inherent in this
approach comes a daunting challenge: to identify and deploy an actionable range of
platforms is “use cases”—practical marketing applications that, fueled by data, may drive
shifting focus— transformative improvements in both marketing effectiveness and efficiency.
from the
management of Today, even while some enjoy modest success in redeploying their existing resources
to the new cross-channel task, most other marketers—saddled with legacy technology
raw information platforms, depleted of expertise by years of underinvestment and structured only to
to the support “traditional” approaches to data usage—are finding they’re woefully
optimization of unprepared for this transformation. For them, a growing data divide is taking shape,
distinguishing those use cases to which data may now be profitably deployed from
granular
those which—though promising in their strategic potential—still represent nothing
consumer more than ideals of how automated, multichannel marketing may someday take
audiences. shape.
3

4. © 2012 Winterberry Group LLC.
This white paper—produced in conjunction with the Interactive Advertising Bureau—
will explore four data-driven use cases (audience optimization, channel optimization,
advertising yield management and targeted media buying) that collectively represent
the foundation of how many are now seeking to leverage the potential of “big”
marketing data. In addition to that analysis, it will demonstrate that capitalizing on
this opportunity will require:
 Rules-driven integration of disparate data sets: The collection, analysis and
segmentation of digital data demands the aggregation and anonymization of
virtually all data, challenging marketers’ fundamental ability to draw distinct
insights from consumers’ cross-channel interactions
 Improved operating infrastructures: Though substantial process and data
structure challenges also exist, a substantial barrier now inhibiting wider
marketing data optimization resides within the marketing organization—
characterized by rigid “silos” and the paucity of data-savvy marketing
operations, IT and sales talent
 A strong network of data-centric technology and service partners: The fastest
and most efficient data aggregation, analysis and throughput solutions require
a strong ecosystem of partners who understand and can integrate seamlessly
with core data assets and supporting technologies
 Marketing data governance: While organizations have long employed policy
experts to advise on the regulatory ramifications of data utilization, many are
coming to see marketing data governance—defining the “rules of the road” for
assigning distinct data sources to different promotional tasks—as equally
important.
4

5. © 2012 Winterberry Group LLC.
Methodology
This white paper explores a series of “use cases” that define how marketers are
commonly deploying multichannel data to improve their advertising and marketing
effectiveness and efficiency. It further highlights a series of trends that are defining
how data is now being used to drive broader advertising and marketing performance
for companies based in the United States.
Developed in research partnership with the Interactive Advertising Bureau—and with
the sponsorship of Acxiom Corporation, IBM, BlueKai, eXelate, Janrain, ShareThis and
V12 Group—the paper’s findings are based on the results of an intensive research
effort that included in-person, phone and online surveys of more than 175 marketers,
agency executives, data compilers, technology developers and other industry thought
leaders around the globe.
Where possible, contributors have been cited by name so as to provide transparency
into the research process and supporting panel. In some cases, contributors have
asked that we omit their name and company information so as to allow them the
freedom to speak with maximum candor.
5

6. © 2012 Winterberry Group LLC.
The Emerging Marketing Data Use Cases
The span of today’s data use cases is broad, reflecting the relative immaturity of the
“digital data” enterprise and the array of pilot solutions that marketers and publishers
are deploying to make use of the growing information resources at their disposal. For
some, a data use case may be as simple as demographic-driven customer acquisition
(as enabled by a rented mailing list); for others, the span of what’s actionable may
include a host of sophisticated display advertising targeting solutions.
Interest in these applications is being piqued by the realization that information may be
used to drive transformative value that spans “demand” and “supply” sides of the
advertising and marketing value chain. Data availability is now allowing advertisers,
agencies and publishers to optimize ad delivery, evaluate campaign results, improve
site selection and retarget ads to other sites. It’s also improving the value of media to
brands by delivering their advertising to better-qualified prospects—making the ad
more efficient, more valuable and providing a more compelling user experience.
Grounded in years of direct response, data use by those marketers that predominantly
leverage offline channels is proving to be just as sophisticated as those applications
that dominate in the online sphere. Ironically, best practices developed in this
“traditional” DR marketing world are often used to establish parameters for the
deployment of digital data, even in those cases where data are being used to enable a
shift in strategic emphasis from direct response to brand engagement. “The industry
has spent a lot of time and money at the bottom of the funnel,” said Jeff Liebl, chief
marketing officer at TruSignal. “Advertising is supposed to be about generating intent,
but the bottom of the funnel is mostly about looking for people who have already
shown interest. I think we’ll see ad dollars shift and a greater focus placed on earlier,
upper-funnel brand awareness activity, targeting people that haven’t necessarily
demonstrated online behavior yet that shouts ‘I’m in market.’”
What follows is a discussion of four selected marketing data use cases—audience
optimization, channel optimization, targeted media buying and advertising yield
management—along with an assessment of fundamental benefits, current maturity
levels, core beneficiaries and long-term potential.
Use Case Fundamental Maturity Core Beneficiaries Long-Term
Advertising Benefit Level Potential
Audience Optimization Effectiveness Low E-commerce Marketers, Digital Advertisers, High
Lead Generation Portals, Publishers (for
traffic acquisition)
Channel Optimization Effectiveness/Efficiency Low E-commerce Marketers, Publishers, Lead High
Generation Portals
Advertising Yield Efficiency Low Publishers High
Optimization
Targeted Media Buying Efficiency/Effectiveness Intermediate Marketers (via Demand-Side Platforms), High
Digital Agencies / Trading Desks
6

7. © 2012 Winterberry Group LLC.
7

8. © 2012 Winterberry Group LLC.
Audience Optimization
Identifying customers and likely prospects through the integration of rich (though
disparate) first- and third-party data sources; managing cross-channel marketing
execution with the goal of engaging those audiences strategically—and in
accordance with consumers’ preferred advertising media.
Fundamental Maturity Core Long-Term
Advertising Benefit Level Beneficiaries Potential
Effectiveness: Identifying Low: Though the technology now E-commerce High: More so than any other
the “right” target exists to capture and deploy large Marketers, Digital use case, the ability to define
consumers is the foundation quantities of information (in the Advertisers, Lead high-potential audiences from
of targeted advertising, and necessary “real-time” windows), Generation Portals, disparate indicators—and then
may be used to improve consensus has yet to coalesce Publishers (for communicate with them across
performance across around the optimal approach to traffic acquisition) a range of media—represents a
branding, engagement and structured integration of this fundamentally new approach
direct response functions data—especially when its sources to managing addressable
span traditional (“PII”) and digital customer markets
(“non-PII”) channels
The plethora of first-party data now being amassed and analyzed by both publishers
and advertisers is being used to build rich audience profiles that, marketers say, can
enhance advertising effectiveness by enabling improved targeting and message
relevancy. Today’s dominant approach calls for the development of unique
customer/prospect profiles, which are then segmented and modeled as the basis for
identifying what are commonly called “lookalike audiences” for follow-up marketing
across channels.
For publishers, third-party data overlays and data exchanges—providing access to a
wealth of additional information generated through online sources—are providing the
opportunity to enhance first-party data with demographic and interest-based
indicators, as well as first-party data from other online publishers. “Companies usually
own very rich first-party data,” said Travis May, head of strategy and operations at
Rapleaf. “Third-party data is especially helpful when there are new customers or early-
lifecycle customers and the data need to be enhanced to be segmented more quickly.”
In one example: Catalina Marketing, which claims to collect and analyze in-store
purchase data covering 80 percent of the U.S. population, is now combining offline and
online sales data to help its consumer goods clients make more intelligent, audience-
centric predictions for in-store promotions. According to Eric Williams, Catalina’s chief
information officer, this approach is generating 8-10 percent coupon redemption rates
(versus 0.5 percent rates for comparable mass-market couponing programs).
“By linking this data, we are creating a total purchase history that will allow us to
categorize and stratify consumers into purchase category buckets and infer what will
be of interest to them before they actually buy,” said Williams.
8

9. © 2012 Winterberry Group LLC.
Channel Optimization
Enabling “right message, at the right time, via the right media” targeting; expanding
the role of consumers in choosing optimal/preferred communications media.
Fundamental Maturity Core Beneficiaries Long-Term
Advertising Benefit Level Potential
Effectiveness/Efficiency: Low: Traditional advertising and E-commerce High: Migration to media-
Allows for the strategic marketing efforts have been Marketers, Publishers, agnostic communication
utilization of media in structured around the Lead Generation strategies stand to enhance
alignment with the inherent deployment of individual channels Portals consumer engagement,
strength of those channels, through distinct campaigns, and promote a robust dialogue
as well as consumer migration to true “media- and reinforce both single-
preferences; engages agnostic” models that seek to purchase behavior as well as
audiences at a richer level match audiences to lifetime customer value
and minimizes investment in optimal/preferred output levers
wasted/suboptimal channel requires process, technology and
efforts data source alignment that most
marketers have not yet
undertaken
The rapid introduction of new addressable marketing channels over the past two
decades—starting with the emergence of foundational digital media such as email,
search and display advertising, and hallmarked today by the maturation of tablets,
smartphones, addressable television and other media—has reinforced consumers’
technological sophistication, and provided them with a new span of control over
marketing content. At the same time, the diversity of promotional options has
introduced a new challenge to both publishers and advertisers: maintain a marketing
dialogue that matches strategic intent to optimal delivery channel, but honors
consumers’ choice with respect to messaging cadence and medium.
Brands that are able to integrate multichannel data across channels—effectively
becoming “agnostic” to the deployment of any single medium—hold the prospect of
creating holistic, near-360-degree views of customer preferences and intent regardless
of channel. The result is more relevant advertising—delivered at the optimal time, via
the consumer’s preferred channels.
Executives across the marketing ecosystem agree that data owners are sitting on
mountains of valuable information that can be used to drive these kinds of media-
agnostic efforts, but say much of the potential of that data is being undermined by
efforts to deploy messages through “sexy” channels, such as social media platforms.
“Marketers are anxious to jump ahead into social and other burgeoning areas of digital
marketing, yet they shouldn’t overlook that they have a tremendous asset right on
their own website that can be used to make these efforts more effective,” said Marc
Kiven, founder of BrightTag. “Imagine being able to walk behind every customer in your
store and see where they go, what they look at and what they touch. This data already
exist… *marketers+ just need permission to use it and the technology to unlock it.”
9

10. © 2012 Winterberry Group LLC.
Advertising Yield Optimization
Maximizing the value of available advertising inventory by identifying and “selling”
high-value audiences across individual publisher properties and delivery media.
Fundamental Maturity Core Long-Term
Advertising Benefit Level Beneficiaries Potential
Efficiency: Allows advertisers to Low: Though technological Publishers High: For a publisher community
avoid investing in media on the advances are rapidly struggling to effectively monetize
basis of simple demographic allowing audiences to be content (both “premium” and
characteristics—where “sold” across distinct online among “long tail” sites that
impressions generally reach a media platforms, the generate less Web traffic), the
large number of suboptimal potential of the approach identification and optimization of
target consumers as a means of demands true cross-channel audience-centric inventory has
capturing good prospects from a yield optimization; most the potential to deliver
larger universe. (To publishers, publishers are very early in substantial revenue
the benefit is all about their efforts to integrate opportunities, possibly even
effectiveness—as optimizing traditional ad inventory supplanting existing approached
yield generates higher (where it exists) into a to advertising packaging and sales
advertising revenues) holistic optimization effort
On the supply side, publishers are moving fast to deploy third-party data overlays
(sourced largely through exchanges) and the services of data management platforms in
an effort to create richer audience profiles designed to maximize their yield (the rates
they may charge for advertising inventory) and improve the value of that ad inventory
for which traffic doesn’t warrant a “premium” sales approach or pricing.
With multiple data streams, typically, feeding internal systems in rapid succession,
publishers said data control, accuracy and processing speed are critical prerequisites
for identifying high-yield audiences across disparate media platforms. “We have two
big relationships with publishers and both recognize the need to control their data
ecosystem in a very robust way,” said David Soloff, chief executive officer of
Metamarkets. “They are carefully overseeing first- and third-party data and usage logs
and trying to uncover tremendous pockets of inventory that may be mispriced or
ignored. It’s great for building ROI.”
One publisher said that the benefits of yield optimization ultimately won’t stop with
more informed pricing of inventory. “Creative versioning,” he said, will allow
advertisers to provide variable, tailored content to different audiences across all of the
publisher’s properties—enhancing the effectiveness of each ad unit (while driving the
publishers’ ability to extract value from that inventory). “We can execute this idea now
on any given property, but we’re working on a way to be able to roll this out across all
of our sites,” the publisher said.
One major challenge, he added, has already surfaced as a barrier to capitalizing on this
potential: the ability and willingness of advertising sales teams to understand, embrace
and communicate the role of these complex ad units.
10

11. © 2012 Winterberry Group LLC.
Targeted Media Buying
Enabling the economical, value-oriented purchase of advertising media; delivering
targeted messages to audiences across a diverse, actionable range of channels.
Fundamental Maturity Core Beneficiaries Long-Term
Advertising Benefit Level Potential
Efficiency/ Intermediate: “Real-time Marketers (via High: Meaningful media-
Effectiveness: The use of bidding” (RTB) tools have Demand-Side buying efficiencies are
automated, real-time media matured substantially over Platforms), Digital already accruing to
buying tools allows for access to the past few years, and are in Agencies / Trading sophisticated users; deeper
audiences at “true” market common use by enterprise Desks value will come through the
pricing—eliminating the need to marketers across verticals coordinated use of these
invest in “eyeballs” that are not applications and the targeted
likely to value a message or messaging/offer tools that
offer; likewise provides a deeper allow for optimization of
platform for customizing message content, timing and
marketing offers or content in a cross-channel integration
move to expand relevance of
those underlying messages
Demand from advertisers for the efficiencies inherent in real-time bidding and the
improved effectiveness that comes through improving brand messaging relevance is
driving more sophisticated data use across both ad targeting and media buying
practices. Demand-side platforms (DSPs) and digital agencies (many empowered, over
the last few years, by the addition of automated trading desk capabilities) are leading
the market in this respect by enabling marketers to identify, “purchase” and target
high-value customers across channels, in rapid timeframes.
In particular, search and display retargeting programs—targeting site visitors who have
abandoned a shopping cart or left a site without otherwise converting—can provide
specific offers based on the visitors’ on-site behavior. “By way of example, one of our
retail clients… wanted to establish dynamic targeting rules as its customers came onto
its site,” said BrightTag’s Kiven. “By splitting its audience into control and test groups,
the retailer was able to understand the differences in behavior of users who saw a
retargeted message versus those who did not.” The results of this more flexible, rules-
driven approach to data collection and integration lets the company shift attention
from top-of-funnel branding efforts and work more closely with its DSP partner to
better manage retargeting bids.
Multichannel data integration is a critical component of improved media-buying
capabilities. According to one agency executive, integrating on- and offline data for one
of the agency’s advertising clients resulted in a nearly 30 percent increase in online
display performance.
11

12. © 2012 Winterberry Group LLC.
The Opportunity Ahead: Trends in Marketing Data Utilization
Rules-driven integration of disparate data sets: While traditional data management
practices were built largely around centerpiece “personally identifiable information”
(PII) elements—usually consumer names and postal addresses—the collection,
analysis and segmentation of online data demands the aggregation and
anonymization of virtually all data sets, challenging marketers’ fundamental ability to
draw distinct insights from consumers’ cross-channel interactions.
Marketers and publishers continue to be wary of using personally-identifiable (PII) data
in the digital realm due to concerns about consumer privacy and data accuracy. “You
can’t make a flawless data background when moving data between multiple devices
because there are too many unknowns when it comes to privacy,” says the CEO of one
DMP company. “With addressable TV, for example, the cable distributors have PII that
they could easily match up with computer background and deliver a custom broadcast
based on a customer’s search history, but no one is willing to bridge that gap yet *in
fear of running afoul of privacy best practices+.”
As a result, data collection, analysis and segmentation processes are being driven (or
constrained, depending on your perspective) largely by the need to aggregate and then
anonymize—remove any “PII” elements—wide swaths of both first- and third-party
data. In response to this inherent complexity, many are taking a cue from the data co-
op models that emerged in the 1990s (largely for use by catalog marketers) and turning
to data exchanges, where participating digital publisher data is blended, segmented by
interests and made available to all contributors to augment their own audience
insights.
Data collection,
analysis, First-party, browser-based data—collected primarily through cookies—is being widely
modeling and supplemented with this third-party data to scale data sets and identify large,
“lookalike” audiences of high-value customers. Available sources span a wide range—
segmentation
from social media registration data (including, at times, insight into income, age and
processes are gender), to transaction-based data that includes activity on shopping behaviors, to
being driven (or general-interest data indicating news and other areas of consumer interest.
constrained,
A debate is raging, though, about the value of third-party data. Some executives warn
depending on that it is becoming increasingly generic and, therefore, less valuable. “Third-party data
your perspective) has become over-commoditized,” said an executive at one media application
largely by the developer. “We are actually seeing a shift to first-party data.”
need to
Not so, said an executive of one data technology company. “Accurate third-party data
aggregate and remains valuable because it provides context, scale and cross-channel consistency. It
then anonymize gives advertisers useful context for messaging to know the demo- and psychographic
wide swaths of elements associated with a person interested in ‘Product X.’ It provides a level of
insight-driven scale that, even in online environments, still isn’t available to advertisers
both first- and
using first-party data alone. And it is the key mechanism for reaching target audiences
third-party data. across channels with consistent messages.”
12

13. © 2012 Winterberry Group LLC.
Actions Speak Louder Than Names
Despite the challenges inherent in the PII/non-PII divide, some data executives
downplay the importance of knowing a prospect’s name and address, arguing that
pixel-driven data—insight into what an individual browser does on a website or a
platform like Facebook—often brings the sought-after targeting capabilities, even
without a consumer name. “A cookie is just as good as an individual ID,” argued an
executive at one large media-buying platform. “Knowing what people do through
trackable cookies can be very sophisticated and pinpoint those who engage or convert
at higher levels by following their behaviors—whether through display, social, a
website or viral video.”
These strategies are being enabled by large, sophisticated machine networks and
algorithms that identify useful signals and patterns of behavior that can’t be found in
PII data alone. Ultimately, many said, the consumer’s name and address isn’t as
important in raw behavioral data to determine propensity to respond. Said TruSignal’s
Liebl: “We take first- and third-party data, put it in our modeling engine and let the
algorithm decide the attributes and segmentations that identify the person as a high-
value customer.”
13

14. © 2012 Winterberry Group LLC.
Improved Operating Infrastructures: The primary barrier to widespread marketing
data optimization resides within the marketing organization, rather than the data
itself. Specifically, legacy operating infrastructures—characterized by rigid
organizational “silos” and the paucity of data-savvy marketing operations, IT and sales
talent—are substantially hindering the maximization of data, processes and systems.
“A big hurdle is how companies are organized and structured. Traditional marketing
data is managed in marketing operations, behavioral data is probably with a VP of
digital marketing and then digital data will be fragmented across those groups or have
its own VP of social media.” – John Zell, VP, Global CRM Solutions, Razorfish
Many enterprises suffer from an embedded culture of “traditional” media
management; even though they may be deploying new digital channels (as led by
distinct planning, creative and delivery teams), they are often managing those distinct
efforts in organizational silos, separate and distinct from the company’s other
marketing channels and data sources.
Panelists reported that it’s uncommon for online and offline channel managers to
share data, and typical for different managers to oversee digital execution channels
such as email, social and search. Moreover, many organizations still rely on their
corporate IT function—which commonly has neither the budget nor the decision-
making authority to steer marketing programs—to manage granular marketing data
applications. In addition, installed legacy systems and architectures (frequently built by
different contractors with the intent of making integration with other platforms
difficult) can’t accommodate the number of channels and volume of data now
available, much less the need to integrate complex, real-time data feeds.
The answer that many forward-thinking companies have developed is to invest in the
development of a data accessibility culture–led by a chief data officer. “Ultimately
there’s going to be a chief data officer (CDO) that exposes the data to you and wrestles
away some of the technology needs from the internal groups,” said Christian Ward,
senior vice president at Infogroup.
“It’s a rare breed Advertising Sales Reps Lack Critical Technology Expertise
of person who
can understand “It’s a rare breed of person who can understand what’s going on technology-wise and
what’s going on tie it to the marketing world.” – Ari Buchalter, COO, MediaMath
technology-wise The second organizational challenge that has limited the monetization opportunities
and tie it to the linked to marketing data is a lack of sales expertise when it comes to data-driven
marketing advertising. Media sales reps, for example, are historically trained to sell inventory by
way of traditionally volume- and demographic-driven variables—estimated magazine
world.”
circulation, say, or television ratings. But to successfully sell “audiences” (as defined by
Ari Buchalter, COO disparate data sources) across channels, reps today must be technology savvy as well
MediaMath as media savvy. “Reps have to get in the dirt more to understand this new ad
14

15. © 2012 Winterberry Group LLC.
technology, and most don’t have a technical background,” explained Dwight Green,
Nielsen’s vice president of digital product leadership.
According to some industry executives, the most effective data-driven sales reps are
coming from the analytics sector because they know how to sell software as a service.
Additionally, staff with digital agency or DSP experience (reflecting an understanding of
trading desks and technology-driven data-use models) will be valuable, as they
understand the specialized buying of data-driven audiences.
Publishers built on traditional advertising sales are using education and training to
better prepare their sales forces to monetize data value through media. “The first thing
we’re doing is building a ski slope of analytics tools that include beginner, intermediate
and expert proficiency levels,” says one broadcast and online publisher. “We’ll build
out the simple tools first and invest in training and education to make it successful.”
15

16. © 2012 Winterberry Group LLC.
A strong network of data-centric technology and service partners: The fastest and
most efficient “big data” aggregation, analysis and throughput solutions require a
strong ecosystem of service- and technology-enabled partners. The burgeoning data
supply chain is proving to be extremely agile in streamlining and processing data for
marketing performance—supported by a new class of open-source tools (such as
Cassandra and Hadoop) as well as maturing data optimization providers that offer
new solutions for managing and redeploying large quantities of disparate data
sources.
“When you’re in a fast-moving environment and you want to create smaller segments,
that’s a machine-solved problem rather than a human-solved problem, and that’s the
problem we’re trying to solve.” – David Soloff, CEO, Metamarkets
“As these machine-learning processes become bigger and more important they will
need to be outsourced. More sophisticated data processing requires more exotic
software that is hard to master independently.” – Stewart Allen, CTO, Clearspring
There’s widespread industry agreement that achieving optimal data collection, analysis
and throughput performance requires a strong ecosystem of technology-enabled
partners, particularly as digital data—which is growing increasingly temporal (or time-
sensitive)—requires faster processing and integration.
The data ecosystem is proving extremely agile in the streamlining and processing of
data for marketing outputs, particularly on the demand side. Most DSPs, for example,
The “DMP have built trading desks that drive speed and efficiency in ad bidding and buying. A
approach” is corps of analytics-focused marketing agencies—grounded in data segmentation, but
grounded deeply often tasked with the execution of those strategies, as well—has emerged to drive
sophisticated audience modeling. And not to be left behind, email service providers
in a service- (ESPs) are adding more analytical services, including A/B testing, to improve their
driven supply clients’ targeting efficiency.
model,
distinguished by Other service-driven vendors, including agencies and data management platforms
(DMPs), have focused on analytics and segmentation to make data more usable for
the overlay of client marketing and advertising. As it matures, for example, the DMP market is
data access, progressively splintering into a number of primary specialty disciplines—focused
analytics and respectively on the aggregation of third-party data and intersection of ad network
technology, as well as “pure-play” models focused around the integration of customer
media data, with a variety of views into the underlying data.
optimization
capabilities (but What providers in all these groups share is a focus on integrating multichannel data
also, some streams to plug into CRM and other systems to provide data owners with a “360-
degree” view of their customers (and customer interactions). This “DMP approach” is
criticize, by a lack grounded deeply in a service-driven supply model, distinguished by the overlay of data
of core data access, analytics and media optimization capabilities (but also, some criticize, by a lack
management of core data management capacity).
capacity).
16

17. © 2012 Winterberry Group LLC.
Technology, meanwhile, is rapidly growing to meet the critical execution
requirements—including rapid throughput, low latency and high-capacity processing
power—that real-time marketing execution demands. New open-source tools such as
Cassandra and Hadoop, for example, provide “virtual” platforms for managing
unwieldy data sets.
Marketing data governance: Data governance has emerged as a critical priority for
virtually every player in the data ecosystem. But whereas organizations have long
employed policy experts to advise on the regulatory ramifications of data utilization,
many are coming to see marketing data governance—defining the “rules of the road”
for assigning distinct data sources to different promotional tasks—as an equally
critical go-forward priority.
Realizing potential value from a vast new array of data sources presents a series of
challenges wholly separate from those associated with process management,
technology or marketing strategy. By comparison, the basic governance questions
associated with data usage— dictating who may access a given data set, and what rules
or rights to data usage, data privacy and data security may be associated with its
deployment—are just as thorny, and present an even costlier potential array of risks.
17

18. © 2012 Winterberry Group LLC.
When it comes to umbrella data governance strategy, panelists were united on one
best practice: Unless associated with one’s own customers or other first-party data
sources, PII data is essentially off-limits for online targeting purposes. The potential
cost of doing otherwise—of running afoul of privacy regulation, or violating the
consumer’s inherent right to choice in managing his or her marketing
communications—are simply too great for most marketers to bear.
“We check with our privacy counsel Requirements of the DAA Self-Regulatory
before we do anything,” said Nielsen’s Principles for Multi-Site Data:
Green. “The penalties are tough and you
must have the right internal teams and  Organizations that collect multi-site
know the laws and acceptable standards data for purposes other than online
that are in place.” For many data owners, behavioral advertising must provide
the solution is to house data internally— transparency and control regarding
“behind the corporate firewall”—even Internet surfing across unrelated
when third-party solutions are used for websites
the purpose of managing data processing,  The collection, use or transfer of
Internet surfing data across websites
analytics or optimization.
for determination of a consumer’s
eligibility for employment, credit
For its part, the marketing data industry is standing, healthcare treatment and
moving to develop, publish and promote a insurance are prohibited
series of universal data-use guidelines in  Organizations must comply with the
an effort to provide self regulatory Children’s Online Privacy Protection
solutions that may assuage consumer or Act (COPPA) regarding the collection
regulatory concerns. The new principles— and use of children’s data
like the Digital Advertising Alliance’s  The Multi-Site Data Principles are
recent Self-Regulatory Principles for Multi- subject to enforcement through strong
Site Data—build upon FTC accountability mechanisms.
recommendations regarding the collection
of Web viewing data and establish a clear framework governing the collection of online
data that also provides consumer choice for the collection of such data.
Data transparency is a critical component of the solution. Industry executives agree
that consumers need to understand how their data is being used before they will begin
to trust brand use of that data. The preferred response for most marketers is to allow
consumers to opt out of some data use practices.
Individual vertical industries, too, are moving to balance their own unique marketing
concerns with the lucrative potential of new data sets and potential consumer
concerns about the use of that information. In the auto industry, for example—where
“data,” for example, could conceivably include detailed information on everyday
consumer whereabouts—the importance of maintaining best practices in all regards is
incredibly important. “The connected car will have a huge impact on our industry,” said
Paula Skier, senior product marketing manager for digital products at Polk. “Through
the combination of in-vehicle technology and smartphones, cars can be the conduit for
creating unbelievable amounts of data—driver and passenger attributes, driving
patterns, location, speed, media consumption, communication with other consumers
18

19. © 2012 Winterberry Group LLC.
and even social networking. But with access to this information, [the industry] has to
demonstrate a benefit to consumers so they are comfortable with and want to
participate in programs leveraging that data.”
Global Marketers Face Restrictive Data Use Environment Abroad
With the expansion of the global marketplace, U.S.-based companies with international
operations face greater privacy and data governance challenges. For example, each
European Union country has its own set of data regulations that, individually and
collectively, are more restrictive than their counterparts in the U.S. For example: “An IP
address is PII in every country except the U.S,” argued Catalina’s Williams.
The result is that execution of data-driven marketing abroad is even more difficult.
Vigilant awareness and compliance with data regulations within each country will be
critical as the industry continues to evolve. Such interactions could be self regulated as
U.S.-based data or governed under safe-harbor rules.
19

20. © 2012 Winterberry Group LLC.
Conclusion
The use of marketing data is evolving as rapidly as the technology driving it.
Today’s immature use cases will become tomorrow’s standard marketing practices.
Strategy will follow technology, as new suppliers push the marketing envelope to
identify and integrate offline and online data streams into broader data assets that can
be analyzed, segmented and modeled—creating audience profiles that cut across
channels.
Core direct marketing skills and practices in data analysis, segmentation and modeling
will continue to provide a solid foundation for emerging digital data use cases—but
must be augmented to account for new techniques and skills required to collect,
analyze, integrate and derive value in the face of these new applications. Meanwhile,
marketers and publishers will continue to grapple with a number of challenges posed
by big data: storage capacity and accessibility; machine-generated insights (i.e.
modeling and algorithms) versus human intuition and skill; consumer choice; and the
role of PII in digital marketing.
But there will be more growth opportunities, as well, as the relationship between top-
of-funnel branding and bottom-of-funnel conversion programs become better defined
in the online world. There’s widespread agreement among marketing industry
executives that consolidation is coming—and that it will encourage more brand
marketers and publishers to mature and grow their deployment of data use cases (and
maybe acronyms, too).
“Within the digital ecosystem we’ll start to see consolidation and horizontal
integration,” said Caribou Honig, a partner with QED Investors. “Point solutions
focused on a single channel will fall to the wayside unless they’re highly superior, and
even then they’ll be integrated with a platform somewhere. Ultimately, your online
display DSP, online video DSP, social DSP and PPC platforms will all reside on a single
platform.”
20

21. © 2012 Winterberry Group LLC.
Appendix A: Our Research Panelists Tell Us What
“Data Developments” They Expect to See as 2012 Unfolds
“Attribution and cross-channel performance aggregation will continue to expand and
become utilized to greater marketing benefit.”
“More advanced machine-learning techniques that incorporate meaningful data
points to predict outcomes. The algorithms are out there, we just need to plug in all
the disparate data sources—context, first- and third- party data, site, geography—to a
stable online ID pool in real time to deliver the right creative to the right person and
the right place for a brand to pinpoint the best prospect.”
“We’ll continue advancing in connecting multichannel marketing and personalization
via a host of services and technologies. In a short period of time, consumers will more
strongly voice preferences on how they choose to accept marketing messages.
Marketers quick to adapt to these preferences will pull away from the pack.”
“The metrics for display advertising need to change. Basic clickstream metrics provide
little to no insight into success/failure. Additionally, a large percentage of online
targeting through multiple platforms will be driven by data on the front end.”
“The convergence, with sufficient anonymization, of large offline data segments into
online platforms. It is an untapped resource and data companies and CRM marketers
are becoming savvier about the opportunities.”
21

22. © 2012 Winterberry Group LLC.
“Offline data is valuable and will bring the tried-and-true maturity of offline market
research and advertising lessons learned to the digital space. It will bring consistency
and scale back into the multichannel advertising equation.”
“The willingness to rework frameworks—especially in the area of offer management—
to reach customers and potential customers effectively and efficiently. Most
advertising agencies are unwilling to recommend this to their clients because it will
ultimately result in loss of revenue.”
“The combination of qualitative and quantitative measurement. Large companies
aren’t yet driven to ask how positive or negative sentiment reflects upon other
numbers. Where are the dependencies? How can this be measured and acted on?”
“The sheer volume of data available will require marketers and their respective
channels and vendors to be able to digest and deal with “big data.” Those who have
access to larger data pools will be exponentially better equipped and can significantly
build out market share.”
“With more robust offline data able to be connected at a sub-zip code level as a geo-
targeting technique, the use of single-threaded cookie attributes as the definitive
targeting methodology will fade. Inferred interest as a sole metric will fade and so
could the privacy issues associated with tracking users.”
22

23. © 2012 Winterberry Group LLC.
Appendix B: A Marketing Data Lexicon
Evolving technology, data and marketing process are ushering in an entirely new
language to define how advertisers, publishers and intermediaries do their work. The
below lexicon—developed predominantly by the IAB Networks and Exchanges
Committee—represent a small selection of the terms appear that appear in this paper
and may be new to various constituencies of the advertising ecosystem:
Term Definition
Ad Network Provide an outsourced sales capability for publishers and a means to
aggregate inventory and audiences from numerous sources in a single
buying opportunity for media buyers. Ad networks may provide
specific technologies to enhance value to both publishers and
advertisers, including unique targeting capabilities, creative generation
and optimization. Ad networks’ business models and practices may
include features that are similar to those offered by ad exchanges
Cookie A small text file sent by a website’s server to be stored on the user’s
Web-enabled device that is returned unchanged by the user’s device to
the server on subsequent interactions. The cookie enables the website
domain to associate data with that device and distinguish requests
from different devices. Cookies often store behavioral information
Data Management A technology-enabled infrastructure for managing the aggregation,
Platform (DMP)* integration, analysis and redeployment of multiple first- and third-party
data sources, particularly online
Demand Side Provide centralized (aggregated) media buying from multiple sources
Platform (DSP) including ad exchanges, ad networks and sell-side platforms, often
leveraging real-time bidding capabilities of said sources. While there is
some similarity between a DSP and an ad network, DSPs are
differentiated in that they do not provide campaign management
services, publisher services nor direct publisher relationships
First-Party Data* That which is sourced by, owned and managed by an entity (or its
direct affiliates on its behalf) independently
Personally User data that could be used to uniquely identify the consumer.
Identifiable Examples include name, social security number, postal address and
Information (PII) email address
Pixel (or Beacon) An HTML object or code that transmits information to a third-party
server, where the user is the first party and the site they are interacting
with is the second party. Pixels are used to track online user activity,
such as viewing a particular Web page or completing a conversion
process
Segment A set of users who share one or more similar attributes
Third-Party Data Data that did not originate from either the publisher or advertiser.
Typically this is used to enhance ad targeting. For example,
demographic data from a third party might be used to help determine
which auto ad (make/model) to display on an auto site
* Defined by Winterberry Group
23

24. © 2012 Winterberry Group LLC.
A global leader in interactive marketing services, Acxiom Corporation connects clients
with their customers through deep consumer insight that enables profitable business
decisions. We incorporate decades of experience in consumer data and analytics,
information technology, data integration and consulting solutions for effective
marketing across digital, Internet, email, mobile and direct mail channels.
Headquartered in Little Rock, Ark., Acxiom serves clients around the world from
locations in the United States, Europe, Latin America and the Asia-Pacific.
For more information, please visit www.acxiom.com.
IBM Netezza data warehouse appliances revolutionized data warehousing and
advanced analytics by integrating database, server and storage into a single, easy-to-
manage appliance that requires minimal set-up and ongoing administration while
producing faster and more consistent analytic performance. The IBM Netezza data
warehouse appliance family simplifies business analytics dramatically by consolidating
all analytic activity in the appliance, right where the data resides, for industry-leading
performance.
Visit http://ibm.com/software/data/netezza to see how our family of data warehouse
appliances eliminates complexity at every step and helps you drive true business value
for your organization.
For the latest data warehouse and advanced analytics blogs, videos and more, please
visit thinking.netezza.com.
24

25. © 2012 Winterberry Group LLC.
The Interactive Advertising Bureau (IAB) is comprised of more than 500 leading media
and technology companies that are responsible for selling 86 percent of online
advertising in the United States. On behalf of its members, the IAB is dedicated to the
growth of the interactive advertising marketplace, of interactive’s share of total
marketing spend, and of its members’ share of total marketing spend. The IAB
educates marketers, agencies, media companies and the wider business community
about the value of interactive advertising. Working with its member companies, the
IAB evaluates and recommends standards and practices and fields critical research on
interactive advertising. Founded in 1996, the IAB is headquartered in New York City
with a Public Policy office in Washington, D.C.
For more information, please visit www.iab.net.
Winterberry Group is a unique strategic consulting firm that helps advertising,
marketing, media and information companies build value. Our services include:
Corporate Strategy: The Opportunity Mapping strategic development process
prioritizes customer, channel and capabilities growth options available to advertising
and marketing industry firms, informed by a synthesis of market insights and intensive
internal analysis.
Market Intelligence: Comprehensive industry trend, vertical market and value chain
research provides in-depth analysis of customers, market developments and potential
opportunities as a precursor to any growth or transaction strategy.
Marketing System Architecture: Process mapping, marketplace benchmarking and
holistic system engineering efforts are grounded in deep supply chain insights and
“real-world” understandings—with a focus on helping marketers and publishers better
leverage their core assets.
Mergers & Acquisitions Due Diligence Support Services: Company assessments and
industry landscape reports provide insight into trends, forecasts and comparative
transaction data needed for reliable financial model inputs, supporting the needs of
strategic and financial acquirers to make informed investment decisions and lay the
foundation for value-focused ownership.
For more information, please visit www.winterberrygroup.com.
25