3. Winterberry Group: Helping Advertising, Marketing, Media and
Information Companies Grow Value
Strategic Consulting
• Corporate Strategy Development
• Market Intelligence
• Marketing Process Optimization
• M&A Transaction Diligence Support
• Investment Banking Services, through
4. Our Agenda
From Information to Audiences
What inspired the research?
What role is data playing in
digital advertising today?
How should we be thinking
about our “data strategy” for
the future?
5. In the Beginning, There Were Subscriber Files…
Name:
Addre
ss:
H. Catalogus
(0-~1980 A.D.)
6. … Which Begat Demographic “Selects,” Data Cards and the First True
Commercial Data Models…
NAME ADDRESS
PHONE
GENDER
AGE
INCOME
H. Catalogus H. Mailinglistus
(-~1980 A.D.) (~1980s)
7. … Which Begat Modeling, Cluster Segmentation, Cooperative Databases
and—With the Arrival of the Internet—E-mail Data…
H. Catalogus H. Mailinglistus
(-~1980 A.D.) (~1980s)
8. … Which, In Concert with the Growth of “CRM,” Gave Rise to
Sophisticated Database Management, CDI and MDM Infrastructures…
“Customer File”: Contact Info,
Persistent Identifiers
CRM, Demographics
“Prospect File”: Demographics,
Credit Scores
Interactions Logs
Transactional / Loyalty Records “Single Source of the Truth”
Public Records
Self-Reported “Intent” Data Mr. John Q. Customer
One Response Rate Way H. Analyticus
Boston, MA 01234 (~1990-2000s)
9. But “Evolution” Isn’t Always Painless; The Emergence of Digital Channels
Has Brought With It a Deluge of New Data Sources
Direct Mail
Call Centers
Catalogs
Retail Transactions
Print Publications
Broadcast Outlets
Email
10. And Those Various Channels Generate—and Rely Upon—a Range of
Information Types
Transactional added from
Psychographic and behavioral purchase records, cooperative
compiled from surveys, analytical databases
models
Offline Social compiled from
Providers
social sites, blogs,
Geo- Social Sites / sharing sites,
Demographic Offline Online
compiled from Compilers Providers
?
publishers, Online Data Types:
databases and • Registrations
other third parties • Cookies (Flash) /
browsing activities
Publisher
Portals / • Social networks
Online • Online purchase data
s
Compilers • In-market purchase intent
Artwork Source: David Harbaugh, Harvard Business Review
11. … And So “Traditional” Database Infrastructures Are Being Asked to
Support Vast New Streams of Unstructured Information
?
Behavioral (Clickstream)
Intent (Opt-In/Registered and
Inferred)
Web Analytics (Geo-/
Technographic)
“Customer File”: Contact Info and
Demographics
“Prospect File”: CRM
Demographics, Credit Scores
Transactional / Loyalty Records
Public Records
H. Digitalus
Self-Reported “Intent” Data (~2009-Today)
12. But The Integration of “Traditional” and “Digital” Data Poses a Set of
Unique Challenges, Owing To Discrepancies Between…
Known Names/Addresses… “Batch” Processing…
m er
n Q. Custo Way
Mr. Joh onse Rate
esp
One R , M A 01234
n
Bosto
… and Anonymous IP Addresses … and Real-Time Deployment
Campaign-Driven Execution… Single-Channel Focus…
… and Continuous Targeting … and Integrated Marketing
13. Today, The “Use Cases” for Marketing Data Differ Substantially Across
Addressable Media
14. Our Agenda
From Information to Audiences
What inspired the research?
What role is data playing in
digital advertising today?
How should we be thinking
about our “data strategy” for
the future?
15. Our Panel: Senior Thought Leaders Across the Data Ecosystem
“Which Best Describes Your Job Role / Function?”
N=176
Source: Winterberry Group survey
16. “To What Extent Are the Following Use Cases Focal Points of Your
CURRENT Data-Driven Marketing Activity?”
Not a focus of our A significant focus of
Source: Winterberry Group survey current data utilization our current data
utilization
17. “To What Extent Do You Believe The Following Use Cases Will Be Focal
Points of Your FUTURE Data-Driven Marketing Activity?”
Not likely to be a focus Likely to be a significant
Source: Winterberry Group survey of our future data focus of our future data
utilization utilization
18. Use Case: Audience Optimization
Identifying customers and likely Fundamental Effectiveness: Identifying customers
prospects through the integration of Advertising and likely prospects through the
rich (though disparate) data sources; Benefit integration of first- and third-party data
managing cross-channel marketing sources
execution with the goal of engaging
those audiences strategically—and in Maturity Level Low: Despite technology advances,
accordance with consumers’ preferred uncertainty around the optimal
advertising media. approach to structured integration of
data
Core E-commerce Marketers, Digital
Beneficiaries Advertisers, Lead Generation Portals,
Publishers (for traffic acquisition)
Long-Term High: The ability to define high-
Potential potential audiences and facilitate
multichannel communication
represents a fundamentally new way of
marketing
19. Use Case: Channel Optimization
Fundamental Effectiveness/ Efficiency: Enabling “right
Advertising message, at the right time, via the right
Enabling “right message, at the right
Benefit media” targeting; expanding the role of
time, via the right media” targeting;
expanding the role of consumers in consumers in choosing
choosing optimal/preferred optimal/preferred communications
communications media. media
Maturity Level Low: Traditional marketing efforts are
channel-specific; “channel agnostic”
internal alignment that most marketers
have not yet undertaken
Core E-commerce Marketers, Digital
Beneficiaries Advertisers, Lead Generation Portals,
Publishers (for traffic acquisition)
Long-Term High: Media-agnostic communication
Potential strategies will enhance consumer
engagement (through dialogue and
purchase behavior)
20. Use Case: Advertising Yield Optimization
Fundamental Efficiency: Maximizing the value of
Advertising available advertising inventory by
Maximizing the value of available
Benefit identifying and “selling” high-value
advertising inventory by identifying and
“selling” high-value audiences across audiences across individual publisher
individual publisher properties and properties and delivery media
delivery media.
Maturity Level Low: Though technological advances
are rapidly allowing audiences to be
“sold” across distinct online media
platforms, the use case demands true
cross-channel yield optimization
Core Publishers
Beneficiaries
Long-Term High: For a publisher community
Potential struggling to effectively monetize
content, the identification and
optimization of audience-centric
inventory has the potential to deliver
substantial revenue opportunities
21. Use Case: Targeted Media Buying
Fundamental Efficiency/Effectiveness: Enabling the
Advertising economical, value-oriented purchase of
Enabling the economical, value-
Benefit advertising media; delivering targeted
oriented purchase of advertising
media; delivering targeted messages to messages to audiences across a diverse,
audiences across a diverse, actionable actionable range of channels
range of channels.
Maturity Level Intermediate: “Real-time bidding” (RTB)
tools have matured substantially over
the past few years, and are in common
use by enterprise marketers across
verticals
Core Marketers (via Demand-Side Platforms),
Beneficiaries Digital Agencies/Trading Desks
Long-Term High: Meaningful media-buying
Potential efficiencies are already accruing to
sophisticated users; coordinated use of
these applications and the targeted
messaging/offer tools will deepen value
22. “To What Extent is Your Company (Or Your Clients) Realizing Value From
the Following Data Sources?“
We (or our clients) are We (or our clients) are
Source: Winterberry Group survey realizing no value from realizing significant from
these data sources these data sources
23. “To What Extent Do You Believe Each of the Following Are Driving
Deeper Interest/Investment in Marketing Data?”
Not a focus of our A significant focus of
Source: Winterberry Group survey current data utilization our current data
utilization
24. “To What Extent Do You See the Following Attributes Driving the
Underlying Usefulness of a Marketing Dataset?”
Not at all important in Critically important
driving the value of a in driving the value
Source: Winterberry Group survey data set of a data set
25. “To What Extent Do You Believe Each of the Following Are Inhibiting
Interest/Investment in Marketing Data?”
Not inhibiting interest / Substantially inhibiting
Source: Winterberry Group survey investment in interest / investment in
marketing data marketing data
26. Our Agenda
From Information to Audiences
What inspired the research?
What role is data playing in
digital advertising today?
How should we be thinking
about our “data strategy” for
the future?
27. The Complexity of Today’s Advertising and Marketing Programs Has
Driven Many to Re-Examine their Internal Operating Silos
What’s at stake? Holistic Marketing
• Data Process Management
• Strategic Resources/Authority
• Creative Assets
Effective People
• Investment Capital
• Knowledge/Expertise Management Brand Mktg.
Digital Direct Mktg.
LoBs Sales
Fin. Int’l Mktg.
IT
28. Requirement: Improved Operating Structures
Holistic Marketing
Efficient, Effective
Process Management
Ad./Mktg. Execution
Strategic
Effective People
Management
Utilization
Process People Technology
of the
Design Mgmt. Deploy.
Supply
Chain
30. Requirement: A Strong Network of Data-Centric Technology and Service
Partners
Real-Time Media Buying
Asset Management
Ad Serving
Scheduling/Routing
Distributed Mktg. Mgmt.
Ad Operations
Budgeting
Business Rules Mgmt.
Creative Production
Creative Optimization
Yield Management
Data Processing
Ad Verification
Web Analytics
Campaign Management
31. Requirement: Marketing Data Governance
PII vs. Non-PII External & Self-Regulation
Transparency Marketing Use Guidelines
32. “To What Extent Do the Following Reflect Your Long-Term (2013 and
Beyond) Priorities for Improving the Usefulness of Marketing Data?”
Improving attribution across channels 4.1
Integrating across different data collection… 3.9
Improving access to granular audience… 3.9
Identifying dedicated data staff and building… 3.9
Improving automation abilities 3.9
Improving data hygiene and quality assurance 3.7
Linking data inputs (and the appropriate rules… 3.7
Extracting greater value from the use of data… 3.7
Simplifying the processes by which third-party… 3.6
Reducing latency / processing data more rapidly 3.5
Storing / warehousing data more efficiently 3.4
Clarifying regulatory barriers to data utilization 3.3
1 2 3 4 5
Not a priority for us Very important priority for
Source: Winterberry Group survey (or our clients) in the us (or our clients) in the
long term long term
33. Thanks to Our Sponsors and Contributors!
Jonathan Margulies
Managing Director
jmargulies@winterberrygroup.com
(212) 842-6031
www.winterberrygroup.com/ourinsights
www.iab.net/marketingdatause
Editor's Notes
-- The span of marketing “use cases” is broad, and the only way to really understand what’s working and what’s not was to poll the entire data ecosystem -- We set out to learn a few things: How mature is your deployment of each of these “use cases” (however you might define them)? How much value are you receiving from your efforts? To what extent do you expect that each of these will be part of your mainstream advertising/marketing strategy in the future?
-- In kicking off our efforts, we turned to a very senior-level panel of marketing data decision-makers -- Included in-person, telephone and online surveys with 176 senior decision makers… spanning advertisers/marketers, publishers, agencies, marketing service providers, technology developers and other folks who are using data or involved in the compilation and processing of information
-- We started, as you might imagine, by asking our panelists to assess a very wide range of potential use cases AS THEY’RE DEPLOYED TODAY—essentially, tell us how intensely they’ve been focusing on almost every use case we had been hearing about -- What you see here, simply, is that there is pretty substantial interest in a wide range of use cases. If we assume that every instance that scored “3” or higher is rising in interest, then at least 10 major data applications are growing in demand, with at least half of those attracting substantial resources
-- Next: we asked the same panelists to take a look into the future and guess which use cases they expected would grow in importance -- For the most part, the lineup of “priority” use cases remained the same… with one exception. Panelists said OFFER OPTIMIZATION would grow substantially in importance… -- And when you think about it, that use case—which really speaks to advertisers’ ability to imbue their online ads with very specific, targeted, relevant content on offers that are personalized to the consumer or cohorts of consumer—really does speak to the great potential of online advertising: the ability to target not just on the basis of MEDIA…. or PRODUCT…. or even “AUDIENCE”… but at the granular intersection of all three of those -- It’s worth drilling down more deeply into the “big four” use cases we identified
-- The first use case is something of a hybrid of the ad and offer targeting cases we asked about, and is very much at the heart of the innovation efforts in the online marketing world today -- All about identifying actionable audiences from disparate data sets, including first- and third-party resources -- Ultimately, the goal is to reach and ENGAGE these audiences (and MICROaudiences, really) across media, using online behaviors as the baseline -- The challenge, though, is in identifying and maintaining a steady dialogue… given limitations with respect to tracking and specific user identification. Does an IP address always constitute a specific “audience” member? How do we track these audiences across “traditional” media? -- These are significant challenges. And even though great investment and effort is going into this use case today, it’s why maturity remains very low
-- The second use case is closely related to the first one, but adds an element of audience DECISION-MAKING to the mix -- Channel optimization is all about developing strategies that seek to deploy the “right media, at the right time… to the right audience,” irrespective of the actual message -- Part of this is and ought to be dictated by the user (through preferences). And part of this resides in true marketing “science,” through predictive preference modeling -- But on both counts, the industry is still immature. Preference centers are common in email marketing, for example… and beginning to gain in prominence on the display side (especially moreso now that the industry and government are coalescing around a series of standards that provide for consumer choice over their “trackability”), but true MULTICHANNEL optimization is a long ways off -- The potential, though, is very high. And it will require deep data integration (and a lot of historical attribution/performance data) in order for this to grow
-- The third use case is the only one specific to PUBLISHERS, rather than advertisers… and it speaks to the integration of data to drive the value of individual media units, essentially allowing publishers to set optimal pricing for their ad units in alignment with the potential of the audiences beneath them -- The real value here is in being able to show true value for an ad unit, which has been elusive for centuries. (John Wanamaker’s: “I know half of my advertising is wasted…”) -- A brewing debate here continues. Between data and technology constituencies who see this as driving substantial efficiencies to the way that ads are sold… to current-state ad sales teams who see this as a threat to their ability to package large ad units for sale -- This is a case where the potential is likely high, but a substantial educational and re-engineering effort will be needed to align the interests of buyers and sellers
-- The final “big picture” use case is really the flip-side of yield optimization… and that’s the buying of media on the basis of likely AUDIENCE through efficient, targeted media platforms -- This is the one case where maturity levels have finally grown to an “intermediate” point… and that’s because, in simple terms, buying “targeted” media has been happening in one form or another for many years. What has emerged over just the past few, though, has been a new array of technologies (like RTB and demand-side platforms) that essentially allow for the automation of very manual, time-consuming purchase decisions -- The great potential here is in continuing to improve the quality/effectiveness of ad TARGETING… and in exporting these tools to a variety of other media to deliver seamlessness across the advertising effort -- At the heart of that, as in the case with the other use cases, is identifying which data elements really add value…
-- And when we asked panelists, it wasn’t surprising to see that FIRST-PARTY resources deliver the most value. Especially those that provide very tailored detail into target audiences and their past purchase behaviors and likely needs/wants as expressed through the social graph -- The challenge for the industry will reside in deriving value from the vast array of THIRD-PARTY resources which are in tremendous supply and we know can add substantial value if integrated appropriately into a holistic analysis
-- That thinking was expressed, too, when we asked what’s driving greater investment in data: the need to make better use of proprietary data, and the need to export those insights across a range of other media applications -- Ironically, that “multichannel integration” point could just as well speak to the potential impact of third-party data
-- Foremost among priorities, though, is the need for data to be accurate, recent and insightful. Specific priorities here will vary substantially across vertical markets, though. What’s “recent” or “insightful” for an auto campaign, for example, will vary substantially for a campaign for CPG products… the lifetime value of the customer, timing of the message, and depth of supporting detail all differing in potential impact
-- Ultimately, though, progress on all of the above fronts will require a new effort to make data a centerpiece of the multichannel advertising strategy
Fade from silos to 4 different pillars holding up marketing infrastructure
Fade from silos to 4 different pillars holding up marketing infrastructure