Data Driven Strategies

1
Data Driven Strategies

• Integrating Quantitative Qualitative

2
Objectives

• Share some new marketing strategies that have
  been successful within the Financial Services
  arena in 2007.
• Get you thinking about breaking down traditional
  barriers that have existed within our space.
• Explore three examples from 2007
• Advancements in ITA marketing
• Segmentation Development and Execution
• Qualitative and Quantitative Integration
• Take learnings and apply to your business.

3
Why Merkle?

• Merkle works with over 50 different financial
  services clients around driving a better
  marketing result
• Our passion is to continue to push the DBM
  envelope and find new areas that can change the
  marketing landscape
• Merkle spends over a 1 MM in RD around DBM best
  practices and thought leadership
• We will share with you two emerging opportunities
  that we believe can drive big gains in your
  marketing results!

4
Breaking through Barriers to Growth

• Mature organizations face barriers to sustained
  revenue growth.
• They key is to understand how to respond to these
  challenges.

• Innovation
• New Audiences
• New Channels
• New Offers.

Revenue
Doing more of the same and playing with
incremental wins to maintain.
5
Biggest Opportunities in DBM

• Marketing Measurement
• Media Mix Optimization
• Enterprise Segmentation
• Brand Engagement Measurement
• Data Sourcing, Evaluation and Integration
• Whole Brain Marketing
• Integrating Quantitative Planning with Creative
  Execution

6
Whole Brain Marketing

• Lots of press and emphasis around Left Brain
  Marketing
• Forrester, What Sticks, Competing on
  Analytics, etc.
• Less talk about the power of research and
  consumer insights - the why
• Our belief is that it is not one or the otherit
  is both working together - Whole Brain
  Marketing!
• Traditionally, marketers have approached, managed
  and hired these two competencies separate from
  one another
• This has led to under-optimized marketing
  performance

7
Current Landscape
In general, we have observed four different
levels integrating the right and left brain
capabilities.
Level 4 One Agency Both capabilities exist
in one team The team works daily with each
other Neither capability Dominates Marketing
strategy is in a unified fashion Accountability
and the Big Idea working together!
Level 3 Virtual Agency Both capabilities
exist through separate groups Groups come
together periodically to share info Strategies
are debated and decided upon in this virtual
agency forum Turf war tends to exist

• Level 2
• Agency and DBM
• Both capabilities exist
• through separate groups
• Groups work independent
• from one another
• Both offer up strategies
• with marketer left to
• determine best approach

• Level 1
• Right Brain Only
• Creative and Big Idea
• people lead marketing
• Little quantitative skills
• Experience and intuition
• are key drivers

8
Research and Analytics

• Data Analytics (often called Quantitative) and
  Primary Research (often called Qualitative) are
  typically disjointed from each other.
• Example DBM company and Market Research
  Company. Very rarely do you go to one company to
  get these.
• Yet both are dependent on each other, both tell a
  story and both are data driven.
• Separating them can and will cause a loss of
  knowledge and create inferior results

9
Research Analytics
DESCRIPTIVE Research Focuses on the why and the
how of audience behavior identifying and
interpreting the meaning behind consumer actions,
decisions, beliefs and values and
understanding their decision making process
Qualitative Insights
PREDICTIVE Analytics Focuses on the who and
what components of direct marketing utilizing
data, facts, information and knowledge to
identify and create statistically valid measures
that drive business decisions
Quantitative Approach
10
The Combined Benefits

• Accurately identifying customer and prospect
  universes and segments
• Understanding purchase influencers
• Gauging reaction to new concepts/products
• Identifying key loyalty and attrition factors
• Optimizing channel expenditures
• Increasing the ability to predict campaign
  performance
• Moving beyond Beat the Control in creative
• Tangible and actionable insights to improve your
  value proposition.
• Ultimately...improved ROI

Qualitative Insights
Quantitative Approach
11
Process Overview
Creative Production
Analytics Test Design Behavioral Attitudina
l Data
Applicable Data
Quantitative Models
Objectives Strategy
Qualitative Insights
Content Solutions
Why?
What?
Who?
How?
Combined approach answers the root questions
12
Whole Brain Marketing Its Impact on ITA
Marketing
13
ITA Marketing

• Over the years, ITA marketing (especially within
  the credit card space) has struggled to find a
  firm spot within the marketing plan
• Major advances in technology, data and analytics
  have the ability to boost the performance of ITA
  programs
• We are observing that many credit cards companies
  are not capitalizing on this and are running
  sub-optimal ITA programs

14
The Credit Crutch

• Many companies have become too reliant on credit
  data!
• Past ITA performance poor
• Data, technology, analytics, creative and
  campaigns are built around pre-approved programs
• Difficult to change
• Assumed to minimized risk and loss
• Hard to get commitment from Sr. Management

15
Disadvantages of Credit Data

• Many disadvantages exist around utilizing credit
  data in 2007
• Increasing Legislation and privacy
• Expensive
• Limited audience
• Diluting offers and messaging to be conservative

16
ITA - Yesteryear Today
ITA Today
Many Major Advancements
Better Data - Predictive
Advanced Analytics
Segmentation
Data Lab vs. Vertical Lists
New Channels
ITA Yesteryear
Faster Learning Cycles
17
Making ITA Work - Approach

• Create an integrated ITA Marketing Team
• Get access to External Data
• Make use of powerful Internal Data
• Segment the Audience
• Powerful analytics
• Insight
• Offer, messaging and creative testing
• Measure, Learn and Improve

18
Integrated ITA Team

• Optimizing your ITA program requires a team of
  highly skilled people working together on a daily
  basis
• Requires both right and left brain marketers
• Marketing ITA leader
• Analytics lead
• Data Expert
• Insight specialist
• Creative Lead
• Program manager
• Campaign manager

19
External ITA Data

• Data is extremely important in building a high
  performing ITA program!
• In the last several years, the availability of
  external third party data has increased
  dramatically
• Summarized Credit Data
• Auto Data
• Mortgage, Homeowner Data
• Transactional Co-op Data
• Demographic Data
• Wealth Data and Models
• Lifestyle Data
• Life Event Data, Etc.

20
ITA Data Matrix
Summarized Credit Statistics
High
Auto Data
Mortgage Data
Universe
Demographic Data
Real Estate Data
Wealth Indicators
ITA Impact
low
high
Econometrics
Transactional
Lifestyle Data
Segmentation Data
Life Event Data
Low
Vertical Data
21
Internal ITA Data

• This is data that is available within your
  existing infrastructure
• Some of the most descriptive and predictive data
  come from your internal data
• Promotional history
• Summarized cardholder data
• Response history
• Derived Data
• Etc.

22
Derived Variable SPENDING VELOCITY
A modeled field estimating how frequently a
household is likely to spend over a period of
time. Higher values indicate high spending
velocity. Value and Ranges1 to 20

• TOP VARIABLES
• Avg loan amount all open revolving trades
• Avg Spending Velocity Index (CR level)
• HOH age Household size Car owner
• LM Finance Card Index
• Avg outstanding balance on all open bank card
  trades
• Avg number of revolving trades 30 or more days
  delinquent or derogatory
• Length of residence
• Percent of profile consisting of open retail
  accounts

ITA Impact Score
HIGH
LOW
23
Derived Variable INFERRED CARDHOLDER
A modeled field that estimates the probability of
the household having bank and/or credit cards.
The higher the ranking, the more likely it is
that the household has a credit card. Value and
Ranges 1-10

• TOP VARIABLES
• Marital status married
• Number of sources
• Merkle Donor Rating
• Avg Auto1 open date
• Percent population inside urbanized area
• Polk flag
• Merkle Wealth Rating
• Age range in HH
• Responders (Group Variables)
• Avg student high credit

ITA Impact Score
HIGH
LOW
24
Prospect Segmentation
Experian
I
III
II
ITA Universe
Credit Universe
Equifax
TU
Segmented Targeting Opportunities Segment I
Universe Expansion Segment II Leverage
non-credit data to optimize the
performance Segment III Maximize the Pre-screen
Program
25
Audience to Offer Expands Universe
Climbers
Urban Singles
Gen Y
Difficult
Underserved Seekers
Market Penetration Index
Poor Credit
Easy
Low
High
Affluent Families
26
Behavioral Segmentation
Descriptive Profiling Characteristics
ITA Segments
Spending Habits
Lifestyle
Travel
Cash transactions Credit transactions
Lifetime Average Spend Lifetime Transactions
Lifetime Quantity per Transaction Trailing
12 mo Spend Trailing 12 mo Transactions
Cars Electronics Home Garden Babies
Kids Apparel Jewelry Health Sports
Outdoors Entertainment
Foreign Travel Domestic Travel Premier
Foreign Travel Premier Domestic Travel
Gen Y
Climbers
Seekers
Urban Singles
Affluent Families
Poor Credit
27
ITA Success Audience to Offer vs. traditional
Offer to Audience Whole Brain Marketing
Profit Drivers
CREATIVE
CONTACT
MESSAGE
CHANNEL
MEDIA
Database Marketing Profit Drivers
2 to 1
3 to 1
3 to 1
5 to 1
6 to 1
Marketing Focus Changes from YOU to THEM
(Audience) Expands Prospect Universe Maximizes
and Optimizes Revenue Establishes More Relevancy
28
Building Executing the Plan Whole Brain
Marketing Profit Drivers
OFFER
AUDIENCE
Database Marketing Profit Drivers
10 to 1
20 to 1

• Need to build communication strategic roadmap
  that optimizes the
• Media (Channel) preference,
• Frequency,
• Channel of interaction,
• Messaging (value proposition) and finally
• The creative that brings all of the above to
  life!

29
Strategy vs. Creative

• Marketing campaigns conceived by blue-sky
  creative thinking are 4X more likely to fail
  than succeed.
• Campaigns based on insights are 15X more likely
  to succeed than fail.
• A statistic quoted by Michael Moon at a DMA
  Symposium in Amsterdam

30
The Proof

• Background
• Client was 100 Credit Data
• Losing share and needed universe expansion
• Tired creative with vanilla offerings
• Result
• Increased marketable universe by over 25 via new
  data,
• Lower Cost, Greater Response, Same Risk via
  segmentation scheme, modeling
• New Segment-specific Offers Messaging with
  comprehensive communications plan
• Through aggressive testing strategies, developed
  optimal communication plan for select segments
• Program deemed roll-out in 3 Months

31
Enterprise Segmentation Development Execution
32
Enterprise Segmentation

• That market is cluttered with segmentation
  schemes and solutions
• Some are robust and effective but most are
  vaporware with little impact
• Merkle believes that, at the highest levels,
  segmentation can and should be a strategic asset
  for a company

33
Segmentation

• Why Segmentation Disappoints
• Excessive interest in consumer identities rather
  than focusing on product features that matter
  most to consumers
• Too little emphasis on actual consumer behavior,
  which reveal attitudes and predict business
  outcomes
• Undue absorption in the technical details of
  devising the segmentation
• Creation of a solution that focuses on insights
  but fails to address actionability
• Lack of vision regarding methods and tactics
  enabling maximization of the segmentation solution

Source First three (3) reasons for
disappointment provided by the following article
Rediscovering Market Segmentation, by Daniel
Yankelovich David Meer, Harvard Business
Review, February 2006
34
Segmentation - Our View

• There is no single best approach to segmentation
• The right approach is one that will satisfy the
  overall objectives and leverage the right
  information
• Merkle believes that leveraging research-based,
  customer and marketable universe data is a best
  practice
• The solution must be both relevant to the
  marketing executive and actionable to the
  marketing manager

35
3 Major Types of Information
Customer
Research

• Primary Research
• Secondary Research
• Syndicated Panels

• Behavioral
• Product Mix
• Transactions
• Usage
• Revenue / Value

?

• Most directly tiedto customer value
• Strongest predictor of future purchase activity
• Product mix provides overall depth of relationship

• Provides insightsinto consumer needs and
  intentions
• Overall market share and total behaviors (not
  just with your brand)
• Self reported information, typically very simple
  to understand
• Get media consumption and awareness /
  consideration metrics

?
?
Universe

• Compiled
• Credit
• Verticals
• Co-ops

• Only source available on total marketable
  universe
• Other two buckets are a subset of this universe
• Bureau and co-op sources can provide some
  behavioral insight

36
Option 1 Start with Attitudes

• Benefit Directly gets to customer intentions,
  needs and attitudes
• Approach
• Drive the initial survey and segmentation guide
  it to leverage more widely available data and
  steer the initial segmentation to potentially be
  more predictable
• Build a unique methodology to map to DB and then
  re-profile and define the segments based on the
  post mapping procedure

Customers
Sample
InitialResearch
AdditionalResearch
Map segments to Universe
Marketable Universe
Prospects
Segmentation using Market Research
Sample
Panel
Seg 1
Seg 2
Seg 3

Seg n
Seg 1
Seg 2
Seg 3

Seg n
?
Build Initial Profiles on Segments
Re-Profile and Re-Define Final Segments based on
Prospect, Customer and 2nd Research Stage
Initial Segments(Sample Only)
Final Segments(Marketable Universe)
37
Option 2 Start with Customer Behaviors

• Benefit Assume that in mature markets,
  companies with decent market share have some
  penetration into all underlying (unobserved)
  market segments
• Given that, this approach can identify and
  describe the differences within these segments,
  especially between prospects and customers
• Approach
• Leverage market research, conduct research
  specific to each segment and drive the
  descriptions and profiles with that insight
  information for the presentation layer
• Starting with customer behaviors gives us a solid
  foundation on which to build a mapping process to
  the marketable universe

Marketable Universe
Customers
Segmentation using Customer Behavioral Data
Map Segments from Customers to Universe
Seg 1
Seg 2
Seg 3

Seg n
Seg 1
Seg 2
Seg 3

Seg n
Build Profiles on Customer Behaviors
Build Profiles on Customer Prospect Data
Customer Segments
Final Segments
Finalize Market Segments based on Learnings from
Research Use this information to color
segments
ConductResearch by Segment
38
Choosing the Approach

• Determining the right approach to strategic
  depends on the client and the situation
• Consider the companys business model and
  competitive landscape
• The maturity of the company and the product play
  a factor immature markets/company lack data and
  market share
• Determine the primary and overall objectives if
  there is a specific objective, perhaps a tactical
  approach would be better
• Consider modifications or customizations to the
  two solutions presented to best fit the situation
  (hybrid approaches)

39
Choosing the Approach
Mature / MarketLeader
Option I
Option II
Industry / Company
Option II
Option II
Immature/ New
Non-Direct
Direct toConsumer
BusinessModel
40
Segmentation Execution
41
Making the Segments Actionable
5
2
3
4
6
1
Coming of Age
Savings Hunters
Growing Families
Upscale Singles
Cosmopolitans
Conservatives
Prospects
of all US Households
13.3
15.6
11.2
27.7
14.3
17.8
of Client Inquiries (funnel Index)
15.8 (119)
9.3 (65)
Inquiries
20.8 (133)
12.4 (111)
29.8 (108)
11.9 (67)
Population
of Client Customers (funnel Index)
Customers
11.6 (125)
15.7 (75)
10.3 (83)
15.6 (99)
31.0 (104)
15.5 (133)
of Client Cancels (funnel Index)
15.7 (100)
Cancels
21.5 (137)
13.0 (126)
30.2 (97)
10.3 (66)
9.2 (79)

• 27lt30 yrs 14gt60yrs
• 24 with Bachelors or higher
• 5.5
• 39
• 39 married
• 21
• 2.9 Years

• 17lt30 yrs 21gt60yrs
• 19 with Bachelors or higher
• 6.7
• 38
• 50 married
• 33
• 4.7 Years

• 17lt30 yrs 15gt60yrs
• 29 with Bachelors or higher
• 7.7
• 48
• 60 married
• 47
• 5.6 Years

• 15lt30 yrs 18gt60yrs
• 37 with Bachelors or higher
• 7.6
• 38
• 58 married
• 41
• 5.0 Years

• 6lt30 yrs 45gt60yrs
• 44 with Bachelors or higher
• 8.9
• 18
• 62 married
• 56
• 7.0 Years

• 7lt30 yrs 40gt60yrs
• 37 with Bachelors or higher
• 8.4
• 22
• 62 married
• 44
• 6.0 Years

Age Education Wealth Rating Children in
Household Married Homeowner Length of Residence
Demographics

• Quoted Premium
• Premium per Driver
• Premium per vehicle
• Stated Previous Insurer
• Competitor 1
• Competitor 2
• Competitor 3
• Competitor 4
• Competitor 5
• Competitor 6

Inquiry Profile
Premium Premium per Driver Premium per vehicle
New Customers Tenure Internet Sale With a
new car with claim Gender
Customer Profile
42
The Funnel Framework

• Primary Framework for Clients
• The two permanent segments attitudinal segments
  (across the funnel lifecycle (down the funnel)
• The Funnel data can be selected based on time
  (most recent 1, 3, 6, 12 months) and geography
  (national, state, DMA)
• Many different metrics can be displayed within
  the boxes (population and indices shown to the
  left)

43
Identifying with Your Segments

• Some basic rules
• Telling them you know things about them with the
  copy or images does not win you their
  appreciation or drive results.
• Remember creative should always keep an
  aspirational (younger, wealthier, action packed)
  vibe.
• Listen to your segments and define them based on
  their needs, attitudes and behaviors.
• Tailor your offers (if you can) and/or your value
  propositions within those offers as your primary
  means of executing segment specific creative
  approaches.

44
Implementation Focus
75
The Offer
50
Impact
Value Prop/ Offer Message
25
Creative Design Copy
Focus
45
Recent Segment Specific Creatives
Based on the new segmentation, Merkle recommended
expanded testing to increase messaging strategies
and personalize the marketing approach Client
selected two segments of inquirers to start the
testing process Cosmopolitans Savings
Seekers Merkle and Clients internal creative
team would each develop new kits to drive
inquirer conversion and beat the longstanding 3
year control. The Result ALL TESTS BEAT THE
CONTROL
46
Valuable Messaging Insights
Cosmopolitans

• Demographics
• Oldest Population (66 over 50yrs old)
• High income group
• Most educated (44 with Bachelors or higher)
• Media Consumption
• Very likely to read newspapers
• Radio formats likely to listen to news/talk,
  sports
• Radio formats not likely to listen to urban, CHR
• Likely to watch Documentary, evening news, golf,
  horse racing, skating, tennis
• Not likely to watch daytime comedies, primetime
  comedies, daytime drama, game shows, reality
  shows

• Insurance and Automotive
• Highest levels of auto insurance coverage
• Attitudes and Perceptions
• Buy based on quality not price
• Loyal to brands they like
• Low consumer confidence feel they will be
  financially worse off within 12 months
• Focused on eating healthy
• Feel TV advertising is too loud and repeated too
  often
• Most likely to trust newspaper, Least likely to
  trust TV
• Likely to purchase items by phone
• Much more likely to engage in foreign travel

47
Valuable Messaging Insights
Savings Hunters

• Insurance and Automotive
• Want the cheap/easy to maintain vehicle
• Attitudes and Perceptions
• Tend to make impulse purchases
• Will switch brands for small discount
• Only save money for specific purposes
• Like humorous TV advertising
• Get least amount of sleep at night compared to
  other segments
• Most likely to say holding true to religious
  faith and beliefs is important
• Most likely to trust TV, least likely to trust
  internet

• Demographics
• Typically 30 55
• Low income group
• Lowest education level (19 with Bachelors or
  higher)
• Media Consumption
• Likely to watch Auto racing, daytime dramas,
  reality shows, primetime comedies, early morning
  news, game shows, primetime films
• Not likely to watch Golf, horse racing, tennis,
  news specials
• Radio formats likely to listen to country,
  urban, rock
• Radio formats not likely to listen to news,
  talk, oldies

48
Quantitative QualitativeWorking Together
49
Quick Story

• CFM Direct Had Great Results
• AOR of College Savings Product in the 529 Family.
• Handled all aspects Planning, Media, Creative,
  Production, Analytics
• Knew the product - had the insights, control
  and skills.
• Identified good stable of Vertical and some
  Compiled Lists
• Had great lead volume from multiple channels
  Schools, Pubs, Web Mail
• Butthe product was not performing the way they
  had planned - it was grow or die

50
Quick Story

• CFM Direct Purchased by Merkle
• Analytic horsepower and data options expanded
  exponentially
• Knew we needed a more precision based marketing
  approach - a way to SHOW the client hot to drive
  results
• Assembled a new Client team complete with
  Research, Creative, Analytics, Data and Search
  SMEs.

51
The Process and Approach
Quantitative
Qualitative
Analyzed Past Performance of all
Variables Lists, Creative, Channel, Value
Propositions, Seasonality, etc.
Delivered full year Precision marketing plan
with detailed learning agenda

• Analytic Approach
• Lower CPA via Look Alike Models
• Grow Universe via New Data Resp Models
• Identify Segments for Messaging

• Creative Approach
• Reviewed Secondary Research
• Created 3 New Key value propositions.
• Conducted Primary Research

Coordinated design of testing with execution to
accelerate learning agenda.
Out of the gateresults beat our existing
controls by 48
52
Qualitative Quantitative Why its Working
Sharing a Vision Strategic Roadmap Understanding
what the Business, Product and Client Requires
Face to Face Meetings Analysts, Creative,
Strategy, Production
Key Points of Integration Analytic Brief,
Analysis Review, Data Brief, Data
Review Creative Brief, Creative Review
Integrated Communications Schedules, Status
Reports, Invoicing, Approvals, Presentations,
Performance Review
53
Summary

• Whole Brain Marketing Bring the two
  competencies of left and right brain closer
  together

Client
Client
DM Agency
DBM Company
DBM Agency
Virtual Agency Good
DBM Agency Best
54
Summary

• Capitalize on the synergy and power of
  integrating research and analytics

Traditional Scenario
New Scenario
Market Research
Market Research
Quantitative Analytics
Quantitative Analytics
55
Thank You!