Partnership Opportunity: Use of comScore

1
Unparalleled Insight into Consumer Behavior
The Professional Respondent Problem In Online
Survey Panels Today Gian M. Fulgoni MRA Annual
Conference June, 2005
2
comScore Networks, Inc.

• Founded in 1999
• Corporate headquarters Reston, VA
• Offices in Chicago, NYC, San Francisco, Seattle,
  Toronto and London
• 300 full-time employees
• 80 employed in sales/servicing survey-based
  research
• Experienced senior leadership team with an
  outstanding record of innovation in the market
  research industry
• Created the only continuous survey panel that
  also passively captures all web-wide behavioral
  and transaction data for a sample of 3.5 million
  people
• Global access to survey 10 Million people, with 5
  Million resident in the US.
• More than 500 clients

3
Recognized For Our Accomplishments
Americas Fastest Growing Private Companies
comScore ranked 97th among the nations fastest
growing privately held firms. November 2004
Fifth Fastest Growing Research Firm Currently
ranked the 30th largest U.S. research firm up
ten spots from previous year. June 2004
Top 100 Innovative Companies Named one of the 100
companies that will shape the next year in
technology. December 2004
Rising Star Named one of only three Rising
Stars among Virginia-based technology
companies. October 2004
High Tech Award Recognized as a technology
leader for successfully developing innovative
products. November 2004
Worlds Largest Database Grand prize winner for
developing the largest NT decision support
warehouse. December 2003
4
Online SurveysMany Positives But Also Some
Emerging Concerns

• Custom survey research is increasingly moving
  online because of the benefits of faster
  turnaround and lower cost
• This is creating an apparently insatiable demand
  for online samples
• But, problems of over-fishing and declining
  response rates are raising concerns about
  respondent quality

5
A Possible Cause of the ProblemMultiple
Suppliers Using the Same Recruiting Site
6
Other Suppliers Are Using the Same Recruiting
Sites Favored By their Competitors
7
The comScore Panel

• comScore has recruited an opt-in consumer panel
  that has been validated to be representative of
  the online population and projectable to the
  total U.S. population.
• The comScore panel consists of millions of people
  who have given comScore explicit permission to
  passively measure all of their online activities
• Validated in comparison to Dept. of Commerce
  sales data
• Validated in comparison to individual site server
  log data
• Validated in comparison to offline surveys
• Panelists are recruited across thousands of sites
  not used by other panel suppliers and they do not
  have to be willing to answer surveys to be
  accepted to the panel

8
How The comScore Panel Benefits You

• A higher quality panel that delivers more value
  to your end clients
• comScore is the ONLY panel that can
• 1. Screen out professional respondents.
• 2. Truly validate the identity of the respondent.
• 3. Contextually pop surveys anywhere on the
  internet based on actual behavior.
• 4. Append surveys with online behavioral data.
• 5. Append surveys with other profile and offline
  information.
• This enables your company to
• 1. Deliver better quality, accuracy, validation,
  and insights.
• 2. Have the assurance and confidence that your
  results and advice are based on sound and
  projectable information.

9
comScores Customer Knowledge Platform A 360
View of Millions of People Observed Behavior and
Survey Measurement

• Designed to be projectable to the total U.S.
  population
• Ernst Young certified for information privacy
  security

DEMOGRAPHICS Self-reported Appended (e.g.
Axiom)
Individual Household Level
Technology Ownership (e.g. PDA, cell)
WEB VISITING VIEWING Web Site/Page Click
Streams HTML Proprietary Content
Search Engine Queries Product Service
Research
ATTITUDES General Behavior-specific
Awareness Intent
New Product Concepts,
etc.
ONLINE TRANSACTIONS Purchases/Subscriptions
Applications/Configuration
s Value/Price Paid
Shipping Handling Promotion
Incentives
OFFLINE PURCHASING Third Party Sources Cars,
Houses, etc Purchase Surveys
Supermarket Shopping Data
Credit Card Purchases
MARKETING STIMULI Banner Ads Referrals
10
Measurement Capabilities Of the comScore Panel

• comScores technology captures all panelists
  online activity, including the details of all
  surveys taken.
• All demographics (including name and address) are
  verified
• Using the comScore panel, its possible to
  measure the number of surveys taken by the
  members of any survey panel, encompassing surveys
  from all panels to which they belong.

11
The Professional Survey RespondentAlive and
Well Online

• comScore research shows that
• Only 0.25 of all Internet households account for
  30 of all surveys taken
• These heavy survey responders each took an
  average of 80 surveys per quarter in 2004
  thats about 1 a day!
• Some, but not all, of the leading panels contain
  too many of these professional respondents
• On average, a member of each of the leading
  panels belongs to at least 8 other panels

12
Some, But Not All, Leading Panels Are Biased
Towards Heavy Survey Respondents(Data Shows
of Each Panel Completing Specified Number of
Surveys From All Suppliers in a Three Month
Period)
A Leading panel
Other Leading Panel
13
The Professional Survey RespondentSome
important skews to consider when doing CPG
research

• comScore research shows that while heavy survey
  respondents are normal in many ways, there are
  important differences that can create problems in
  certain surveys. Compared to the average person,
  professional survey respondents
• Are significantly less likely to be employed full
  time and are less well educated
• Are significantly more likely to be influenced by
  advertising
• Shop much more frequently at grocery stores but
  are less likely to be warehouse/club store
  shoppers
• Are not as likely to consider themselves to be
  impulse shoppers
• Are significantly more likely to use coupons when
  they shop
• Are significantly more likely to think that
  buying food and/or household products with a
  brand name that you can trust is important
• Underestimate how many survey panels they belong
  to, as well as grossly underestimate how many
  surveys they actually complete in an average
  month

14
But, professional respondents are also similar to
the general population in some ways

• As likely to think of themselves as
• Politically active
• Spiritual
• Creative
• Organized
• Comfortable w/change
• Action-oriented
• Positive/hopeful about US Economy
• Concerned about environment
• Participate in many activities w/similar
  frequency
• Exercise
• Go to movies
• Read a book
• Eat out at a restaurant
• Go to religious services
• Go out with friends

• Spend similar amounts of time grocery shopping
• Spend similar amounts of money on groceries
• As likely to use a list for groceries
• As likely to have a budget for groceries and
  stick to it
• Were as likely to vote in 2004 election and who
  they voted for was similar
• Are similar in age
• Similar household size and composition
• Similar residence profiles

15
The Screen-out Challenge

• comScore research shows that professional
  respondents
• Underestimate how many survey panels they belong
  to
• Grossly underestimate how many surveys they
  actually complete in an average month (or they
  dont want you to know!)
• This makes it very difficult to screen out
  professional respondents simply by asking them
  how many surveys theyve taken.

16
Some Ways to Minimize the Inclusion of
Professional Survey Respondentsin an Online
Survey Panel

• Avoid recruiting panelists at sites favored by
  other panel suppliers
• Dont recruit only those panelists who want to
  participate in continuous survey programs
• Use large numbers of occasional survey
  responders also
• Limit the aggregate financial incentive you offer
  panelists by limiting the number of surveys you
  invite them to complete
• And be sure to take into consideration that your
  panelists are likely to also be compensated by
  their membership in other survey panels

17
Advantages of comScores Survey Panel

• Unsurpassed accuracy, quality and panel hygiene
• - ARF-endorsed sampling methodology
• - Random online recruiting integrated with RDD
  calibration sample avoids the bias of recruiting
  at only a few large sites
• - Patent-pending electronic monitoring of online
  behavior allows confirmation of respondent
  demographics
• - comScores ability to measure households
  participation in all survey programs permits the
  identification and exclusion (as required) of
  professional respondents
• Contextual pop survey based on actual behavior
• - Increases accuracy of screening by reducing
  reliance on respondent memory and minimizing
  recall errors
• - Permits real time surveying of consumer
  behavior
• - Address the latency issue

18
Gian M. FulgoniChairmancomScore Networks, Inc.

• Gfulgoni_at_comscore.com
• 312 775 6481