Behavioral Targeting Online: A Case Study

1

• Behavioral Targeting Online A Case Study

2
Outline

• Introductions Behavioral Targeting and its
  role in online media
• How to measure and value Behavioral Targeting.
• WSJ/American Airlines Case Study Campaign Results
• composition
• effectiveness (awareness)
• cost efficiency
• How Behavioral Targeting will evolve

3
What is Behavioral Targeting

• Behaviors provide the clearest window into
  peoples desires and interests

4
Revenue Science Leaders in Behavioral Targeting
WEB PUBLISHERS
SAMPLE AUDIENCE SEGMENTS
ADVERTISERS

• Car Buffs
• Health Nuts
• Travel Seekers
• Mutual Fund Aficionados
• The Leisure-Minded
• Engaged Investors
• Frequent Travelers
• Affluent Users
• Gamers
• Hardcore Sports Fans
• SUV Shopper
• End of Year Rebates
• Hooked on Convertibles

5
How does Behavioral Targeting work?
Visitors Are Assigned To Publisher created
segments Segments
Visitors Navigate

• What people say
• Registration
• Search
• What people do
• Website
• Categories and Pages
• Words
• How often they do it
• Frequency
• Recency

On Future Visits, Targeted Ads Are Served To
Segment Members Regardless of Where They Go On
The Site (Instead Of House or ROS)
6
How to Measure Behavioral TargetingAudience
Composition

• True metric of reaching the right audience
• Measures core value proposition of targeting
• Audience Composition Qualified / Total
• Widely used in print media
• How does a leading airline make campaign
  allocation decisions for advertising in the print
  world?

7
Case Study

• Collaboration between WSJ.com, Revenue Science,
  Global Airline Carrier and its Advertising Agency
• Conducted audience composition and advertising
  effectiveness study using Dynamic Logic tools
• Used Test/Control methodology
• Brand Metrics Control Unexposed, Test IBT
  Exposed
• Brand Metrics Control Unexposed, Test RON
  Exposed
• Audience Composition Control RON, Test IBT

8
Behavioral Targeting Delivers Significant Lift
in Effectiveness for Branding Campaigns
Business Travelers 1 Business Trips Per Year

• Higher Composition across the board (115 to 145
  lift)
• Better targeting can deliver higher compositions
  than naïve targeting
• BT gives you brand lift over RON (on 5/6 metrics)
• In target brand lift is higher on 6/6 metrics
• When you get the right people, they remember the
  message more.
• The BT campaigns are more cost effective than RON
  even if you dont count the brand lift!

Result 115 lift inaudience comp.
9
Behavioral Targeting Delivers Significant Lift
in Effectiveness for Branding Campaigns
Business Travelers 5 Business Trips Per Year

• Higher Composition across the board (115 to 145
  lift)
• Better targeting can deliver higher compositions
  than naïve targeting
• BT gives you brand lift over RON (on 5/6 metrics)
• In target brand lift is higher on 6/6 metrics
• When you get the right people, they remember the
  message more.
• The BT campaigns are more cost effective than RON
  even if you dont count the brand lift!

Result 145 lift inaudience comp.
10
American Airlines Campaign Results Right
Message Target Audience

• Message Association for the targeted business
  travelers experienced statistically significant
  lifts for all three messages.
• The following table compares the results for the
  Qualified Business Travelers
• Total Respondents RON 820 IBT 373
• Qualified Respondents RON 216 IBT 210

Statistically significant difference between
Control and Exposed group at a 90 confidence
level Lift
(Exposed-Control)/Control Please Note Lift
percentages are calculated from the actual
Control and Exposed group scores, while the above
presented scores are rounded, if attempting lift
calculation using the scores above there will be
differences from the actual lift due to rounding
11
Economic Case

• CPT (cost per thousand targets) CPM/Composition
• ROS CPT CPM 35/0.263 133
• Behavioral Targeting CPT 60/0.563 106
• RON is 25 more expensive than Behavioral
  Targeting

12
Where is Behavioral Targeting Going

• A major piece of the online mix
• Standards for finding and comparing audiences
• Network Buys and better privacy protection
• Advertiser contribution to behavior

13
Thank You

• Omart_at_revenuescience.com