Transcript combination of TV and Internet advertising

Multi-Screen Impact: Is There
One Metric to Rule Them All?
Multi-Screen Impact Case Study for
Leading Entertainment/Film Brand
Joan FitzGerald, comScore
Anant Mathur, Starcom Mediavest
Measuring Impact of Media is
More Challenging than Ever
• New Media and New
Devices Creates More
Opportunity
– To create brand
experiences that
consumers enjoy
– To create compelling
programming to build
and retain audiences
• Marketing Mix
Modeling is not an
optimum solution
– Media responses are not
like promotions and
coupons
– Impacts are multiple, not
necessarily immediate,
and time periods are not
clearly defined
Single-Source Research to Uncover
Multi-Screen Effects
• Exposure to multiple media from the same individuals
• Break out groups of consumers who may
use media in dramatically different ways
– Media hasn’t been “mass marketing” in a long time
– Consumers who experience your brand using mobile are different
than consumers who only encounter the brand through TV
• Results potentially after each week of campaign, not
after 24 months of prior data
– Answer questions about what marketers should do now
Single Source Dataset
Major Entertainment/Film brand Case Study
• Impact of TV and Internet advertising on website visitation for a
Major Entertainment/Film brand
• Based on opt-in consumer “Test Panel” of 3,700 Persons
developed in 2010
– 25,000 person panel in progress
• Persons-level Internet usage matched to household-level
television viewing using blind, third-party match
• Data sources
– TV ad exposure: ad schedules matched to TV “set top box” viewing data
– Internet ad exposure: comScore technology
– Consumer response: website visitation via comScore technology
The Question of Multi-Screen
Impact
• Does the combination of TV and Internet advertising have a
greater impact than each alone?
– Does 1 + 1 = 2? Or Does 1 + 1 = 2.X?
• Is the impact of exposure greater for the target consumer
group?
• How much does the timing (recency) of the exposure matter?
• Can we determine the “equivalent” number of Internet banner
ads that equal the impact of a 30 second TV advertisement?
Results: Impact Measures
The Impact of Multiple Media is Greater
than Each Medium Alone
1 + 1 = 2.X
• TV and Internet advertising in combination had higher impact
on website visitation than TV advertising alone or Internet
advertising alone
128
100
Results
indexed to TV
only
advertising
exposure
77
Index of Behavior Probability with 3-Day Exposure Window
TV Only
Internet Only
TV and Internet
Results apply to
Case Study brand
and do not
necessarily apply to
other brands
More Recent Exposures Had Greatest
Impact on Behavior
• The closer the ad exposure to the behavior, the greater the
impact
• The greatest “recency” effect shown for the combination of TV
and Internet advertising
0.17
TV Only
Internet Only
0.12
TV and Internet
0.07
3-Day
7-Day
14-Day
Results apply to
Case Study brand
and do not
necessarily apply to
other brands
Target Consumers Had Greater Response
• Consumers in the brand target group had the greatest response
• Whether exposed to TV only advertising, Internet only
advertising, or both TV and Internet advertising
TV and Internet
Target
Internet Only
Not Target
TV Only
0
20
40
60
80
100
120
140
Results apply to
Case Study brand
and do not
necessarily apply to
other brands
Similar Targeting Impact Seen for Digital
Advertising and CPG Brands
Lift in In-Store Sales
*Source: comScore AdEffx Offline Sales Lift Norms for CPG, March 2011
Non-purchased based targeting includes, but is not limited to, the following types of buys: contextual, audience, run of site//run of network, etc.
**Source: comScore Audience Advantage, which leverages sophisticated predictive targeting algorithms that are created using anonymous panel and census data sources; targeting was
deployed on the Microsoft Network sites.
Note: Retail sales is measured by linking the comScore panel of 1 million U.S. Internet users to their retailer loyalty card data from dunnhumby, which provides a measure of the panelist’s in-store
buying activity.
#ARFAM6
It’s Possible to Quantify “Media
Equivalencies”
• For this Case Study brand, 5 Internet Only advertising exposures had the
equivalent impact of 1 TV Only advertising exposure
• The impact of exposure to both TV and Internet was higher than TV and
Internet alone, even at higher frequency levels for Internet
1 TV Only
Exposure
5 Internet
Only
Exposures
1 TV and 1
Internet
Exposure
Results apply to
Case Study brand
and do not
necessarily apply to
other brands
Using Both Internet and TV is More
Powerful than Using Each Medium Alone
• For this Case Study brand
– TV and Internet combined had a greater impact than TV or
Internet alone
– Ad exposures closer to the behavior had greater impact
– Target consumers had a stronger response than non-target
consumers
– Media “equivalencies” can be calculated and used in media
planning
Next Steps
More Questions
• What about sequencing of exposures?
– TV first? Internet first?
• What about different Internet advertising types?
– Such as banner ads compared to video ads?
25,000 Multi-Screen Panel
• 8,000 current panelists
– Target is 25,000 opt-in panelists with access to mobile, TV, Internet
• Building in 2011 and 2012
• Single-source cross-platform measurement
Thank you
Joan FitzGerald
[email protected]
Anant Mathur
[email protected]