Bill Harvey, author of TACODA's recent behavioral vs. contextual targeting studies, explains the significance of the results.

An independent study released in April of the impact of behavioral targeting vs. contextual targeting revealed that behavioral targeting identified and reached 50.3 percent more imminent buyers of Panasonic plasma TV's than contextual targeting, and that the cost to reach each potential buyer was 50 percent less than contextual targeting. The findings were phase two of ongoing research conducted by Next Century Media and Insight Express in conjunction with TACODA Inc. We sat down with Bill Harvey, founder of Next Century Media and author of the report, for insight.

Dawn Anfuso: In an earlier study you surmised that online audiences respond to behaviorally targeted ads at a higher rate than contextually placed ads because of clutter on context pages and the surprise of being served a behavioral ad in an unexpected place. What about the new study reinforces those suppositions?

Bill Harvey: In research there are two types of measures: evaluative and diagnostic. Evaluative measures tell you what happened and diagnostic measures tell you why they happened. The same Panasonic campaign was measured in both the earlier study to which you're referring, and in the current study. The current study is evaluative and shows that the campaign had higher advertising effectiveness (according to a number of standard metrics) when behavioral targeting was used than when contextual targeting was used. The prior study was diagnostic and showed that the reason why these evaluative results came out this way was because people tended to look more and longer at the same ads when they had been behaviorally targeted than when they had been contextually targeted. Therefore the two studies are in perfect agreement.

We are now more confident about our hypothesis that it is clutter of competing ads in the same category in contextual, versus two things in behavioral targeting -- lack of that clutter enhanced by the surprise of seeing an ad that has nothing to do with the surrounding editorial content -- which drive the tendency for the audience to look more and longer at the same ads in the behavioral targeting situation.

Anfuso: What audience attributes helped TACODA decide who was a potential Panasonic plasma TV buyer?

Harvey: Tacoda has been a practitioner in behavioral targeting for longer than most anyone else, having virtually invented the technique in its current form, and therefore has had more opportunity to experiment with endless alternative ways of defining targets and then observing the results. This has led to the use of many non-obvious and non-intuitive targeting approaches that constitute Tacoda's "secret sauce" and raison d'etre. Therefore, we can't reveal the target criteria used. The advertiser and agency involved in each Tacoda campaign of course are fully informed of these details and in fact play an important role in the customization of targets to specific campaigns.

Anfuso: A press release claims that it costs 50 percent less to reach potential Panasonic buyers via behavioral targeting vs. contextual placements. How was that equation calculated?

Harvey: We looked at the relative waste of the two targeting methods. The cost per thousand (CPM) of behavioral targeting was 24 percent less than the contextual placement, yet it delivered 50.3 percent more imminent purchasers of plasma televisions. Therefore the CPM against imminent buyers was 50.6 percent of the CPM of contextual targeting. Behavioral targeting was twice as cost effective as a channel for Panasonic to reach its target audience.

Anfuso: The new study showed a 168.9 percent advantage for behavioral targeting over run of network (RON) in terms of increasing the likelihood of buying the Panasonic's brand of plasma TV. What does that mean?

Harvey: The sample seeing the same Panasonic ad via RON was matched in every way to the sample seeing that ad via behavioral targeting. Everything was identical except the method of targeting. After the campaign, 10.6 percent of the RON sample said that they would be most likely to purchase a Panasonic plasma TV. The comparable stat for behavioral targeting was 28.5 percent-- 28.5 percent is 168.9 percent higher than 10.6. Behavioral targeting left almost three times as many people likely to buy Panasonic rather than another brand.

Anfuso: Behavioral targeting also generated a 63.1 percent lift in unaided brand awareness, which was 3.45 times greater than contextual targeting. Can you explain how you arrived at those conclusions?

Harvey: Contextual targeting raised unaided brand awareness by 18.3 percent. Behavioral targeting raised it by 63.1 percent-- 63.1 is 3.45 times greater than 18.3. Behavioral targeting had a much greater effect on raising unaided brand awareness.

Anfuso: The new study focuses on response to a specific brand, Panasonic. Do you think the results are projectable for any other brands?

Harvey: There's every reason to believe that they are projectable to other brands. Panasonic is not an isolated case. Three brands in different categories were measured in the recent eye tracking study. Many brands in many categories were analyzed in the industry case study review we conducted before that. There is a consistent pattern of results here. No matter which metric you go by, whether it's ROI or purchase intent or awareness or clickthrough or whatever, we are seeing behavioral targeting demonstrating an ability to reach targets at lower cost and also get their attention -- whether it's due to relative lack of clutter, or to non-contextual surprise, or both -- which drives noticing, consideration and a greater likelihood of a positive response.

There are three factors -- lower cost, higher proportion of targets and a noticing factor -- all advantaging behavioral targeting. The first two factors were part of the plan. The third factor -- greater noticing -- had never been anticipated and is an unexpected bonus in behavioral targeting. Apparently, internet users have grown used to seeing ads for products related to the web page -- contextual targeting -- so when they see unrelated products, but ones for which they have been recently shopping -- that's what behavioral targeting is -- not only are they automatically interested in the product (because they have been recently shopping for it and probably still are) the ad stands out and they engage with it. The same thing might happen with contextual targeting except that there, typically, a number of ads for the same product category clamor for the user's attention and the user's automatic reaction is to shut down and stay focused on the editorial.

Additional resources:
Learn about the results from phase one of the study.
Read the press release.

Bill Harvey has spent over 35 years as a pioneer and innovator of media research with special emphasis on New Media. At the start of his career, Mr. Harvey (as the 24 year-old strategy head of the American Research Bureau, now Arbitron) invented the Area of Dominant Influence or ADI, an audience-based definition of television markets that Nielsen emulated as the DMA, and which was called by Sales & Marketing Magazine "the most widely used marketing tool in the world today."

Mr. Harvey has been involved in every major New Electronic Media trial in the United States since 1975 and many other trials around the world. He has brought more advertisers and agencies into more Interactive TV trials than everyone else in the world combined.

Today, Mr. Harvey is active on three Advertising Research Foundation committees and his current passionate media interests continue to include maximizing and measuring Return On Marketing Investment (ROMI), addressable media, measuring media effectiveness, mobilizing set top box based audience data, and accelerating the profitable and psychologically appropriate Business-To-Consumer use of all digital media for two-way communications and relationship building.