• by Joe Mandese  @mp_joemandese, February 23, 2010

 

It may have seemed like a no-brainer when Publicis' VivaKi unit leaked the findings of an ambitious study to find an alternative to the current online video advertising standard - the pre-roll unit - that consumers might simply prefer select which ads they are shown online, but the real surprise is how much that preference affects the effectiveness of online video advertising. According to detailed findings released Monday at the Interactive Advertising Bureau's Leadership Conference, it is nearly five times more effective.

"That's crazy good," gushed Tracey Scheppach, the Starcom executive who oversaw the 16-month study, which collaborated with a variety of online video purveyors, including Hulu, who's AdSelector method was deemed the best new online video advertising format, albeit somewhat modified. Instead of simply enabling users to select, which creative from a given advertising brand they could watch prior to an online video session, the modified approach enabled users to pick brands from different product categories that might be more personally relevant to a user.

The result, Scheppach said during a briefing with Online Media Daily late last week, was that "unaided awareness," the criteria she prizes most among advertising recall criteria, rose 386% over conventional pre-roll ads evaluated during the test.

"If you love them, set them free," Scheppach said of the finding, paraphrasing a popular song by Sting. By them, of course, she means the consumer's ability to choose which ads they see, a concept that has been assumed to boost relevance and recall, but which has never before been validated to the extent of the new research. Specifically, she said, the average findings of the study found that when people select their own ads, they have an unaided awareness of 68% vs. only 14% for standard pre-roll ad exposure.

Another key finding of the research, and one that Scheppach said was personally even more surprising for her, was that when given the option, people selected their own online video ads 45% of the time.

"I was surprised that 45% of them chose, primarily because we didn't use any sophisticated addressability to it. We basically just took three different ads and put them up there," she explained.

Asked why the other 55% of users did not proactively pick the ads they were exposed to, Scheppach said half of those respondents said it was because none of the three ads they were able to choose from were personally relevant to them.

"The sophisticated part is that we have to figure out how to serve the most relevant ads to them," Scheppach said, adding that would be the focus of the next phase of research, which falls under the broader banner of "The Pool," a series of research studies VivaKi has embarked on to improve the effectiveness of online advertising for its clients, and to boost Publicis' P&L in the process.

Among other things, the findings of the Pool's research, and how it is applied, are the intellectual property of Publicis, and the agency holding company is developing a strategy for how to exploit that. Scheppach said that's still formative, and that the agency ultimately would like to see the entire industry benefit from the findings and utilize the techniques developed through the Pool, but she said the initial application would likely be for Publicis' agency clients, and that it remained to be seen whether other agencies and marketers would pay Publicis for the privilege of its insights. Among the possible scenarios she said, was the idea of spinning the AdSelector system off into a separate, free-standing business that would service the entire industry. Ad agencies have successfully incubated similar technology solutions in the past. DoubleClick, which the No. 1 ad server in the online industry, which is now owned by Google, for example, was originally incubated at Poppe Tyson, a now defunct ad agency that was part of Bozell, another now defunct ad agency, that has been absorbed into Interpublic.

Scheppach said the commercial exploitation of the AdSelector findings would likely take time to develop, and that the next phase of the research would focus on methods for improving "addressability" and finding ways to get a greater percentage of online users to select the ads they are exposed to.