Facebook And Nielsen Play To Dominate Ad Effectiveness Research


Advertising effectiveness in media is a longtime research cesspool. For decades, advertisers have demanded data to “prove” their placements work, while allowing publishers and a small cottage sub-industry of market research vendors to use heavily skewed samples (often composed of a title’s most engaged readers, or a site’s most loyal users) to demonstrate such effectiveness.

Now, Facebook and Nielsen are getting into the act. CNET’s Caroline McCarthy reports Facebook, with a huge user base but under growing pressure to demonstrate profitability (its Beacon ad partner program provoked a privacy backlash and is being shut down today), is rolling out Nielsen BrandLift as part of a broader partnership:

It will use opt-in polls on Facebook’s home page to gauge user sentiment around advertisements, measuring “aided awareness, ad recall, message association, brand favorability, and purchase consideration.” It’ll roll out in the U.S. to a number of test partners this week and to all advertisers over the next few months. There will be “hundreds” of BrandLift tests in that time, the release explained.

People who take these polls can differ in profound ways from the vast majority of Internet users who are too busy for them, raising questions of how widely the results can be trusted. And depending on how the service is rolled out, there is risk for Facebook in crossing a user annoyance tipping point. [more]

Flawed sampling is not the only problem with ad effectiveness research: measures like “aided awareness” (picking brands off a list rather than in free recall) and brand favorability can strongly overstate or understate whether an ad really worked, and whether consumers will really buy after seeing it. Advertisers and publishers justify their reliance on these survey measures on the semi-plausible theory that the sampling and questionnaire flaws cut equally, so the results are valid for comparisons between brands and to observe trends over time.

Regardless, Facebook and Nielsen are a smart combination. Facebook is privy to social network data about its 300 million users that would make a sociologist (or the CIA) drool, and data has become the Internet’s most reliable goldmine. Nielsen, whose original brand equity in television ratings is threatened by the reshuffle of the media landscape, is wise to hitch itself to a social networking colossus like Facebook. Their challenge is to make the brand-poll features appealing rather than annoying, and to fend off questions about whether their survey results represent more than an oddly spam-hungry consumer subsample of social networking diehards.