I got some results from Platform A for the behavioral advertising campaign we’re running for my book, The Numerati.
The study, involving some 12 million ad impressions, looks at three different audiences—business strategists, book-lovers, and general public—and measures their responses to three different types of ads, one focused on economic impact of the Numerati, the second on how their work will affect our lives, and the third on how scary this all is. (sample text: “Meet the Numerati… They’ve already met you.”)
I’ll paste the entire text below, for those who are interested. Most of the results fit into what you’d expect. Each audience responded more favorably the ads designed for it. One interesting note: Business strategists are very parsimonious with their clicks. So while in real terms they clicked less than other groups, they responded strongly to certain ads when measured against their own index.
The key piece of data—how likely these ads are to make them buy the book—should be coming later this month.
Insights from the Campaign Clicks
The campaign ran three sets of creatives, one with a business focused message, one designed for a general audience and one which attempted to scare the viewer with what the Numerati knew about them. These three ads were each shown to three different audiences - people interested in literature, business strategists and the general public. Finally a control campaign was run with a public service announcement ad to each of the three audiences as well.
The Numerati ads generated higher click-rates than the public service ad - People clicked on these ads at an index of 165 compared to the PSA.
Targeted audiences generated the lift - The targeted Numerati ads generated a much higher response with the right audiences, while showing the ads to uninterested audiences had little effect.
* General audience who saw the Numerati ad - index 106
* Book lover audience who saw the Numerati ad - index 199
* Business strategist audience who saw the Numerati ad - index 325
The ad creative designed for the general population generated the highest click rates - The indexed click rates for all respondents were:
* General ad - index 269
* Scary ad - index 244
* Business ad - index 100
When the right ad is shown to the right audience, the response rate increases - If we use the general ad click rates as a baseline we can determine the effect of serving targeted ads to different audiences. For example when the business audience sees the ads the click rates index quite differently for each creative.
* Business creative - index 161
* Scary creative - index 66
* General creative - index 60
*Note - Why use indexes? We know that different audiences click on ads differently. People who enter lots of online sweepstakes and have the time to play online games often click on ads at a higher rate than the general population. People such as business strategists are often some of the lowest clicking audience online. Instead of comparing peoples absolute click rates, which would only show that general audiences which contain many "sweepstakers" click at a higher rate, we want to look at how audiences click compared to their own norms. This tells us if an audience is more drawn to a particular ad than any other and normalizes the historic differences in click proclivity different shown by different people.