Toby Grabriner wrote in his 8/25/06 iMedia Connection article, Boost ROI with Audience Screening about pixel-tagging pages and segmenting audiences for better targeting and ad serving. Tony knows this technology well as he is he CEO of [X+1], formerly Poindexter. This excerpt does a good job of characterizing the potential of behavioral targeting:
…if you combine the best attributes of behavioral targeting with a number of different technologies including progressive optimization and the more advanced audience profiling engines you can accomplished what I will call audience screening. Audience screening allows the advertiser to identify the audience represented from an impression on a network or a portal and determine if that audience member is more or less likely to act in response to an advertisement than the general audience. If the audience member is regarded as highly desirable, then the ads are exposed. If the audience member is not deemed highly desirable, then they are not exposed to the ad and the next sequential audience member is evaluated for desirability and match to the potential customer base for the advertiser.
What Tony described is actually known as event-based targeting. Using a pixel on pages on a network or site, people are cookied so that the ad server can determine if that individual has been to a page or experienced a specific event in the past. Such a characterization of an individual, or event experience, puts them on a desirability list for targeting for an advertiser.
The issue here is that the individual is anonymous. We don’t know who they are. In fact we don’t know much of anything about them other then the fact that they carry a third-party cookie associated with an event. Could even be an event experienced by a different person who used the same computer. Someone did something desirable and received a cookie. Later they were targeted based on that cookie with an ad.
What about that third-party cookie?
Jupiter says that over 40% of third party cookies get deleted every month from adware, ant-spyware and user-deletions. So that means that only 60% of the time, behavioral, event-based targeting will be effective. The only we to re-establish the cookie is for an individual to re-experience the event and be re-cookied. Of course, they are cookied as if they are a new person, not recognized as the former person targted before. They remain anonymous.
Not to knock it too much. Behavioral targeting works for prospecting. It gets expensive and many retailers and marketers have found that the cost-benefit mathematics are not there. But the concept is. If the cookie deletion problem was not so rampant it might strengthen the model. If third party ad servers were more free with the data and provided the targeting features as a part of their services (not making it an incremental cost) it might strengthen the offer. But for now behavioral targeting like Boomerang, Tacoda and Advertising.com stays off to the side and is not mainstream online advertising. It get’s used but its impact is marginal.
_uacct = “UA-980395-1”;