| > When we all know that our data of who we are and our actions is just as valuable as the ads we are forced to watch. In a world where Facebook just makes up video impressions and neither they nor Google let you really audit their inventory, it is basically impossible to tell the difference between (1) seasonal effects, (2) data-enabled ad targeting, and (3) income-related effects. In other words, if you show an ad at the right time to a rich person, say by only targeting iPad users shortly after a new Apple product release, which is known from the user agent and does not require behavioral tracking at all, your ad will perform as well as a typical interest-targeted ad but with lower cost. Which is to say, your data is valuable in the marketing sense that it gets people to spend more on targeted ads, but not necessarily in some secular sense that it actually increases returns. After fundamentals (like what you are selling) and the creative, most of the evidence points to trends in how rich (and poor) people use technology, like iPad users versus entry-level Android users, as being the greatest predictor of an ad's returns. Ads customers don't really know if user profile information like their browsing behavior improves ROI because it is secularly important for the ad or because it is also correlated with greater income. It is unmeasurable for ad customers, so people claiming that they know are just lying. At least among the ad tech companies I know, I'm told the user profile data gathering story is very valuable for ad buyers and investors, but fundamentally they do not use any behavioral data in ad targeting because they did not observe an increase in clicks - for their guaranteeably low income user channels. |