Turns out good ol’ John Wanamaker was more than right:
Half the money I spend on advertising is wasted; the trouble is I don’t know which half.
– John Wanamaker
Google admitted that 56.1% of all the impressions served on Google’s Display Network — that is, served by a serious company, not by an un-named organisation running a dubious ad network from the middle of nowhere — could never have been seen.
But that’s only half of the story. Or 56% of it, if you wish. The more interesting part, and the one nobody talks about, is: what do you track? Tracking systems will tell you that a user “has seen” a certain banner ad and will speculate that even though he did not bother to click on it, he was in fact influenced by it when, on a later date, he encountered another ad or did a search on Google, clicked and then converted.
But can we trust this to be true, especially when Google is selling not merely Adwords on the search results, but a large part of the banner ads as well via Doubleclick? Another good question could be: is this guess done on ads that the users have seen, or on those served to them? A further one: why do we “attribute” part of the merit to banner ads nobody clicked on, but fail to do the same with print, radio or tv ads?