In practice, I get the impression that the emperor isn't wearing a whole lot: It's really difficult to demonstrate that any particular ad campaign generates a p

In practice, I get the impression that the emperor isn't wearing a whole lot: It's really difficult to demonstrate that any particular ad campaign generates a positive return at all, let alone that any particular technique justifies its cost, or that something that worked well the first time someone tried it will continue working well once everyone is doing it. The whole space looks to me like a morass of wicked high variance, confounding variables for days, spillover effects, blah blah blah. Combine this with marketers wanting to be more data-driven - and needing to report numbers to their own bosses - and you've got a situation ripe for exploitation. I expect the adtech companies could sell pretty much anything, as long as it smells like numbers.