And as an aside, conversion measurement is der rigueur in digital advertising— for obvious reasons, companies don’t publish it. It is the basis for A/B testing which uses large sample sizes to see which presentation is more likely to make people do or not do something. It’s easier than ever to tell where someone was exposed to something and the click trail they took to end up on a purchase screen. I don’t, nor have I ever worked in digital marketing, but this information is extremely easy to find. Additionally, the Freakonomics guys are no strangers to vibing their way through topics they know nothing about and mistakenly assuming their brief thought experiments and unchallenged researchers have uncovered something useful. It’s not like they’re never right, but they’re no stranger to unforced errors in their analysis and you’d be wise to not rely on them as a source.