As the story goes, Hillary wasn't willing to spend a lousy $5 mill on analytics, Trump was, and so he won.
(Kinda like the Orban cannon that brought down the Byzantine Empire? perhaps an overstretch comparison)
Anyway, the author provides several links to debunk this story & Cambridge Analytica's effectiveness. Of course folks on the winning side can say pretty much anything, but that don't make it true.
Here, there and everywhere.
More important, what do folks suggest Hillary should have done? Kriegel, her analytics head, focused on *testing* effectiveness of different channels, not just going with gut feelings. You can say her tests were wrong, or perhaps more intuition was needed, but what's the rational approach rather than "guess your ass off"?
Nate Silver has done a nice 8-part series of campaign retrospective, which points out in a number of places that "conventional wisdom" wasn't wisdom at all and not borne out by the facts, where in a couple places it was. But if anything the campaign taught us, we're quite susceptible to nonsense and seldom know what we really know.