I don't believe in a conscious conspiracy within Facebook to defraud advertisers. I absolutely believe in a pervasive, semi-conscious but unspoken feeling that asking deep questions about the effectiveness of Facebook advertising or the validity of ad metrics would raise conclusions that would be definitively bad for Facebook, so it doesn't happen.
I interned on Atlas (FB's enterprise advertising platform purchased from MS a while back) and this was the impression I got. Everything has a dollop of feel-good verbiage piled on top. We're not selling ads, we're "building more meaningful connections between businesses and their customers."
The initial value proposition of online ads was click = confirmed intent. When not enough people click, they construct increasingly convoluted notions of purchasing seconds of mindshare at a time. Of course, at a certain point these purchases become as untestable as a billboard on a highway, a big difference being that highway billboards don't prop up industries, ecosystems, and even geographical regions like online advertising does.
If someone calls shenanigans on online ads, it will be far more catastrophic than if they did the same on billboards/radio/TV ads.
+1 "construct increasingly convoluted notions..." to rationalize or support anything is a huge red flag in this case and, imo, life in general. Understanding the tax system requires trudging through increasingly obscure rules? It's probably a broken, inefficient system. Health care billing system rules are convoluted, vague, and complex? Yeah, it's probably pretty broken. Need a convoluted, complex explanation to rationalize your support for Peter Thiel despite his endorsement of certain political figures? Again, a huge red flag.
Sort of like that YouTube auto-play-by-default bullshit. I know that as a user I've accidentally had it plow through dozens of videos -- and ads -- while I was AFK, but it probably looks good on somebody's metrics...
Many sites also have auto-replay looping enabled. Add to that, Facebook's "you thought you were clicking to pause/stop, but it's actually full-screen now."
Any internal team at least slightly aware of the problem would not have a good performance review for the year if their biggest project was to reduce revenue by x% due to efforts to address this.
