Not really: ads have a budget of attention within an envelope that Engagement (the team whose work is described here) is trying to maximise. There’s detailed es

Not really: ads have a budget of attention within an envelope that Engagement (the team whose work is described here) is trying to maximise. There’s detailed estimates to how much ads take away but it’s minimised and isolated from that conversation completely. There are many issues, some addressed there; some more nuanced (figuring out that is a good post or comment is genuinely hard with crude metrics because clickbait and flamewars look very much like compelling content); some that I can’t mention publicly. There are non-scalable ideas for solutions (my PhD was around complex network so check Jure Leskovec’s research for ideas like isolating cliques and defining them as meaningful or not). One idea that I’ve floated, but less than I should, and many other have too, but more than they should (or at least not articulated well around this problem of sorting good from bad content to inform the News Feed algorithm): have secret negative and positive reactions per post and comment. The fact that Likes are public makes them performative, both very compelling (it’s who you want your friends to see in you) and counter-productive. But giving people the ability to say: "look, I will comment on every flame bait that I see, because I can’t help myself, but please, remove from from my feed" would have helped argue against what is still a very linear model of weighting acting that are uniformly seen as "engagement". Source: I was the DS looking into those questions in 2015.