Hypothesis: more addictive may well not actually be more profitable. I wouldn’t be at all surprised if social media engagement was Pareto-distributed, with the most active people quite possibly not the most lucrative customers for advertisers to advertise to. This is immaterial if effectiveness-of-advertising is Goodharted into ”page impressions”, but a pile of Nash Equilibria which only works because multiple parties are Goodharting against noisy proxies can collapse abruptly.
So, kinda like… most people never click on any Facebook ads, but there is a small fraction of morons that click on all of them… but there is no simple known method to make the morons stay on the website longer without also making everyone else stay longer? (Exaggerated to make a point.) Sounds possible.
(Epistemic status: guesswork)
Hypothesis: more addictive may well not actually be more profitable. I wouldn’t be at all surprised if social media engagement was Pareto-distributed, with the most active people quite possibly not the most lucrative customers for advertisers to advertise to. This is immaterial if effectiveness-of-advertising is Goodharted into ”page impressions”, but a pile of Nash Equilibria which only works because multiple parties are Goodharting against noisy proxies can collapse abruptly.
So, kinda like… most people never click on any Facebook ads, but there is a small fraction of morons that click on all of them… but there is no simple known method to make the morons stay on the website longer without also making everyone else stay longer? (Exaggerated to make a point.) Sounds possible.
Whales, to use the metaphor used by casinos.