> I worked at OkCupid 8 years ago... OkCupid's problem, and the problem at Match Group in general, is that you learned something about 100% of OkCupid users in

> I worked at OkCupid 8 years ago... OkCupid's problem, and the problem at Match Group in general, is that you learned something about 100% of OkCupid users instead of 0.01% of randomly-sampled real world people, and maybe that something that you learned was 200% wrong. And then, you think your answers are 1,000% more correct because 100% - 0.01% = 99.99%. Granted that logic is great for equity investors. It just isn't necessarily great for users. Nobody pushed back on the assumption that online dating was a representative sample. Nobody pushes back on whether the right questions were being asked. Match didn't even choose the right engagement metrics, choosing things like message reply rate which can rise while active users fall. The style of how these metrics were chosen and what analysis was done papered over serious issues. And all the people who would have, by now, the maturity and tenure to admit they were wrong are gone.