[runT1ME]

What fascinates me about the online dating market is that the optimal user experience is indirectly proportional to a customer's lifetime revenue in any site that is ad based or subscription based.

If a company was purely motivated by profit, I imagine that some clever use of ML would reveal an optimal ($ wise) timeframe for matching a user with their eventual mate (assuming their matching algorithm was perfect), and it wouldn't be 'as soon as possible'.

Too soon and you're missing valuable revenue, too late and the user gets frustrated and quits. Do companies do this? Probably not. As matching algorithms become more advanced and more people use online dating though...

[goodside]

Nobody at OkCupid thinks like this. We want people to enjoy the site so they're inclined to pay for premium membership, which they won't do if it sucks. We want people to return to the site if they're single in the future, and to recommend it to their single friends. Optimizing for ad impressions over user experience isn't a road we want to go down.

[gonehome]

This reminds me of a conversation Larry and Sergey had when they had first come up with page rank and were trying to sell it to existing search engines.

The companies complained that their algorithm was too good and that users would leave the search portal too quickly without seeing any advertising.

It was obvious to Page and Brin that this was stupid - pushing them to start Google.