> Retention is a terrible metric for a dating app. The "perfect" dating app would have a 0% retention rate since the first person you meet would be your "ideal partner".

A terrible metric for the users. For the shareholders, the perfect dating app is one where the users are strung along for their whole lives, paying for a subscription but never finding anything permanent.

This is why you fundamentally cannot fix dating apps without changing the incentives around dating apps. The only incentives that modern apps have today is to, as you said, string users along, just frustrated enough that they can't leave, but not frustrated enough to quit outright.

There are ways to fix it, but you have to fundamentally restructure how the app works. Instead of subscriptions, the app has to be paid upon success. Did the matches lead to marriage? Great, they get paid. Otherwise nope. Until incentives are aligned there is literally no financial incentive for them to do otherwise.

There is one app I'm aware of that works this way, but I won't link it so I don't end up sounding like a shill account. But as a user I really hope they succeed, because the current generation of dating apps aren't in anyone's interest except the people who run them.

The only other option would be to have one run by a non-profit organisation.

This is mentioned in the article:

> I don't think retention is a good metric for a dating app, unfortunately that's how VC evaluate performance of B2C apps. When we think about dating it is more about quality than quantity.

For long-term relationships, yes. For the people that exclusively use them for hookups and similar, those will "always" be there.

That's not a business model. Creating opportunities to mate, not to therapy. ;)

Managed Dissatisfaction, FOMO, Abundance: And the users want it. That's what the data shows. Proof: essentially, every app is a dating app covered up as "social features."

Apple just entered the dating app market.

> Managed Dissatisfaction, FOMO, Abundance: And the users want it. That's what the data shows

Correction: the users don't "want" it, but many of them aren't strong or wise enough to resist those features.

The (presumably usage) data can't show what the users want, only what they do, and those often aren't the same thing.

I think users "want" it the same way smokers "want" cigarettes.