I went through 542 dead mental health startups before building mine. The payer decided almost everything.

I'm building a CBT diary app, and before sinking years into it I wanted to know why this niche has so many corpses in it. So we went through everyone who left the market between 2000 and 2026: 542 digital mental health companies — shutdowns, bankruptcies, acquisitions, quiet dissolutions. Up to 18 coded fields each, at least two independent sources per company, and the report is generated from the dataset, so every fraction in it can be recomputed.

The strongest signal is who signs the check. Where the user pays, 53% of the companies are dead; where an institution pays, 21%. The groups are nearly the same size, 258 against 254, so it isn't a denominator trick. B2C died 2.2 times more often than B2B, and apps sold as a one-time purchase died at 85%, 22 of 26.

The finding that landed on me: a medical co-founder does nothing on its own — 47% exits with one, 47% without. My own team has one.

What this does not show: it's a graveyard, not a random sample, so the shares compare groups against each other and don't predict failure. Funding is disclosed for 59% of the companies, and 67% of the sample is US and UK.

Full report, seven findings with denominators:

https://mentalium.me/en/research/mental-health-startup-graveyard/

If you're building in health right now — who pays in your model, and how long did it take you to find that out?