4 ms·
Hey everyone, I have a weird hobby of collecting data on startup failures (autopsies). I just finished analyzing 1,600+ failures (burning $500B+ in capital) an
by loot-dropdude 8mo ago
Hey everyone,
I have a weird hobby of collecting data on startup failures (autopsies). I just finished analyzing 1,600+ failures (burning $500B+ in capital) and put it all on a site I built just to share the findings.
The biggest surprise? We're often told startups die because of "No Market Need." But in this dataset, that was only the #9 cause (36%).
The real killers were Product Problems (85%) and Competition (82%)—which totally contradicts the old "startups commit suicide, they don't get murdered" saying.
I built this with no ads, no sign-ups, and no paywall. Just raw data for anyone who wants to avoid these potholes.
Hope it helps you build something that lasts.
- robocat 8mo agoDo you know any links about whether successful founders learn about failures or not? From my limited experience, founders concentrate on their current risks. I also think founders are good at filtering out the deluge of failure porn that is mostly irrelevant to their business: there's millions of ways to fail. Yes, it helps to learn the patterns. But I'm unsure learning even famous examples like "The Osbourne effect" helps a founder. (reëdited) What was your emotional drive for creating this?