90% of startups fail. This appears to be an accepted statistical fact. But is it really? I believe it’s possibly one of the most misleading statistics out there. When I was writing a post on How To Properly Evaluate the Risk and Rewards of Startups I had this 90% statistic in my head and thought it was going to be really damning evidence against the success of startups. Turns out it’s not as it seems.

Finding the Statistical Source

I turned to Google to see if I could figure out where the 90% of startups fail narrative came from. The best source I could find was Failory mentioning the statistic and attributing it back to a Startup Genome report from 2019 (thanks Wayback Machine for caching it!). Unfortunately, Startup Genome cites Small Biz Trends claiming this 90% number, which I can’t find. More Googling led to this 2023 Small Biz Trends article which is in line with what I already knew and does not support a 90% failure rate. Within the first 5 years of a startup’s life around 50% will fail. After 10 years around 65% will fail.

Failory in their article seem to add a bit of context around what their number means and lumps this definition in with all startups:

Early-stage (idea stage) startups bear the highest risk and failure rates. It’s hard to claim accuracy about failure rate statistics for those kinds of projects because a large chunk fly below the radar.
They don’t raise capital from funds or other entities who maintain a dataset - most early-stage businesses are funded by the founders, their families, and friends.
Many early-stage startup projects don’t even register a legal entity – you don’t need one to test an idea. You need one once you start making money.

I suppose a 90% failure rate makes sense if applied to every idea or thing that was ever conceived. My son thought he’d invent bar soap in a bottle, is that a failed startup? I have had a LOT of ideas that I have half-built and then tabled. Are those failed startups? If I were to follow this definition, my failure rate would be closer to 99%. I just don’t think this is a good definition of what a failed startup is.

What Shocks, Grabs Attention

So which number is right? I guess it all depends on your definition of what a startup is. It also depends on if you can actually find the research that supports this 90% failure rate. I think the claim is flimsy and lacks the solid data or research to back it up. But that number is great for pushing your own narrative. People have latched onto it regardless of its validity. Y-Combinator uses it to “prove” that they are superior in their startup success rate. Parents use it to discourage their aspiring entrepreneurial children. It’s a shocking statistic.

LendingTree and Small Biz Trends both get their numbers directly from the U.S. Bureau of Labor Statistics and their evaluations come closely in-line to each other. “After five years, 48.4% have faltered. After 10 years, 65.1% of businesses have failed” (source: LendingTree). I think it’s also valuable to consider what failure even means. Yes 65% of businesses failed after 10 years, but 10 years for a startup seems to quite a good run in my opinion.

This Should Be Great News for Entrepreneurs

So why am I putting so much emphasis on debunking that 90% failure rate? A 50% failure rate within the first five years is still a big number! Well I’d like to tell you the risks aren’t quite as huge and scary as they may seem. Going from 90% to 50% is an extra 40% of success ready and waiting for you to take hold of. The odds are a lot closer to being in your favor. Stop letting people crush your dreams by randomly stating that you have a 10% success rate.