Startup Failure Rates

The cluster focuses on the high failure rates of startups, frequently citing statistics like 90-99% failure and debating the low odds of success compared to lotteries or other gambles.

πŸ“‰ Falling 0.4x Startups & Business
3,386
Comments
20
Years Active
5
Top Authors
#3796
Topic ID

Activity Over Time

2007
61
2008
94
2009
138
2010
171
2011
144
2012
236
2013
241
2014
173
2015
188
2016
212
2017
224
2018
165
2019
159
2020
153
2021
164
2022
230
2023
249
2024
201
2025
173
2026
10

Keywords

forbes.com AI dot.com YC HN techcrunch.com AT hbswk.hbs medium.com fundera.com startups startup succeed success failure rate failure fail business failures extraordinarily

Sample Comments

gfsdgfsdgfeds2 β€’ Nov 3, 2023 β€’ View on HN

How is it "very likely" when most startups fail?

kubb β€’ May 10, 2025 β€’ View on HN

It’s kinda like startup math. Most will fail but the few successes will offset the failures.

ocdtrekkie β€’ Jan 13, 2017 β€’ View on HN

Isn't the failure rate of startups like 99% already?

krzrak β€’ Dec 30, 2017 β€’ View on HN

Seems not that bad, considering that 90% startups fail [1][1] https://www.forbes.com/sites/neilpatel/2015/01/16/90-of-star...

stevage β€’ Apr 4, 2023 β€’ View on HN

It probably won't succeed, because most startups don't succeed.

riffraff β€’ Feb 8, 2021 β€’ View on HN

aren't most startups unlikely to succeed?

vm β€’ Apr 29, 2017 β€’ View on HN

Most startups go out of business. Raising $ doesn't guarantee success.

1vuio0pswjnm7 β€’ Nov 29, 2023 β€’ View on HN

Nevermind "AI". Fact is, most startups are doomed. Only small percentage succeed.

toomuchtodo β€’ Dec 7, 2024 β€’ View on HN

It is highly unlikely anyone here will ever have to worry about this problem, the odds are worse than winning the lottery [1] [2]. ~90% of startups end up in failure [3], for example.[1] https://medium.com/@13032765d/the-chance-of-becoming-a-billi... | <a href="https://archive.today&

philip1209 β€’ Feb 15, 2017 β€’ View on HN

Considering the failure rate of startups, I think the data shows that we just have a representative sample :-)