Back to overview

Internet Entrepreneurship

Failure Rates in Internet Startups Have Not Improved Much Since 1999

Persistence of a pattern that optimism keeps obscuring

Olusegun Adeyemi 999 views 836 likes
Failure Rates in Internet Startups Have Not Improved Much Since 1999

The commonly cited statistic is that 90 percent of startups fail. What is less commonly discussed is that this figure has remained roughly consistent across every decade of internet entrepreneurship. The tools have changed. The infrastructure has matured. The failure rate has not shifted in any meaningful direction.

What changes between eras is the cost of failure

In 1999, failing meant losing investor capital measured in millions. In 2012, the lean startup methodology reduced early-stage burn rates significantly. A founder in 2015 could launch an MVP for under 20,000 dollars that would have cost 2 million dollars in 1998. That is a real improvement. But it also meant more people entered, so the absolute number of failures increased even as the cost per failure dropped.

The framing problem in startup journalism

Coverage of internet entrepreneurship concentrates heavily on founding stories and funding rounds. The closure of a company rarely generates the same volume of press as its launch. A reader consuming startup media would reasonably conclude that most ventures succeed, because the failures are structurally underrepresented in the content ecosystem.

This is not a conspiracy. It reflects what readers click on. But it produces a distorted baseline expectation that affects how new founders assess their own odds.

One concrete way to check your assumptions

Search Crunchbase for companies that raised seed rounds in any given year from 2010 to 2016. Track how many have a recorded acquisition, IPO, or are still operating. The numbers are clarifying.