The most-cited statistic in startup literature is CB Insights' failure post-mortem analysis, and the most-cited line in it is that around 70% of failed startups "ran out of cash." Founders read this and draw the obvious lesson: raise more, spend less, extend runway. It's the wrong lesson — or at least, it's a dangerously incomplete one.
Running out of cash is how startups die. It is almost never why they die. Autopsying a company and concluding "cash depletion" is like autopsying a patient and concluding "cardiac arrest." Technically accurate. Diagnostically useless.
What the data says underneath the cash line
Look at the causes that sit upstream in the same research: poor product-market fit shows up around 43% in CB Insights' updated analysis, with bad timing (~29%) and unsustainable unit economics (~19%) close behind. Independent re-analyses that control for validation quality argue that market-need-adjacent failures — insufficient demand, mispricing, being outcompeted on a value proposition customers didn't prefer — account for something like 60–70% of startup deaths once you trace symptoms back to causes.
One more number worth sitting with: in one dataset of 431 failed companies, the median company had raised $11M before dying, and the median time from final fundraise to death was 22 months. These weren't companies that failed for lack of capital. They were companies that used capital to postpone a strategy reckoning — and the capital ran out before the reckoning was faced.
Translation: these are strategy failures
Strip the startup vocabulary and every top-of-list cause is a classical strategy failure. "No market need" is a where-to-play failure. "Outcompeted" is a how-to-win failure. "Pricing and cost issues" is an economic-model failure. "Bad timing" is a market-diagnosis failure. None of these is an effort failure — nobody in the failure datasets worked too little. They worked hard against unexamined assumptions.
Why this matters more after the startup stage
Here's the part relevant to readers of this site: if you run a $1M–$25M company, you've survived the fail-fast filter — which creates its own trap. Established companies don't collapse in eighteen months; they erode over five years. Revenue keeps arriving from the legacy engine while the strategic assumptions underneath it quietly expire. The startup graveyard's lesson applies with a lag: the market eventually invoices you for every strategy question you've deferred. At startup stage the invoice arrives fast; at your stage it compounds first.
The practical difference is that you, unlike a pre-revenue startup, have data: real customers, real cohorts, real margins by segment. The diagnostic work that startups can only guess at, you can actually run. Most founder-led companies never do — which is why the Diagnose phase so often returns a constraint the founder didn't walk in with.
Founder action: Write down the three assumptions your revenue most depends on — about why customers choose you, what they'd pay, and who else they consider. Date the last time each was tested against data rather than remembered. Anything older than a year is a candidate cause of death, currently in progress.
Sources & notes:
- CB Insights, "Why Startups Fail: Top Reasons" — post-mortem analysis; updated figures cite ~70% ran out of cash, ~43% poor product-market fit, ~29% timing, ~19% unit economics.
- Failure-pattern re-analyses (incl. Wildfire Labs' review of $2B+ in startup failures): median raise $11M, median 22 months from last fundraise to death; market-need-adjacent causes estimated at 60–70% when tracing symptoms to root causes.
- Figures are directional; post-mortem data suffers from self-report and attribution bias — founders rarely write "we chose a bad strategy" in their own obituary.