Chapter 04 of 06
431 startup post-mortems. 25 years of project failure data. What actually kills organizations — and why "ran out of cash" is almost never the real answer.
New here? Start with the overview for context on why failure matters.
Based on: CB Insights · Frontiers in Psychology (2024) · Sourcing Innovation (2024) · Viima · Jesse Nieminen
The cause of death is rarely the cause of death
CB Insights analyzed 431 VC-backed companies that shut down in the post-zero-interest-rate shakeout. "Ran out of capital" appeared in 70% of post-mortems. But it's almost always the final cause of death, not the root. It's the how. The why is almost always something else.
Measurable deterioration shows up months before the cash runs out. Partnership activity drops 44% in the 12-24 months before death. Two-thirds of companies with headcount data were already shrinking 6 months before shutdown. The company was dying; the bank account was the last to know.
72% of dying companies showed declining health scores in the year before death — average drop of 15%.
— CB Insights Mosaic score analysis, 431 companies
What actually kills companies
Built something nobody wanted enough to pay for
Symptom. See above for the real reason.
Right idea, wrong moment. Market wasn't ready.
Timing was right, but someone else executed better.
Competency gaps at the founding team level.
Of high-potential failures specifically. People problems first.
Source: CB Insights, Top 20 Reasons Startups Fail (original report + 2024 update, 431 companies)
What the academic research says
A 2024 peer-reviewed study in Frontiers in Psychology examined 50 documented startup failure accounts using a modified Critical Incident Technique and Spencer's behavioral competence framework. The focus was on CEO perspectives — the people who watched their companies die.
The study's key finding: "The failure of a startup is rarely attributable to a single factor." Multiple competency gaps combine. But two deficits show up consistently across nearly every case.
The root failures are behavioral — not strategic, not financial. They're about whether the people running the org had the specific skills to navigate to product-market fit.
— Szathmári et al., Frontiers in Psychology, 2024
The most common competency gap. Teams that couldn't consistently ask the right questions of the right people — customers, experts, competitors — built in the dark. They gathered data that confirmed what they already believed rather than challenging it.
Not customer service as a department — as a mindset. The genuine curiosity to understand what users actually experience vs. what they say. Teams without this built products that looked right on paper but solved problems users had already found workarounds for.
Secondary deficits: technical expertise gaps, analytical thinking limitations, inflexibility in response to market signals.
25 years of not learning
Sourcing Innovation compiled failure statistics from the major consulting and research firms going back two decades. The pattern is striking: the numbers are almost identical year to year. And the firms publishing these reports are often the same ones being paid to fix the problem.
88% of business transformations fail to achieve original ambitions
85% of AI projects fail. 87% of R&D initiatives never reach production.
17% of large IT projects threaten company existence. 70% of digital transformations fall short.
Only 29% of IT implementations succeed. 19% are "utter failures."
Large IT projects deliver 56% less value than predicted.
The firms publishing these reports for 25 years are often the same firms being hired to solve the problem. If they actually fixed it, the reports would stop. The consistency of failure rates over a quarter century suggests this may be structural — not solvable with better tools or methodologies. Understanding this changes how you approach transformation.
What separates the 8% that survive
Companies using structurally separate teams for existing business vs. emerging bets succeed at breakthrough innovation 90% of the time. Organizations without that structure: 25%. The gap is not talent. It's permission and structure.
Organizations with genuinely aligned innovation strategies and pro-innovation cultures show 30% higher enterprise value growth. Culture is not soft — it's the primary variable.
Only 22% of top-performing innovation teams base project approval on future revenue projections. Weaker performers: 75%. The best teams approve based on learning potential, not projected ROI.
Top innovation performers support open collaboration across teams and external partners 77% of the time. Weaker performers: 23%. Insular teams fail faster and learn slower.
Naming what killed something — honestly, specifically — is stewardship. It honors the people who worked on it, the donors who funded it, and the communities it tried to serve. An unmarked grave helps nobody.
The two competency deficits that kill startups — information-seeking and customer orientation — are the same ones that kill missions programs. Did we actually listen to the people we were trying to reach? Did we keep asking questions, or did we stop once we had a model we believed in?
The 25-year failure rate doesn't budge because organizations keep outsourcing the diagnosis to people with an incentive not to find the real answer. Do your own autopsy. Name what actually happened.
Key Takeaways
Sources