A reading of the public post-mortems founders write when their companies fail finds that the causes cluster around a small number of mistakes, and that most of them are not technical.

The recurring causes

Three appear more than any others. The first is a market that was smaller than the plan assumed, usually because the founder confused a problem people had with a problem people paid to solve. The second is cash: running out of runway because a round closed later than expected, not because the business was broken. The third is a founding team that could not agree on the hard decisions, a failure that arrives late and is rarely diagnosed as the cause.

What is absent

Technology is rarely the cause. The companies that write post-mortems almost always had a working product. The failure was that the product was for a market that did not exist at the price it required, or that the company could not reach the market before the money ran out.

The uncomfortable implication

The honest reading of these documents is that most startup failure is preventable and boring. The decisions that kill companies are made early, in the choice of market and the management of cash, and the technical execution that occupies most of a founder's day is rarely the thing that decides the outcome.