Of the 431 venture-backed companies that shut down between 2023 and early this year, 70% gave the same immediate reason: they ran out of capital (CB Insights, March 2026). It is the tidiest explanation available, and close to the least useful. Running out of money is where the story ends. It is rarely where it went wrong.
One layer down, the picture changes. Poor product-market fit appears in 43% of the post-mortems, bad timing in 29%, unsustainable unit economics in 19% (CB Insights, March 2026). Some of that is the market simply not being there, and no management saves a company selling something nobody wants. A good part of it is not a market problem at all. It is an operating problem: a real product that never became a company running at a margin, with a team that held and a founder who could see round the next corner.
The failure is usually slow, not sudden
The same analysis tells the deeper story. In the year before they closed, 72% of these companies showed a measurable decline in their health scores (CB Insights, March 2026). Two-thirds were shedding staff in the final six months. The median company ran for 22 months between its last raise and its last day.
These are not businesses that hit a wall. They are businesses that leaked for the better part of two years while everyone hoped the next quarter would settle it.
The decline is usually quiet. There is no single dramatic failure. A planning rhythm slips. A capable hire is never properly brought into the work. A number sits on a dashboard that nobody actually owns. A decision waits because only one person can make it and that person is in another meeting. By the time any of it reaches the cash position, the damage is a year old and the runway has gone with it.
A founder is a specialist, promoted overnight
Here is the part early-stage investing structurally underprices. The quality that makes a founder worth backing and the quality that makes a company survivable are two different competencies. They sit in the same person far less often than the cheque assumes.
A founder is, almost by definition, a specialist. They are exceptional at the thing the company sells: the technology, the product, the commercial insight that opened the gap in the first place. The investment thesis rests on exactly that. Then the act of founding promotes them, in a single step, into the hardest management job in the business. They run the whole thing, with no assessment beforehand of whether they can, and no training once they are in the chair.
Being a brilliant specialist tells you very little about whether someone can run a company. Management capability does not arrive with the title on the door. There is a line in management thinking that puts it precisely: the specialist ended, and the manager never began. The business is left with a founder who is genuinely excellent and an empty seat where the operator was supposed to be. That is not a flaw in the person. It is the predictable result of asking one set of skills to do the job of another, with nothing built in between.
The constraint is rarely the product
In a just-launched business, the binding constraint is almost never the thing everyone is watching. The product is usually good enough to start. The market is usually real enough to enter. What decides the first eighteen months is the operating capability of the founder and whatever they manage to build around themselves in that window.
That is an uncomfortable place to put the risk, because it is the one variable the pitch is not designed to reveal. A founder can be magnetic on stage and have no idea how to run a Monday. The two have almost nothing to do with each other, and only one of them keeps the company alive.
There has never been more help, or less change
The early-stage world has answered the problem with more support than at any point in its history. Accelerator cohorts, playbooks, office hours, board decks, warm introductions to people who have done it before. Founders are not short of information about how a company should be run.
Information is not the part that is missing. Knowing how a company should run and being able to run one are different things, in the way that knowing how an engine works does not make you a mechanic. The first is structured information. The second is a skill, installed by doing the work under guidance until it holds under pressure. A cohort hands over the first and quietly assumes the second will follow. Often it does not. The founder leaves with a sharper vocabulary and the same working week they walked in with.
This is the gap that decides outcomes, and it is the one almost nothing in the standard support stack is built to close.
What the question should be at the term sheet
For an investor, this reframes the decision before the cheque clears. You have already settled that the founder is brilliant and the market is real. The risk that remains, the one diligence almost never reaches, is whether this person can build the daily disciplines that turn a strong idea into a company that runs without constant heroics.
That risk does not appear in a pitch. It is not in the data room. It surfaces eighteen months later, in the quiet leak. So the question worth asking is not “is this founder impressive.” It is “what is being done, on purpose, to install the operating capability this business will need by month twelve, and who is doing it.”
The disciplines themselves are unglamorous and specific. A documented picture of who does what, and why. Delegation that matches authority to responsibility, so the founder stops being the only point at which a decision can be made. A management rhythm that surfaces problems while they are still small. A short set of numbers that someone genuinely owns rather than merely reports. The strategy broken down far enough that it lands in individual weekly calendars instead of living in a deck. None of it is exotic. All of it has to be built into the working week by the founder, in their own hands, or it does not survive the first real shock.
I built the operating side of a fibre business from nothing during its growth phase: the process maps, the standards suite, the dashboards, the academy that trained the field engineers. It launched on schedule and scaled from a single town into many inside the first year, and the people who had backed it noted the governance. What decided it, every time, was getting the disciplines built into how the week actually ran before the growth outran them.
Capability can be installed, on purpose
The capital going into early-stage businesses is betting, ultimately, on what the founder knows and who they are. The survival of the business depends on something narrower and far more buildable: what the founder can actually do once the company has staff, customers, and a cash cycle that does not wait.
Treating that as an unknowable quality of the individual is the expensive mistake. It is not unknowable, and it is not fixed. It can be installed, deliberately, through the founder’s own hands, in the months before the leak starts rather than the months after it turns terminal. The investors who do best over the next few years will be the ones who stop hoping their founders grow into operators, and start making sure they are equipped to.
References
- CB Insights. “The top 9 reasons startups fail.” March 2026. Read Article
