The Ecosystem That Keeps a Failing Startup Alive
Entrepreneurial ecosystems are the subject of this series: the networks of investors, accelerators, universities, public agencies, advisers, and support organisations that shape what founders can learn, fund, and do. This opening focuses on a specific question within that system: why can a startup remain active after the evidence for its viability has weakened? The answer matters because it helps readers distinguish an evidence-responsive pivot from a continuation that merely postpones a decision, and judge whether an ecosystem is helping founders or quietly keeping them stuck.
A startup can be alive and already non-viable
A startup does not fail only at the moment its bank account reaches zero or its legal entity closes. In the study at the centre of this series, failure includes the earlier condition in which a venture no longer demonstrates viable market or technological potential, or no longer justifies the founder’s continued investment of time and resources. That definition moves attention to the interval before formal closure: the period when a founder is still working, still explaining the plan to stakeholders, and still receiving reasons to continue.
“Failing fast” is often used as if it meant moving quickly at any cost. Here it has a narrower and more useful meaning: the timely alignment between non-viability and termination. A venture fails fast when the decision to stop arrives while its remaining resources can still be released and redirected. A venture fails slowly when that alignment breaks down—when the business is no longer earning a credible claim on continued effort, but remains alive through another grant, another programme, another round of persuasion, or another attempt to make the story work.
The difference is not moral. Persistence is not automatically a virtue, and closure is not automatically a success. A difficult venture may need time to test a serious hypothesis. It may also need to change its core idea in response to useful criticism. A separate study of 316 academic-led teams reported that teams often changed the core of their business idea after negative feedback, and that core changes—not peripheral changes—were positively associated with commercialization. That finding does not prove that every core pivot works. It does establish why “continuing” cannot be judged by duration alone: the important question is whether continued work changes the venture’s evidence and logic, or simply extends its life.
The timing of termination matters beyond the founder. Prolonged commitment can absorb money, labour, attention, and emotional capacity that could otherwise move to more promising initiatives. The entrepreneurial-ecosystem literature has often emphasized startup creation and growth, but a dynamic ecosystem also depends on the movement of resources out of ventures that no longer justify them. A founder’s time, accumulated knowledge, and team experience can become available for another project after a closure. When a venture enters a “living-dead” state instead, those resources remain locked in a business that is neither clearly learning nor clearly ending.
That is why this series does not begin with the question, “Why did the founder refuse to quit?” It begins with a harder systems question: what does the surrounding ecosystem make persistence mean? A founder may continue because the market has not yet delivered decisive evidence. They may continue because a stakeholder’s endorsement looks like validation. They may continue because an investor can make leaving expensive. Or they may continue because the support that claims to be available requires them to reveal a failure they are ashamed to name. These are different mechanisms, even when they produce the same outward result: a company that keeps operating while its options narrow.
Amsterdam as a revealing case, not a universal verdict
The primary study examines Amsterdam’s entrepreneurial ecosystem through the experiences of founders whose ventures failed and the stakeholders who supported, funded, educated, or governed them. The research combines interviews with 30 failed-venture founders and 25 ecosystem stakeholders, observations at entrepreneurship events, secondary material, and a feedback event in which emerging findings were discussed. The stakeholder group included educators, accelerator and incubator staff, investors, government officials, support organisations, advisers, and professional-service providers. That range matters because a founder’s termination decision is rarely shaped by one relationship alone.
The study’s strength is not that it produces a universal percentage of startups that fail slowly. It is that it lets different parts of the ecosystem be seen at the same time. Founders describe what it feels like to be under pressure or to withdraw from a network. Stakeholders describe the programmes and signals they believe they provide. Secondary sources and observations help the researchers compare those accounts with the wider environment. The result is a qualitative explanation of how a decision that appears personal can be socially embedded.
Amsterdam is useful here because the case contains the very supports that are supposed to make startup decisions faster: entrepreneurship education, accelerators, investors, events, and public programmes. The problem is therefore not simply a missing resource. It is a mismatch between what the ecosystem says it is doing and what its relationships make possible when a venture declines. A system can be rich in launch support yet poor at helping a founder recognize that the launch hypothesis has failed. It can celebrate experimentation while making the admission of failure socially costly. It can offer advice while giving some institutions or advisers reasons to prefer survival over an honest exit.
The authors identify four mechanisms that explain this mismatch. Narrative contradiction arises when lean-startup language tells founders to test and stop, while cultural and media narratives say that good entrepreneurs persevere and never give up. Signal distortion occurs when acceptance, funding, or portfolio membership looks like proof of viability even though the institution may be responding to its own incentives. Authoritative pressure appears when investors or other powerful actors make termination costly or practically unavailable. Support pipeline breakdown occurs when founders withdraw in shame or cannot find advice that is independent, confidential, and suited to a distressed startup.
These mechanisms should not be read as an accusation that every investor, accelerator, or public agency is acting in bad faith. The point is subtler and more consequential: institutions can produce effects that contradict their public purpose without intending to create them. An accelerator may fill a cohort because it has capacity and obligations to demonstrate activity. A government programme may preserve a struggling company to protect jobs or avoid the appearance that public money failed. An investor may favour another pivot because closing crystallizes a loss. Each action can be understandable from the institution’s position while still making it harder for the founder to receive an unambiguous signal or choose an orderly exit.
The question this series will keep asking
The central subject, then, is the relationship between founders and the ecosystem around them—not startup failure as an abstract personality test. The core theme is ecosystem-mediated slow failure: the way incentives, narratives, signals, power, and support design can preserve entrepreneurial activity after they stop serving the founder’s need for a clear decision. That theme gives us a way to examine the ecosystem without losing sight of the people and ventures it is supposed to support.
A useful test begins with four questions. What evidence would show that the venture is becoming more viable rather than merely more elaborate? Who benefits when the venture continues, and who bears the cost of another month of work? Can the founder hear negative information from someone who is independent enough to say it and trusted enough to be believed? Finally, is termination a real option, or only a word used in public rhetoric while the contracts, incentives, and social expectations point elsewhere?
Those questions also clarify why a pivot and slow failure can look similar from the outside. Both may involve a new product, a revised market, another funding request, or a fresh experiment. The difference lies in the decision environment. A useful pivot responds to credible critique, changes the hypothesis that the evidence has challenged, and leaves room to stop if the new test fails. A continuation trap can use the language of learning while treating every negative result as a reason to try again, especially when institutional signals remain positive and the founder lacks independent guidance.
The Amsterdam evidence does not give us a universal deadline for closing a company. It gives us a better way to see the decision: as a point where market evidence meets institutional interests, cultural expectations, and unequal power. The academic-entrepreneurship study adds a bounded reminder that external critique can lead to commercially useful change; the ecosystem study shows how the same broad vocabulary of feedback and support can become ambiguous or inaccessible when a venture is in decline.
The next step is to locate this decision in the venture’s lifecycle. Before asking which ecosystem mechanism delayed termination, we need to know what counts as declining viability, how a resource moves from one venture to another, and where a founder’s apparent choice becomes a living-dead state. That is the decision window in which “fail fast” either becomes a disciplined release of resources—or remains only a slogan.