How to Judge an Ecosystem by What Happens at the End

This chapter turns the slow-failure model into questions about ecosystem design. The focus is not a universal programme for rescuing startups, but the practical conditions that help founders interpret signals, seek independent advice, and leave when a venture no longer justifies continued effort. It also asks what an ecosystem should measure if healthy outcomes include informed growth, constructive disengagement, and re-entry—not only more companies, cohorts, or funding rounds.

Design decline support around the founder’s actual conditions

The first design question is signal transparency. An institution should make clear what its endorsement means and what it does not mean. An accelerator can say that a team was selected for a learning programme. A public agency can say that a grant supports technical development or employment. A university can say that a project has potential for further commercialization work. An investor can explain whether another round is intended to test a revised hypothesis or preserve an existing position. This does not remove uncertainty, but it prevents founders from treating every institutional decision as a verdict on market viability.

Transparency also requires stating what evidence would change the recommendation. A programme that funds another period of experimentation should identify the uncertainty it expects the period to resolve. A stakeholder who recommends a pivot should explain which negative result the pivot addresses. If the institution cannot say what would count against continuation, the recommendation is functioning more as encouragement than as decision support. This is a design implication of the evidence, not a remedy tested by the Amsterdam study.

The second question is independence. Founders should be able to obtain a view from someone whose return, portfolio, cohort, or public standing does not depend on keeping the venture alive. Independence does not mean that investors, accelerators, or universities are useless. Each can offer knowledge and resources that an outsider lacks. It means that a founder should not have to ask one interested party to provide the only judgment about whether the founder should keep sacrificing time and money.

A support system can make independence visible by separating roles. The person who funds a test need not be the person who helps the founder decide whether to close. A programme that teaches experimentation can establish a separate, confidential channel for decline. A public service can work with advisers who understand technology ventures rather than assuming that every distressed founder needs the same generic intervention. These arrangements may create additional cost and coordination. Their value is that they reduce the hidden coupling between advice and institutional survival.

The third question is confidentiality and trust. The study reports founders withholding bad news because they feared rumours, felt ashamed, or believed that an entrepreneur should solve problems alone. A system designed around proactive disclosure will miss the founders who are least able to perform confidence. Decline support therefore needs a lower-friction route: private contact, clear boundaries around information, and an expectation that seeking help is a normal part of responsible decision-making rather than an admission of personal failure.