The most useful document in an early venture is one almost nobody writes: a page, written before starting, listing the conditions under which you would stop.
Without explicit kill criteria written in advance, ventures persist on inertia long after they should have ended. The cost of that persistence is rarely the money. It is the years.
Why the decision cannot be made later
By the time the evidence is in, you are not the same person who could evaluate it. You have told people what you are building. You have invested months. You have an identity attached to it. Every one of those makes an honest reading of the evidence harder, and all of them arrive after the decision would ideally be made.
The person best placed to judge whether the evidence is sufficient is the one who has not yet invested anything. That person only exists before you start, which is why the criteria have to be written then.
What good criteria look like
They are specific, dated, and falsifiable. 'Not enough traction' is not a criterion, because it can be argued with indefinitely. 'Fewer than two buyers have paid or committed in writing by day forty-five' is a criterion, because on day forty-five it is either true or it is not.
Each criterion should name a number, a date, and the evidence that settles it. If you find yourself unable to specify the evidence, that usually means you have not decided what success looks like either.
“The person best placed to judge the evidence is the one who has not yet invested anything. That person only exists before you start.”
A workable default set
By day forty-five, at least two buyers have paid for a manual delivery or committed in writing to pay for the built version.
By day ninety, at least one customer is paying and using the product without you present.
The person best placed to judge the evidence is the one who has not yet invested anything.
By month six, at least five customers, and the most recent three did not come from your personal network.
Turn what you know into what you own.
Vibepreneur builds structured ventures from professional expertise, with positioning, launch assets, and growth systems included.
Join the WaitlistThat last one catches the most common quiet failure: a venture that works entirely because the founder knows people, and has no evidence it works otherwise.
The clause that makes it survive
Criteria written by a founder can be revised by the same founder, which usually means they get revised at exactly the moment they would have triggered.
The fix is cheap: give the page to one person who will hold you to it and who has no stake in the outcome. Tell them the dates. Ask them to ask. That single social commitment does more than any amount of personal resolve.
Stopping is not failing
The framing that helps is that the criteria are not about whether you can succeed. They are about whether this particular problem is the right one to spend the next three years on.
Stopping at day forty-five with a clear negative answer and eleven weeks of expensive market knowledge is a good outcome. You now know something specific about an industry that almost nobody else knows, and the next attempt starts from a much better position. Several of the ventures in our showcase were their founder's second or third attempt.
The bad outcome is not stopping. It is the two years spent not deciding.
See when to kill the idea for the emotional side, and the ninety day path for the plan these criteria attach to.