Kill criteria: how ventures leave the portfolio

A venture is easy to start on the factory and, for that reason, easy to keep alive past the point of usefulness. The discipline that makes a portfolio honest is not starting ventures — it is ending them, on a schedule, in writing, in public.

The charter carries the kill threshold

Every venture begins with a one-page charter. It names the hypothesis, the buyer, the price, and two things most plans leave out: a kill threshold and a review date. The kill threshold is a number decided before there is any emotional investment — a level of usage, revenue, or evidence that the bet is working. The review date is when we check the number against reality.

Three outcomes, decided on the date

On the review date a venture does one of three things, and only one:

• Scale — the threshold was cleared. It gets more direction, more agent throughput, and a new charter with a higher bar.
• Hold — it is close, or the signal is genuine but slow. It gets one more review window, and only one.
• Kill — the threshold was missed. It moves to the archive, publicly, the same week.

The point of deciding in advance is to remove the argument. When the date arrives, the number already exists; the only work is honesty.

Killing in public

A killed venture does not disappear. It goes to the archive with its lifespan, one honest line on why it ended, and what it left behind — an audience, a corpus, a module, a lesson. The residue is often the most valuable thing the venture produced, and it is reusable by the next one.

A portfolio with no archive is not disciplined; it is either very young or not telling the truth. The archive is where the method proves it is real.