Read Achilles markRead Achilles

Product Management

Kill Criteria, Written in Advance

The hardest meeting in a young company is not the one where a product is proposed. It is the one, eighteen months later, where everyone in the room privately knows the product should end and nobody can find the…

♥ 0 Sign in to like 0 comments

This post’s retention & record terms are stated in full in the colophon below.

The hardest meeting in a young company is not the one where a product is proposed. It is the one, eighteen months later, where everyone in the room privately knows the product should end and nobody can find the sentence that ends it — because that sentence had to be written at the beginning, by people who were in love, for people who would be in grief. I have sat in that meeting as founder and as undertaker. Kill criteria are how you spare the second meeting its worst hour: conditions, agreed at launch, under which the product stops. [author assertion]

Founders resist them for an honourable reason: commitment is real, and criteria look like doubt wearing a spreadsheet. But watch what the absence of criteria actually produces. Without an agreed test, the decision to end a product is re-litigated from first principles every quarter, and first principles lose to arithmetic every time — the incumbent roadmap, the key account, the sunk cost with a roadmap of its own. The product does not survive on its merits. It survives because ending it requires a verdict nobody was ever empowered to deliver. That is not commitment. It is an auction where the status quo always holds the winning bid and never has to raise it. [author assertion]

Written in advance, the criteria change the meeting's physics. The question stops being *should we kill it?* — a question about the team, the founder, the story — and becomes *did the thing we agreed to watch do the thing we agreed would matter?* A negotiation with your past self is still a negotiation, but it is one where the terms were set before anyone had anything to defend. I have come to think of kill criteria as the roadmap's assumptions, published, with dates on them — the same discipline, aimed at the ending instead of the beginning. [author assertion]

Two practical rules, bought at the usual cost. First, write the criteria in the launch document, in the same font as the ambitions: the adoption floor, the date by which it must be met, the specific evidence that would count as the market's answer rather than the team's effort. Effort is always maxed and always moving; it is the one metric that can never acquit or convict. Second, appoint the undertaker at launch — a named person, not the product's champion, whose job on the review date is to read the criteria aloud. Champions should argue. Somebody else should hold the document. [author assertion]

None of this shortens any product's life by a day. What it shortens is the twilight: the quarters in which a dying product eats the attention a beginning needed, defended by reasonable people making reasonable arguments in meetings with excellent minutes. The first product got to be a beginning when the company had nothing to protect. Kill criteria are how the company protects the *next* beginning — from the first success's entirely reasonable demand to never be succeeded. [author assertion]

So, founders and product leads: for the product you launched most recently — what is written down, today, that would end it, and who holds the document? If the honest answer is "nothing, and no one", you have not made a commitment. You have scheduled a harder meeting. [author assertion]

Ben Cowan

All posts by Ben Cowan → · Follow Ben Cowan by RSS

The Board 20 essays 0 corrections Agent citations: —— — citation attached ≠ claim validated As of 16 November 2026 Full board →

Provenance

About this version

How this post was verified, labelled and shared — in plain language first. The complete machine-readable record is kept below, word for word, for anyone who wants to check it.

Version
1 · in good standing
Author
Ben Cowan · verified at publication
Published
13 Nov 2026
Claims
6 labelled — 6 author assertion
Citations
No external citations
Distribution
Shared with the author’s consent — including to agents acting for readers, under the author’s per-post opt-in
Content fingerprint
ba566364…d6e67e
Full verification record

The verbatim graded record of this version: identifiers, claim classes, citation targets, hashes and standing, exactly as issued.

This record is also published in structured text for agents and tools — the same fields, word for word. View structured text

House editorial — Read AchillesAuthor verification: verifiedStanding: governingHash: ba5663644e9edf0456f025519222ec31740f025b5d52b643779a87aeefd6e67e

Version & standing

Post post-kill-criteria-written-in-advance-d08ee6 · version 1 (post-kill-criteria-written-in-advance-d08ee6-v1)

@ben-cowan · verified state at publish: verified

Published: 2026-11-13T14:29:41.292Z · Standing: governing

Claims & citations

  • c-1: author_assertion
  • c-2: author_assertion
  • c-3: author_assertion
  • c-4: author_assertion
  • c-5: author_assertion
  • c-6: author_assertion

    Disclosure & trust

    When your words reach the end of their retention period they are destroyed, on the schedule above. What the platform keeps after that — indefinitely, with no expiry — is a skeleton of acts: that an act occurred, its class and time, the bundle and section it touched, its change class, a hash that identifies the record without containing it, and the keys that link a correction to what it corrects. The skeleton is not your words and cannot reproduce them. Your elections to shorten a period, to ask ephemerally, or to skeletonise now govern your words; they do not shorten, and cannot remove, the skeleton — it is the record that lets the platform prove what it did with your work after the words are gone. The platform will never produce your skeleton by choice against you; like any record it can be reached by a court order the platform cannot refuse, and if that happens you will be told wherever the law allows and the production will state, on its face, that it is a skeleton and that the words expired on this schedule.

    When your words reach the end of their retention period they are destroyed, on the schedule above. What the platform keeps after that — indefinitely, with no expiry — is a skeleton of acts: that an act occurred, its class and time, the bundle and section it touched, its change class, a hash that identifies the record without containing it, and the keys that link a correction to what it corrects. The skeleton is not your words and cannot reproduce them. Your elections to shorten a period, to ask ephemerally, or to skeletonise now govern your words; they do not shorten, and cannot remove, the skeleton — it is the record that lets the platform prove what it did with your work after the words are gone. The platform will never produce your skeleton by choice against you; like any record it can be reached by a court order the platform cannot refuse, and if that happens you will be told wherever the law allows and the production will state, on its face, that it is a skeleton and that the words expired on this schedule.

    Integrity & distribution

    Content hash: ba5663644e9edf0456f025519222ec31740f025b5d52b643779a87aeefd6e67e

    Distribution: opted_in

    Canonical presentation: https://blog.readachilles.com/post-kill-criteria-written-in-advance-d08ee6/

    What we retain, and why — the retention disclosure

    The retention disclosure for this version is the disclosure line of record, stated in full in the colophon fields above.