Write your kill criteria before you write any code

Every founder I respect has a private moment — usually month 8, sometimes month 14 — when they realize the thing they're building is not going to work. The good ones close it down within a month of that realization. The bad ones spend another year pretending the data is ambiguous.

What separates them isn't intelligence or willpower. It's a single piece of paper they wrote on day one: their kill criteria.

What kill criteria are

Kill criteria are conditions written in advance that, if met, will cause you to walk away from the idea. Not "consider walking away." Not "discuss with the team." Walk away. Pre-committed. Notarized in your own head before you've spent a dollar or an emotion on the thing.

The reason they have to be written in advance is that the version of you who's three months in is not the version who should be deciding whether to continue. Three-months-in you has sunk costs, an identity tied to the venture, and a hundred reasons that "this metric isn't quite the right one to look at." Day-one you is sober. Day-one you should be making the kill decision on three-months-in you's behalf.

How to write good ones

Bad kill criteria are vague: "if we can't get traction, we'll stop." What's traction? When do you check? Who decides? You'll trivially defeat this when the time comes.

Good kill criteria are unambiguous and binary. Try this format:

BY [DATE], we will have [SPECIFIC METRIC] of at least [NUMBER]. If we do not, we shut down.

Examples that work:

• By June 1, we will have 10 paying customers at $99/mo or more. If we do not, we shut down.
• By the end of August, our 30-day retention will be at least 35%. If it is below 25%, we shut down.
• By 90 days post-launch, our blended CAC will be under $80. If it's over $200, we shut down.

The hardest one to write

The criterion most founders refuse to write is the one about themselves. Some examples:

• If I've stopped looking forward to Mondays after 4 months, we shut down.
• If my co-founder and I have had three escalated arguments about strategy in a quarter, we shut down or restructure.
• If I am not the right person to do customer development for this market by month 6, I quit or hire.

Most founders won't write these because they feel like prophecies. They're not. They're insurance. You're paying a small cost now (writing them down) to avoid a much larger cost later (the year you can't admit you've stopped believing).

How to use them

Pin them somewhere visible. Re-read them on the first of every month. When a kill criterion fires, the only acceptable responses are: shut down, or write a public memo explaining what changed and why the criterion doesn't apply. The memo is the friction. Friction is the point.

The most expensive year in a founder's life is the year they spent ignoring the data because they hadn't said in advance what data would matter. Spend an hour on this now. The hour pays back a thousand fold.