Pre-mortem

What it is

A structured way to find a plan’s weak points while you can still fix them. Instead of asking “what could go wrong?”, you declare that it did go wrong and ask everyone to explain the failure.

When I use it

Any decision that is expensive to reverse: a large purchase, a hire, a launch, a change to how a team works. For small, reversible decisions it is overkill. Just decide and adjust.

How

1. Set a date in the future. Accept that the project has failed by then.
2. Each person writes down reasons on their own, before any discussion.
3. Combine the lists. Rank each reason by likelihood and by damage.
4. For the top three, set a tripwire: an early signal you can observe, and what you will do when you see it.
5. Keep the list. Compare it with what actually happens.

A worked example

Before building this website I ran a pre-mortem on it. The top failure was “it fills up with AI-generated text and stops sounding like me.” The tripwire: nothing is published until I’ve rewritten it myself. The full example is in the essay Write the failure first.

How it fails

It fails when the most senior person speaks first and everyone else edits their list to match. It also fails when the list is written and never read again. A pre-mortem is only as good as the tripwires that come out of it.