Pre-Mortem
Gary Klein’s trick is to hold the funeral before the wedding. Gather the team a week before launch. Tell them the project has failed — catastrophically, six months from now — and ask each person to write the obituary. What killed it? Pass the obituaries around. The reasons surface immediately, because everyone has been quietly noticing them. They just had no socially acceptable way to say so while the plan was still being defended.
This is inversion applied to planning. You can’t reliably enumerate the paths to success — there are too many and most are uncertain. You can usually enumerate the plausible paths to failure, and most projects die from a small handful of them. Naming them in advance won’t guarantee survival, but it converts unknown unknowns into known risks you can actually mitigate.
The real work the technique does is social. People in groups have private worries they will not voice publicly — what Cass Sunstein calls preference falsification. The pre-mortem is a structured permission slip to break the silence. The first person says “the API team won’t deliver on time,” and everyone else nods, because they all knew. The forward question — will this work? — invited the cheerful conformity of people who wanted to look committed. The inverted question makes the pessimist useful.
!! Post-mortems explain failure. Pre-mortems prevent it.
The cheap version is to do it alone, on paper, before any commitment. Spend ten minutes writing the failure post-mortem of a decision you haven’t made yet. The exercise is short and the friction is low and the answers are usually clarifying enough to change what you do.
The discipline is doing it before, not after. Post-mortems explain failure. Pre-mortems prevent it. Both work; only one is cheap.