The premortem vs. postmortem

The launch has failed.

Customers hated it.

The supplier vanished.

The budget doubled.

Six months have passed and everyone is explaining why the disaster was obvious.

Now do that before spending the money.

A premortem gives doubt permission to arrive while it can still be useful. People who would sound disloyal saying “this plan is weak” can safely answer “what caused the imagined failure?” The grammar moves criticism out of the present and removes some of its social sting. It also changes the task. People stop arguing whether the plan is good and start inventing a causal history. That produces details: the customer who walked, the dependency that snapped, the early warning everyone explained away. Specific failure stories can be tested; “general execution risk” merely sits in a register looking sombre. One points towards a phone call, fallback supplier or cancelled assumption. The other waits patiently to become a sentence in the postmortem.

Postmortems are easier because the corpse cannot cancel the session. Every warning becomes wise after the event; every decision receives context; responsibility dissolves into learnings. Lovely documentation, shame about the business.

The premortem earns its keep through timing. Ask for the history of failure while assumptions can still be tested, responsibilities changed and exits kept open.

Then record the answers. Otherwise the ritual becomes another hour of imaginative gloom followed by the original plan in a fresher font.

A disaster imagined early is cheap. Reality charges full price.

Keep the imagined failure history beside the project plan, with an owner and review date against each warning. Otherwise nobody will remember which disaster they were clever enough to predict.