VOL. I
NO. —
DOSSIER REGISTRY
DISP-289FILED: AUG 18

Pre-Mortem at the Planning Counter

Gary Klein's pre-mortem gives teams a simple way to surface hidden risks before optimism, hierarchy, and launch pressure turn them into expensive surprises.

Tools Worth Filing4 min read

KEY TAKEAWAYS FOR COGNITIVE LOGGING

  • A pre-mortem makes dissent easier by placing the failure in a hypothetical future.
  • The method is most useful when it produces owners, mitigations, and decision thresholds.

The workbench tool today is not complicated, which is why it works. A pre-mortem asks a team to imagine that a project has already failed. The launch happened, the money was spent, the deadline passed, and the result was poor. Each person then writes down the most plausible causes of that failure before the group clusters the answers.

Gary Klein popularized the method because ordinary planning meetings are biased toward optimism. People want to look constructive. Junior staff may not want to challenge senior sponsors. Specialists often assume someone else has handled the risks outside their own lane. By declaring a hypothetical failure up front, the pre-mortem gives everyone permission to name the thing they are worried about.

The digest cites Klein and Daniel Kahneman for the claim that pre-mortems surface materially more risk factors than standard reviews. The exact percentage should be checked against the original context before becoming a hard metric, but the mechanism is intuitive. The exercise changes the social contract. Instead of asking “who wants to be negative?” it asks “what would have had to be true for this to go wrong?”

The simplest version takes less than an hour. State the project goal. Ask everyone to silently write the obituary for the project. Have each person read their reasons aloud. Cluster repeated themes. Pick the risks that are plausible, severe, and not already owned. Then assign mitigations or decision thresholds.

The last step is where weak teams often stop early. A list of risks is not a plan. A useful pre-mortem produces changes: a smaller launch, a kill criterion, a monitoring dashboard, a customer interview, a security review, a rollback path, or a named owner for an unresolved dependency.

Pre-mortems are especially valuable for AI projects because demos can hide operating risk. A model may look capable in a narrow test and still fail when permissions, edge cases, latency, cost, privacy, support handoffs, or adversarial input appear. Asking “the agent rollout failed spectacularly; why?” will often surface the real blockers faster than another demo review.

They also help with personal planning. Before accepting a job, shipping a feature, raising a round, or starting a health routine, imagine the failure. The answer may reveal a missing constraint: not enough time, no feedback loop, weak incentives, unclear authority, or a hidden dependency on someone else’s behavior.

The pre-mortem is not pessimism. It is inversion with a meeting invite. Good planning is not pretending the road is clear. It is noticing where the wheels are likely to leave the track while there is still time to steer.

FILED EVIDENCE (VERIFIABLE SOURCES)

FILE CODEDOCUMENT DESCRIPTION
REF-101Performing a Project Premortem - Gary Klein, Harvard Business Review
REF-102Inversion Mental Model: Solving Problems by Thinking Backwards - Neurolaunch