Working Start something skill
Run a pre mortem
Run a pre-mortem that imagines failure, surfaces independent causes, prioritizes controllable risks, and converts insight into owned prevention and contingency actions.
Create permission to name failure without treating concern as disloyalty.
When to use
- Use before launch, migration, contract, event, policy, investment, or irreversible change.
- Do not use the exercise to identify or blame hypothetical individuals.
Procedure
- Define decision, plan, horizon, success criteria, participants, facilitator, and psychological-safety rules.
- Present the same factual plan and assumptions to everyone.
- Ask participants independently to imagine the plan has failed and write plausible causes and early signals.
- Collect causes before group discussion to reduce anchoring and hierarchy effects.
- Cluster by assumptions, people, process, technology, dependency, customer, finance, security, safety, legal, and external conditions.
- Prioritize by plausibility, impact, detectability, control, and decision timing.
- Convert top causes into prevention, detection, contingency, owner, due date, and evidence.
- Feed actions into the plan and risk register; review when assumptions or conditions change.
--- name: run-a-pre-mortem category: start description: Run a pre-mortem that imagines failure, surfaces independent causes, prioritizes controllable risks, and converts insight into owned prevention and contingency actions. Use when a plan is important enough to challenge before execution. --- # run-a-pre-mortem Create permission to name failure without treating concern as disloyalty. ## When to use - Use before launch, migration, contract, event, policy, investment, or irreversible change. - Do not use the exercise to identify or blame hypothetical individuals. ## Procedure 1. Define decision, plan, horizon, success criteria, participants, facilitator, and psychological-safety rules. 2. Present the same factual plan and assumptions to everyone. 3. Ask participants independently to imagine the plan has failed and write plausible causes and early signals. 4. Collect causes before group discussion to reduce anchoring and hierarchy effects. 5. Cluster by assumptions, people, process, technology, dependency, customer, finance, security, safety, legal, and external conditions. 6. Prioritize by plausibility, impact, detectability, control, and decision timing. 7. Convert top causes into prevention, detection, contingency, owner, due date, and evidence. 8. Feed actions into the plan and risk register; review when assumptions or conditions change. ## Done - A pre-mortem report records scenario, independent causes, clusters, priorities, signals, controls, contingencies, owners, and decisions - Participant-coverage, independence, hierarchy, risk-link, owner, due-date, and plan-update checks verify useful action