Why a post-mortem comes too late
The review after the failure is honest but useless for this project: the reasons are found when there is nothing left to fix. At the same time, planning mistakes are predictable — optimism, overestimating one's strength, underestimating dependencies. The pre-mortem, described by Gary Klein in the book Sources of Power, exploits this predictability: if the team knows how similar projects fail, let it say so before the start.
Step 1: assume that the project has failed
The first two minutes — the formulation: "In a year we learn that the project has failed". It is important to speak about the failure as an accomplished fact, not as a probability: psychologically this removes the defense and turns on the search for reasons, not for excuses.
Step 2: generate the reasons in silence
The next ten minutes — each participant writes the reasons of failure on sticky notes without discussing them out loud: this protects from the leader effect, when everyone repeats the first voiced version. Then the reasons are grouped by topics, the duplicates are removed, and the team votes for the most probable ones. Usually 5-10 reasons are collected, and the main one is often not the one that was named before the session.
Step 3: score the risk = probability × damage
The remaining three minutes — the scoring. Each reason gets points: probability from 1 to 5 and damage from 1 to 5. The product gives the risk score; the top-3 by score go into the mitigation plan. The selection threshold is the line below which a risk is accepted consciously, not "forgotten".
Step 4: the mitigation plan and the "DO NOT do" checklist
For each of the top-3 risks one action is written down: the owner, the deadline, the trigger that starts the action, and what changes in the project. Separately, the "DO NOT do" checklist is composed — the actions the team consciously refuses: not to expand the scope at the start, not to change the contractor in the middle, not to launch without acceptance. This is the protection from the known reasons of failure that the team has already seen in other projects.
What you get in the end
The result of the fifteen-minute session is four artifacts: the list of the reasons of failure, the risk matrix with the scores, the mitigation plan for the top-3 and the "DO NOT do" checklist. The pre-mortem is the natural continuation of the cause analysis: if the Fishbone looks for the reasons of a failure that has already happened, the pre-mortem looks for the reasons of a failure that has not happened yet. About turning the found risks into verifiable solutions — in the article "From contradiction to test".
Questions and answers
How is a pre-mortem different from a usual risk discussion?
A risk discussion rests on the known concerns, while a pre-mortem starts from the fact of the failure and makes the team look for its reasons before the start.
Who invented the pre-mortem method?
The psychologist Gary Klein described it in the book Sources of Power (1998): the team 'remembers' a failure that has not happened yet.
How long does the session take?
15 minutes is enough: formulating the fact of the failure, the silent generation of reasons, grouping and voting by probability.
What does "risk = probability × damage" mean?
Each risk is scored by probability and by damage on a 1-5 scale; the product gives the score, the top-3 go into the mitigation plan.
What is the "DO NOT do" checklist?
The list of actions the team consciously refuses to do in order not to repeat the known reasons of failure.
Does a pre-mortem ensure the success of the project?
No: the method reduces the blind spots of planning but does not cancel uncertainty — some reasons will become visible only in the process.
Where is the pre-mortem used on the site?
At the research stage before the launch: the analysis of reasons (Fishbone) and of contradictions is complemented with checking the plan for failure.
What does the team get in the end?
The list of the reasons of failure, the risk matrix, the mitigation plan for the top-3 and the 'DO NOT do' checklist for the start.