How to Run a Pre-Mortem Before a Big Decision
A post-mortem examines a failure after it happens. A pre-mortem, a technique developed and popularized by the psychologist Gary Klein, does the same investigation before you commit — while there is still time to act on what you learn. The move is deceptively simple: you assume the decision has already gone badly wrong, and you work backward to explain why. That small shift in framing pulls out risks that a hopeful, forward-looking discussion reliably misses.
Why imagining failure works better
The mechanism behind the pre-mortem is an effect sometimes called prospective hindsight: treating an outcome as though it has already happened tends to make people noticeably more effective at generating specific, concrete reasons for it than simply speculating about what might happen. “Why did this fail?” is a more productive question to sit with than “What could go wrong?”, because the first frames failure as a settled fact you now have to explain in detail, while the second only invites vague, hedged possibilities that are easy to wave away in the moment. The shift is entirely in the framing of the question, not in any new information anyone in the room has — which is part of why it is such a cheap technique for the value it returns.
It also fixes a social problem that pure information could never fix on its own. Once a team has rallied behind a plan, voicing doubts starts to feel disloyal, and quiet skeptics stay quiet rather than risk seeming negative or unsupportive — the classic recipe for groupthink, where a group converges on unwarranted confidence precisely because no one wants to be the one to break the momentum. A pre-mortem makes doubt the explicit assignment rather than an act of dissent. When everyone in the room is instructed to explain a disaster, the person who quietly spotted the fatal flaw weeks ago finally has not just permission but an explicit prompt to say it aloud, and it no longer reads as undermining the plan — it reads as doing the exercise correctly. By forcing the group to confront concretely how the plan could collapse, a pre-mortem also punctures the overconfidence that makes ambitious plans feel safer in the room than they actually are once they meet reality.
Running one, step by step
- Gather the people and present the plan. Bring together everyone close to the decision and lay out the chosen course as it currently stands, in enough detail that everyone is imagining the same plan rather than their own private version of it.
- Set the scene of failure. The facilitator says something like: “Imagine it is a year from now. We went ahead with this decision, and it has been a complete disaster. Take a few minutes and write down every reason you can think of for why it failed.” The specificity of the framing matters — a vague “imagine things went badly” does not produce the same effect as a concrete, dated scene of total failure.
- Write independently and in silence. Each person lists their reasons on their own first, before anyone speaks aloud. Independence is essential here — it keeps the loudest or most senior voice in the room from anchoring everyone else’s answers, and it preserves the genuine diversity of concerns that different people, seeing the plan from different angles, will have noticed.
- Collect every reason. Go around the group and gather all the failure causes onto one shared list, without debating, ranking, or dismissing any of them yet. The aim at this stage is coverage, not judgement; a reason that seems unlikely to one person may be the exact blind spot of everyone else in the room.
- Strengthen the plan. Now review the full list together. Sort the risks by how likely and how damaging each would realistically be, and for the serious ones, decide on a concrete mitigation, an early warning signal worth watching for, or — if a risk is grave and hard to mitigate — a real change to the decision itself rather than a cosmetic fix bolted onto the original plan.
A worked example
Imagine a small team about to launch a new subscription product in six weeks. Everyone is excited, the demo works well in every internal run-through, and the planning meeting is drifting toward a rubber-stamp approval. Instead, the lead calls a pre-mortem: “It is launch day plus one year, and this product flopped. Why?” Working silently first, the team writes, then pools a shared list: onboarding was too confusing and users churned within the first week; the pricing sat awkwardly between two competitors and ended up appealing to neither their price-sensitive nor their premium customers; a key third-party integration was not actually ready at launch, so the earliest and most vocal reviews were harsh; support was swamped by the initial wave and response times cratered right when first impressions were forming; and the launch date collided with a rival’s larger release and got buried in the noise.
None of these five risks surfaced during the earlier, optimistic planning session, even though every person in the room had the information needed to raise at least one of them. Acting on the list, the team simplifies onboarding down to three screens, runs a quick pricing test against the two nearest competitors before locking a number, pulls the fragile integration out of the initial launch scope rather than shipping it half-ready, staffs up support for the first month specifically, and checks on the competitor’s rumored launch timeline before finalizing a date. The decision to launch stands — the pre-mortem is not a tool for talking a team out of things — but it is now a materially sturdier version of that same decision, hardened against its five most likely and most damaging ways of failing.
Common ways a pre-mortem goes wrong
The technique is simple enough to run badly without anyone noticing. The most common failure is skipping the silent, independent writing step and going straight to open discussion; without it, the exercise quietly reverts to whichever confident voice speaks first setting the agenda, which is exactly the groupthink dynamic the technique exists to prevent. A second failure is treating the exercise as a box to check rather than an input to act on — producing a thorough list of failure modes and then filing it away unread is arguably worse than not running the exercise at all, since it creates the appearance of diligence without the substance. A third is reaching for a pre-mortem on every decision regardless of size; running a thirty-minute group exercise on a choice that is cheap to reverse if it goes wrong is a poor trade of everyone’s time, and the effort is better reserved for the decisions described below.
Running a pre-mortem alone
The technique does not require a group, though a group with genuinely different perspectives tends to surface a wider range of failure modes than one person working alone can manage. For a personal decision — taking on a large financial commitment, ending a long-standing arrangement, moving somewhere new — the same structure still works. Set aside the same block of time, write the same failure scene at the top of a page, and force yourself to generate at minimum eight to ten distinct reasons before you let yourself stop, since the first three or four that come to mind are usually the ones you have already half-consciously considered; the value is disproportionately in the reasons that only surface once the obvious ones are already on the page. Then apply the same final step: for each serious risk, decide on a mitigation, a warning sign to watch for, or a change to the plan.
Turning the list into priorities
A pre-mortem can easily produce fifteen or twenty candidate failure causes, and treating all of them as equally urgent defeats the purpose of having prioritized at all. Once the list is collected, sort each item along two rough dimensions: how likely it seems, and how damaging it would be if it happened. The items that are both plausible and severe deserve a real mitigation before you proceed. The items that are severe but unlikely are usually worth a cheap early-warning signal rather than a full fix — something you can watch for that would tell you the unlikely scenario is starting to unfold. The items that are neither likely nor especially damaging can often just be noted and set aside; a pre-mortem that tries to fix everything on the list ends up diluting effort away from the two or three risks that actually mattered. If the list of serious risks is long enough that ranking them by eye feels unreliable, a quick pass through a weighted scorer — weighting each risk by its likely damage rather than treating every listed concern as equally worth acting on — can make the priority order more defensible than gut feel alone.
When to use it
A pre-mortem earns its keep specifically on decisions that are consequential and hard to reverse — the one-way doors, in the language of the companion piece on reversible versus irreversible decisions, where a hidden flaw discovered after the fact is genuinely expensive rather than a minor inconvenience to fix. It typically takes twenty to thirty minutes and needs no special tools or software, only the discipline to imagine failure honestly and the independence step to keep any one voice from steering the room. For a small, easily reversible decision, the same thirty minutes is usually better spent just trying the thing and observing what actually happens. Make the pre-mortem a standing step before you commit to anything genuinely big and hard to walk back, and you will consistently catch problems while they are still cheap to fix, rather than after they have already cost you the year the exercise asked you to imagine.