Pre-Mortem Technique: Failing the Plan on Purpose, First

Why imagining a plan has already failed gets people to say things a normal risk review never quite manages to surface.

The plan looks solid. Everyone around the table nods along. Then someone asks the group to imagine it's a year from now and the project didn't just struggle — it failed completely — and to write down why. The room goes quiet for a second. Then the objections nobody wanted to raise five minutes earlier start coming out, one after another.

That exercise is a pre-mortem: assuming failure has already happened and working backward to explain it, instead of asking people to predict what might go wrong. The idea traces back to a psychology finding from the late 1980s about how people reason differently about certain outcomes versus uncertain ones — and the way that finding usually gets retold is a little more impressive than what the original researchers actually found.

What the Pre-Mortem Technique Actually Does

You'll often see the technique credited with a specific number: that treating an outcome as certain rather than merely possible improves people's ability to correctly explain it by 30%. That figure comes from how the technique's popularizer later described a 1989 study on what researchers call prospective hindsight — imagining a future event as if it had already occurred. The study itself found something narrower: people given a certain outcome to explain generated more reasons for it than people given an uncertain one. The researchers' own conclusion was careful about what that meant — more reasons is "seeing" more, they wrote, which isn't automatically the same as seeing better.

That distinction matters for what a pre-mortem is actually good for. It's not a crystal ball that makes your predictions more accurate. What the certainty framing does, by the research above, is make it easier to generate concrete, specific explanations once failure is treated as settled rather than hypothetical. In teams, the framing may also make criticism easier to voice, since participants are explaining an assumed failure rather than directly attacking the current plan — a plausible additional advantage the cited studies didn't set out to test.

What the Research Shows — and Where It's Contested

The more direct test of the actual technique comes from a later study that compared the pre-mortem head-to-head against other common review methods. 178 participants evaluated a hypothetical crisis plan using one of five approaches: a plain critique, listing pros and cons, listing only cons, running a full pre-mortem, or no structured review at all. The pre-mortem group's confidence in the plan dropped roughly twice as much as the pro-con and cons-only groups' did — and after each group worked on fixes for the problems they'd found, confidence rebounded the most in the pre-mortem group.

That's a genuinely useful result, and it's worth being precise about what it does and doesn't show. The study measured self-reported confidence, not calibrated accuracy — the plan was hypothetical and never actually run, so there's no real outcome to check whether the resulting confidence level was appropriately low or still too high. It also used a single setting and hasn't been extensively replicated; one review of the technique's evidence base noted that few empirical studies of pre-mortems exist at all. The finding that a pre-mortem reduces confidence in a plan more than asking for cons alone is real and specific. Whether that reduction produces better-calibrated confidence, or a measurably better plan, remains unsettled.

Worth noting: the "30% more accurate" figure that circulates alongside this technique traces to a description of the 1989 study, not a direct finding of it. The original researchers measured more reasons generated, not more correct ones. That doesn't make the technique useless — it makes the honest pitch narrower than the popular one.

How to Run One Properly

The technique has a specific structure, and skipping steps is the most common way it gets watered down into a regular risk discussion:

1. State the failure as fact, not possibility. Not "what might go wrong" — "it is now six months from now, and this has failed completely." The certainty framing is the entire mechanism; softening it back into a hypothetical undoes the effect the research actually measured.

2. Have everyone write their reasons independently first. Silently, on paper, before any discussion. This is the step most groups skip, and it's the one that prevents the loudest voice in the room from anchoring everyone else's answers before quieter objections get a chance to form.

3. Share one reason per person, in rounds, until every independent list is covered. That’s the tested procedure. A practical adaptation worth considering: start the rounds with the most junior or newest participant rather than the team lead — whoever speaks first shapes what feels safe to say next, so starting with the person who has the least social cost to naming a problem may surface more of them. That ordering choice is our own suggestion, not something the cited research tested directly.

4. Consolidate and rank by how damaging and how likely each reason is. Not every failure reason deserves the same response — this step turns a pile of worries into a short list worth actually acting on.

5. Generate a fix for each of the top reasons before the meeting ends. The confidence rebound in the research happened specifically after this step. A pre-mortem that stops at the problem list captures the confidence reduction but skips the corrective-planning stage — the part of the research where confidence actually rebounded.

Running a pre-mortem is a structured way of forcing new information into a decision before you're committed to it — which is close kin to a skill we test directly. The Stick or Switch Test measures whether you update appropriately once new information arrives, rather than defending a plan out of momentum:

Fail It on Purpose First

A pre-mortem doesn't make anyone smarter or more accurate about the future — the honest version of the research doesn't support that claim, however often it gets repeated. What it reliably does is make people generate more concrete explanations for failure before the plan is locked in rather than after it's failed for real. In a team, it may also make concerns easier to voice. That's a smaller claim than "30% more accurate," and it's also one the evidence actually backs.

The rest of the Decision Making section covers other structured ways to catch a decision before it goes wrong, including how the Eisenhower Matrix handles a different kind of planning failure. Treat running a pre-mortem the same way you'd treat any other brain training habit — worth doing before it matters, on our full library of brain tests.