Most decisions fail at the frame, not the choice. By the time you are comparing Option A to Option B, the real decision — whether this is the right class of problem, whether your stated criteria match what you actually optimize for, whether you are solving a symptom — is already closed. You are optimizing inside a frame you never examined.
The pre-mortem is the simplest inversion available: before you commit, assume the decision failed spectacularly. Work backward. Explain why. The exercise is decades old, well-documented, and almost universally skipped — because skipping it feels like progress, and running it feels like betrayal of the outcome you already want.
Why frames close before choices open
Human decisions compress quickly. A founder hears “we need a CTO” and within a week is comparing candidates — never asking whether the company needs a co-founder, a senior engineer, an agency, or six more months of founder-led technical work. A family hears “parent is struggling at home” and within a month is touring facilities — never asking whether the struggle is medical, social, grief, or role reversal.
The compression is not stupidity. It is relief. A closed frame produces action. An open frame produces discomfort. Agreement — from colleagues, from AI, from your own narrative — closes the frame faster.
The pre-mortem inversion
Gary Klein popularized the pre-mortem for organizational decisions. The structure is simple:
- Assume it is 18 months later. The decision failed.
- Everyone writes down why — independently, before discussion.
- Compare stories. The overlap is your risk map.
What makes this powerful is timing. Before commitment, failure stories are allowed. After commitment, the same stories feel like disloyalty. Organizations punish the person who keeps asking “what if we are wrong?” once the CEO has said “we are doing this.”
The pre-mortem gives cover. It is not pessimism — it is a scheduled exercise. That scheduling is the entire trick.
Klein grounded the method in prospective hindsight: imagining that an outcome has already happened, rather than that it might. Mitchell, Russo, and Pennington found that this certainty frame increased people's ability to correctly identify reasons for future outcomes by about 30% — which is why “it failed” generates different risks than “what if it fails.”
Prospective hindsight vs. ordinary critique
Source: Mitchell, Russo & Pennington (1989), cited in Klein — Performing a Project Premortem (HBR, 2007) — Relative reason-identification indexed to ordinary future-tense critique (=100).The grammatical shift is the tool. Past-tense failure makes dissent safe enough to say out loud before commitment punishes it.
Veinott, Klein, and Wiggins later found that a premortem reduced plan overconfidence more than a simple critique or a pros/cons list. Reduction of overconfidence is not the same as fixing a wrong frame — but it is the cheapest way to reopen the room.
What a pre-mortem still misses
A pre-mortem assumes the decision you named is the decision you are making. It stress-tests execution inside the frame. It rarely asks:
- Whether the real decision is different from the one on the agenda
- What incentive makes failure stories feel disloyal to tell out loud
- Which question would change the decision if answered honestly
- Whether you have seen this pattern before in a different costume
That is where a full decision audit extends the examination — not replacing the pre-mortem, but running before it, to verify you are pre-morteming the right thing.
From the field
In a generalized eldercare audit, the stated decision was “which assisted-living facility.” The examination surfaced that the real decision was whether the parent wanted to move at all — a question the family had closed because asking it felt like abandoning hope of keeping them home. The facility comparison was elaborate, sincere, and orthogonal to the decision that mattered.
That is a frame failure. No amount of facility research fixes it.
The debt beneath an unexamined frame
Some organizations do not fail at a single bad choice. They fail by carrying decision debt — unresolved decisions that keep charging interest as delay, rework, founder load, and inconsistent work. PathwaysHQ’s Decision Debt Audit names three useful types (Decision Debt Audit):
Matthews’s process discipline fits here: judge the decision by the quality of the reasoning at the time, not only by the later outcome — and treat postponement as a decision you still made, with a bill that keeps compounding.
A pre-mortem stress-tests a named decision. Decision-debt language finds the decision you never named. Most-examined wins require both.
Same stalled organization. Two examinations.
Assume it failed — reasons become speakable before commitment
Avoidance, framing, or ownership — which interest keeps charging
Often different from the agenda item everyone was optimizing
Run it — or have it run on you
If you are mid-decision and the frame feels settled, run a pre-mortem tonight. Assume failure. Write three reasons. Do not share them until you have three.
If the frame itself feels suspiciously settled — if momentum is doing the deciding — audit the decision before you compare options. The most-examined decision wins not because examination guarantees success, but because unexamined frames guarantee preventable failure.
Read confidence vs. evidence for the measurement side of the same problem.