Pre-Mortems and Checklists

Most bad outcomes are not caused by problems nobody could have foreseen. They are caused by problems someone could have named, but nobody said out loud in time — because the team was optimistic, or busy, or reluctant to be the one raising doubts. The two tools in this lesson are cheap, unglamorous, and remarkably effective at pulling those problems into the open before they do damage.

Neither requires expertise or new technology. A pre-mortem takes fifteen minutes. A checklist fits on one card. What they require is the discipline to use them precisely when confidence is highest and the temptation to skip them is strongest.

The pre-mortem

A post-mortem asks why something failed after the fact. A pre-mortem asks the same question before you start. You gather the people involved, assume the decision has already been carried out and failed badly, and ask everyone to write down why. The framing matters: instead of "what might go wrong," which invites polite reassurance, you assert that it did go wrong and ask for the story.

This works because of what researchers call prospective hindsight — imagining an outcome as already certain makes people generate far more concrete, specific reasons than open-ended speculation does. It also lowers the social cost of dissent. Voicing doubt about the boss's plan feels disloyal; explaining a failure the group has been told to assume feels like helpfulness.

The checklist

A checklist is a written list of the steps that must not be skipped on a complex, repeated, high-stakes task. Its purpose is not to teach the task — experts already know it — but to catch the small, dull, easily forgotten steps that get dropped under pressure, fatigue, or interruption.

Good checklists are short, usually five to nine items, and cover only the steps that are both essential and commonly missed. They come in two forms: read-do, where you perform each step as you read it, and do-confirm, where you do the work from memory then pause to verify nothing was omitted.

A product team is about to launch a feature they are sure about. In the pre-mortem, they assume it flopped and write reasons: the onboarding confused new users, a payment edge case failed silently, the launch email went to the wrong segment. Three of these were fixable in a day. Next, they turn the recurring launch mistakes into a nine-item do-confirm checklist — verify analytics fire, test the payment edge case, confirm the email segment — run before every release. The next launch skips none of them.

Checklists are not free, and applied to the wrong task they cause harm. A one-off creative decision, a novel strategic bet, or a genuinely ambiguous problem has no stable list of steps to confirm; a checklist there is theatre that manufactures false confidence. Overlong checklists are worse than none — people learn to tick boxes without reading them. The tools fit routine, repeatable, high-stakes work with known failure modes, not open-ended judgment.

Gary Klein described the pre-mortem technique in the Harvard Business Review in 2007 as a way to strengthen a plan before committing to it. The checklist has a longer and better-documented record. Surgeon and writer Atul Gawande popularised it in his 2009 book The Checklist Manifesto, drawing on aviation, where pre-flight checklists are standard.

Gawande led development of the World Health Organization Surgical Safety Checklist, a 19-item list covering steps around anaesthesia, incision, and the end of surgery. A study by Haynes and colleagues published in the New England Journal of Medicine in 2009 introduced it across eight hospitals in varied countries and reported that the rate of death fell from 1.5% to 0.8% and inpatient complications from about 11% to 7%. Later studies found smaller or mixed effects, a reminder that a checklist works only when a team genuinely uses it rather than signing the box.

• "A pre-mortem is just pessimism." It is structured, time-boxed, and ends in fixes, not worry.
• "Checklists are for people who lack expertise." They exist because experts under load forget dull, essential steps.
• "Longer checklists are safer." Overlong lists get ignored; brevity is a feature.
• "If we adopted the checklist, we are covered." The benefit comes from real use, not from the document existing.

Take a high-stakes task you repeat and draft a do-confirm checklist of no more than nine items — only steps that are essential and often missed.

Think Like a Maester: Before you commit, assume the plan has already failed and make everyone tell you why.

1. How does a pre-mortem differ from a post-mortem, and why does the framing help?
2. What is prospective hindsight, and why does it produce more specific reasons?
3. What is the difference between a read-do and a do-confirm checklist?
4. Which kinds of task are a poor fit for a checklist?
5. What did the 2009 NEJM study associate with the WHO Surgical Safety Checklist, and what caveat matters?

Pre-mortems and checklists are cheap tools that prevent avoidable mistakes. A pre-mortem imagines the decision has already failed and asks why, surfacing fixable problems and lowering the cost of dissent. A checklist protects routine, high-stakes tasks from the dull steps experts skip under pressure. Both help only when actually used, and only where they fit.

• Learning from Failure
• Game Theory and Strategic Choices