The Pre-Mortem: Imagining Your Plan Has Gone Wrong Before You Start

A pre-mortem is a short exercise you do before a plan begins. You imagine that it has already gone wrong, then write down the most likely reasons. It works because explaining a failure feels easier than predicting one. "What might go wrong?" invites a shrug and "probably fine". "It went wrong; what happened?" invites a list.

This essay compares the pre-mortem with its older cousin, inversion, and shows both on one small plan. It also gives a five-minute version, and says where the method helps less than you might hope.

What is a pre-mortem?

A post-mortem looks back at something that has ended and asks what went wrong. A pre-mortem borrows the question and moves it earlier. You take a plan that hasn't started yet and write as though you're looking back on its failure.

The core instruction is to say, "It's next Sunday, and this went badly. Why?" Then you write reasons in the past tense, as if reporting them: the tent leaked, we arrived after dark, nobody had the corkscrew.

The idea is not new. Seneca, in his letters, suggested picturing misfortunes in advance so that they would be less of a shock if they arrived. He was after steadiness of mind more than good logistics, but the habit is recognisably the same. Modern planners have given it a tidier shape, and that shape is the one I describe here.

One misconception is worth correcting early. A pre-mortem is not pessimism, and it is not a prediction. You are not claiming the plan will fail. You are borrowing the vantage point of someone who already knows how it ended, because that seat has a better view of the weak spots.

Why is it easier to find problems when you assume they exist?

I don't have a study to point to here. What follows is my own reasoning, and it's the explanation I find most convincing. There are three parts.

Looking for flaws feels disloyal. When a plan is fresh and you're fond of it, "will this work?" quietly turns into "please say yes". Objections feel like complaints. In a pre-mortem, failure is the premise, so naming a weak point is simply doing the task. Nobody is being negative, because the exercise requires it.

A question about the past asks for a story, not a number. "How likely is it that the cake gets squashed?" is hard to answer. "The cake was squashed; how?" is easy: it slid off the back seat on the first roundabout. Minds are better at causes than at probabilities.

Proofreading offers a familiar parallel. If a friend hands you a page and says "there are three typos on this", you will probably find them. If they say "check this over", you may read straight past all three. Knowing the problem exists changes how closely you look.

How is a pre-mortem different from inversion?

Inversion is the better-known cousin. You ask, "What would make this fail for certain?" and then avoid those things. If you wanted a miserable camping trip, you'd bring no water, arrive in the dark and not check the forecast. So you bring water, arrive in daylight and check the forecast.

The two methods overlap, but they aren't identical:

• Inversion produces deliberate ways to fail. It's a thought experiment about the opposite of what you want, and it tends to give general principles.
• A pre-mortem produces accidental failures, the ones that happen to a sensible person with good intentions. It tends to give specific, local worries, such as "the borrowed tent might be missing pegs".

I tried both on the same plan in my notebook, and the lists barely overlapped. Inversion gave me sensible rules. The pre-mortem gave me odd, particular things I could act on. Neither replaced the other. If I had to pick one for a small, ordinary plan, I'd pick the pre-mortem, because its answers come already shaped like tasks.

Benjamin Franklin described a related habit, though it is a different tool. He suggested dividing a sheet into two columns, pros and cons, and weighing them against each other. That is a pros-and-cons method, not a pre-mortem. What they share is the idea that a worry written down is easier to handle than a worry carried about.

A worked example: a birthday dinner for six

Here is an invented example, kept tidy on purpose so the steps show clearly. Suppose you're cooking a birthday dinner at home for six people on Saturday evening. The plan is a roast, two sides, a cake and a table laid for six.

Now imagine it is Sunday morning and the dinner went badly. Write the reasons quickly, without judging them:

1. Dinner was served at 9 pm because the roast and the sides needed the oven at the same time.
2. One guest turned out not to eat meat, and this came up at 6 pm.
3. Only four of the chairs were comfortable.
4. The cake was fine, but there was no room left in the fridge for it.
5. The cook spent the whole evening in the kitchen and barely spoke to the guest of honour.
6. The candles were in a drawer nobody could find.

Look at the list as a whole before you fix anything. The obvious items are the ones you would probably have caught anyway: chairs, candles, fridge space. The useful ones are the surprises. Reasons 1 and 5 are the same discovery in two forms. The plan quietly assumed the cook could be in two places at once: attending the oven and attending the party.

That is the real yield of a pre-mortem. It mostly surfaces hidden assumptions, not risks. A risk is something you already knew might happen. An assumption is something you didn't know you were relying on. Once it's written down, it's cheap to fix:

• Choose a main dish that finishes in the oven without attention, or one that can be made earlier and reheated.
• Ask one guest to look after drinks, and hand the evening's smaller jobs to whoever offers.
• Ask about dietary needs on Thursday rather than finding out on Saturday.

None of these costs much, and every one is easier before the day than during it.

The same method works on a camping trip. If the trip "went badly", plausible reasons might include a borrowed tent missing a few pegs, a stove with a nearly empty gas canister, or arriving after dark with nobody sure where the water was. The cheap fixes are to pitch the tent in the lounge room first, test the stove for a minute, and leave an hour earlier. As with the dinner, the value lies in the fixes being small and the list having turned up the job before it mattered.

The five-minute version

Set a timer and keep a cheap notebook to hand. The total is five minutes.

1. Name the plan in one sentence (1 minute). Write it down, along with the date you'll be judging it. "Saturday's birthday dinner for six, judged on Sunday morning."
2. Write the failure reasons (2 minutes). Begin with "It went badly because…" and list as many as come, in the past tense. Don't rank or edit. Aim for quantity; six to ten is plenty.
3. Circle the surprises (1 minute). Mark the two or three that you would not have thought of without this exercise, or that would be hardest to repair on the day. Ignore the ones you already had under control.
4. Make one change for each circled item (1 minute). A small change is enough: a text message, a swap, something added to the shopping list.

Keep it light. If the list starts to feel gloomy, stop and move to step 3. The point is to find a few cheap fixes, not to enlist every possible disaster. A pre-mortem that makes you dread the weekend has overshot.

If you're planning with other people, the same steps work aloud. Give everyone a minute to write their own list in silence before comparing. That way the first confident voice doesn't set the whole list, and quieter people tend to notice different things.

When a pre-mortem helps, and when it doesn't

It depends, and here's on what.

How costly a failure would be. For a plan that is cheap and easy to reverse, such as an ordinary Tuesday dinner, a pre-mortem is more ceremony than it's worth. Just cook and see. It earns its five minutes when the plan involves other people, a fixed date, or something hard to redo.

How familiar the plan is. For something you've done many times, your habits already hold most of the answers. For something new, or a familiar plan with one new ingredient, such as a borrowed tent, it tends to pay for itself.

Whether you can adjust partway through. If you can fix things as you go, a pre-mortem matters less. If decisions are locked in early, as with booking, buying and posting invitations, it matters more.

There are also limits worth stating plainly:

• It can only surface what someone involved can imagine. It won't catch the problem nobody has thought of.
• It has no way to tell you which items are truly likely. Your circling is a judgement, and it will sometimes be wrong.
• The list can sometimes be used to avoid starting. If it never turns into one small action, it has become worry in a tidier format.

I'm still learning where the edges are, and I'd treat anything I say about it as "this is what worked for me, with small plans".

A question to carry about

If the pre-mortem works by switching a question from "will it?" to "how did it?", what happens if you switch the other way? Imagine the dinner went wonderfully, then ask what made it so. I suspect that list would hold different things: perhaps the pleasant surprises you'd otherwise never notice, like someone who stayed late to wash up or a dish that kept well. I haven't tested it enough to say. It would be worth one cup of tea and one notebook page to find out.