Picture the meeting where the plan was approved. Everyone nodded. The timeline looked tidy, the budget looked reasonable, and the person who had doubts kept them to themselves, because nobody enjoys being the one who deflates the room.

Six months later the plan has collapsed. In the debrief, someone says the thing that stings most: we all kind of knew.

That sentence shows up in postmortems everywhere. The doubts existed. They just had no socially or psychologically safe place to land before the failure made them obvious.

The premortem is a deliberate way to give those doubts a place to land. It works on a team, and it works just as well when the only person you need to convince is yourself.

What a premortem is

The technique comes from the psychologist Gary Klein, who described it in a 2007 Harvard Business Review article, "Performing a Project Premortem." Daniel Kahneman later praised it in Thinking, Fast and Slow as one of the few practical ways to counter overconfidence in planning.

The instruction is simple. Before you begin, imagine it is some time in the future and the project has failed. Not struggled, not been delayed. Failed, clearly and completely. Then write down the reasons why.

A postmortem asks what went wrong after the damage is done. A premortem asks the same question while there is still time to do something about the answer.

Why a small change in framing works

The usual way to check a plan is to ask, "What could go wrong?" It sounds sensible, and it mostly produces polite, shallow answers. Two things get in the way.

The first is optimism bias. People systematically expect their own plans to go better than the evidence says they will, and a plan we've already invested in gets a protective glow. Asking "could this fail?" invites us to defend it.

The second is social. In a group, once a plan has momentum, raising an objection can feel disloyal. The premortem changes the assignment. Everyone is now supposed to find the flaws, so doing it becomes a contribution rather than a betrayal.

There is also a cognitive mechanism underneath. In a 1989 paper, Deborah Mitchell, J. Edward Russo, and Nancy Pennington reported that people generate more and better reasons for an outcome when they imagine it has already happened than when they imagine it might happen. They called this prospective hindsight. The widely cited headline is that it improved people's ability to identify reasons for future outcomes by around 30 percent, though that figure comes from a specific lab setup, so treat it as a sign of the direction rather than a promise.

The likely reason is that "might happen" keeps your mind in abstract, hypothetical territory. "Has happened" pushes it toward concrete stories. Once you must explain a failure as a fact, your brain starts building specific causal chains. The vendor missed the deadline, so we skipped testing, so the launch bug reached customers. A chain like that is something you can act on.

A small example

Say you're planning to move apartments next month while also starting a new job. The usual plan looks like this: pack on weekends, movers on the 30th, first day of work on the 2nd.

Now run the premortem. It's the 3rd of next month, and the move was a disaster. Why?

Written quickly, without censoring, the answers might look like this:

  • The movers rescheduled and I had no backup.
  • The old landlord's walkthrough was a day before I'd finished cleaning, so I lost part of my deposit.
  • Internet wasn't installed, and I had to work from a cafe in my first week.
  • I packed the one box with chargers and documents at the very bottom of the truck.
  • I was so exhausted by the second that I made a bad first impression.

None of these are exotic. Every one of them is cheap to prevent today and expensive to discover on the day. That's the point of the exercise. Most failures aren't surprises. They are ordinary risks nobody wrote down.

What the exercise is not

A premortem is not pessimism, and it isn't a ritual of doom. The goal is not to talk yourself out of the plan. Done well, it usually leaves people more confident, because the vague dread of "something might go wrong" gets converted into a short list of named problems with named fixes.

Vague worry is draining. A specific risk with a countermeasure is almost restful.

It is also not a prediction. You aren't claiming the failure will happen, and you don't need to be accurate about probabilities. You're widening the set of possibilities your attention covers, which is the thing planning-stage optimism narrows.

Doing it alone

Klein's original version is built for teams: everyone writes privately first, then reasons are shared round-robin so quiet voices count as much as senior ones. Alone, you lose the variety of perspectives but gain something else, which is the freedom to be blunt with no one watching.

The one trap solo is stopping at the first two or three reasons. The most obvious causes arrive first, and they're the ones you were already half-aware of. The useful ones tend to show up around the fifth or sixth item, when you've run out of the easy answers and have to think harder.

Your next moves

Pick a real plan you're about to start, such as a project, a trip, a hard conversation, a move, or a launch, and do this today:

  • Set a ten-minute timer and write the failure as a fact. Open with a sentence like "It's [date], and this went badly." Past tense matters, because it's what triggers the prospective-hindsight effect.
  • List at least seven reasons before you stop. Push past the obvious ones. If you only have three, keep going until the list starts to feel slightly uncomfortable.
  • Mark each reason as controllable, partly controllable, or not. Don't spend effort on the third group, except to decide whether a fallback exists.
  • Turn the top three into one concrete action each. "Book a backup mover by Friday" counts. "Be more careful" does not. Give each a date.
  • Put a review date on the calendar halfway through the plan. Reread your list then. Some risks will have passed, and some will have grown teeth.

If the plan involves other people, send them the first step and ask them to do it privately before comparing notes. You'll likely hear risks you never would have thought of.

Why writing it down matters

A premortem you only run in your head tends to evaporate. The reasons feel clear in the moment and are gone by Thursday. And the entire value of the technique lies in being able to return to the list later, when the plan is under way, optimism has crept back in, and a risk is quietly coming true.

A written list also gives you something rarer: an honest record of what you feared before you knew how things turned out. Looking back, you can see which worries were real and which weren't, and that gradually sharpens your instinct for the next plan.

That's the quiet reason I keep premortems somewhere I can find them. In Pagebox, I make a page per plan, the failure list at the top and the three actions beneath it as a checklist. Because it opens in under a second and is there offline, the halfway review actually happens instead of getting buried. If you want a place for that, you can try it at pagebox.lumenlabs.works. But the method works on the back of an envelope, so start there.