The plan launches soon. Here is how to run a pre-mortem in 20 minutes

Do this before the meeting ends

You do not need a new tool or a long prep. You need paper, a timer and about twenty minutes. Here are the steps from the book, in order.

1. Say the plan in one or two sentences so everyone pictures the same thing. Pick a future date when you would know if it worked.
2. Say the failure as fact. "It is that date. The plan failed. Not a near miss. A clear failure."
3. Set a timer for three to five minutes. Everyone writes, alone and in silence, every reason it failed.
4. Go around the room one reason at a time until the lists run dry. The leader goes last.
5. Sort every reason into two columns: things we control and things we don't. If the first column is thin, go around again and ask only for reasons that are on us.
6. Pick the two or three most worrying reasons we control. For each, decide one change to the plan or one early warning sign.
7. Write down who owns each change and when you will check.

The step most guides leave out

Step 5 is the one to protect. In a 2021 set of studies with 2,345 people, pre-mortems did make people less overconfident. But people tended to blame things outside their control, like the weather or bad luck. And the poster's bottom line was blunt: no difference in taking action to improve results.

If your list says "the market turned," there is nothing to fix. You feel humbler and do the same thing anyway. So require some reasons that are on you.

Why the silence matters

Picture the usual last meeting. The leader asks, "Any concerns?" Someone mentions a slow vendor. The engineer who thinks the timeline is fantasy says nothing.

Now say it already failed and hand out paper. The room gets busy. People who said nothing fill half a page. The book puts it in one line: it turns doubt from disloyalty into the assignment.

Silent writing first keeps the boldest voice from setting the list. The leader last keeps the boss's worries from becoming everyone's worries. Keep the tone on the plan, not on people. "The handoff broke down" helps. "Sam was too slow" starts a fight.

Who to invite, and running it online

Most pre-mortem guides tell you to invite more than the core team. The book agrees, with a sharper rule: invite the people who will actually do the work, not only the people who approved it. They usually know where the plan is thin.

If the team meets online, the silent step still matters. Have everyone write in a private note first. Then paste the lists in at the same moment, so nobody edits their worries after seeing the boss's.

After the launch, check the list

A pre-mortem is not done when the meeting ends. The book's log for this tool has six lines: the plan, the failure reasons you control, the ones you do not, the changes you made, your confidence before and after, and what actually went wrong.

Fill in the last line on the check date. Did the trouble come from your list, or from somewhere nobody looked? Either answer makes the next pre-mortem better. And it is the only way to learn if the meeting changed anything, since, as the book notes, no large trial shows that pre-mortems lead to better project results.

Alone, in ten minutes

No team? A small 2020 study found no difference between solo and group pre-mortems. Write the date and "It failed" at the top of a page. Write for five minutes. Mark each reason C for in your control or O for outside it. Get at least three C's, then change one thing today.

The book uses a freelancer example. "The economy slowed" is an O. "Never found a third client" and "No cushion when a payment came late" are C's. Each one points to something to do before the start date.

Where to go after the meeting

This is the problem Mental Models That Actually Work was written for, and the full card is in The Pre-Mortem: Imagine It Failed, Then Ask Why. That chapter also takes apart the famous 30 percent claim. It traces to a 1989 study that counted reasons, not correct ones. Gary Klein's own 2007 article is short and worth a read.

Steelman and Red Team covers the limit of a pre-mortem: it only brings out doubts already in the room. That chapter shows how to borrow the minds that see the plan differently.

One-Way and Two-Way Doors helps you decide if this launch even needs the full treatment. If it is cheap to roll back, see our page on one-way vs two-way door decisions.