Premortem Example: A Complete Worked Premortem, Step by Step

The best contract of your year is on the table. One client, six months, guaranteed volume, a logo worth framing. The pen is in your hand, and part of you is already spending the money.

That is exactly the moment a premortem earns its keep. I have written about premortem analysis as a method, and the premortem framework page walks you through it interactively. But reading about a premortem and watching one get done are different things, the way a recipe and a meal are different things. So this article is one complete worked example, start to finish.

First, a note on where the method comes from. The premortem was devised by the cognitive psychologist Gary Klein, who described it in Harvard Business Review in 2007. It rests on research by Deborah Mitchell, J. Edward Russo, and Nancy Pennington into what they called prospective hindsight: imagining an event as if it has already happened. In their study, that single shift increased people's ability to identify reasons for future outcomes by about 30 percent. Certainty, even imagined certainty, unlocks reasons that mere possibility keeps hidden.

Now, the example.

The decision

Imagine you are an independent consultant. A large client offers a six-month engagement at a discounted day rate in exchange for guaranteed volume. Stable income, deep work, a marquee name for the portfolio. Every part of you wants to say yes, and confidence is high.

High confidence is not a reason to skip the exercise. It is the reason to run it.

Step one: assume it failed

A premortem begins with a sentence that feels unnatural to write:

It is six months from now. The engagement has ended, and it went badly. You are exhausted, the relationship has soured, and your pipeline is empty.

Notice the tense. Not "it might fail". It has failed. You are no longer predicting; you are explaining. That small grammatical shift is the whole trick.

Step two: write the failure list

Set a timer for ten minutes and write down every reason the failure happened, without filtering. Here is the full list this scenario produces:

1. Scope crept, because "guaranteed volume" was never defined as deliverables.
2. Your champion left the company in month two, and nobody else owned the relationship.
3. The discounted rate started to feel unfair as the workload grew, and quiet resentment leaked into the work.
4. Your other clients drifted away, so when the engagement ended, the pipeline was empty.
5. Payment terms of sixty days created a cash crunch in month three.
6. You stopped marketing entirely, because you felt safe.
7. A reorganisation froze the project midway, and the contract had no pause clause.
8. You could not push back on bad requests, because you were afraid of losing the whole contract.
9. You burned out, because you left no margin between sprints.
10. They offered to renew at the same discounted rate, and you could not afford to say no.

Ten minutes. Ten failure paths. Before the exercise, the mind held exactly one story: the good one.

Step three: sort by likelihood and control

Not all ten deserve equal attention, so sort them with two questions. How likely is this? How much of it is in my control?

The champion leaving is outside your control, but you can contract around it. Scope creep is both likely and controllable. Pipeline decay is nearly certain unless you deliberately prevent it. Those three go to the top. The reorganisation is unlikely but cheap to protect against with a single clause. The rest you note and accept, because a premortem is not about eliminating risk. It is about knowing which risks you are choosing.

Step four: strengthen the plan

Here is what changes in the contract and the calendar:

1. Deliverables are defined per month, with a simple change-request process, so scope has a gate.
2. The agreement is tied to the engagement, not to the champion, and a pause clause is added.
3. One day each week stays reserved for other clients and your own marketing, priced into the plan.
4. Payment terms move from sixty days to the 20th of the following month.

You still say yes. But it is a different yes: smaller in fantasy, larger in structure. Run the tape forward, and if the champion does leave in month four, or the reorg does arrive in month five, the contract absorbs it instead of the relationship.

That is what a premortem does. It does not make the decision for you, and it does not predict the future. It works like a mirror held at an unfamiliar angle, showing you the parts of the plan you have been avoiding looking at. The judgment stays yours. The blind spots get smaller.

If you want to run your own, the premortem template gives you the blank structure, and a decision journal is the natural place to keep what you wrote, so hindsight cannot edit the story later.

And if you remember only one thing, remember the tense. Do not ask what could go wrong. Say it has gone wrong, then explain why. Your imagination is a better risk analyst than your optimism.