Pre-Mortem: How to Surface the Risks Nobody Says Out Loud

What is a pre-mortem?

A pre-mortem is a risk exercise developed by psychologist Gary Klein. Instead of asking “what could go wrong?”, it asks you to assume the worst has already happened: “It’s a year from now. This project failed. What went wrong?”

That small change matters. “What could go wrong?” invites people to defend the plan. “It failed — why?” invites them to explain a fact. Research on prospective hindsight — imagining an event has already happened — found that it helps people come up with more reasons for an outcome than simply asking whether it might happen.

A post-mortem tells you why a project died. A pre-mortem asks the same question while the patient is still alive.

Why new projects need a pre-mortem most

In my experience, most reviews of a new business start from the same quiet assumption: this will work. The plan is built to show why it succeeds, the numbers are built on that plan, and the review checks whether the numbers add up.

That works reasonably well for a business you already run. You know where it breaks.

It works badly for a business the company has never run before. Nobody in the room has seen it fail, so nobody knows where it breaks. That is exactly when you need to assume failure is possible — likely, even — and ask what would cause it. Over the years I’ve come to think this habit is most of what people call deal know-how: not predicting success better, but being specific about how something could fail.

Why nobody says the risk out loud

Here’s a pattern I’ve seen more than once on new projects: several people in the room feel uneasy about the same thing, and nobody says it. Later, that is the thing that goes wrong.

The reasons people stay quiet are rarely about the risk itself. They’re about the room:

• The organization is rigid. Objections feel like disloyalty to a plan the leadership already likes.
• “What if I’m wrong?” Raising a risk you can’t prove feels like putting your judgment on the line.
• The Math of Silence. If you say nothing, your score is zero. If you speak up and you’re wrong — or even right but unwelcome — your score goes negative. Silence is the rational choice.
• Nobody senior has said it. If the people above you aren’t raising it, raising it yourself feels out of place.

A pre-mortem works because it changes that math. When the assignment is “list the reasons this failed,” naming a risk isn’t dissent. It’s doing the exercise. Nobody has to be the person who objected.

A public example: Target’s expansion into Canada

You don’t have to take my word for how this plays out. One of the best-documented cases is Target’s move into Canada.

In 2011, Target paid about $1.8 billion for the leases of a Canadian discount chain and planned to open more than 120 stores within roughly a year. It was a new country, a new supply chain, and new systems — a business Target had never run.

What went wrong was not a secret inside the company. According to former employees interviewed by Canadian Business in 2016:

• The product data loaded into the new system was accurate only about 30% of the time. Stores had empty shelves while warehouses overflowed.
• Before launch, senior staff had concerns about the timeline. But in the meeting meant to raise them, nobody clearly said the plan wasn’t ready. As one former employee put it, “Nobody wanted to be the one to say, ‘This is a disaster.’”
• Status meetings kept ending on an upbeat note. The leases were already signed, so many people felt speaking up wouldn’t change anything.

In January 2015, less than two years after the first stores opened, Target announced it would close all 133 Canadian stores. It expected a pre-tax loss of about $5.4 billion, and more than 17,000 people lost their jobs.

Every risk that sank the project was known to someone in the building. A pre-mortem — “It’s 2015 and we’ve pulled out of Canada. Why?” — would not have guaranteed a different decision. But it would have put “our data isn’t ready” and “we’re opening too many stores at once” on the table as the assignment, not as somebody’s career risk.

What a pre-mortem would have caught

I described one of my own mistakes in survivorship bias: we bought a second business because the first one — in the same industry — had done very well. The second one struggled to cover the interest on its acquisition loan.

If we had run a pre-mortem — “Two years from now, this acquisition is a problem. Why?” — I think two answers would have come up within ten minutes:

1. “The first deal rode a customer’s growth. This one doesn’t have that tailwind.”
2. “Its major contracts are re-tendered every few years. If we lose one, revenue drops and the loan doesn’t care.”

Both were knowable at the time. Neither was in the review, because the review was built to explain why the deal would work.

How to run a pre-mortem (45 minutes)

1. Set the scene (2 min). The leader says: “It’s [date, 12–24 months out]. This project failed — clearly and badly. We’re here to understand why.”
2. Write alone (5–10 min). Everyone writes their reasons independently, in silence. This is the most important step — it stops the first or most senior voice from setting the direction.
3. Go around the room (15 min). Each person reads one reason at a time, round-robin, until the lists are empty. The leader goes last. No debate yet — just capture.
4. Group and rank (10 min). Cluster similar reasons. For each cluster, mark how likely it is and how bad it would be.
5. Assign actions (10 min). For the top risks: what do we change in the plan, what do we check before committing, and what early warning sign will we watch? One owner each.
6. Revisit (later). Put the top risks on the agenda of the first two progress reviews.

A practical rule: the most senior person speaks last. If the leader says “I’m most worried about X” first, you’ll get a room full of X.

5 risk areas I always check before a deal or project

When I review a deal or a new project, I go through the same five areas. I call it the Pick Enough Risk Scan. They’re also the backbone of the template below.

1. Scale. Is this big enough to matter, and small enough to survive if it fails? How does its size compare to what we usually run?
2. Money at risk. How much is committed, how much could we lose in the worst realistic case, and what does the financing assume? (Loans don’t adjust when revenue does.)
3. The owner’s knowledge and experience. Has the person running this done something like it before? If not, who in the room has — and are they actually involved?
4. Fit with what we already do. Does it share customers, capabilities, or assets with our existing business — or is “synergy” a word in the deck with no mechanism behind it?
5. Policy and legal risk. What regulation, licensing, permits, or policy change could block it, delay it, or change its economics?

Notice that none of these ask “will the market grow?” Market forecasts are where optimism hides. These five ask what happens to us if things go badly.

Pre-mortem template (copy and paste)

PRE-MORTEM: [Project / Deal name] — [Date]
Participants: [Names]

1. THE SCENARIO
   It is [date, 12–24 months from now]. This project has failed.
   What does "failed" mean here? [e.g., missed target by 50%, wrote off investment, exited]

2. REASONS IT FAILED (write alone first, then share)
   - 
   - 
   - 

3. CHECK THE 5 RISK AREAS
   | Area                           | What could fail? | Likelihood (H/M/L) | Impact (H/M/L) |
   |--------------------------------|------------------|--------------------|----------------|
   | Scale                          |                  |                    |                |
   | Money at risk / financing      |                  |                    |                |
   | Owner's knowledge & experience |                  |                    |                |
   | Fit with existing business     |                  |                    |                |
   | Policy & legal                 |                  |                    |                |

4. TOP 3 RISKS -> ACTIONS
   | Risk | Change the plan / Check before committing / Watch | Owner (one person) | By when |
   |------|---------------------------------------------------|--------------------|---------|
   |      |                                                   |                    |         |

5. EARLY WARNING SIGNS
   - [Signal] — if we see it, we [action]

6. REVISIT ON: [date of first progress review]

Pre-mortem vs. risk assessment vs. post-mortem

	Pre-mortem	Risk assessment	Post-mortem
When	Before you commit	Before and during	After it ends
Question	“It failed — why?”	“What could go wrong?”	“Why did it fail (or work)?”
Strength	Gets unspoken risks out	Structured, complete	Real data
Weakness	Depends on honest participation	Easy to fill in defensively	Too late to change this project

They don’t compete. A pre-mortem is a fast way to feed a better risk assessment.

Frequently Asked Questions

The bottom line

Most project reviews are built to explain why something will work. A pre-mortem asks the opposite question on purpose — and gives everyone in the room permission to answer it. Assume it failed, ask why, and fix what you find while it’s still cheap.

Related reading:

• Survivorship bias: why success stories mislead you at work
• Satisficing: why good-enough decisions beat perfect ones
• How to write meeting minutes executives actually read