Pre-mortem

TL;DR

Pre-mortem: Before launching a project, announce that it has already failed. Ask everyone to write down why. The exercise surfaces risks that normal planning suppresses — and does so before, not after, they materialize.

What Is the Pre-mortem?

The Pre-mortem was developed by cognitive psychologist Gary Klein and first described in his 1998 book Sources of Power. The technique inverts the conventional planning review: instead of asking a team to assess risks to a plan they have collectively developed, it announces — hypothetically — that the plan has already been executed and has failed, then asks participants to generate the reasons why.

The psychological basis is what Klein calls "prospective hindsight." Research by Deborah Mitchell, Jay Russo, and Nancy Pennington showed that people who imagine a future outcome as if it has already occurred generate a richer and more complete set of explanations for it than people who try to predict what might happen. The past tense creates imaginative freedom that the conditional future does not.

In a typical planning review, team members face strong incentives to avoid raising concerns. Voicing pessimism about a plan that has already consumed months of effort, that the team leader is committed to, and that requires everyone's cooperation to execute well can feel disloyal, demoralizing, or politically costly. The Pre-mortem removes this inhibition by legitimizing pessimism: the explicit premise is that failure has occurred, so identifying causes is not disloyal but analytical.

The result is a richer failure map produced before the consequences of failure are real — which is the only time when the map is fully actionable. This distinguishes the Pre-mortem from the Post-mortem, which analyzes the same failure modes after they've already cost time and money.

How It Works

Step 1: Assemble the relevant team
  — Include people with different vantage points on the project:
    technical, commercial, operational, customer-facing.
  — The diversity of perspectives is what makes the exercise
    productive.

Step 2: State the premise
  — "Imagine it's 18 months from today. We executed this plan.
    It has failed — completely, clearly, undeniably.
    We are gathering to understand why."

Step 3: Individual writing (5–10 minutes, no discussion)
  — Each person independently writes down every reason they
    believe the failure occurred.
  — Silence prevents anchoring on the first person's answer.
  — This is where the most valuable insights are generated.
    People write things they wouldn't say in open discussion.

Step 4: Round-robin collection
  — Each person reads one reason. Facilitator records all.
  — Continue until all unique reasons are surfaced.
  — No immediate debate — just collection.

Step 5: Synthesize and act
  — Group similar reasons.
  — Categorize by: probability, severity, and controllability.
  — High-probability + high-severity + controllable = address now.
  — Assign ownership and mitigation actions.

The key rule: The premise must be stated as fact ("the project has failed"), not as hypothesis ("imagine if the project failed"). The past tense activates prospective hindsight; the conditional does not.

Real-World Examples

Example 1: Hospital Medication Safety

Gary Klein originally developed the Pre-mortem while consulting with hospitals on medication safety. A typical failure mode: a patient receives the wrong medication due to a chain of small errors in the ordering, dispensing, and administration process. Conventional risk analysis identified obvious single-point failures. But medication errors in well-run hospitals are rarely caused by single failures — they are caused by combinations of small, individually innocuous errors that collectively produce harm.

When Klein ran Pre-mortems with nursing and pharmacy teams — telling them to imagine that a serious medication error had occurred with a specific patient scenario — the teams generated error chains that conventional analysis had consistently missed. Nurses described workarounds they had developed that violated protocol but seemed harmless in isolation. Pharmacists described labeling confusions that were "obviously fine" in normal conditions but catastrophic under time pressure. The Pre-mortem surfaced the systemic vulnerabilities that individual reviews had missed.

The hospital system implemented several process changes based on Pre-mortem findings. Medication error rates decreased measurably in the following year — before any errors of the type identified had actually occurred.

Example 2: Product Launch Risk Assessment

A consumer tech startup was preparing to launch its first hardware product after two years of development. The engineering team was confident; the marketing plan was approved; the launch date was set. The CEO proposed running a Pre-mortem two months before launch.

Premise: "It's four months from now. The launch happened. It was a disaster — returns are high, reviews are negative, press coverage is damaging. Why did this happen?"

The team's collective output identified 22 failure modes. The three that produced immediate action: (1) A logistics partner had not been stress-tested at the projected launch volume — if demand exceeded 5,000 units in the first week, fulfillment would break. (2) The product required a firmware update on first use, but the update process was fragile on slower Wi-Fi connections — affecting an estimated 20% of consumers. (3) The product packaging included a QR code linking to setup instructions, but the link had never been tested on non-US phone number formats.

All three were fixed before launch. The launch was successful. The CEO noted that none of these issues had appeared in any of the pre-launch reviews — they emerged specifically because the Pre-mortem created psychological safety for the engineer who knew about the Wi-Fi fragility but hadn't wanted to be "the person who delays the launch."

Example 3: Pre-mortem in Investment Due Diligence

A private equity firm had completed months of financial modeling, customer interviews, and management assessment on a potential acquisition. The investment committee was strongly favorable. One partner proposed a 30-minute Pre-mortem before the final vote.

Premise: "It's three years from now. We completed the acquisition. The investment has lost 40% of its value. What happened?"

Two insights emerged that the due diligence had not adequately surfaced. First, the company's primary revenue source was a single contract with a Fortune 100 customer representing 42% of revenue — a concentration risk that had been noted but not stress-tested. A partner who had previously worked at the Fortune 100 customer in question noted that the company's procurement team had recently reorganized, and the executive who championed the contract had retired. Second, the management team had no experience with the integration of a small acquisition they had announced six months earlier — which was already running 3 months behind schedule.

The Pre-mortem didn't kill the deal — the investment thesis was sound — but it led to two specific contract provisions: a seller escrow tied to the customer contract renewal, and an integration milestone schedule with financial penalties if the pending acquisition wasn't complete within 90 days. Both provisions were accepted. The renewal occurred; the integration was completed on schedule. The investment performed well.

When to Use It

✅ Before the launch or execution of any significant project, decision, or investment. The Pre-mortem is most valuable when there is still time to act on its findings.

✅ After a plan has been developed and committed to, but before it is executed. The commitment creates the psychological investment that makes the Pre-mortem necessary; the pre-execution timing makes its findings actionable.

✅ When a team has strong cohesion and enthusiasm for a plan — conditions where groupthink is most likely to suppress legitimate concerns.

✅ For any decision where multiple people have relevant knowledge but incomplete overlapping views. The Pre-mortem pools distributed knowledge in a way that individual reviews cannot.

❌ After execution has begun. Once a project is underway, most Pre-mortem insights have become either problems to address or risks to monitor — but the specific window of maximum actionability has passed.

❌ When the team hasn't genuinely committed to the plan. The Pre-mortem works by exploiting the psychological investment in a plan. If the plan is still in flux, participants haven't formed the attachment that creates the need for permission to be pessimistic.

Model Combinations:

Combine with	Effect
Inversion	Pre-mortem is inversion operationalized for teams; they complement each other
MECE	Use MECE to ensure the Pre-mortem covers all failure domains without gaps
Scenario Planning	Pre-mortem within each scenario produces a more complete risk map

Common Misuses and Limitations

Misuse 1: Treating the Pre-mortem as confirmation exercise. If the facilitator frames the exercise as "let's check that our plan is solid," participants will generate a weaker, more sanitized set of failure modes. The premise must be stated as fact — "the project has failed" — without qualification.

Misuse 2: Skipping the independent writing phase. The most valuable insights come from the silent individual writing phase, where participants write what they actually think rather than what they're comfortable saying in the group. Jumping straight to group discussion anchors on the first speaker's view and suppresses the quiet observations that make the Pre-mortem valuable.

Misuse 3: Generating failure modes without acting on them. A Pre-mortem that produces a list of 20 failure modes and then isn't followed by specific mitigation assignments has done half the job. The output must be categorized, prioritized, and converted into specific actions with owners and timelines.

Limitation — can produce false confidence if done superficially: A Pre-mortem that surfaces only obvious, already-known risks is not worthless, but it doesn't achieve the primary goal of uncovering unknown risks. The quality of the Pre-mortem depends on the team's willingness to surface uncomfortable insights.

Inversion: The Pre-mortem is the team-facilitated, structured application of inversion thinking. Where inversion is typically an individual practice, the Pre-mortem leverages collective knowledge.

Scenario Planning: Scenario planning maps multiple futures; a Pre-mortem within each scenario produces a more complete risk assessment.

Confirmation Bias: The primary cognitive bias the Pre-mortem is designed to counteract — by creating an explicit permission structure for disconfirming the prevailing plan.

FAQ

How is a Pre-mortem different from a risk assessment?

A conventional risk assessment asks 'what could go wrong?' from a forward-looking perspective — participants are still mentally in the posture of defending the plan. The Pre-mortem assumes failure has already occurred, which activates 'prospective hindsight' — the psychological mechanism that makes imagined past events easier to explain than predicted future ones. Research shows this produces a richer, more complete failure map than conventional risk assessment.

How long should a Pre-mortem session take?

For most project launches, 60–90 minutes is sufficient: 10 minutes for setup and framing, 10–15 minutes for independent writing, 30–40 minutes for round-robin collection and discussion, and 15–20 minutes for synthesis and action assignment. For very large or complex projects, a 3-hour session with more thorough synthesis is worthwhile. The minimum viable version — even 20 minutes of silent writing by two or three people — produces value.

What's the best resource for learning more about the Pre-mortem?

Gary Klein's Seeing What Others Don't (2013) is the best primary source, covering the technique's development and evidence base. Klein's original HBR article 'Performing a Project Pre-mortem' (September 2007) is shorter and freely available. Daniel Kahneman discusses the Pre-mortem favorably in Thinking, Fast and Slow (2011) as a rare example of an organizational technique that genuinely reduces overconfidence.

Apply This Model with AI

Describe your project and its current plan in MindMax. The AI will facilitate a structured Pre-mortem — prompting you to generate failure modes across all relevant domains, categorize them by probability and severity, and convert the highest-priority items into specific mitigation actions.

🚀 Run a Pre-mortem in MindMax →

What Is the Pre-mortem?​

How It Works​

Real-World Examples​

Example 1: Hospital Medication Safety​

Example 2: Product Launch Risk Assessment​

Example 3: Pre-mortem in Investment Due Diligence​

When to Use It​

Common Misuses and Limitations​

Related Models​

FAQ​

Apply This Model with AI​

Further Reading​