In 2007, cognitive psychologist Gary Klein published a three-page article in the Harvard Business Review with a deceptively simple idea. Before launching a major project, gather the team and tell them the project has already failed. Spectacularly. Then ask each person to independently write down what went wrong.
Klein called it the pre-mortem. The technique exploits what psychologists call prospective hindsight: the documented observation that people identify approximately 30% more causes of failure when imagining a negative outcome has already occurred than when asked to predict what might go wrong. The empirical finding, first published by Mitchell, Russo, and Pennington in 1989, has been replicated multiple times.
Daniel Kahneman endorsed the technique in his 2010 McKinsey Quarterly dialogue with Klein. McKinsey has since published four articles recommending pre-mortems for capital projects, board governance, CFO decision-making, and bias reduction. And yet procurement — a function that signs contracts worth millions with multi-year lock-in periods and irreversible consequences — has almost no documented adoption of the technique.
The original concept: how prospective hindsight works
The core insight from behavioral psychology is not complicated. When people are asked to predict future risks, several cognitive biases interfere. The planning fallacy makes them underestimate what can go wrong. Groupthink suppresses dissenting views. Overconfidence, particularly in teams that have succeeded before, narrows the imagination of possible failure modes.
The pre-mortem defeats these biases with a grammatical trick. Instead of asking "what might go wrong?" in the conditional tense, it asks "what did go wrong?" in the past tense. This small frame shift activates prospective hindsight: the brain searches differently for causes of an event it believes has already happened than for causes of an event that might happen.
Klein's original technique, and the version McKinsey recommends, follows a precise five-step process. The structure matters. Skipping steps produces a brainstorming session, not a pre-mortem.
Present the full sourcing plan, contract structure, or supplier strategy. Everyone starts from the same baseline.
The leader declares the initiative has failed. Not might fail. Has failed. This is the grammatical shift that activates prospective hindsight.
Each person writes reasons privately. No discussion. This prevents senior voices from calibrating what junior members say.
Share all reasons, group related causes, identify the top two or three showstoppers.
For each showstopper, build specific mitigations into the plan before proceeding. Do not move forward until every showstopper has a response.
What pre-mortems catch that risk registers miss
Surfaces known, politically safe risks. Senior voices dominate. Produces lists of generic items (supplier delay, budget overrun) without specificity. Items rarely change the decision to proceed.
Surfaces risks team members know but would not volunteer. Independent listing prevents hierarchy bias. Produces specific scenarios: "Supplier X's single production line in Thailand floods in October, component delivery stops for 8 weeks."
McKinsey documents a case where a company built two plants using entirely new, unproven technology. The investment failed and was scrapped. A pre-mortem would have surfaced the technology risk as a showstopper and prompted the company to build a pilot plant first before committing to full-scale construction. The technique costs 60-90 minutes. The alternative cost the full capital investment.
The UK Ministry of Defence provides the strongest procurement-specific example. In 2024, the MoD contracted for cyber pre-mortem workshops as part of its CARP transformation program. The contract, valued at £819,600, explicitly describes one-to-two day facilitated sessions where "cyber threat scenarios have already played out" and teams work backward to identify mitigations. A government procurement body is paying nearly a million pounds for a technique most CPOs have never encountered.
Where the technique fails in procurement: doing it wrong
The most common misapplication is confusing a pre-mortem with a risk brainstorming session. When the facilitator says "what could go wrong?" instead of "this has failed, tell me why," the prospective hindsight effect does not activate. The session becomes a standard risk workshop that surfaces the same known risks the risk register already contains.
A second failure mode is running the pre-mortem after contract award rather than before. The technique is a decision gate, not a post-hoc review. By the time a contract is signed, the pre-mortem's value has evaporated: the team is now invested in proving the decision was correct rather than imagining its failure. McKinsey explicitly positions the pre-mortem in the period after deciding to proceed but before committing resources.
The third failure is skipping the independent listing step. When team members discuss failures aloud instead of writing privately first, the most senior person's items set the range of acceptable answers. Junior buyers who suspect a supplier relationship is problematic, or who noticed a clause that doesn't work in practice, stay silent. The pre-mortem's entire value proposition — surfacing what people know but won't say — depends on anonymity in step three.
How to run a procurement pre-mortem: the operational guide
The pre-mortem as a procurement decision gate requires three adaptations from Klein's original project-management framing:
1. Run it before contract award, not before project launch. The procurement equivalent of "project launch" is contract signature. The pre-mortem belongs in the period between supplier selection and final commitment. McKinsey's guidance is clear: the technique is for after deciding to proceed but before committing resources. In procurement terms: after the sourcing decision, before the signature.
2. Include the supplier's team when possible. Parabol, a project management platform, recommends running pre-mortems "when working with new contractors on projects, so each party can set expectations." A joint buyer-supplier pre-mortem surfaces misaligned assumptions about delivery timelines, resource requirements, and governance that neither side would identify alone. This is the most underused variant.
3. The two showstopper rule. A pre-mortem typically produces 15-30 failure causes. Most are real but manageable. The exercise's value is identifying the two or three causes where mitigation must be built into the contract or sourcing strategy before signing. If mitigation for a showstopper is not feasible, the sourcing decision itself may need revisiting. This happens rarely. When it does, it saves far more than the pre-mortem costs.
What this means for procurement teams
Identify one major contract negotiation in the next quarter where the commitment is above $2 million or the duration exceeds three years. Schedule a 90-minute pre-mortem after supplier selection but before contract signature. Invite the sourcing team, the business stakeholder, and legal. If the supplier relationship is collaborative, invite the supplier's delivery lead.
The industrial evidence is clear: McKinsey found that suboptimal contract terms and poor contract management erode approximately 9% of annual revenues in sourcing. A pre-mortem that catches one contract gap that would have cost $180,000 on a $2 million commitment pays for itself several hundred times over. The technique costs 90 minutes and a meeting room. The alternative costs what risk registers have been missing for thirty years.
Frequently asked questions
What is a pre-mortem in procurement?
A pre-mortem is a structured exercise where a procurement team imagines that a major contract or sourcing initiative has already failed, then works backward to identify what went wrong. Developed by cognitive psychologist Gary Klein in 2007, the technique exploits prospective hindsight: the documented observation that people identify 30% more failure causes when imagining a negative outcome has already occurred than when asked to predict what might go wrong.
How is a pre-mortem different from a risk register?
Traditional risk assessment asks "what might go wrong?" and produces known, politically safe risks. A pre-mortem announces the project has already failed spectacularly, then asks each person to independently list reasons why. This frame shift surfaces risks that team members know but are reluctant to voice: over-optimistic timelines, unproven supplier technology, or unrealistic resource assumptions.
Has any procurement organization actually used pre-mortems?
Yes. The UK Ministry of Defence contracted for cyber pre-mortem workshops as part of its CARP transformation program (contract value £819,600). McKinsey also recommends pre-mortems for large capital projects, citing a case where a company built two plants with unproven technology — a pre-mortem would have surfaced the risk and prompted building a pilot plant first.
How long does a procurement pre-mortem take?
A typical pre-mortem session takes 60-90 minutes. The process: brief the team on the plan (10 min), announce failure and have each person independently list reasons (10-15 min), share and cluster causes (15-20 min), prioritize showstoppers (10 min), then integrate mitigations into the plan (15-20 min). McKinsey recommends half-day to full-day sessions for major capital projects.
Data sources
- Klein, G. — Performing a Project Premortem, Harvard Business Review (2007)
- McKinsey — Bias Busters: Premortems: Being Smart at the Start (2019)
- McKinsey — Four Ways to Assess Projects and Keep Them on Track
- UK Ministry of Defence — CARP Cyber Pre-Mortem Contract
- Channel Futures — How a Pre-Mortem Can Save You From Disaster
- Gary Klein — Pre-Mortem Method Overview
- McKinsey — Contracting for Performance: Unlocking Additional Value (2018)
- Parabol — How to Run a Pre-Mortem Meeting