| name | apply-premortem |
| description | Use when a plan is finalized but not yet launched, to surface failure modes the team hasn't acknowledged before commitment is locked in. |
| source | Gary Klein, "Sources of Power" (1998); Klein "Performing a Project Premortem" Harvard Business Review (2007); adopted by Google Ventures and standard in decision-making research |
| tags | ["risk-analysis","decision-making","project-planning","cognitive-bias","mental-models","team-process"] |
| related | ["apply-devils-advocate-technique","apply-dual-process-diagnostic","apply-reference-class-forecasting","apply-sunk-cost-discipline","apply-hindsight-bias-correction"] |
| verified | true |
Apply Premortem
Before launching a plan, imagine it has already failed catastrophically — then work backwards to identify every cause of that failure while you can still act on them.
Why This Is Best Practice
Adopted by: Google Ventures (Sprint methodology), U.S. Intelligence Community (analytical red-teaming), and recommended in standard decision-making curricula at Harvard Business School, Wharton, and by Nobel laureate Daniel Kahneman as the single most effective debiasing technique available before a decision.
Klein's original research showed premortems increase the ability to identify reasons for future outcomes by 30% over standard prospective analysis. Kahneman in "Thinking, Fast and Slow" (2011) states: "The premortem is a prophylactic against overconfident planning." Google Ventures adopted it as a standard Sprint exercise, exposing critical failure modes in product plans before any code is written.
Planning suffers from two documented biases: optimism bias (overestimating success probability) and groupthink (dissent is socially costly after a plan is endorsed). The premortem defeats both simultaneously — by framing failure as a given, it gives permission to voice concerns without appearing disloyal, and by forcing specificity about how failure happened, it bypasses the vague "what could go wrong?" question that teams answer with platitudes.