Asking why a plan failed produces objections that asking whether it might fail does not.
Why it matters when the plan changes
Once a group has committed to a plan, raising doubts costs social capital and reads as disloyalty. The pre-mortem, described by Gary Klein in Harvard Business Review, 2007, removes that cost by changing the question: the failure is stipulated, so naming a cause is analysis rather than dissent. It counters a group's tendency toward confirmation bias, the pull to favour information that supports a view already held, which a committed plan produces.
The tension is that it works only if what it surfaces is recorded and acted on. Run as a workshop ritual whose output is a flip chart, it becomes a way of demonstrating that concerns were heard rather than a change to the plan. The value lies in converting each named cause into an owner, a mitigation and a review date, so a stated risk carries a deadline instead of dissolving into the record of a meeting that happened.
In practice
A leadership team asked whether it has concerns about a launch raises two, politely. The same team, told the launch has failed and asked why, produces eleven causes in twenty minutes, including the sequencing conflict between two functions that nobody had been willing to name while the plan was still being celebrated.
Evidence
A project pre-mortem asks a team to imagine a plan has failed and to generate the reasons, before the plan begins.
Gary Klein, Performing a Project Premortem, Harvard Business Review (2007)People search for, favour and recall information confirming a view they already hold, which is what a committed plan produces.
Confirmation bias, Wikipedia (2026)
What it cannot tell you
A pre-mortem produces plausible reasons a plan might fail, not a measure of how likely each one is or how much it matters. It does not verify which causes are real, rank them by consequence, or guarantee that anyone acts on what surfaces. The exercise ends at a list; what happens to that list is a separate question the method does not answer.
Questions
The team is told the plan has failed comprehensively, a technique Gary Klein described in Harvard Business Review in 2007, then each person writes down the reasons privately before discussion. The lists are collected, causes are grouped, and each significant one is given an owner, a mitigation and a check date.
Because asking for concerns invites people to weigh whether raising one is worth the cost. Stipulating the failure removes that calculation, since a concern is no longer a prediction but an explanation of a stated outcome, countering the confirmation bias described in Wikipedia's 2026 entry on the term.
Because the first spoken objection anchors the discussion and the most senior voice anchors it hardest. Independent written lists capture what people actually think before the room converges, which is where the least comfortable and most useful causes tend to be.
After a plan is defined enough to fail in specific ways and before it is committed, which is a narrow window. Too early and the causes are generic; too late and the exercise documents concerns about something that can no longer be changed.
Producing a list and stopping. Without an owner and a date on each significant cause it becomes evidence that concerns were invited rather than a change to the plan, and the causes named will reappear at the milestone exactly as described.