A project meeting can become strangely quiet once everyone senses that the plan has already been approved. People still see possible problems, but raising them feels like criticising the person who proposed the work. A premortem creates a different starting point: imagine the project has gone badly, then explore plausible reasons.
Atlassian’s premortem exercise provides a structured way to discuss risks before a project begins and revisit them later. Its value comes from the conversation and follow-through, not from predicting the future accurately.
Give people time to think independently
Define the project and the imagined point of failure. “The launch has missed its intended outcome” is more useful than “everything went wrong.” Let participants write possible causes before the most senior person begins speaking. Otherwise, the first confident explanation can become the answer everyone edits around.
Ask for observable circumstances. “Communication was bad” is difficult to act on. “The supplier changed the delivery date, but the installation team was not informed before booking travel” identifies a dependency and a handoff that can be examined.
Turn the strongest concerns into decisions
Consider a fictional training launch dependent on a new booking system. The team identifies a risk that staff will not have access in time. A useful response might assign an owner to confirm accounts, set a test date and define an alternative booking route. The concern has now become a piece of work with a visible checkpoint.
Do not treat every imagined failure as equally likely or spend the entire schedule defending against remote possibilities. Discuss impact, plausibility and what the team can influence. Some risks can be reduced; others need monitoring or an explicit decision about whether to proceed.
Close the session with a short record: the risk, the early signal, the response and the owner. Revisit it when assumptions change, especially after a supplier delay, staffing change or altered scope.
A premortem should make disagreement easier to express without making pessimism the only acceptable contribution. The question is not who can invent the worst outcome. It is which overlooked condition deserves attention while the team still has time to change the plan.