What you will learn
- Reconstruct what happened from evidence.
- Distinguish a plausible explanation from an established cause.
- Design a limited change and a fair test.
Reconstruct the event before judging it
A useful failure review starts with an expectation and an observable result. Write what you intended, what actually happened, and when you first noticed a difference. Replace ‘I am terrible at organizing’ with ‘Only six of the twenty registered participants attended.’ The second statement identifies a problem you can investigate. Keep messages, dates, counts, and examples where appropriate, while protecting other people’s private information.
Then record what you knew at the time. A choice can have been reasonable with the information available even when its result was disappointing. Conversely, a lucky outcome does not make a careless process reliable. Reviewing the decision separately from the outcome helps you avoid rewriting history.
Compare explanations with evidence
List at least three possible contributors before choosing one. Low attendance might reflect an inconvenient time, unclear directions, or a competing local event. Ask what evidence you would expect under each explanation. If people registered enthusiastically but repeatedly asked where to go, directions deserve attention. If they understood the location but could not leave work, changing the map will not solve the main problem.
A single event rarely proves a cause. Contact a few participants with a neutral question: ‘What affected whether you could attend?’ Include people who came as well as those who did not. Their answers are useful clues, not a representative survey of everybody.
Choose a change that addresses the mechanism
‘Try harder’ contains no operational lesson. ‘Send a reminder with the exact entrance and a contact number the day before’ explains what changes and why it could matter. Distinguish what you can change directly, what requires agreement, and what is outside your control. Bad weather may be uncontrollable, but a cancellation message can still be planned.
Do not change everything at once if you want to learn which adjustment helped. For a small, low-risk repeat event, keep the basic format and improve one likely bottleneck. In a safety-sensitive situation, address all necessary protections immediately; learning is never a reason to leave people exposed to a known hazard.
Close the review with a test
Specify the next occasion, the new behavior, and the evidence you will collect. For example, ask registrants to confirm they can find the entrance, then record both confirmation and attendance. Attendance might still vary for unrelated reasons, so also examine whether location questions decreased. This gives you a measure closer to the proposed explanation.
AHRQ’s debrief guidance connects an accurate event review with changes to future practice. Adapt that structure to everyday work without treating every disappointment as a formal investigation. A good review ends with an owner and a next action, while preserving what already worked.
Fictional case: Mei’s empty workshop
Mei, a library assistant, arranged an evening repair workshop. She blamed her presentation after only four people arrived. Reviewing registration messages showed that several people had asked whether they should use the locked main entrance. She could not conclude that this explained every absence, but it was a correctable obstacle.
For the next workshop she added an entrance photograph and asked a colleague to welcome visitors. Attendance increased modestly, and nobody reported being unable to enter. Mei kept the practical change while acknowledging that the different date also affected the result.
Write a one-page failure review
- Describe the expected and actual result without labels about your character.
- Create a brief timeline and separate known facts from guesses.
- Compare three possible explanations using available evidence.
- Choose one justified change, an owner, and a date to check what happened.
Further reading
- AHRQ: Reviewing the Team’s Performance—DebriefOfficial guidance on reviewing events and changing future practice; the everyday examples here are original.