Run a post-incident review without assigning blame

A useful incident review reconstructs the sequence and improves the operating procedure rather than treating an individual mistake as the complete explanation.

A practical procedure

1. Collect the timeline, observable symptoms and decisions from the people involved.

2. Identify missing cues, unclear ownership or untested changes that contributed to the incident.

3. Assign one concrete improvement with an owner and a rehearsal that can verify it.

Worked example: a fictional session

A fictional wrong-price incident leads to a shared active-offer check before product changes, rather than an instruction to the host to 'be more careful'.

Check the limits

A review should separate confirmed contributing factors from explanations proposed after the event.