When a service fails, teams focus on restoring it. The exact sequence of symptoms, actions and decisions can become unclear surprisingly quickly. A simple incident note keeps facts in one place and makes it easier to learn after systems are stable.

Start a timeline in UTC

Record when the issue was detected, when user impact began if known, key observations, mitigation steps and recovery time. Note uncertainty clearly instead of inventing precision. Link to logs, dashboards and tickets rather than copying secrets into the document.

Describe the customer impact

Write what users actually experienced: failed requests, missing access, slow transactions or interrupted internal workflows. Avoid confusing a system metric with its business impact. If the blast radius is unknown, say that and explain what will be checked.

Capture important changes

Record deployments, configuration changes, patches, DNS updates or infrastructure events that might be relevant. A change near the incident is not automatically the root cause. Keep observed facts separate from hypotheses.

Document mitigation and recovery

Explain which action reduced harm, how the team verified recovery and whether any temporary workaround remains. Assign an owner and a date for follow-up work rather than ending the record at service restored.

Review without blame

Look for process and system improvements: clearer alerts, safer release steps, improved backup testing or better access procedures. Avoid using the review to search for a single person to blame.

  • Incident title, status and severity level
  • Detection and start timestamps in UTC
  • Observed customer and workflow impact
  • Mitigation actions and who performed them
  • Recovery verification checks and timestamps
  • Action items with owners and target dates