Investigate and follow up

Investigate, record a decision, and verify the result

1. Open the issue from Site Health or the alert link. Read the affected function, current status, and the time of the evidence.
2. Inspect the supporting check results or browser/application evidence. Check Activity for related changes and earlier occurrences.
3. Use I'll review this when available to take responsibility. Add a team note with the investigation context.
4. Make the appropriate fix in the system that owns the problem, or use an available MetricPoints action after reviewing its scope and confirmation.
5. Review fresh evidence after the action. An action being accepted does not prove that the website recovered.

Review decisions and recovery mean different things

For a notice with Complete review, record whether the condition is expected or addressed and explain why. The decision remains in history. Stop reviewing releases your assignment; it does not close the notice.

Do not use a completed review as proof that an unavailable check passed. Monitor recovery evidence separately. Reopen a notice if the decision needs another look; ongoing checks can also reveal new failures.

Keep a useful handoff

Record the affected page or function, what you checked, what changed, and what remains unverified. Avoid passwords, tokens, submitted form data, and private customer information in notes. Historical evidence remains useful after an incident is resolved or archived.

For repeated alerts, review notification routing and reminders before changing the monitor itself.