Posting an incident with multi-stage updates

When something is broken, the goal is fast, honest communication. Status Pages structures that with a four-stage update template — every stage triggers an email to subscribers and pushes to the public page in real time.

To create an incident:
1. Status → "+ New Incident"
2. Title (e.g. "Elevated error rates on the API")
3. Severity — Minor / Major / Critical
4. Affected components — pick from the components you've defined
5. Initial status — usually "Investigating" with the first update
6. Click "Publish Incident"

The four-stage update flow:
• Investigating — "We're aware of an issue and looking into it" (first update)
• Identified — "We've identified the root cause" (when you know what broke)
• Monitoring — "A fix is in place; we're watching for stability" (after deploy)
• Resolved — "The incident is resolved" (when confidence is high)

Each update:
• Posts to the public page within seconds
• Triggers a transactional email to every confirmed subscriber
• Updates the affected components' health (Investigating may flip API to "Major Outage", Resolved flips it back to "Operational")
• Stamps your name and timestamp so the audit trail is clear

Postmortem (after Resolved):
• Open the resolved incident → "Add postmortem"
• Free-form text — root cause, timeline, what changed, what's next
• Postmortems are visible on the public page under the incident
• Customers and partners read them; they're a trust-building artifact

Tone guidance:
• Lead with what's affected ("API requests are returning 500s")
• Avoid jargon ("Aurora replication lag" → "our database is slower than usual")
• Don't blame third parties unless you're certain it's their issue
• Update at least every 30 minutes during an active incident, even if there's no news ("Still investigating, no new findings")