Incidents · Free+
Incident management
When Stripe goes weird, when a customer reports a bug your probes can't see, when your DB migration runs long — open an incident, post updates as you triage, and beat silence with a narrative.
Why you need it
Probes answer "is it Up right now?" Incidents answer "what's going on, what are you doing about it, and when did it end" — the questions customers actually ask. Open one in five seconds when stress is high, and your status page stops being a silent wall of green pills.
- Four-state lifecycle that mirrors how you actually triage. Investigating → Identified → Monitoring → Resolved. Each transition broadcasts to subscribers automatically, so updating the state is updating your customers.
- Markdown timeline, append-only on purpose. Post a correction instead of a stealth edit — subscribers have already read the original. Honest narratives build more trust than tidy ones.
- Per-component impact matrix. Mark checkout as Degraded while the rest stays Operational — the public banner tells the truth about scope instead of going all-red on a partial outage.
- Internal vs public updates. Take notes for your team in the same timeline; flip the public toggle off and they stay off the status page. One incident, two audiences, one source of truth.
- Severity that gates the loud channels. Major fires SMS and turns the banner red; Minor stays amber and quiet; Maintenance goes blue and won't dent your SLA.
Where it pays off
The default tool for any customer-facing event your team wants on record:
- Third-party dependency outage. Stripe is intermittently failing but technically returning 200s — your probes don't catch it, your checkout does. Open Minor, mark checkout Degraded, post updates as Stripe posts theirs.
- Internal change going sideways. DB migration taking three times its estimate. Open Major, post the 15-minute heartbeat customers want, resolve when the rollback lands.
- Customer-reported issue with no probe coverage. A feature flag misfired and 2% of users hit a broken flow. Open Minor, narrate the fix, prove you saw it before they emailed support twice.
- Augmenting an auto-incident. The probe engine opened "HTTP probe X is Down" — rename it to language an actual customer understands, add the human context.
Not the right tool when: the event is invisible to customers and you only need to alert your own team — that's an alert, not an incident. For planned downtime, use Scheduled maintenance.
Available on Free+. Already on StatusPulse? See the full config in Help →
Related
Try Incident management in StatusPulse
5 probes, 1 status page, forever. No credit card. US or EU host — you choose.