Updated 2026-07-02
An incident opened but you never heard about it. Before assuming something is broken, work through this checklist — it covers the overwhelming majority of cases.
1. Does the check actually have a channel attached?
A check with no alert channels attached will still record incidents in its history — it just won't notify anyone. The dashboard shows a warning badge on checks like this. See Email and webhook alert channels.
2. Is the email channel verified?
An email channel can't send alerts until it's confirmed with the 6-digit code sent to it at creation. An unverified channel is effectively silent.
3. Was the alert folded into an hourly digest?
If your account sent a lot of alert emails within the last hour, further individual alert emails are collapsed into a single digest instead of arriving one at a time. Check for a digest summary email before assuming nothing was sent at all.
4. Was the channel auto-disabled?
A webhook channel that fails 10 deliveries in a row is disabled automatically, and the account owner is emailed about it. Check the channel's status and last error in the dashboard — if it's disabled, fix the receiving end and re-enable it with one click.
5. Was the incident still confirming?
If the check requires more than one region to agree before it's marked down, a single-region failure shows as pending, not down — and pending states don't page you. This is by design (see How multi-region confirmation works) but it does mean a transient regional issue that never gets confirmed by a second region won't generate an alert, only a recorded blip.
6. Was a maintenance window active?
Alerts are suppressed by default during an active maintenance window covering that check. Check whether one was scheduled (or left running longer than intended) at the time.
7. Did the alert delay push the notice later than expected?
If the check's alert policy has an alert delay configured, Status waits that long after confirmation before sending the first notice, in case the check recovers on its own first. Check whether the incident resolved inside that window.