Before you start
Native on-call requires Pro. An owner or admin configures it in Settings → On-call. Invite the people who will respond and configure the delivery destinations they will use before creating the rotation.
Create a rotation
Enter a recognizable name, choose the timezone and rotation length, then select members in the intended order. Members take the primary position in order and hand over each rotation period. Review the resulting schedule before relying on it for coverage.
Create an escalation policy
A policy determines who is paged, the delay before paging, and how many times the sequence repeats. Select the first rotation, configure the delay and repeat setting, and save.
Attach the policy to the monitors it should cover using the monitor's available on-call controls or the API's escalation-policy assignment endpoint. A saved rotation or policy alone does not establish coverage for every monitor.
Respond to an alert
Open On-call to review escalation alerts and acknowledge the one you are taking responsibility for. Acknowledgement is not evidence that the service recovered. Investigate the affected monitor, keep the incident timeline updated, and confirm recovery separately.
Verify coverage
Use a controlled monitor to exercise a failure, verify the intended responder receives the alert, and acknowledge it. Check timezone, membership, destination, policy assignment, and delay settings when the wrong person is notified or no page arrives.
Review rotations after team changes. API automation for schedules, policies, assignment, and acknowledgement is documented in the API reference.