Create an Escalation Policy

This page walks through building a policy from scratch — the policy-level settings first, then the steps that do the paging.

Create the policy

1. Open Incident Management → Escalation Policies and click Add Escalation Policy.

2. Enter a name and description.

3. Set the wait time — minutes before the first step fires. Leave it at 0 to page immediately.

4. Set the repeat count — how many extra times to run the whole policy from step 1 if nobody acknowledges. 0 runs the policy once.

Add steps

Each step names its recipients and how long to wait before escalating.

1. In a step, add one or more recipients:

   • Users — pick a user, then choose how to notify them: email, SMS, or voice. Select at least one method. SMS and voice are disabled (with a tooltip) unless your plan includes them and the user has a verified phone number.
   • On-call schedule — pick a schedule; the step notifies whoever is on call when it fires — possibly several people. Requires the Pro plan.
   • Slack channel — pick a connected channel and set its fallback order (0 is primary). Connect Slack first under Slack Integration.

2. Set Escalate after — minutes to wait for an acknowledgement before the next step fires.

3. (Optional) Turn on retries for the step: set a retry interval and a max retries count to re-page the same recipients until someone acknowledges.

4. Add more steps as needed. The connector between steps spells out what happens next — “Escalate after 10 minutes • then escalate to step 2”, or “then repeat the policy from step 1”, or “then stop”.

5. Click Save.

Retry interval must be shorter than escalate-after

Retries only run during a step’s escalate-after window. If the retry interval is at least as long as that window, the first retry never fires before the policy advances — the editor flags this with a warning. Set the retry interval shorter than escalate-after.

A worked example

A policy that pages on-call first, then widens if no one answers:

1. Wait 0 min — step 1 fires as soon as the incident opens.
2. Step 1 pages the on-call schedule plus a backup user by email and SMS. Escalate after 5 min.
3. No ack in 5 minutes → Step 2 posts to a Slack channel and pages the manager by voice. Escalate after 10 min.
4. Still no ack → with repeat count 1, the policy runs from step 1 once more.

As soon as anyone acknowledges, the policy stops.

Related

• Escalation Policies overview — the model and timing.
• On-Call Schedules — what a schedule recipient resolves to.
• Slack Integration — required for channel recipients.