Leaving Opsgenie: a checklist for small teams

Published on: November 17, 2026, by: João Alves

Atlassian stopped selling Opsgenie as a standalone product in June 2025, and support ends in April 2027. If your team still runs on it, you have to move, and the official route is Jira Service Management or Compass.

For a big incident-response setup that route makes sense. For a small team it is often more than you need, because many small teams only ever used part of Opsgenie. This checklist helps you find out which part you use, move each piece somewhere that fits, and switch over without a week where nobody is on call.

1. Take an inventory

Before deciding anything, write down what is actually configured. In Opsgenie, go through:

  • Schedules and rotations: who is in each, in what order, the handover day and time, and the time zone.
  • Overrides already booked for the coming weeks, such as holidays and swaps.
  • Escalation policies: who gets notified next if the first person does not respond, and after how long.
  • Integrations: which monitoring and alerting tools send alerts in.
  • Notification rules: whether people actually get phone calls, SMS or push notifications.
  • Heartbeats and anything else that alerts when something goes quiet.

A spreadsheet with one row per item is enough. You will use it for the rest of the move.

2. Sort everything into three groups

  • Things that page people. An alert arrives and someone's phone rings. These need a real paging tool.
  • Things that only schedule. The rotation decides whose week it is, but alerts actually land in Slack or email and nobody is woken up.
  • Things nobody uses. Old schedules, integrations for services that no longer exist, the escalation policy from 2021.

Be honest about the second group. If the last time Opsgenie phoned anyone was a test, that rotation is a schedule, not a paging setup.

3. Pick a destination for each group

  • Paging: Jira Service Management, Compass or another incident platform. Move only the services that really need it, which often means fewer seats.
  • Schedule only: something light. A tool like RotaHog gives you the rotation, reminders, swaps and a link anyone can open, without per-seat pricing. See how it compares with Opsgenie.
  • Unused: delete it. Moving it would just bring the clutter with you.

4. Move the schedules

For each rotation you are keeping:

  1. Copy the order of people exactly, so nobody gets two turns in a row at the switchover.
  2. Set the start date to the next handover, not today.
  3. Re-create the overrides already booked: a substitute for each covered week.
  4. Check the time zone, especially for teams split across regions. One rotation per region is usually simpler than one clever one.

5. Run both for one full cycle

Keep Opsgenie running while the new setup goes through a full rotation, so everyone has had at least one handover on the new system. Compare the two each week. If they disagree about who is on call, find out why before you switch off the old one.

6. Update every place that points at Opsgenie

  • Runbooks and incident docs that say "check Opsgenie for who is on call"
  • Slack channel topics and pinned messages
  • Wiki pages, onboarding docs and the support team's escalation guide
  • Scripts and dashboards that read the on-call person from the Opsgenie API

The last one is easy to miss. If something reads who is on call automatically, point it at the new tool's API before switching off. RotaHog's REST API returns the current turn for any rotation.

7. Switch off, well before the deadline

Once a full cycle has gone smoothly, disable the Opsgenie integrations, then the schedules, then cancel. Aim to be done months before April 2027, not in the last week, when everyone else is migrating too and nobody has slack for surprises.

Want the schedule part done for you? RotaHog builds a fair rotation from a list of names, shares it as a link and reminds whoever is up. Free, no signup.

Make a on-call rotation

More guides