We've identified that Incidents are the hardest event type to ingest into Pulse, and that as a consequence, most users don't have a way of measuring the Time to recover and Change failure rate Accelerate metrics.
To help with this, we'll detect Incidents every time you revert a pull request. Git providers offer native support for this operation through their UIs. We'll also be able to understand which pull request is being reverted, giving us a way to infer how long the Incident took to calculate the Time to recover.

Pulse can also detect Incidents from arbitrary pull requests. You can configure a prefix that identifies a branch name or pull request title to match the naming convention of what is a rollback or hotfix in your workflow.

For users that are already collecting Incidents, this feature will be disabled by default, but you can opt-in at any time.