Playbook: Open an Incident Huddle

Spin up a cross-functional incident huddle for a live production issue. Run this whenever an incident needs synchronous discussion across Engineering, Product, and MX, or when an outside team (for example Google Play Console) has paged the company.

Prerequisites

You need the following before you start:

  • Slack access to the #help-eng-open channel.

  • Permission to start a workflow in #help-eng-open.

  • The Jira incident ticket — create it first if one does not exist.

Steps

  1. Post a single top-level message in #help-eng-open with the incident ticket, a one-line description, and an @here so on-call sees it.

  2. Start a Slack huddle from the same top-level message so the huddle is anchored to the incident thread.

  3. Move all follow-up discussion into the message’s thread to keep the main channel scannable.

  4. Pull in the responders the incident requires:

    • Executives from Engineering, Product, and MX for company-wide pages.

    • Domain consultants for the affected surface — for example the mobile leads for Flutter incidents.

    • Intake when the report came from an external partner or user.

  5. Capture decisions, owners, and ETAs.

  6. Once root cause and remediation are agreed, post a summary back to the top-level #help-eng-open message so anyone outside the huddle can catch up.

Verification

  1. The incident ticket links to the Slack thread.

  2. The thread contains a decision summary with owners and ETAs.

  3. The main #help-eng-open channel has a single top-level message per incident, not scattered discussion.

Rollback / Undo

This procedure has no rollback — you cannot un-page a huddle. If a huddle was started by mistake, post a correction in the thread and close the huddle.