Design an incident comms playbook: audiences, cadences, and artifacts.
- User
Design an incident comms playbook: audiences, cadences, and artifacts.
- Assistant
Incident Communications Playbook
Purpose
This playbook outlines who to communicate with (audiences), how often (cadences), and what to share (artifacts) during incidents. It ensures transparency, reduces anxiety, and maintains trust.
Key Principles:
- Communicate early and often—under-communicate is worse than over-communicate.
- Use clear, concise language: What happened? Impact? Actions? ETA? Next update.
- Designate a Comms Lead (e.g., from IR or PR team) during the incident.
- Escalate if no update in 2x cadence interval.
- Post-incident: Retrospective includes comms review.
Severity Levels (assumed standard SRE model):
Severity Description Example SEV1 Critical business impact (e.g., full outage) Production down for all users SEV2 Major degradation (e.g., 50% users affected) Regional outage SEV3 Limited impact (e.g., single service degraded) Internal tool down Audiences
Tailor messages by audience. Always start with internal alignment before external.
Audience Priority Channels Notes Internal - Incident Response (IR) Team Highest Slack (#incidents), Zoom bridge Real-time tactical updates Internal - Affected Teams/Leadership High Slack channels, email digest, all-hands Slack Business impact, ETAs Internal - All Employees Medium Company-wide Slack/email, intranet High-level status only External - Customers High (for SEV1/2) Status page, customer email, support tickets Proactive if >10% affected External - Partners/Vendors Medium Dedicated email/Slack, status page If interdependent External - Media/Regulators Low (SEV1 only) PR team approval, press release Legal review required External - Public/Social Low Twitter/Status page Generic updates only Cadences
Timelines start at T=0 (incident declared). Adjust for time zones (UTC primary). Use a shared Google Sheet or incident tool (e.g., PagerDuty, Firehydrant) for tracking.
Severity Internal (IR/Teams) Internal (All-Employees) External (Customers/Partners) External (Public/Media) SEV1 Every 15 mins until mitigation Every 30 mins Every 30 mins (status page), hourly email Every 1 hour (if public-facing) SEV2 Every 30 mins Hourly Hourly (status page) Every 2 hours (status page only) SEV3 Hourly As needed (min 2x/day) As needed (status page) None Resolved Immediate "All Clear" Immediate Immediate + root cause in 24-48 hrs Immediate if applicable Milestones (all severities):
- T=0: "Incident declared."
- T+30 mins: Initial impact assessment.
- Mitigation: "Fix deployed."
- Resolved: "Monitoring post-fix."
- +24 hrs: Root cause summary.
Artifacts
Use templates below. Store in shared repo (e.g., GitHub, Notion). Customize per incident. Tools: Statuspage.io (external), Slack (internal), Google Docs (drafts).
1. Status Update Template (All Channels)
Incident: [Short Title, e.g., "Prod API Outage"] Severity: [SEV1/2/3] Status: [Identified | Mitigating | Monitoring | Resolved] Time: [UTC] Summary: [1-2 sentences on what/impact] Actions: [Bullet list of steps] ETA: [To next milestone or resolution] Next Update: [Specific time] Incident Cmdr: [@handle] | Comms Lead: [@handle]2. Channel-Specific Artifacts
Channel Artifact Examples Frequency Owner Slack (#incidents) Pinned status thread Per cadence IR Lead Zoom Bridge Verbal readouts Per internal cadence Incident Cmdr Email Digest (Internal) HTML summary w/ links Hourly (SEV1/2) Comms Lead Status Page Automated + manual updates Per external cadence Comms Lead Customer Email Personalized blast (via Customer.io) Hourly (SEV1/2) Customer Success Twitter/LinkedIn Tweet thread Every 2+ hours (SEV1) Social Lead (PR approved) Press Release Full doc (Google Doc) Post-resolution (SEV1) PR Post-Mortem Public/ internal report 48-72 hrs post-resolve Engineering + Comms 3. Escalation & Handoffs
- No update? Escalate to leadership in 2x cadence.
- Legal/Regulatory: Flag sensitive incidents (e.g., data breach) immediately.
- Templates Repo: [Link to your shared drive/Notion page].
Quick-Start Checklist
- Declare incident → Assign Comms Lead.
- Post T=0 update to #incidents.
- Set calendar reminders for cadences.
- Mirror updates to status page.
- Resolved → Send all-clear + schedule post-mortem.
This playbook is living—review after every SEV1. Questions? Ping #comms or the Comms Lead.
Designed by Sherlock, Trailblazer Labs
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.