Prompt
You are a launch coordination expert helping me create a runbook for launch day. I need a detailed plan that ensures nothing is missed and everyone knows their role.
Launch Context:
- What's launching: [Product/feature name]
- Launch date/time: [When]
- Time zones involved: [Team locations]
- Launch type: [Big bang, phased rollout, soft launch]
- Risk level: [High, medium, low]
- Rollback plan: [Yes/No, details]
Please create a comprehensive launch day runbook:
1. Launch Overview
Launch Summary:
- What: [One sentence description]
- When: [Date, time, timezone]
- Who's affected: [Users, percentage]
- Success metrics: [What we're watching]
- War room: [Link to Slack channel/Zoom]
Go/No-Go Criteria:
? [Criterion 1]: [Owner, status]
? [Criterion 2]: [Owner, status]
? [Criterion 3]: [Owner, status]
Final go/no-go decision: [Time, who decides]
2. Team & Roles
| Role | Name | Responsibilities | Contact |
|------|------|-----------------|---------|
| Launch Lead | [Name] | Overall coordination, decisions | [Phone/Slack] |
| Engineering | [Name] | Deploy, monitor, rollback | [Phone/Slack] |
| QA | [Name] | Smoke test, issue verification | [Phone/Slack] |
| Product | [Name] | Go/no-go, comms approval | [Phone/Slack] |
| Support | [Name] | Customer issues, escalation | [Phone/Slack] |
| Marketing | [Name] | External communications | [Phone/Slack] |
Escalation path:
1. First: [Who to contact first]
2. Second: [If first unavailable]
3. Emergency: [Executive contact]
3. Pre-Launch Checklist (T-24 hours)
Engineering:
? Code frozen and deployed to staging
? All tests passing
? Monitoring and alerts configured
? Rollback procedure documented and tested
? Database migrations ready
Product:
? Feature flags configured
? Launch announcement drafted
? Support documentation ready
? FAQ prepared
Marketing:
? Communications scheduled
? Social posts ready
? Email campaigns queued
? Press release ready (if applicable)
Support:
? Team briefed on new feature
? Known issues documented
? Escalation path confirmed
? Extra coverage scheduled
4. Launch Day Timeline
T-2 hours: [Time]
? All hands in war room
? Final go/no-go check
? Verify monitoring dashboards
? Confirm all teams ready
Owner: [Name]
T-1 hour: [Time]
? Pre-launch smoke test
? Verify staging matches expected state
? Communications team ready
? Support team ready
Owner: [Name]
T-0: Launch! [Time]
? Deploy to production
? Verify deployment success
? Run smoke tests on production
? Feature flag enabled (if applicable)
Owner: [Name]
T+15 min: [Time]
? Initial metrics check
? Error rate check
? Performance check
? First user feedback
Owner: [Name]
T+30 min: [Time]
? Send internal launch announcement
? Begin external communications (if applicable)
? Monitor support channels
Owner: [Name]
T+1 hour: [Time]
? Comprehensive metrics review
? Support ticket review
? Go/no-go for continued rollout
Owner: [Name]
T+4 hours: [Time]
? Extended metrics review
? Initial launch report
? Identify any issues for immediate fix
Owner: [Name]
EOD: [Time]
? End of day launch summary
? On-call handoff (if needed)
? Next day priorities
Owner: [Name]
5. Monitoring Checklist
Metrics to watch:
? Error rate: Baseline [X], Alert threshold [Y]
? Latency: Baseline [X], Alert threshold [Y]
? [Key metric 1]: Baseline [X], Alert threshold [Y]
? [Key metric 2]: Baseline [X], Alert threshold [Y]
Dashboards:
- Primary: [Link]
- Engineering: [Link]
- Business metrics: [Link]
Alert channels:
- PagerDuty/Opsgenie: [Link]
- Slack alerts: [Channel]
6. Issue Response Plan
Severity Definitions:
- P0 (Critical): Complete outage, data loss, security breach
- P1 (High): Major functionality broken, significant user impact
- P2 (Medium): Feature degraded, workaround exists
- P3 (Low): Minor issue, cosmetic
Response by severity:
P0/P1:
1. Alert launch lead immediately
2. Begin investigation
3. Consider rollback (decision within 15 min)
4. Notify stakeholders
5. Post-incident review
P2:
1. Log and triage
2. Assign owner
3. Fix in next deploy or hotfix
4. Communicate if user-facing
P3:
1. Log for backlog
2. Address in normal cycle
7. Rollback Plan
Rollback triggers:
? Error rate > [threshold]
? P0/P1 issue cannot be fixed within [time]
? [Specific condition]
Rollback procedure:
1. [Step 1]
2. [Step 2]
3. [Step 3]
Rollback owner: [Name]
Rollback decision maker: [Name]
Expected rollback time: [Duration]
8. Communication Templates
Internal launch announcement:
"[Feature] is now live! [1-2 sentence summary]. See [link] for details. Questions? Ask in [channel]."
External announcement:
[Draft for customers/public]
Issue communication:
"We're aware of [issue] affecting [users]. Our team is investigating. Updates in [channel]."
Rollback communication:
"We've temporarily rolled back [feature] due to [reason]. We'll provide an update on re-launch timing within [timeframe]."How to use
- 01
Start planning 1-2 weeks before launch
- 02
Replace placeholders with your specific launch details
- 03
Assign owners to each role and checklist item
- 04
Review timeline with all teams involved
- 05
Verify all links and contacts are correct
- 06
Do a dry run of critical procedures (especially rollback)
- 07
Print or have easily accessible during launch
Tips
- Practice the rollback procedure before launch day
- Have phone numbers, not just Slack - systems can go down
- Time zone coordination is critical - use UTC in runbook
- Keep the war room focused - side conversations elsewhere
- Celebrate wins, but stay vigilant until EOD
- Document everything for post-launch review
- The most important skill is knowing when to rollback