Progressive Delivery & Advanced Deployment
Canary deployments, GitOps, progressive traffic shifting, and blue-green deployment automation.
Job to be done: When we need to release a critical feature, I want to deploy it behind flags with a gradual rollout strategy, so I can catch issues in production with a subset of users before they affect everyone.
You will instrument canary and blue-green rollouts with feature flags, then implement policy-driven gates that allow you to ship multiple times per day with automated rollback if metrics degrade.
What you’ll implement
These are the roadmap epic features, organized as a starter backlog.
Execution guide
Practical guidance aligned to the Execution Kit Definition of Done.
Outcome
Rollouts are safer and faster via feature flags, canaries/blue-green, and policy-driven approvals.
Before to After Transformation
Features go live immediately on deploy, no gradual rollout, difficult to disable broken features
# Before: Feature = Deployment
Release process:
1. Merge feature to main branch
2. Deploy to production
3. Feature immediately visible to ALL users
4. If feature breaks: Emergency rollback required
5. Rollback affects ALL recent changes
Problems:
- Can't test in production with subset of users
- Can't disable feature without rollback
- Large blast radius for every change
- Fear of releasing on Fridays
Example incident:
- New checkout flow deployed at 2 PM
- Bug discovered affecting 10% of transactions
- Only option: Full rollback (30 minutes)
- Rollback also removes yesterday's bug fix
- Total impact: 50,000 affected usersDeploy anytime, enable features gradually, instant kill-switch for broken features
# After: Progressive feature rollout
Release process:
1. Deploy feature behind flag (OFF by default)
2. Enable for 5% of users (canary)
3. Monitor metrics for 1 hour
4. Gradually increase: 10% to 25% to 50% to 100%
5. If issues: Disable flag instantly (no deployment)
Benefits:
- Test in production safely
- Instant kill-switch (< 1 second)
- Targeted rollout (by region, user tier, etc.)
- Deploy 10x per day without fear
Example incident (now):
- New checkout flow deployed at 2 PM (flag OFF)
- Enable for 5% of users at 3 PM
- Bug discovered affecting 1% (500 users)
- Disable flag instantly
- Fix deployed in 10 minutes
- Re-enable for 5% to verify fix
- Total impact: 500 users (100x reduction)Symptoms
Prerequisites
Implementation steps
- Pick a feature flag system and conventions
- Define canary/blue-green rollout path
- Update release checklist
- Add policy checks (freeze windows, consent/data residency)
- Instrument key journeys
- Run first canary
- Expand to 3–5 services
- Standardize flag lifecycle (cleanup)
- Document playbooks
Definition of Done
- Teams can ship features behind flags
- At least one canary/blue-green rollout is successful
- Policies are documented and applied
- Practice integrated into team workflow
- Practice integrated into team workflow
Metrics
- % rollouts using progressive delivery
- Time to disable a feature
- Change failure rate
- Deployment frequency
Failure modes
Ownership
- Provide flag/rollout tooling
- Maintain policy packs
- Define rollout criteria
- Own flag lifecycle
What good looks like (by org scale)
- Feature flags for risky changes
- Canary rollout for critical services
- Country-aware targeting + policy-driven gates
References
Resources
Templates and related materials for this kit.
Related capabilities
Capabilities tracked under this epic in the roadmap.
- Automated Canary Deployments>= 70% of production deployments use canary strategy (10% traffic -> 50% -> 100%) with automated rollback on error spike.
- GitOps Deployment Model>= 80% of deployments managed via GitOps (ArgoCD, Flux) with Git as single source of truth for desired state.
- Automated Progressive Traffic Shifting>= 70% of deployments use automated traffic shifting with SLO monitoring at each phase (Flagger, Argo Rollouts).
- Blue-Green Deployment Automation>= 60% of stateless services support instant blue-green cutover with automated traffic switching and rollback capability.
Related kits
Other kits in the same milestone or with similar DORA impact.