POC v3 Playbook
Purpose: Production-ready. Approved infrastructure. Real data, real auth, real users.
Duration: 2 weeks
When to use: After POC v2 proves the product is useful.
Inputs Required
- Working POC v2 with user validation
- Feedback log from v2
- Infrastructure approval (or clear path to it)
- Access to real data sources
Principles
Focus: Is it ready for real use?
Customer-approved infrastructure. Now you care about where it runs. Get sign-off on cloud, hosting, data storage.
Real data. No more synthetic data. Connect to actual data sources. Handle real edge cases.
Real auth. Users can log in and out. Proper access controls. Session management.
Real security. Basic security review. No obvious vulnerabilities. Audit logging if required.
v3 = Production. This is the version that scales. v4+ continues from here.
Process
Week 1: Foundation
- Set up approved infrastructure
- Implement proper auth
- Connect real data sources
- Security basics (HTTPS, input validation, access controls)
Week 2: Polish & Launch
- Address critical v2 feedback
- Error handling and edge cases
- Basic monitoring/alerting
- User documentation (minimal, in-app if possible)
- Launch to broader user base
Outputs
- Production POC v3
- Users can log in, use it daily, rely on it
- Handoff documentation (if transferring to another team)
- Decision: continue scaling (v4+) or done?
Tips
- Don't rebuild everything. Port what works from v2, fix what doesn't.
- Infrastructure approval can block you. Start that conversation during v2.
- "Production" doesn't mean "perfect." It means "reliable enough for real use."
- If v3 scope is ballooning, you're adding features, not shipping production. Cut scope.
- v3 is where you pay down technical debt from v1/v2. Budget time for it.
After v3
v3 is production, but it's not the end. Scale continues with v4, v5, etc. Each version builds on the last.
The pattern holds: previous version = requirements for next version. Keep iterating.
See also: Sprint Planning for ongoing development