Smithery is now a part of Arcade.dev! Read more in our announcement here
Rapid MVP development and feature prioritization for fast product validation...
Ship MVPs in 1-2 weeks, not months.
Start with the smallest thing that proves/disproves your riskiest assumption.
MVP = Minimum Viable Product, not Minimum Pretty Product.
Categorize every feature idea using this matrix:
| Category | Definition | Action |
|---|---|---|
| P0 | Must have for core value proposition | Build now |
| P1 | Important but can wait for v1.1 | Ship after validation |
| P2 | Nice to have, delight features | Ship v2+ |
| Out of Scope | Not needed for validation | Defer indefinitely |
P0 (Week 1):
P1 (Post-validation):
P2 (Future):
Out of Scope:
Definition: Manually deliver the service before building automation
When to use: Core value is service delivery, not technology
Examples:
Time: 1-3 days | Validates: People want the service
Definition: Interface looks automated, but humans operate it behind the scenes
When to use: Automation is expensive/complex to build
Examples:
Time: 1 week | Validates: Users engage with interface
Definition: Explain the product + collect emails (no product yet)
When to use: Testing demand before building anything
Examples:
Time: 1-2 days | Validates: People sign up for waitlist
Definition: One feature that solves one problem
When to use: One feature delivers 80% of value
Examples:
Time: 1-2 weeks | Validates: People use the core feature
Definition: Combine existing tools instead of building custom
When to use: You can cobble together a solution with existing tools
Examples:
Time: 2-5 days | Validates: Workflow works end-to-end
Is this a web app?
├─ YES
│ ├─ Need real-time?
│ │ ├─ YES → Next.js + Supabase + WebSockets
│ │ └─ NO → Remix + PostgreSQL
│ └─ Simple CRUD? → Supabase + React
└─ NO
├─ Mobile app? → React Native + Expo
├─ API only? → Express + PostgreSQL
├─ Chrome extension? → Vanilla JS + Chrome APIs
└─ CLI tool? → Node.js or Python
Recommended MVP Stack:
Speed-Focused Tools:
❌ Building features "just in case" → Build when 3+ users request ❌ Perfect design before launch → Ship functional, ugly is fine ❌ Over-engineering architecture → Monolith is fine for MVP ❌ Custom authentication → Use Clerk, Auth0, or Supabase Auth ❌ Building admin panel first → Use database GUI (Retool/Supabase) ❌ Mobile app before web → Web first, always ❌ Scaling for 1M users → Build for 10 users, scale when needed
Days 1-2: Design + Setup
Days 3-4: Build
Day 5: Polish + Deploy
Days 1-2: Iterate
Days 3-4: Get Users
Day 5: Analyze & Decide
Decision Point:
Pre-Launch (1 hour):
Launch Day:
Week 1 Post-Launch:
MVP validation criteria:
If you hit these: You might have something. Build P1 features.
If you don't: Pivot or move on. Don't invest months in something nobody wants.
Successful minimal starts:
All started much smaller than you think.
Related Skills:
product-strategist - For validating product-market fitfrontend-builder - For building the UIapi-designer - For backend API designdeployment-advisor - For deployment decisionsgo-to-market-planner - For launch strategyRelated Patterns:
META/DECISION-FRAMEWORK.md - Platform selection decision treeRelated Playbooks:
PLAYBOOKS/build-mvp.md - Step-by-step MVP build procedure (when created)For experienced users:
