Idea Validation

Idea Validation

Guide founders from raw idea to validated hypothesis with confidence to commit or kill.

Workflow

1. Diagnose Stage

Ask: "Where are you in the journey?"

| Stage | Signals | Next Step | |-------|---------|-----------| | Raw idea | "I'm thinking about..." | → Go to Step 2 | | Hypothesis | Has problem + solution defined | → Go to Step 3 | | Interviewed | Talked to 5+ people | → Go to Step 4 | | Testing | Has MVP/landing page | → Go to Step 5 |

2. Shape the Hypothesis

Help founder articulate three components (Kevin Hale/YC framework):

1. Problem: What burning pain exists? Who has it?
2. Solution: What's the experiment/MVP?
3. Insight: Why will YOU win? (unfair advantage)

Unfair advantages to probe for:

• Founder-market fit (lived the problem, domain expertise)
• Market timing (20%+ annual growth)
• 10x better product possibility
• Built-in distribution/word-of-mouth
• Network effects potential

Output: One-sentence hypothesis in format: > "[Target customer] struggles with [problem] because [root cause]. We believe [solution] will work because [insight/unfair advantage]."

3. Validate the Problem

Quick Validation (60 min) — Tom Bilyeu Method

For rapid signal-checking before deep interviews:

Problem Hunt (15 min) Search Reddit, Twitter, Facebook groups, forums for:

• "Why is there no..."
• "I wish someone would make..."
• "I've tried 5 different solutions..."
• "Paying $X but it doesn't even..."

Collect 10 specific complaints. Can't find 10? Different problem.

Size Check (10 min) Google Trends + Keyword Planner:

• future predictions.

5. Test Before Building

Choose validation method based on idea type:

| Idea Type | Validation Method | |-----------|-------------------| | B2B SaaS | Sell on mockups/Figma before code | | Consumer app | Simulate populated experience in Figma | | Marketplace | Validate BOTH sides separately | | Dev tools | Leverage personal network first, then SEO content |

Nikita Bier principle: Do things that don't scale. Make it the BEST it could ever be manually. If it works, then automate. If you only launch the "scalable" version, you won't know if it failed due to concept or execution.

Landing page test (if applicable):

• Create page: "Finally, [solution] for [specific audience]"
• Run $20-50 in ads
• Conversion benchmarks:
• <1% = Kill immediately
• 2-5% = Test further
• 5%+ = Build immediately

6. Decision Checkpoint

GO Signals

• People reaching out unprompted
• Strong reactions in interviews ("I need this")
• Pre-sales or LOIs on mockups
• Clear pattern in 20+ conversations
• You have an unfair advantage

PIVOT Signals

• Adjacent problem keeps coming up
• Different customer segment more excited
• Core insight valid, solution wrong

KILL Signals

• Lukewarm reactions after 20+ interviews
• Have to "convince" people of the problem
• Can't find 10 complaints in 15 min searching
• No unfair advantage identified
• Market too small (<1K searches, <$1B TAM)

Remember: If you have to ask "Do I have product-market fit?" — you don't. When you have it, you know.

Common Traps

See references/traps.md for the 7 deadly traps and how to avoid them.

Reference Files

• references/frameworks.md — All 5 validation frameworks in detail
• references/case-studies.md — 8 founder stories (Vanta, Flexport, etc.)
• references/interview-guide.md — Question scripts + what to listen for
• references/traps.md — 7 traps with detection and avoidance

---

Adapted from Linas Beliūnas's The One-Person Unicorn founder skill set.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: samuelcastro
• Source: samuelcastro/startup-skills
• License: MIT

Install and usage instructions live in the source repository linked above.