Section 06 Validation Experiments & Hypotheses
Lean experiments to de-risk RecipeRoots assumptions on problem, solution, pricing, and channels. Total projected cost: $5,000. Timeline: 8 weeks.
🔴 Hypothesis Framework (10 Key Hypotheses)
Tested via interviews, landing pages, and prototypes. Precision-engineered for RecipeRoots' emotional heritage focus.
#1: Problem Existence (Recipe Loss) 🔴 Critical
We believe that adults 35-55 with aging parents
Will actively seek tools to preserve fading family recipes
If they recognize handwritten cards/stories are at risk.
We will know this is true when we see 65%+ of interviewees rank it top-3 pain + 6%+ landing page signups.
Details & Metrics
Risk: 🔴 Critical (core problem fails if invalid)
Evidence: Supporting: Google Trends ("family recipe preserve") up 40% YoY; Contradicting: None; Gaps: No primary data.
Experiment: 25 interviews + landing page (1k visitors).
Success:
| Metric | Fail | Min | Success |
|---|---|---|---|
| Interview confirmation | <45% | 45-65% | >65% |
| Landing signup | <3% | 3-6% | >6% |
Next if Validated: Solution tests. If Invalidated: Pivot to genealogy-only.
#2: Problem Existence (Story Loss) 🔴 Critical
We believe that family historians 35-55
Will prioritize capturing recipe stories/memories
If they value emotional heritage over just ingredients.
We will know this is true when we see 70%+ confirm stories as key loss in interviews.
Details & Metrics
Risk: 🔴 Critical
Evidence: Supporting: Ancestry.com 3B genealogy searches; Gaps: Recipe-specific.
Experiment: Interviews.
Success: 70%+ story priority.
Next if Validated: Story capture prototype. If Invalidated: Drop stories, focus digitization.
#3: Solution Fit (AI Photo-to-Recipe) 🔴 Critical
We believe that 35-55 adults
Will use AI to digitize handwritten recipes
If it extracts ingredients/steps accurately in seconds.
We will know this is true when we see 75%+ prototype satisfaction.
Details & Metrics
Risk: 🔴
Evidence: Supporting: OCR apps 80% accuracy; Gaps: Handwritten variance.
Experiment: Wizard of Oz.
Success: 75%+ useful rating.
#4: Solution Fit (Story Capture) 🟡 High
We believe that target users
Will complete guided story prompts
If prompts evoke memories easily.
We will know this is true when 60%+ add stories to prototypes.
Details
Risk: 🟡
Success: 60%+ completion.
#5: Willingness to Pay (Premium Sub) 🔴 Critical
We believe that engaged users
Will pay $7.99/mo for unlimited AI/print
If it saves 10+ hours vs manual.
We will know when 15+ pre-orders at price.
Details
Risk: 🔴
Success: $59/year viable.
#6: Pricing (Physical Books) 🟡 High
We believe that users
Will pay $40-80 for custom cookbooks
If high-quality heritage design.
We will know when 25%+ interest post-prototype.
Details
Risk: 🟡
Success: 25%+ buy-in.
#7-10: Channels & Secondary (Summary)
- #7 Channel (Emotional Ads): Facebook Mother's Day ads >8% CTR if "Interview Grandma now". Success: CAC <$10.
- #8 Channel (Genealogy): Ancestry forum posts yield 10% signup. Success: 200 leads.
- #9 Retention: 40% return for 2nd recipe. Success: NPS >40.
- #10 Sharing: 30% share to family. Success: Viral k>1.0.
Risk: 🟢 Medium. Evidence: Genealogy market $3B validates channels.
Experiment Catalog (12 Experiments)
| # Experiment | Hypothesis | Method | Timeline/Cost | Success Criteria |
|---|---|---|---|---|
| 1: Discovery Interviews | #1, #2 | 25 Zoom calls w/ 35-55 adults (LinkedIn recruit, $50 incentive) | 2w / $1,250 | ✅ 65%+ problem confirm | ❌ <45% |
| 2: Landing Page Test | #1-3 | Carrd page "Preserve Grandma's Secrets", FB/Google ads (1k visitors) | 2w / $800 | ✅ >6% signup | ❌ <3% |
| 3: Wizard of Oz MVP | #3-4 | Manual AI+human digitization/story capture for 15 users | 4w / $0 (time) | ✅ 75% satisfaction, NPS>30 | ❌ <60% |
| 4: Van Westendorp Pricing | #5-6 | Survey 100 users on price sensitivity | 1w / $300 | ✅ $7.99 optimal | ❌ <$5 |
| 5: Pre-Order Test | #5 | Stripe on landing for annual sub | 2w / $200 | ✅ 15+ orders | ❌ <5 |
| 6: Fake Door Features | #3,4 | Buttons for "Voice Record" / "Cookbook" on landing | 1w / $400 | ✅ >20% clicks | ❌ <10% |
| 7: Channel Ads Test | #7-8 | FB (Mother's Day), LinkedIn (genealogy), Reddit | 3w / $1,500 | ✅ CAC <$12 | ❌ >$25 |
| 8: Competitor Interviews | #1 | Why Paprika/Ancestry not enough? (20 users) | 2w / $500 | ✅ Gaps confirm moat |
| 9: Story Completion Test | #4 | Google Form prompts to 50 users | 1w / $100 | ✅ 60%+ full stories |
| 10: Referral Test | #10 | Email WoZ users "Share w/ family" incentive | 2w / $200 | ✅ k>1.1 |
| 11: Retention Email | #9 | Follow-up to WoZ: "Add another recipe?" | 1w / $0 | ✅ 40% open/click |
| 12: Cookbook Mockup | #6 | PDF preview, gauge interest | 2w / $300 | ✅ 25%+ buy intent |
Experiment Prioritization Matrix
| Experiment | Hyp | Impact | Effort | Risk if Skipped | Priority |
|---|---|---|---|---|---|
| 1: Interviews | #1,2 | 🔴 Critical | Med | Fail | 1 |
| 2: Landing | #1-3 | 🔴 | Low | Fail | 2 |
| 3: WoZ MVP | #3,4 | 🔴 | High | Fail | 3 |
| 4: Pricing | #5,6 | 🟡 High | Low | Subopt | 4 |
| 5: Pre-Order | #5 | 🔴 | Med | Fail | 5 |
| 7: Channels | #7,8 | 🟢 Med | Med | Ineff CAC | 6 |
8-Week Validation Sprint
Week 1-2: Problem | 3-4: Solution | 5-6: Pricing/Channels | 7-8: Decide. Total cost: $5K. Owner: Founder + VA.
Minimum Success Criteria (Go/No-Go)
| Category | Metric | Must Achieve | Nice-to-Have |
|---|---|---|---|
| Problem | Interview confirm | 65%+ | 80%+ |
| Problem | Landing signup | 6%+ | 10%+ |
| Solution | WoZ satisfaction | 75%+ | 85%+ |
| Solution | NPS | 35+ | 50+ |
| Pricing | Pre-orders | 15+ | 30+ |
| Pricing | Price acceptability | $7.99 | $9.99 |
| Overall | Hypotheses passed | 8/10 | 10/10 |
Go: All Must + clear path. Conditional: 80% metrics. No-Go: <80% (pivot/exit).
Pivot Triggers
- #1 Problem Weak: <45% confirm → Pivot: Pure genealogy integration.
- #2 Solution Fail: <60% satisfaction → Simplify: No AI, manual only.
- #3 Low WTP: <$5 → Freemium-only, ads revenue.
- #4 High CAC: >$20 → Organic: Blogs/churches.
Contingencies
Pivot budget: $2K extra. Re-test top adjacent: Cultural recipes for immigrants.
Experiment Documentation Template
Track in Notion/Airtable. Review weekly.
Total Validation Cost: $5,000 | Projected ROI: De-risk $400K build