RecipeRoots - Family Recipe Preservation

Model: x-ai/grok-4.1-fast
Status: Completed
Cost: $0.053
Tokens: 147,932
Started: 2026-01-03 22:33

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:

MetricFailMinSuccess
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

ExperimentHypImpactEffortRisk if SkippedPriority
1: Interviews#1,2🔴 CriticalMedFail1
2: Landing#1-3🔴LowFail2
3: WoZ MVP#3,4🔴HighFail3
4: Pricing#5,6🟡 HighLowSubopt4
5: Pre-Order#5🔴MedFail5
7: Channels#7,8🟢 MedMedIneff CAC6

8-Week Validation Sprint

Wk1
Wk2
Wk3
Wk4
Wk5
Wk6
Wk7
Wk8
Interviews/Landing
WoZ MVP
Pricing + Channels
Go/No-Go

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)

CategoryMetricMust AchieveNice-to-Have
ProblemInterview confirm65%+80%+
ProblemLanding signup6%+10%+
SolutionWoZ satisfaction75%+85%+
SolutionNPS35+50+
PricingPre-orders15+30+
PricingPrice acceptability$7.99$9.99
OverallHypotheses passed8/1010/10

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

## Experiment: [Name] **Date:** [Start - End] **Hypothesis:** #X ### Setup - What we did - Sample: N - Tools: [e.g. Carrd, Zoom] - Cost: $X ### Results | Metric | Target | Actual | Pass/Fail | |--------|--------|--------|-----------| ### Key Learnings - Insight #1 - Quote: "..." ### Next Steps - Product change - Follow-up exp

Track in Notion/Airtable. Review weekly.

Total Validation Cost: $5,000 | Projected ROI: De-risk $400K build