Case study - Sanskrit coaching
A detailed case study applying the validation framework to the idea: Demand for Sanskrit coaching by parents for their kids in India. This is treated like a real validation experiment with concrete execution steps, numbers, and decision points.
Checklist
- Convert this into a real go-to-market plan
- Build a full operating system (agents + tools + workflows)
- Design a multi-agent system to run this business automatically
- Build a full product architecture (app + curriculum + AI tutor)
- Build a detailed 90-day execution plan with daily tasks
- Design the actual landing page + ads + scripts
- Build actual code + repo (LangGraph/CrewAI)
- Create SKILL.md files for each agent (production-ready)
- Design dashboard + automation flows (Zapier/Make)
Case study: Validating “Teach Sanskrit to kids (India)”
Step 1: Define the hypothesis
Hypothesis
Urban Indian parents (age 30–45) will pay ₹1,500–₹3,000/month for structured Sanskrit learning programs for their children (ages 6–12), focused on speaking, chanting, and cultural understanding.
Sub-hypotheses
- Parents value cultural connection, memory/cognitive benefits, and academic advantage
- Preferred format is online live classes (weekends)
- Kids engagement is higher with storytelling and chants than grammar-heavy teaching
Agentic AI used
- Research agent. Scraped YouTube comments on Sanskrit channels, Reddit parenting threads, and Amazon reviews of Sanskrit books.
- Output. Strong demand signals for “easy Sanskrit for kids”, “spoken Sanskrit”, and “cultural education”.
Step 2: Build a landing page (demand capture)
Built using
- Framer
Landing page structure
Headline: “Help your child speak Sanskrit in 30 days — fun, interactive classes rooted in Indian culture”
Sections
-
Benefits:
- Improves memory and pronunciation
- Builds cultural connection
- Enhances focus
-
Demo:
- Short video: kids chanting simple shlokas
- Conversational Sanskrit clips
-
Pricing:
- ₹1,999/month
- 8 live sessions
CTA
- Book free trial class
- Reserve seat (limited batch)
Agentic AI used
- Copywriting agent. Generated headline variants.
- UX agent. Suggested adding parent testimonials (mock initially) and highlighting limited seats.
Step 3: Simulate real demand
Setup
- Razorpay payment link integrated
-
Two flows:
- Free trial booking
- Paid seat reservation
Strategy
If a user pays, show: “Next batch starts soon — we’ll confirm your slot shortly”.
Results (week 1 test)
| Metric | Value |
|---|---|
| Visitors | 1,200 |
| Trial bookings | 96 (~8%) |
| Paid reservations | 21 (~1.75%) |
Insight: Strong validation signal — parents are willing to pay upfront without the full product.
Agentic AI used
- Funnel agent. Identified drop-off at the pricing page.
- Insight. Parents wanted curriculum clarity and teacher credibility.
Step 4: Traffic experiment
Channels tested
- Instagram ads (parents)
- YouTube shorts (kids chanting Sanskrit)
- WhatsApp groups (schools, communities)
Ad variants
Variant A: “Make your child speak Sanskrit in 30 days”
Variant B: “Improve your child’s memory and focus with Sanskrit”
Results
| Variant | CTR | Conversion |
|---|---|---|
| A | 2.1% | 1.2% |
| B | 3.4% | 2.3% |
Insight: Performance framing (memory and focus) performed better than language learning framing.
Agentic AI used
- Ad optimization agent. Shifted budget to Variant B based on early performance.
Step 5: Validation decision
Thresholds defined earlier
- Conversion > 2%: borderline overall (1.75%), but strong trial conversion (~8%)
- 20+ payments: met (21 achieved)
- Demand > capacity: starting to show
Decision
Proceed to a pilot batch.
Step 6: Pilot execution (service-first model)
Approach
Start with live classes (no tech build).
Batch
- 20 kids
- Weekend classes
- Duration: 4 weeks
Delivery
Zoom plus a WhatsApp group.
Agentic AI used
- Curriculum agent. Generated lesson plans for conversational Sanskrit, chants, and games.
- Engagement agent. Suggested quizzes and parent feedback loops.
Pilot results
| Metric | Value |
|---|---|
| Completion rate | 85% |
| Parent satisfaction | High |
| Referrals | 30% of batch |
Key insight
Parents valued engagement and kids’ enjoyment over academic rigor and depth.
Step 7: Scale strategy
Phase 1
- Add more batches
- Hire 2 instructors
Phase 2
- Record sessions
- Build a hybrid model
Phase 3
- App plus structured curriculum
- Gamified Sanskrit learning
Agentic AI used
- Operations agent. Created SOPs for instructors.
- Hiring agent. Defined screening criteria (Sanskrit fluency and teaching ability).
Final system view
Hypothesis → Landing Page → Ads → Payments → Pilot → Feedback → Scale
Key learnings
1. Parents will pay (validated)
The cultural plus cognitive angle works.
2. Positioning matters more than product
“Memory and focus” positioning performed better than “learn Sanskrit”.
3. Service-first is powerful
No upfront build cost and immediate validation through delivery.
4. Iteration is everything
Messaging changes significantly affected conversion.