The 3-Month
ConsumerTech Sprint
A program that compresses three years of founder pain into ninety days — with the safety rails of a school and the urgency of a seed-stage company.
The classroom is the market.
The textbook is the live dashboard. The exam is whether a stranger will use what was shipped this week.
Problem-solving is the only durable skill
Ideas die, markets shift. The founder who decomposes a scary problem into small solvable ones wins in any decade.
Real numbers beat real theory
Every concept is taught only against a live number in the student's own venture. No abstract cases.
Speed is a teacher
A team that ships weekly learns 12× faster than one that ships quarterly. The cadence itself is our biggest intervention.
We don't grade Day-1 idea quality. We grade the slope of the learning curve — how fast a team turns ignorance into evidence, and evidence into traction.
Break it until it bleeds.
We never let teams fall in love with a solution. They shatter a scary ambition into testable sub-problems — each a one-week experiment.
A wish becomes seven answerable questions. See the weekly cadence ↗
The Forge Loop
The exact loop a real seed-stage startup runs — repeated 12 times. We build the muscle, not the diagram.
Strip the problem to physics until it's solvable.
Cheapest experiment that produces a number wins.
Revenue = users × conversion × price × frequency. Fix the weakest.
Write the launch press release first. No excitement → don't build.
Three months, three jobs.
Each month ends in a hard gate — not a grade. Sample week ↗ · Demo Day rubric ↗
Discover
Earn the right to build
Core Q: Is this a real, urgent, monetizable problem?
- ✓ 50+ customer conversations
- ✓ Sharp ICP + painful job-to-be-done
- ✓ Demand evidence (waitlist / pre-order / LOI)
- ✓ Bottom-up TAM/SAM/SOM
- ✓ MVP scope locked to one hero use case
Build & Launch
Ship embarrassing, ship fast
Core Q: Will strangers use what we built?
- ✓ Working MVP in users' hands
- ✓ Public launch on a real channel
- ✓ Instrumentation live (activation, retention)
- ✓ First 100–500 real users
- ✓ First monetization experiment
Scale & Prove
Find the loop, prove the economics
Core Q: Can we grow repeatably and charge?
- ✓ One growth loop instrumented
- ✓ Retention improving cohort-over-cohort
- ✓ End-to-end unit economics modelled
- ✓ Monetization validated
- ✓ Demo Day to investors & operators
The 12-week clock (+ Week 0 onboarding)
Taught through their own numbers.
Modules unlock exactly when a team needs the concept to solve a live problem — never as standalone lectures.
TAM · SAM · SOM — why / what / how
Bottom-up only. A huge population is not a huge market — we drill the India-1 / 2 / 3 reality. Worked example ↗
A credible SOM you can win beats a fantasy TAM you can't.
End-to-end P&L & unit economics
Every team runs a live P&L from Week 5 and fixes the single weakest driver. Full template ↗
💰 Pricing in a price-sensitive market
recurring
commission
volume
Low absolute WTP, high sensitivity — but UPI makes micro-transactions frictionless. Frequency & volume often beat margin.
✂️ Cost-cutting & jugaad ops
- • AI-native + no-code → 3 people do what needed 15
- • Community & organic before paid (paid CAC kills consumer in India)
- • Pre-sell before you produce — variable over fixed cost
- • Runway = days of learning bought. Spend on evidence, not vanity.
Two scoreboards, always on.
We track the venture's health and the student's growth in parallel — and obsess over leading indicators.
- • DAU / core action
- • Signups + activation
- • User convos logged
- • Cash & CAC by channel
- • WAU + WoW growth %
- • D1 / D7 retention
- • Activation rate
- • Experiments → decisions
- • M1: validation score
- • M2: 100–500 users, D7
- • M3: loop + LTV/CAC
- • 10k users / ₹25L ARR
- • breakout / funding
- • operator placements
- • capability delta
Cohort retention should improve every launch
Each new cohort retains better than the last — the signal of real learning.
Weekly active users — week-over-week
Consistent positive WoW growth is the best early signal of fit.
Anti-gaming rule: "active user" and "revenue" have one program-wide definition, audited at every gate. Buy installs or log ghost users → you fail the gate. We reward honest retention over vanity reach.
Deliberate trade-offs vs top accelerators.
Where we align with YC / Antler / Entrepreneur First / Techstars — and where we choose to differ.
| Choice | We chose | Trade-off accepted | vs accelerators |
|---|---|---|---|
| Cohort & time-box | Fixed 3-mo, hard gates | Some need more/less time | aligns YC/Techstars |
| Founder stage | Freshers, pre-idea | Higher idea/team risk | closer to Antler/EF |
| Economics | School fee, no equity | Less aligned upside | differs from YC 7%/$500k |
| Curriculum | Just-in-time, structured | More teaching overhead | more than YC |
| Targets | Prescriptive (10k, ₹25L) | Risk of teaching to metric | more prescriptive |
| Geography | India-first, on-ground | Less global network | vs SV-remote |
Designing for the whole distribution
Accelerators ignore the bottom of the cohort. We can't — so from Week 8 we run two explicit landing pathways. Same rigor, different finish line; nobody graduates feeling like a failed founder.
funding / accelerator / Shark Tank
costed modernization plan
₹12–14 LPA roles
can model any business
Nine capabilities the program must supply.
Each with one ecosystem partner (examples). Partner sources ↗
AI build velocity
Ship weekly with a tiny team
Product analytics
Read retention honestly
Growth & distribution
#1 thing teams fail at
Payments / monetization
Usage → revenue fast
India channels (WhatsApp)
Reach India-2/3
Legal / setup / compliance
Incorporate, equity, IP
Investor access
Warm capital for breakouts
Mentor & operator network
Real scar tissue
Founder wellbeing
Burnout is a top failure
Where it breaks — and how we catch it.
Weak idea / loving the solution
→ Validation gate · kill-criteria · budgeted pivot week · mentor red-teams
Co-founder conflict / breakups
→ Founder Contract (W0) · role clarity · vesting analog · mediation
Vanity / gamed metrics
→ One metric definition · cohort-retention focus · anti-gaming audits
Distribution failure
→ Distribution-first curriculum · channel partners · 100 real users first
Burnout / mental health
→ Sustainable cadence · weekly wellbeing pulse · coaching · protected rest
Middle 60% feels like failures
→ Operator & family-business tracks · portfolio of work · employer pipeline
Mentor quality variance
→ Mentor SLAs · structured office-hours · mentor NPS · vetted bench
Working-capital trap (commerce)
→ Pre-sell / made-to-order · micro-grants · asset-light wedges first
Over-engineering / slow ship
→ Ship-weekly rule · AI tooling · "embarrassing MVP" norm
Job-transition stigma (60%)
→ Reframe operator as elite · employer network · comp benchmarking
How I'd teardown a startup idea.
One reusable lens — the 6-Lens Teardown — applied live to Willow & Studojo. Each lens is scored 0–5 by evaluator judgment, grounded in cited market facts. Method, scores & sources ↗
intensity
& wedge
economics
👗 Willow
apparel for tall womenRadar = evaluator score, 0–5 per lens (judgment, not survey). See each score & its source ↗
Verdict: Real painkiller, great niche wedge — but apparel returns (25–40%) & working capital are the killers.
🎓 Studojo
AI student co-pilotRadar = evaluator score, 0–5 per lens (judgment, not survey). See each score & its source ↗
Verdict: Too broad (4 products in 1), low student WTP, heavy ChatGPT substitution. The wedge must be one acute job.
If a student has no idea — three I'd recommend today
Not random bets — each is a wedge, not a platform (Section 02 discipline), with a clear payer and a real Indian tailwind. Market figures cited. Sources ↗
Vernacular voice-AI tutor for outcomes
spoken English & competitive exams
What it is: a voice-first tutor that talks in the learner's mother tongue and drills spoken English or one exam (SSC / banking / state PSC) through daily speaking practice + instant correction — not video lectures.
Who pays: aspirational parents already spending on coaching; priced on the outcome (fluency, a cleared exam, a job), e.g. pay-after-result.
The wedge: "speak English with confidence in 60 days" for one job-seeking segment — then expand exam by exam.
WhatsApp AI commerce assistant
for small Indian sellers
What it is: an AI agent that lives inside WhatsApp — builds the seller's catalog, answers buyer questions, takes orders, sends UPI payment links, and chases reorders. No new app to learn.
Who pays: the seller — a small subscription or per-order fee, paid where they already run the business.
The wedge: own ONE seller vertical first (home bakers, boutique resellers) where catalog + reorders repeat — then go horizontal.
AI health-insurance claims copilot
for India's newly-insured
What it is: an assistant that reads a family's health policy, says in plain language what's actually covered, and walks them through filing a cashless or reimbursement claim with the right documents.
Who pays: B2B2C — an insurer, hospital or TPA wanting fewer rejected claims & happier patients; or a family subscription.
The wedge: one insurer's or one hospital network's patients first, where claim volume is dense.
Build the engine in India. Swap the fuel per market.
🇮🇳 Designed for India
- • Distribution via WhatsApp, campus & creators — not paid ads
- • Pricing for low WTP; UPI micro-transactions, frequency, B2B2C
- • Capital-efficient jugaad ops; pre-sell before producing
- • India-1/2/3 segmentation — population ≠ market
- • On-ground execution, Indian mentors & investors
🌍 To go global, fix
What's portable: the pedagogy — problem decomposition, the Forge Loop, evidence-based gates. What localizes: distribution, pricing, regulation, network.
Ninety days. Real users. Real numbers.
From idea discovery to product-market fit — with the rigor of a school and the urgency of a startup.