Problem & Users
Who hurts, and how
Two groups with adjacent unmet needs โ users who want home food without cooking, and cooks who want flexible income without full-time domestic work.
The core tension
Urban India's Gen Z eats Zomato daily but feels the guilt. They know home food is healthier, cheaper, and more satisfying โ they just don't want to cook it. Meanwhile, millions of skilled home cooks (mostly women) have no flexible platform to earn from their biggest skill. The gap is a matchmaking problem, not a cooking problem.
User personas
๐งโ๐ป
Arjun, 26
Software engineer ยท Koramangala, Bangalore
Primary user
- Spends โน800โ1,200/day on Zomato, feels bad about it
- Wants dal-chawal but can't cook, doesn't have time
- Tried hiring a full-time cook โ too expensive, too much commitment
- Values health, freshness, and knowing what's in his food
๐ฉ
Priya, 29
Working couple ยท Powai, Mumbai
Primary user
- Both partners work; nobody has time to cook on weeknights
- Have a toddler โ restaurant food feels risky and expensive
- Wants a consistent cook who knows their preferences
- Willing to subscribe if it solves the weekly meal problem
๐ฉโ๐ณ
Meera, 42
Home cook ยท Koramangala, Bangalore
Supply side
- 15 years of cooking, known for her Punjabi food
- Currently does full-time domestic help โ โน8,000/month fixed
- Wants flexibility to work near home, pick her hours
- No platform that respects her skill or pays fairly for it
๐ฉ
Sunita, 35
Part-time cook ยท Bandra, Mumbai
Supply side
- Does 2โ3 homes a day, earns โน400โ500/day inconsistently
- No visibility into upcoming bookings, no way to manage her schedule
- Wants to earn more from fewer, better-paying jobs
- Worried about safety when going to unknown homes
Market signals
400M+
Urban Indians who order food online
$8B
India food delivery market 2024
60M+
Gig-eligible home cooks in India
20%
YoY growth, online food services
โน0
On-demand home cook platforms today
72%
Gen Z prefer home food to restaurant food*
*Indicative โ based on market research trends, not primary research
Solution & Positioning
What Rasoi is โ and isn't
Rasoi is not a food delivery app. It is not a domestic staffing agency. It is a new category: on-demand cooking-as-a-service, priced by time, delivered at your home.
How it works
๐ฑ
Open Rasoi
Browse cooks near you
๐ฉโ๐ณ
Pick a cook
By cuisine expertise
โฑ๏ธ
Choose duration
1hr / 2hr / 3hr
๐
Tell them what you want
Open request, any dish
๐
Cook arrives
In ~40 min
๐ฝ๏ธ
Fresh meal, done
Your kitchen, your food
Competitive landscape
| Attribute |
Zomato |
Urban Company |
Full-time cook |
Rasoi โฆ |
| Fresh, home-cooked food | โ | โ | โ | โ |
| On-demand (book today) | โ | โ | โ | โ |
| Affordable for a family of 4 | โ | โ | โ | โ |
| Cooked at your home | โ | โ | โ | โ |
| No long-term commitment | โ | โ | โ | โ |
| Worker protections built in | โ | โ | โ | โ |
| Subscription option | โ | โ | โ | โ |
The key insight
Zomato and Rasoi solve the same underlying need (not wanting to cook) but in entirely different ways. Zomato is a food product platform; Rasoi is a service platform. The comparison users will make is "2 people on Zomato (โน400โ500/meal) vs Rasoi 2hr cook for 2 (โน419 all-in, multiple dishes, fresh)." Rasoi wins on value, freshness, and the feeling of real home food โ categories Zomato can't compete in.
Feature Analysis
Every feature has a reason
Each product decision was made to solve a specific tension โ between user needs, cook needs, and platform economics. This is the thinking behind each one.
1
Time-based pricing
1hr โน249 ยท 2hr โน399 ยท 3hr โน549 (travel included)
Core
Instead of pricing per dish, Rasoi charges by session duration. The cook's contract is unambiguous โ they cook for the agreed time, within the user's request. No disagreements about what counts as "one dish."
Product thinking: Dish-based pricing creates infinite edge cases ("is a dal + roti one dish or two?"). Time is objective and enforced by the platform's tracking. This also mirrors how every other service category works in India โ tutors, maids, masseurs โ making the mental model instantly familiar to both sides.
2
People count selector
1 ยท 2 ยท 3โ4 ยท 5โ6 ยท 7+ with add-on pricing
Core
Cooking for 7 people requires more ingredients, more effort, and more skill than cooking for 1. Add-ons (โน0 for 1โ2 people, up to โน220 for 7+) make the solo/couple use case very affordable while ensuring cooks are paid fairly for the actual work of a larger meal.
Product thinking: Flat pricing for all group sizes either overcharges solo users (they won't book) or undercharges large groups (cooks feel exploited and churn). Transparent, additive pricing builds trust on both sides. The breakdown shown to users โ base + people add-on + travel + convenience โ signals platform honesty rather than hidden fees.
3
Open request field
"What would you like to eat?" freeform input
Core
Instead of picking from a menu, users describe what they want in plain language. "Dal makhani, roti, and something sweet for 4 people, medium spice." The cook interprets and executes.
Product thinking: Menu-based booking sounds user-friendly but creates operational problems โ cooks can't update menus in real-time, users face paradox of choice, and dish expectations mismatch reality. An open request mirrors how you'd talk to any real cook: "make something nice for dinner." Paradoxically, removing the menu increases user confidence because "cook anything I want" is a stronger value prop than "choose from 25 dishes."
4
Cuisine type tags
North Indian ยท South Indian ยท Punjabi ยท Bengali โ not dish lists
Discovery
Cook profiles show cuisine expertise types rather than specific dishes. Users know that a "North Indian cook" implies dal, roti, chole, rajma, paneer dishes โ implicitly and without being told.
Product thinking: Dish lists are cognitively overwhelming and create a false promise of specificity that leads to unmet expectations. Cuisine types tap into India's deeply ingrained regional food identity โ people know immediately what a "Kerala cook" can make. This reduces cognitive load, speeds up matching, and sets realistic expectations on both sides.
5
Worker protection model
Max 3 jobs/day ยท 6hr cap ยท 30 min buffer ยท Travel paid 100%
Supply quality
Platform-enforced limits protect cooks from overwork. A tired cook makes bad food. Worker protections are not just ethical โ they are a direct lever on quality consistency and cook retention.
Product thinking: Gig platforms that extract maximum labor from workers face supply burnout, quality degradation, and eventual PR crises (Urban Company and Zomato delivery partners have both faced this). Rasoi's protections are displayed prominently โ both to cooks (as benefits) and users (as quality guarantees). Cook dignity becomes a marketing claim. Sustainable supply = sustainable platform.
6
Real-time tracking
6-step progress: Confirmed โ Ready โ On way โ Arrived โ Cooking โ Done
Trust
"When will they arrive?" is the highest-anxiety moment in any on-demand service. Live tracking with animated steps and countdown ETA reduces this anxiety and signals platform quality.
Product thinking: Borrowed deliberately from Zomato and Ola's playbook because it works. Trust in an on-demand service is built through visibility. A user who can see the cook is "2.3 km away and heading toward you" is 60% less likely to call support and 40% less likely to cancel. The tracking experience is as important as the food itself for first-time users.
7
Rasoi Pass (subscriptions)
Starter โน499 ยท Regular โน899 ยท Pro โน1,299 / month
Retention
Zomato Pro makes ordering a habit. Rasoi Pass does the same โ convert the platform from a one-time use to a weekly household routine. Subscribers are less price-sensitive per session and have higher LTV.
Product thinking: Per-session pricing invites constant comparison ("should I order Zomato instead?"). Subscription pricing creates a mental budget โ "I have 6 sessions to use this month" changes behavior. The save percentage (up to 38% on Pro) creates genuine value while locking in predictable MRR for the platform. The key insight: subscribers aren't buying sessions, they're buying a lifestyle change.
8
Cook-side dashboard
Earnings, availability toggle, 60-second job request timer, worker protection badges
Supply quality
A full-featured cook experience โ not an afterthought. Cooks see their earnings clearly, control their availability, and have a real-time job acceptance flow with a 60-second timer (no infinite pressure).
Product thinking: Most gig platforms treat the supply side as infrastructure. Rasoi treats cooks as the product. If cooks have a poor experience, they leave โ and without supply, there is no platform. The 60-second timer protects cooks from feeling locked in; the earnings transparency builds trust; the worker protection badges make the platform's values visible. Quality cooks stay when they feel respected and in control.
Business Model
How Rasoi makes money
Four revenue streams built around a core principle: the platform earns when cooks earn. No hidden fees, no travel cut, no surprise deductions.
โ Commission (primary)
20%
Platform takes 20% of session price. Cook earns 80%. Aligns incentives โ when the cook earns more (longer sessions, more people), the platform earns more too.
โก Convenience fee
โน25
Flat per booking, 100% platform. Covers payment processing, insurance, and platform overhead. Small enough not to cause friction.
โข Travel fee (pass-through)
โน0
Platform takes zero from travel fees. โน30โ80 goes 100% to the cook. This is a worker-first choice โ and a trust-building signal to both cooks and users.
โฃ Subscriptions (MRR)
โน499+
Monthly recurring revenue with better LTV, lower CAC per booking, and habit formation. Subscribers lock in Rasoi as a household routine.
Unit economics โ typical booking
| Item | User pays | Cook earns | Platform earns |
| 2hr session (base, travel incl.) | โน399 | โน319 (80%) | โน80 (20%) |
| People add-on (3โ4 ppl) | โน80 | โน64 (80%) | โน16 (20%) |
| Convenience fee | โน20 | โ | โน20 |
| Total | โน499 | โน383 | โน116 |
Platform take rate: 23.2% of GMV ยท Cook take rate: 76.8% of GMV ยท Travel included in base price
Scale projections
| Stage | Daily bookings | Cities | Monthly GMV | Platform revenue |
| Neighborhood validation | 10 | 1 area | โน1.8L | โน39K |
| City launch (Bangalore) | 100 | 1 | โน18L | โน3.9L |
| Multi-city (5 cities) | 500 | 5 | โน90L | โน19.6L |
| Seed milestone | 1,000 | 5 | โน1.8 Cr | โน39.3L |
Why this model is defensible
The 80/20 split is significantly better than Zomato's delivery partner economics. This creates a retention moat: cooks recommend Rasoi to other cooks (word-of-mouth supply acquisition). The travel fee pass-through โ which costs the platform nothing โ signals respect and adds a further incentive to stay. Good supply attracts demand; demand validates supply. The flywheel is cook-first.
Success Metrics
What good looks like
The North Star is household habit formation โ not bookings, not revenue, not ratings. A household that books Rasoi weekly has solved their home-food problem.
โญ North Star
Weekly Active Households โ households that completed at least one booking in the last 7 days
Supply health
Active cooks
Cooks with โฅ1 completed job this week
Jobs / cook / week
Avg completed sessions per active cook
Cook churn rate
% of cooks inactive after 30 days
Supply coverage
% of hours with 3+ available cooks per zone
Demand health
Booking completion
Booked โ cook arrived and completed session
Day-7 retention
% of users who book again within 7 days
Day-30 retention
% of users active 30 days after first booking
Subscription conversion
% of users who buy Rasoi Pass within 30 days
Quality
Avg session rating
Post-session user rating (1โ5). Target: โฅ4.5
% 5-star sessions
Sessions rated 5 stars. Target: โฅ60%
Support ticket rate
Support contacts per 100 bookings. Target: <3
Business
GMV
Total gross booking value (user pays)
Take rate
Platform revenue / GMV. Target: ~22%
CAC
Cost to acquire one booking user
LTV / CAC
Target: โฅ3ร for sustainable unit economics
Why not "bookings" as the North Star?
Bookings can be gamed โ a promotion drives one-off bookings that don't reflect real product-market fit. Weekly Active Households is harder to fake: a household that keeps coming back has genuinely solved their home-food problem with Rasoi. That's the behavior worth optimizing for.