How to Start a Cloud Kitchen in India
A cloud kitchen is a restaurant without a dining room: you cook and you deliver, and nobody visits. That removes the two biggest costs of opening a restaurant and replaces them with a different problem, which is that nobody walks past a cloud kitchen and decides to try it.
Summary
This guide covers what changes when you run a cloud kitchen instead of a dine-in restaurant, including location selection, multi-brand operations, and menu design for delivery.
It explains why the model does not remove the cost of discovery, just shifts it into commission and marketing, and how to model per-order economics before starting.
What the model changes
Removing the dining room removes customer-facing interiors and the need for a prime-footfall location. That is a large reduction in capital and in fixed monthly cost, and it is the reason the format has grown so quickly.
What it does not remove is discovery. A high-street restaurant pays for discovery through rent — people see it. A cloud kitchen pays for discovery through aggregator commission and marketing. The cost does not vanish; it changes from a fixed monthly amount into a percentage of every order, which behaves very differently as you grow.
Choosing a location that nobody sees
Site selection is still consequential, just on different criteria. What matters is delivery reach rather than visibility.
- Density of residential and office demand within a short delivery radius, since delivery time drives both ratings and repeat orders
- Adequate electrical load, water supply and drainage — the same requirements as any commercial kitchen
- Whether the premises can be licensed for food production, which not every industrial or basement space can
- Rider access, including somewhere they can wait that is not your kitchen floor
- Rent low enough that the format's core advantage is not given away
Multi-brand operations
Most established cloud kitchens run several virtual brands from one kitchen. The logic is that different brands surface for different searches on the aggregator platforms, so three brands occupy more of the results a hungry customer sees than one brand does.
The discipline this requires is real. Brands should share ingredients and equipment while appearing genuinely distinct to a customer, and each needs separate accounting — sales, food cost and contribution per brand. Without that, a weak brand hides inside a healthy total and you keep operating it because nobody can see it is losing money.
Menu design for delivery
The constraint that shapes everything is the twenty to thirty minutes between the pass and the customer's door. Dishes that are excellent on a plate and poor after that journey will generate bad ratings no matter how well you cook them.
Design for it deliberately: favour dishes that hold temperature and texture, avoid anything that goes soggy, keep components separate in the packaging where it helps, and test every dish by ordering it to your own home at a realistic distance before it goes on the menu.
Packaging is not a minor line item here. It is the last thing you control before the customer forms an opinion, and it is the most common cause of complaints that have nothing to do with the cooking.
The economics to model before starting
Cloud kitchen margins are tighter than they first appear because commission, packaging and marketing all scale with orders. Model contribution per order rather than gross revenue.
Take your average order value, subtract food cost, packaging, commission and delivery-linked charges, and what remains has to cover rent, salaries and utilities before anything is profit. Running this at your realistic order volume rather than an optimistic one tells you quickly whether the concept works, and it frequently shows that a higher average order value matters more than more orders.
Reducing dependence on aggregators
A cloud kitchen that exists only on aggregators has no customer relationship and no protection if commission terms change or your ranking drops. Building a direct channel from the start is materially easier than retrofitting one later.
The practical route is packaging: a QR code and a WhatsApp number on every order that goes out reaches a customer who has already eaten your food and knows whether they want it again. That is the cheapest acquisition available to you, and unlike aggregator visibility, nobody can take it away.
Frequently asked questions
How much does it cost to start a cloud kitchen in India?
Substantially less than a dine-in restaurant, because customer-facing interiors and a prime location are not required. The costs that remain are kitchen equipment, licences, a modest premises, opening stock, packaging and working capital — plus marketing, which replaces rent as your route to discovery.
Is a cloud kitchen profitable in India?
It can be, but margins are tighter than they appear because commission, packaging and marketing all scale with order volume. Model contribution per order — average order value less food cost, packaging, commission and delivery charges — at a realistic volume before committing.
Can I run multiple brands from one cloud kitchen?
Yes, and most established operators do, because different brands surface for different searches on aggregator platforms. It requires separate accounting per brand, otherwise a loss-making brand hides inside a healthy overall total.
What licences does a cloud kitchen need in India?
Broadly the same as any food business — an FSSAI licence covering the premises, GST registration, and local trade licence and fire clearance as applicable. Confirm the premises can be licensed for food production before signing, since not every industrial or basement space can be.
Try Servyn free
QR ordering, GST billing, kitchen display and inventory in one platform. No card required, no hardware to buy — tell us about your restaurant and we set it up with you.
Get Servyn free