A cloud kitchen — delivery only, no dining room — is genuinely cheaper to open than a restaurant. No prime frontage, no seating, no waiting staff, a fraction of the fit-out. That is the part every guide covers.
The part they skip is that removing the dining room also removes every walk-in customer. One hundred per cent of your revenue now arrives through a channel, and if that channel takes 30%, it is not a cost of marketing — it is the largest line in your business after food.
Start with the arithmetic
Take a Rs 1,000 order from a kitchen running a healthy 30% food cost.
| Line | At 30% commission | At 0% commission |
|---|---|---|
| Order value | Rs 1,000 | Rs 1,000 |
| Food cost (30%) | −Rs 300 | −Rs 300 |
| Packaging | −Rs 50 | −Rs 50 |
| Commission | −Rs 300 | Rs 0 |
| Delivery (your rider) | Rs 0 — platform delivers | −Rs 120 |
| Gross contribution | Rs 350 | Rs 530 |
Before rent, gas, electricity, wages and the platform subscription. Illustrative — use your own numbers.
Both columns are positive, which is why cloud kitchens exist. But the left column has to cover rent and wages out of Rs 350 a time, and it does not improve as you grow — the commission scales with you exactly. The right column improves with every order, because the only fixed cost is the subscription.
What it costs to set up
Ranges vary enormously by city and ambition, so treat these as shape rather than budget:
- Premises — a small back-street unit with three-phase power and proper drainage. No frontage needed, which is where the saving is.
- Extraction and gas — usually the single largest fit-out item, and the one people underestimate.
- Cooking equipment, refrigeration, and prep surfaces.
- Packaging, which is a recurring cost and a real one: Rs 30–60 an order adds up fast.
- Licences — food authority, municipal NOC, possibly fire and environmental.
- Photography for the menu. Not optional for a business with no shopfront.
The licences are the same as a restaurant
No dining room does not mean lighter regulation. You need the food licence from your provincial authority, an NTN, provincial sales-tax registration, and the municipal permissions for the premises. See the registration guide for the detail.
The real risk: you have no customers of your own
A restaurant with a street presence acquires customers by existing. A cloud kitchen acquires none. Every order is bought — from an aggregator’s app, from your own marketing, or from a following you already had.
This is why so many cloud kitchens are effectively subsidiaries of the aggregator they list on: it owns the demand, and the moment the ranking changes the orders stop. The kitchens that survive build a direct channel early, even while the aggregator is still the majority of volume.
- Put a card in every bag with a direct order link and a reason to use it.
- Run one channel you own — Instagram, WhatsApp, or a listing on a low-commission platform.
- Track what share of orders is direct. If it never rises, you are renting your business.
Multiple brands from one kitchen
Running two or three virtual brands from one kitchen is common — a biryani brand and a burger brand sharing a fryer and a rider. It raises utilisation, which is the whole game when the rent is fixed.
Be careful that each brand is genuinely distinct and that quality does not fall as the prep list grows. Two brands done well beat five done indifferently, and customers do notice when the "different" restaurant arrives in the same packaging.
Common questions
How much does it cost to start a cloud kitchen in Pakistan?
Substantially less than a restaurant because there is no frontage or seating, but the extraction, gas and refrigeration are the same. Shared kitchen spaces that rent by the station cut the fit-out cost dramatically and are a sensible way to test a concept.
Do cloud kitchens need a food licence in Pakistan?
Yes — the same licence as any other food business, from the food authority of the province the kitchen is in, plus an NTN, provincial sales-tax registration and municipal permissions. Having no dining room does not reduce the requirements.
Are cloud kitchens profitable in Pakistan?
They can be, but commission is the deciding factor because 100% of revenue arrives through a channel. At 30% commission the contribution per order has to cover rent and wages from a much smaller base, and it does not improve with scale — the commission grows with you.
Can I run more than one brand from one cloud kitchen?
Yes, and it is common — it raises utilisation of a fixed rent. Keep the brands genuinely distinct and watch that quality does not slip as the prep list grows.
Do I need my own riders for a cloud kitchen?
On a commission aggregator, no — the fleet is part of what the commission pays for. On a flat-fee platform you deliver, so you need at least one rider. Which works out cheaper depends on your volume.