Cloud kitchens: how delivery-only restaurant brands make money

A cloud kitchen, also called a ghost kitchen or dark kitchen, prepares meals that customers order online and receive at home. There is no dining space, no waiters and usually no signboard on a main road. The kitchen can sit in a low-rent area, since customers never visit.

The brand exists on delivery apps and its own app or website. The customer sees a restaurant listing, and orders arrive on a tablet in the kitchen. A rider from the app or from the company collects the food.

Revenue comes from the food sold, minus discounts. The costs are ingredients, kitchen staff, rent, packaging, delivery and the commission paid to the order app. Most of the saving comes from skipping the dining room: smaller space, fewer staff and no front-of-house costs.

To earn a profit, the kitchen needs enough orders to cover its fixed costs. A large share of the money is made or lost on order volume per kitchen, which is why companies talk about orders per day and kitchen utilisation. The usual unit economics checks apply, and the gross and contribution margin show what is left after food and delivery costs.

• Revenue: food sales after discounts.
• Variable cost: ingredients, packaging, delivery and app commission.
• Fixed cost: rent, kitchen staff, equipment and technology.
• Contribution: what remains per order after variable costs.

A kitchen costs the same whether it makes one cuisine or several. By running several brands from the same kitchen, such as a biryani brand, a pizza brand and a dessert brand, the company spreads rent and staff over more orders. Each brand appears as a separate restaurant on the app, which gives customers more choices.

The same chefs and equipment can often serve different menus, and shared ingredients cut waste. The risk is complexity: more menus mean more training, and a bad kitchen can harm several brands together. As a live case, StopDown covered Rebel Foods, which runs brands such as Faasos, Behrouz Biryani and Oven Story Pizza, when its board cleared stock options for its co-founders.

Apps such as Swiggy and Zomato charge a commission on each order, which is a percentage of the order value. The rate depends on the deal and on how the app promotes the restaurant. For a business with thin margins, a high commission can take much of what is left after food costs.

Because of this, cloud kitchen companies try to raise the share of orders from their own app or website, where no commission is paid. They also negotiate rates, and build strong brands so that customers search for them by name. Dependence on one or two apps for most orders is a risk that investors watch.

A cloud kitchen needs a food safety licence or registration from the Food Safety and Standards Authority of India (FSSAI) for each kitchen, and aggregator apps ask to see it before listing the brand. It also needs local trade, fire and health permits, which differ by city.

On GST, restaurant services supplied through an app fall under Section 9(5) of the CGST Act, which makes the app responsible for collecting and paying the tax on those orders. Rules on rates and credits are technical and change, so the details should be checked against the latest notifications and with a tax professional, as of November 2026.

Delivery-only brands have to pay commissions, and they cannot see customers face to face. Some companies are adding dine-in outlets, takeaway counters or franchised stores to build the brand, earn better margins on some sales and reach customers who prefer to eat out. The move mixes the cloud kitchen model with the traditional restaurant model.

This also changes the numbers. Dine-in adds rent and staff cost back, so the company gives up part of its cost advantage for brand strength and for a mix of channels.

The biggest risks are low order volume, high commissions, rising food and rider costs and weak customer loyalty. Because customers choose on an app, switching costs are low, and a bad rating hurts quickly. A kitchen that is open but not busy loses money, since rent and staff are fixed.

Another risk is food safety. One failure in a shared kitchen can affect several brands. Strong systems for quality checks and training are a core part of the business, not an extra.

A restaurant chain pays for locations that attract walk-in customers, with seating, signage and front-of-house staff. A cloud kitchen skips these, and its cost base is closer to a factory: kitchen space, cooks and packaging. Its marketing happens on apps, where ratings, discounts and photos matter more than a prime location.

The trade-off is customer relationship. A chain meets guests in person, while a cloud kitchen mostly meets a rating and a delivery address. That is one reason some companies add dine-in or franchised outlets later.

A few figures show whether a cloud kitchen company is improving. Orders per kitchen per day shows how busy each kitchen is. Average order value shows what customers spend. Contribution margin per order shows what is left after food, packaging, delivery and commission. Together, these decide when the company can cover its fixed costs.

Also look at the share of orders from the company's own channels, because those carry no app commission, and at how many kitchens are profitable on their own. A company can grow revenue quickly and still lose money if each new kitchen takes a long time to fill. The loss should narrow as kitchens mature.

Finally, check how the company counts revenue. Some report the full order value and some only what they keep after the app's share, which makes headline numbers hard to compare.

Check whether the story talks about orders, kitchens, revenue or losses, and whether the figures are per brand or for the whole company. Look at how many brands share a kitchen, the share of orders from the company's own channels and any move into dine-in or franchising. Revenue growth means little without a falling loss. This is general information and not advice, and tax and licence rules should be checked with a professional.