Cloud Kitchen vs Packaged Food: Picking a Model

A founder with a good recipe has two ways to build a food business in India. Open a delivery-only kitchen and sell hot meals through the aggregators, or put the product in a pack and sell it through marketplaces, quick commerce and retail. They look adjacent. They are not.

The Money Works Differently

A cloud kitchen sells at an average order value of roughly Rs 300 to Rs 500. Out of that, the aggregator takes a base commission commonly in the 18 to 25 percent band, plus a per-order platform fee that both major platforms raised to about Rs 17.58 in March 2026, plus payment gateway charges of around 1.9 to 2 percent, plus 18 percent GST on the platform’s service and payment fees, plus long-distance charges on orders beyond about 4 km, plus a small TDS deduction. The realistic leak is 25 to 35 percent of order value. Food cost typically takes another 28 to 35 percent. Rent, labour, packaging and marketing live in the remainder.

The compensation is cash. Revenue is recognised daily, settlement cycles are short, and you carry only a few days of raw material as inventory.

A packaged brand never sees the MRP. Quick commerce and marketplace commissions, fulfilment fees, warehousing, returns and advertising come out before you count anything, and in general trade the distributor and retailer margins do the same job. Most food brands realise somewhere in the region of 55 to 70 percent of MRP as net revenue before cost of goods. Then working capital is heavy. You fund a full production batch, transit stock, stock sitting in platform warehouses, and a payout cycle that can run 30 to 60 days. And unlike the kitchen, that stock ages.

Capex and Licensing

A metro cloud kitchen is commonly a Rs 10 to 15 lakh setup, with equipment alone at Rs 6 to 9 lakh, plus deposit, fit-out, exhaust and fire compliance, and a point of sale system. It also needs a municipal trade licence, fire clearance where applicable, GST registration and state-specific approvals alongside the FSSAI licence.

The packaged route can start with no plant at all. A contract manufacturer holds the manufacturing licence for the product category, and you appear on the pack as the marketer. Your money goes into recipe development, packaging artwork and tooling, the first production run and stability testing.

On the FSSAI side, thresholds were revised with effect from 1 April 2026 through an amendment notified on 10 March 2026. Basic registration now covers turnover up to Rs 1.5 crore, a state licence covers Rs 1.5 crore to Rs 50 crore, and a central licence applies above Rs 50 crore. Confirm your own position against the current notification, since these limits moved recently and can move again.

Where Demand Comes From

Cloud kitchen demand is rented. Discovery happens inside the aggregator app through search rank, category placement, images, ratings, delivery time and paid promotion. You rarely own the customer relationship or the phone number. Your levers are rating, prep time, packaging quality and ad spend, and every one of them is priced by the platform. Building direct ordering through your own app or WhatsApp is the only route to ownership, and it is slow work.

Packaged demand is built once and reused. A listing that ranks, a shelf position that holds, a repeat purchase cycle you can measure. You spend on content, advertising, sampling and trial packs. Discovery is spread across marketplaces, quick commerce, modern trade and your own site, so no single platform can reprice your whole business overnight. The trade-off is a longer purchase cycle and a slower feedback loop.

Shelf Life Decides Your Reach

A kitchen product has a shelf life measured in minutes and a service radius of roughly 4 to 6 km. Growth means more kitchens, and each one carries its own lease, staff, licence and break-even point. Growth is linear and it consumes capital.

A packaged product with six to twelve months of life travels anywhere a courier goes. Growth becomes a distribution problem rather than a real estate problem. But shelf life also caps you at the other end. Platforms enforce inbound remaining-life gates, and FSSAI requires stock delivered to a consumer to carry at least 30 percent of shelf life or 45 days remaining. Stock that does not move ages into a write-off, which is a category of loss a kitchen never has to think about.

Why Some Brands Run Both

Plenty of Indian food businesses run both sides. Haldiram’s and Bikanervala built large packaged businesses off restaurant recipes. Wow Momo sells across its outlets and puts frozen momos into quick commerce and modern trade. Blue Tokai and Third Wave Coffee run cafes and sell packaged coffee through their own sites and marketplaces. Chaayos sells packaged tea alongside its cafes. Rebel Foods operates one of the largest cloud kitchen networks in the country across multiple brands.

The logic is consistent. The kitchen creates the taste memory and the brand proof, and somebody who liked the dish will buy the pack. The pack then monetises that memory in every city where you will never open a kitchen. The kitchen doubles as a cheap product development lab, because you can test a recipe on paying customers within a week, where a packaged trial costs a production run and three months of waiting.

The failure mode is running both before either works. Two operating models, two cost structures, two compliance regimes and one team. Sequence it. Get one to positive contribution, then let it fund the other.