The Cloud-Kitchen Playbook

Most people think a delivery brand wins on its food. It doesn't — at least not only. The brands that scale win on a system: a documented way of cooking, packing, and reading demand that produces the same result on a quiet Tuesday lunch and a flooded Friday night, in the first kitchen and the fifth. The recipe is the easy part. The playbook around it is the moat.

This is how we think about running data-driven, delivery-first kitchens at Indian Food Cartel — the operating principles behind every brand we launch.

What a cloud kitchen actually is — and isn't

A cloud kitchen is a delivery-only kitchen with no dine-in space. There's no host, no tables, no décor budget. Orders arrive through delivery apps and the brand's own channels, and the entire operation is built around one thing: the box that lands at the customer's door, 20–30 minutes after they tapped "order."

That reframing changes every decision. A restaurant spends on location footfall and ambience; a cloud kitchen spends that same money on kitchen utilisation, packaging, and consistency. You're not optimising an experience someone walks into — you're optimising one that travels to them. Get that distinction wrong and you'll build a beautiful kitchen that quietly loses repeat orders.

The system is the product, not the menu

A great dish gets you the first order. A great system gets you the next hundred. When founders obsess over an inventive menu but skip the operating system underneath it, they end up with food that's brilliant when the chef is on shift and inconsistent when they're not.

Reverse the priority. The product you're actually building is the repeatable way that food gets made and delivered — the SOPs, the packing standard, the prep discipline. Nail that for one brand and you've built something you can reuse: a second brand mostly inherits the same system, the same kitchen, the same suppliers. The menu is what customers see; the system is what makes it scale.

Write the SOPs before you scale

A recipe in someone's head is a liability. A recipe written down — grams, times, sequence, plating, pack — is an asset that any trained cook can reproduce. Standard Operating Procedures are simply that discipline applied to everything: how the biryani is layered, how long it rests, how the container is sealed, how the station is cleaned.

The test is brutal and simple: could a new cook, on a busy Friday, produce your signature dish to standard from the documentation alone? If the answer is "only if a senior cook is watching," you don't have a system yet — you have a person. People take leave, change jobs, and have off days. Documentation doesn't.

This is also where hygiene and compliance live. FSSAI registration and food-safety SOPs aren't paperwork you do once — they're part of the operating standard that protects the brand every single shift.

Packaging is part of the dish

Here's the uncomfortable truth of delivery: the customer never tastes the food you plated. They taste the food that survived the ride. Gravy that pooled, rice that steamed itself soggy, a lid that leaked into the bag — none of that was on your pass, but all of it is on your rating.

So treat packaging as an ingredient, not an afterthought. Vent where steam needs to escape. Compartmentalise what shouldn't mix. Seal what travels on a two-wheeler over Indian roads. And test it the only way that's honest — pack a real order, send it on a real 30-minute route, and open it like a customer would. The brands that win delivery are the ones that engineered the last 30 minutes as carefully as the first.

Let order data run the kitchen

Delivery hands you a gift that dine-in never did: a live, item-level read on demand. Order volume by hour, prep and dispatch times, rejection and cancellation rates, ratings by dish, and the margin each item actually contributes. Most kitchens ignore most of it.

Use it. Prep to the demand curve so you're not throwing away food at close or running out at peak. Rank dishes by contribution margin, not just popularity — a bestseller that barely clears cost is not the hero it looks like. And prune ruthlessly: the slow, fiddly, low-margin items that drag the line are exactly the ones the data gives you permission to cut. A tighter menu is faster, more consistent, and usually more profitable.

One kitchen, many brands: the economics

This is where the model compounds. Rent, core equipment, and base staffing are largely fixed — they cost roughly the same whether the kitchen ships 3,000 orders a month or 8,500. Running a second or third brand from the same kitchen spreads those fixed costs across far more orders, which is why the cost-per-order falls so sharply as utilisation rises (see the table below).

But the leverage only works if the brands genuinely share the system — overlapping prep, suppliers, equipment, and SOPs. Three unrelated brands with three separate prep lines just rebuild your overhead three times. The art is choosing brands that reuse the same operating spine while looking, to the customer, like distinct experiences.

A simple sequence to launch and scale

The order of operations matters more than the speed:

1. Nail one brand. Tight menu, documented SOPs, packaging tested on real routes.
2. Instrument it. Watch prep times, ratings, rejections, and item margin from day one.
3. Prune and tune. Cut what the data says is slow or low-margin; tighten what works.
4. Then add a brand that reuses the same kitchen, prep, and suppliers — and repeat the loop.

Do it in that order and growth is mostly replication. Do it out of order — a second brand bolted onto an un-systematised first — and you've just doubled the chaos.

---

The cloud-kitchen opportunity isn't really about food being delivered instead of served. It's about building one disciplined system and earning the right to run it many times over. That's the playbook we run at IFC — and it's why we think about kitchens as systems first, and menus second.