How to Start a Cloud Kitchen in Dubai

A cloud kitchen has no dining room, no walk-in trade, and often no sign on the door. It is a production line whose entire revenue arrives through a delivery app. The appeal is obvious — low entry cost, no front-of-house. The trap is structural, and it is hiding in the commission rate of the very platforms that make the business possible. This is Chapter 8 of the book — and like every industry chapter, it follows the same 8-section template.

1 · What This Business Actually Is in Dubai

A cloud kitchen — also called a dark kitchen or ghost kitchen — is a food operation built to do one thing: fulfil delivery orders. No front-of-house. No customer-facing staff. No location a diner would ever seek out. The brand lives on an aggregator app — Talabat, Deliveroo, Noon Food, or Careem — and the kitchen's only job is to produce and dispatch, the same way every time.

In Dubai, the model takes one of two shapes. The first is the owner-operated stand-alone kitchen — a unit in an industrial or mixed-use building, typically in areas like Al Quoz, the JAFZA-adjacent zones, DIP, or Ras Al Khor, with full fit-out, DET licence, Municipality permits and Civil Defence approval — the same permitting stack as a restaurant, with none of the walk-in revenue to soften it.

The second, more common for first-time operators, is the shared or commissary kitchen. Operators such as Kitopi, iKcon and others rent out shared production space, often managing the aggregator integrations, the packaging and the logistics. You bring the concept, menu and marketing. It cuts capital and operational complexity — but stacks a revenue share on top of the aggregator commission.

Read that twice. Two parties — the commissary and the platform — both take a percentage before the brand owner sees a single dirham.

2 · Capital Required

• Trade licence (DET, mainland): AED 12,000–18,000 for year one.
• Dubai Municipality food permit: AED 2,000–5,000, renewable annually, tied to the kitchen address.
• Fit-out (own kitchen): An industrial unit fitted as a commercial kitchen — extraction, fire suppression, stainless steel surfaces, refrigeration and prep equipment — typically AED 80,000–200,000. Civil Defence sign-off is mandatory.
• Shared kitchen (commissary): Entry deposit AED 5,000–20,000 plus monthly production fees. Revenue-sharing terms must be negotiated carefully — understand the exact structure before you commit.
• Equipment: Even in a commissary, the brand owner often provides specialist equipment for the cuisine: AED 20,000–60,000.
• Packaging: Often underestimated. Branded boxes, bags, labels, sauce cups — AED 8,000–20,000 for an initial stock order, then a fixed cost on every single order.
• Staff and visas: Three to five kitchen staff = AED 12,000–25,000 monthly in payroll, plus AED 4,000–7,000 per person in visa costs.
• Marketing & aggregator promotions: AED 5,000–15,000 per month in the first six months.
• Working capital: Six months of operating costs, minimum.

A lean own-kitchen operation can be started for AED 250,000–450,000. A commissary-based brand can enter for AED 80,000–150,000 — but the cheaper door buys you a correspondingly tighter margin for the life of the business.

3 · Licence & Jurisdiction

A cloud kitchen operating from its own mainland facility requires a DET trade licence with a food production or catering activity, a Dubai Municipality food establishment permit tied to the kitchen address, and Dubai Civil Defence sign-off on the fit-out.

Food handler occupational health cards and Person-in-Charge food safety certification are required whether or not a single customer ever sees the place. The Municipality inspects a cloud kitchen on exactly the same basis as a restaurant. The absence of a dining room does not reduce the compliance obligation — it removes the revenue and keeps the rules.

If the kitchen operates from a commissary, the commissary operator typically holds the primary Municipality permit and your brand is listed as a sub-operator. Do not assume that arrangement covers you. Confirm it explicitly, with both the commissary and the Municipality, before you trade. The distinction between "food preparation," "catering," and "restaurant without seating" is not pedantry — it decides which permit category you land in.

4 · The First Three Customers

Here is the paradox: the distribution channel is built for you — aggregator apps are the mechanism by which every order arrives — but discoverability inside those apps is earned, never given. A new listing on Talabat, with no ratings, no reviews and no promotional spend, is not a business. It is invisible.

Aggregator launch promotions. Talabat and Deliveroo offer onboarding packages: reduced commission for the first 30–90 days, a slot in the new-listings section, sometimes a co-funded discount campaign. Use every one. Your first 50 orders are worth subsidising heavily — they generate the review base that makes your next 500 discoverable at a normal margin. Contact platform restaurant partner teams during setup, not after launch.

Social media seeding before the listing goes live. A cloud kitchen has no physical location to send anyone to, which makes social proof more important, not less. A pre-launch Instagram or TikTok account documenting the concept, the kitchen and the food can build 500–2,000 interested followers before the first order is accepted. Those followers become your first customers — and their content becomes your first reviews.

Community WhatsApp groups. Dubai runs on community WhatsApp groups — building groups, school groups, neighbourhood groups. A genuine, non-spammy introduction with a first-order discount code produces a cluster of orders from one tight geographic patch — which aggregator algorithms read as local demand and reward with visibility.

5 · Profit Margins & the Cash Cycle

The unit economics come down to three numbers: food cost as a percentage of order revenue, aggregator commission as a percentage of order revenue, and the volume of orders needed to cover fixed costs. Work those out for your concept, your menu and your chosen platforms before you launch. Do it on a spreadsheet, not in your head.

Aggregator commissions at the time of writing run at roughly 25–35 per cent of order value; confirm current rates directly with each platform. On a AED 70 order, the platform receives AED 17.50–24.50. Of the remaining AED 45.50–52.50, food cost takes 28–35 per cent of the original order value, leaving a gross contribution before fixed costs of roughly 30–40 per cent of order revenue in a well-managed operation.

Out of that contribution come rent, payroll, utilities, packaging and platform marketing. A cloud kitchen generating AED 80,000–120,000 per month in gross order value, at 30–35 per cent net contribution, is covering perhaps AED 24,000–42,000 towards fixed costs. For a lean operation with controlled overheads, that may reach breakeven in 12 to 18 months.

The cash cycle itself is clean — platforms settle weekly or fortnightly, no long B2B receivables, no annual rent cheque unless the lease demands one. The dominant cash risk is different: the front-loaded marketing spend required to drag order volume up from zero.

6 · Top 3 Mistakes

Treating aggregator commission as a line item rather than the core business model question. Too many founders set their prices, then calculate their margin, then notice the commission has eaten it, then cut prices in promotions to chase volume — which makes the margin worse. Run the sequence the other way. Start with your fixed cost per order. Add the net contribution you need. Add the aggregator commission. Then set your menu price. If that price is not competitive at that number, the model does not work, and no amount of promotion will fix it.

Launching with too many menu items. A cloud kitchen that launches with 30 items is slower, wastes more, and trains staff more poorly than one that launches with 12. The aggregator algorithm does not reward breadth of menu — it rewards order volume and positive reviews. A focused menu of 10–15 items done consistently well beats a sprawling one that forces corners to be cut. Expand once operations are stable, not before.

Underinvesting in photography and platform presentation. The customer's entire purchasing decision is made from a phone screen. A listing with poor photography and no reviews competes at a structural disadvantage against established listings carrying hundreds of five-star reviews. Budget AED 3,000–8,000 for professional food photography before launch. That is not optional marketing spend. It is infrastructure.

7 · Who's Already Winning

The cloud kitchens that sustain order volume in Dubai are all doing the same thing: solving a specific problem for a specific customer, reliably and repeatedly. The most durable single-brand concepts are built on a cuisine or dietary niche the existing market underserves — macro-balanced meal prep for gym members, Levantine comfort food for a homesick expat demographic, plant-based versions of a mainstream cuisine, a regional speciality no delivery-first brand is covering. Winning at "general burgers" or "pizza" as a new entrant is genuinely hard. Winning at a specific, underserved niche is structurally more achievable.

Multi-brand operators — running two to four delivery brands from the same kitchen, spreading fixed costs across higher order volume — often achieve better economics than single-brand operations. The risk is operational complexity: running four brands from one kitchen at lunch hour demands tight menu engineering and disciplined ticket management, or it collapses.

The operators who fail quickly are the ones who treat the aggregator listing as the business plan. Discovery, reviews, re-order rate, aggregator marketing — none of it is automatic. It is active, ongoing work, and the founder who will not do it watches the listing sink in the rankings while the fixed costs carry on, indifferent.

8 · Verdict

A cloud kitchen is the most accessible point of entry in Dubai's food sector by capital requirement. It is also the most aggregator-dependent, which means the unit economics are controlled, in large part, by a third party whose commission structure can change. The lean entry costs are real. So is the ongoing margin pressure that follows them through the door.

Who should proceed: a founder with a specific culinary niche, a realistic model of the unit economics, and the operational discipline to manage consistency across high delivery volumes.

Who should pause: anyone who sees the low setup cost and assumes the ongoing business will be correspondingly easy. It will not be.

Bottom line: A cloud kitchen lives or dies by its order economics — work those numbers before you choose a menu name.