Cocina oculta: what it actually costs, and what they tell you it costs
A cocina oculta costs 8,000 to 45,000 USD to open and 900 to 3,500 USD per month to run, depending on the setup — rented corner, shared hub, or your own build — yet the number that decides profitability sits elsewhere: in the app commission, which runs 18% to 33% of the ticket, and in the three costs no vendor puts in the quote — packaging, in-app advertising, and your own courier when the algorithm buries you. The MYTH says you open with 5,000 dollars and charge dining-room prices; the REALITY is that it only works when your menu is built to a food cost at or under 32% per dish, when your Google Business Profile exists with a real address, and when you accept that you pay for the first quarter while the algorithm learns who you are.
Early 2026 brought the same scene to Bogotá, Lima and Mexico City: twenty square meters, no tables, three brands coming off one griddle, and owners who got there because somebody sold them a restaurant without the restaurant's pain. Half of that promise holds. You skip the dining room, the server, the air conditioning, the linen, the bathroom queue. The other half is where the money leaks, because everything saved on floor space transfers whole to an intermediary who charges for bringing you the customer and who can change the rules on any given Tuesday without telling you.
Numbers first, theory after. A rented corner inside a restaurant that already operates starts at 600 to 1,200 USD monthly in mid-sized Latin American cities, and that figure climbs to 1,400-3,500 USD inside a specialized hub with extraction, cold rooms and a dispatch area. Startup investment — equipment, licenses, branding, product photography for the apps — does not drop below 8,000 USD even buying used, and a container or converted warehouse comfortably reaches 35,000 or 45,000. Everyone quotes those figures the same way. Trouble starts on the line below the contract.
I got this wrong for years, and I will say it plainly: I used to recommend delivery as filler, something to keep the kitchen moving in dead hours, with nothing else in the business touched. That worked in 2019. By 2026 the channel is no longer filler, it is a market with its own search engine, its own ranking rules and its own price war, and walking in without understanding that is like opening on an avenue where you control neither the sign nor the window. So this piece is about prices, yes, but prices with the digital engine attached: a cocina oculta's cost is set as much by the rent invoice as by where your brand lands when somebody types «sushi near me» at eight thirty at night.
Side-by-side comparison
| Cocina oculta (dark kitchen) | Physical restaurant with dining room | |
|---|---|---|
| Startup investment | ✕8,000 - 45,000 USD by setup | ✓60,000 - 250,000 USD with build-out |
| Monthly rent | ✕600 - 3,500 USD (20-45 m²) | ✓2,500 - 12,000 USD (120-250 m²) |
| Commission per sale | ✕18% - 33% of ticket on apps | ✓2.5% - 3.5% on card payments only |
| Opening payroll | ✕3 - 5 people, no front of house | ✓9 - 16 people with floor and till |
| Break-even | ✕4 - 8 months if ticket clears 11 USD | ✓14 - 26 months regional average |
| Who owns the customer | ✕The app: no phone, no email for you | ✓You: own database, bookings, repeat visits |
| Average ticket 2026 | ✕9 - 16 USD, highly discount-sensitive | ✓18 - 34 USD with drink and dessert |
| Traffic engine | ✕Rappi, Uber Eats, DiDi and iFood ranking | ✓Google Maps, 5★ reviews and street |
What does it cost to open a ghost kitchen in 2026?
Opening a ghost kitchen costs between USD 8,000 and 45,000, and holding it open runs USD 900 to 3,500 a month, with prices measured as of August 2026 in mid-sized Latin American cities.
The USD 8,000 floor buys a rented corner inside a restaurant that already operates, with used equipment and the bare acceptable minimum in licensing, branding and product photography for the apps. The USD 45,000 ceiling shows up when you build out a container or an adapted warehouse, with your own extraction, cold room and dispatch area. Between those two numbers sit three different models, not three budgets for the same model, and mixing them up is the most expensive opening mistake in this category. The market explains the appetite: cloud kitchens will move USD 83.5 billion in 2026, growing at a 9.7% compound annual rate through 2034, according to Fortune Business Insights.
What each investment range actually buys?
The USD 8,000 to 15,000 range buys a rented corner:
monthly rent of USD 600 to 1,200 inside a working kitchen, a used griddle or fryer, a shared hood, a health permit under the host's name, and a photo package covering 40 to 60 dishes that the apps demand before they rank you. It does not include your own extraction or a cold room, and that is its real ceiling. From USD 15,000 to 30,000 you move up to a specialized hub, where rent jumps to USD 1,400-3,500 a month and brings certified extraction, dedicated refrigeration, a dispatch room with courier stations and, in the good ones, an order management system that unifies your three or four apps. Above USD 30,000 you build: container or adapted warehouse, electrical and gas work, licenses under your name, and the asset stays yours.
Effective commission runs 28-33%, not the contract's 18%
The number that decides whether a ghost kitchen makes money never appears in the build-out budget: it sits in the fine print of your app contract. Paper says 18% or 22% depending on country and category, and that figure is the FLOOR. Stacked on top come the co-funded promotion the app suggests every time your brand slips down the ranking, the two-for-one you accept on a slow Sunday, and the courier fee once you join the delivery program. Effective commission, read straight off the statements, lands around 28-33% of gross ticket. Run it against your dish: with a 30% food cost and an app taking 31, nothing survives for rent, gas or payroll. Build the digital menu backwards, starting from the final price the platform can carry, and then find the dish that fits inside it. Four variables explain almost the entire gap between an USD 8,000 build and a USD 45,000 one.
Four factors that move the price, and what each one weighs
First comes extraction: if the host location already has it certified, you save between USD 3,000 and 9,000 in construction. Second, the licensing regime: operating under the host's permit costs nothing but ties you down; pulling your own runs USD 400 to 2,500 depending on the city and leaves you portable. Third, packaging, the most underestimated cost in the model and the easiest to measure: USD 0.45 to 1.10 per order in 2026 for something that arrives neither cold nor leaking, which is up to USD 660 a month at twenty orders a day. Fourth, how many virtual brands you plan to run off one griddle, because each brand adds photography, a menu and its own recipe sheets. High-performing ghost kitchens reach margins of 10 to 30%, against 3-5% for a traditional restaurant, according to OysterLink 2025, and that is exactly the number people quote when they offer you a corner.
Why the promised 10-30% margin rarely shows up?
It is true and it is a minority. The full sector runs on a 3 to 9% net margin (Statista), and whoever lands at the bottom usually got there the same way:
they built the menu for the dining room and copied it straight onto the app. A USD 9 dish with 2.70 in raw material, 0.80 in packaging and 2.90 in effective commission leaves 2.60 before rent, gas and payroll, so at USD 900 of rent you need 346 monthly orders just to break even. That arithmetic belongs BEFORE the signature, not after the third month. Nearly nine out of ten virtual brands in the United States —86.9%— run on top of a physical kitchen already selling through another channel, while only 13.1% operate purely online, according to Locmatic's State of Virtual Restaurant Brands 2024. That is not caution, it is arithmetic: the kitchen is already paid for, the cook is already on payroll, and delivery orders slide into hours when the burners sat cold.
Virtual brands: 86.9% run hybrid, and cash flow explains it
Which is why 32% of restaurant expansion strategies in 2025 went through virtual brands, according to Technomic. If you already own a location, your ghost kitchen starts with overhead absorbed and the break-even point cut in half. If you do not, you are building a whole restaurant without a street-facing window, and somebody should say that plainly before you sign for the container. Four concrete levers take you from 31% down to 24% effective commission, and none of them requires a fight with the platform. Start by negotiating the rate against committed twelve-month volume: platforms give up 2 or 3 points when you guarantee orders, and almost nobody asks. Next, pull every dish under 55% contribution margin out of the digital catalog, because the app charges the same for selling what leaves you nothing. Third on the list, lock packaging with a supplier on quarterly volume: USD 0.95 drops to 0.62 per order, and those 33 cents are a full margin point.
How to negotiate the contract and cut effective commission?
The fourth lever moves 15 to 25% of volume to direct WhatsApp ordering from customers who already bought, where commission is zero. Within the Masterestaurant method, Diego F.
Parra always starts with the third one, since it can be executed the same Monday. Where your brand lands when somebody searches «sushi near me» at eight-thirty on a Tuesday night is a cost as real as rent, even though no separate invoice ever arrives. Delivery apps moved USD 110 billion in 2024, up 15.5% year over year, according to Business of Apps, and DoorDash alone reported USD 21.3 billion of marketplace GOV in the fourth quarter of 2024. That volume gets distributed by ranking, and ranking is bought with co-funded promotions already buried inside your 28-33% effective rate. I got this wrong for years, recommending delivery as filler to keep the kitchen busy during dead hours: by 2026 that channel has its own search engine and its own rules.
Your app ranking is part of the cost
Open your commission dashboard, add up ninety days of promotions and divide by gross sales. That percentage, not the contract's, is your real price. EFFECTIVE commission, not nominal. The contract says 18% or 22% by country and category, and that figure is the floor. Stacked on top: the co-funded promo the app suggests so you do not drop in ranking, the weekend two-for-one you accept on a slow Saturday, and the courier fee once you join the delivery program. Add it up and the effective commission running through real statements sits at 28-33% of gross ticket. If your dish carries a 30% food cost and the app takes 31, nothing is left for rent, gas or payroll. Build the digital menu backwards: the price the app can carry first, then the dish that fits inside it. PACKAGING.
Where the money hides that nobody quotes you?
The most underestimated cost in the whole model, and the easiest to measure:
0.45 to 1.10 USD per order in 2026 for packaging that arrives hot and upright, climbing to 1.60 for soups, ice cream or anything needing a double seal. On a 12-dollar ticket that is up to 9% gone before you count the protein. Saving here is an expensive trap: cheap packaging that arrives with the sauce spilled buys you a 1★ review, and one 1★ review inside the app outweighs three weeks of paid placement. IN-APP ADVERTISING. Rappi, Uber Eats, DiDi Food and iFood sell position. They call it boost, featured, sponsored, depending on the market, and it behaves exactly like Google Ads: you bid to appear at the top of the category when a hungry customer opens the app. A new kitchen in a competitive zone burns 180 to 450 USD monthly on this during ramp-up, and whoever does not spend it simply does not exist, because ranking rewards order volume and you have no volume yet.
Where the money hides that nobody quotes you — in practice?
Classic cold start, and it has a price tag. THE CUSTOMER IS NOT YOURS. This is the structural gap, and it appears on no spreadsheet.
In a physical restaurant you capture the name, the birthday, the preferred table. In a pure cocina oculta the app hands you an anonymous order and keeps the relationship. That is why I push so hard on the Google Business Profile and on direct WhatsApp ordering: every customer who migrates from the marketplace to your own channel returns 3 to 8 dollars of margin per order, and it is the only lever that cuts your effective commission without negotiating with anybody. PRINTED MENU AND QR MENU, BOTH. If your cocina oculta has even a pickup counter or a window, keep the printed menu alongside the QR: the printed menu is experience control — order rhythm, dish narrative, the suggested side that lifts the ticket — while the QR is the complement that updates prices without reprinting, feeds delivery and gives you analytics on what people look at before ordering.
Where the money hides that nobody quotes you — key points?
Dropping the printed menu to «save on printing» is among the decisions that have destroyed the most average ticket in the past three years.
Each format has its role and neither replaces the other.
Head to head: dark kitchen versus physical restaurant
What the ghost kitchen salesman tells youMYTH
- «Open for 5,000 dollars and recover in three months»
- «You do not need a Google listing, the apps feed you»
- «Commission is 18%, it says so in the contract»
- «Three brands on one griddle triples your sales»
- «Packaging gets absorbed by the delivery fee»
- «Charge dining-room prices, people pay the same»
What the books say at quarter closeMasterestaurant
- A decent build starts at 8,000 and break-even lands in month 4-8
- Without a Google Business Profile with a real address you lose 30-40% of branded demand
- With in-app ads, forced promos and courier fees, effective commission hits 33%
- Three brands on one griddle stretch dispatch time and sink your rating
- Decent packaging runs 0.45 to 1.10 USD per order and you pay it
- App prices need a 15-22% markup or your margin evaporates
Side-by-side comparison
| Cocina oculta (dark kitchen) | Physical restaurant with dining room | |
|---|---|---|
| Startup investment | ✕8,000 - 45,000 USD by setup | ✓60,000 - 250,000 USD with build-out |
| Monthly rent | ✕600 - 3,500 USD (20-45 m²) | ✓2,500 - 12,000 USD (120-250 m²) |
| Commission per sale | ✕18% - 33% of ticket on apps | ✓2.5% - 3.5% on card payments only |
| Opening payroll | ✕3 - 5 people, no front of house | ✓9 - 16 people with floor and till |
| Break-even | ✕4 - 8 months if ticket clears 11 USD | ✓14 - 26 months regional average |
| Who owns the customer | ✕The app: no phone, no email for you | ✓You: own database, bookings, repeat visits |
| Average ticket 2026 | ✕9 - 16 USD, highly discount-sensitive | ✓18 - 34 USD with drink and dessert |
| Traffic engine | ✕Rappi, Uber Eats, DiDi and iFood ranking | ✓Google Maps, 5★ reviews and street |
The figures that rule this model
“We launched two brands in 24 square meters and billed 9,400 dollars the first month, then closed with 380 in profit because commission, forced promos and packaging ate 3,100. Diego made us do three things: raise app prices 19%, pull the four dishes above 34% food cost, and build the Google listing with the pickup window address. Month three we billed 8,100, less than before, and profit was 1,940 dollars. Today 28% of orders come straight through WhatsApp and on those we pay no commission.”
How to cost a cocina oculta without fooling yourself
Take today's dining-room price and add 15% to 22%. That is your working ceiling. Now subtract the effective commission — use 30%, not the nominal figure in the contract — and packaging, which runs 0.45 to 1.10 USD per order. What remains is real revenue. If your food cost exceeds 32% against that number, the dish does not go on the digital menu, however good it photographs. Payroll, rent and utilities never load onto the dish: they belong to break-even, calculated separately.
Almost everyone skips this step, and it is the cheapest one on the list. Register the kitchen's real address, mark the service as delivery or curbside pickup, upload twenty original product photos, set the exact primary category and the delivery radius. A complete listing carries four times the conversion likelihood of an incomplete one, per Google's own documentation. And it gives you something no app will: showing up when somebody searches your name, or your cuisine nearby, with zero commission attached.
Set aside 180 to 450 USD monthly for ramp-up, split between in-app placement — which buys category position — and a tight two or three kilometer radius on Google and Meta pointed at your delivery zone. Kill the in-app spend once organic order volume holds your position for three straight weeks, and move that money to the direct channel. This is where you decide whether your cocina oculta rents somebody else's traffic forever or builds its own.
Every bag leaves with a sticker carrying the direct WhatsApp number and a concrete reason to use it: free dessert, delivery ten minutes sooner, or a price two dollars under the app. Chase 5★ reviews with a name and a face, because Maps and the marketplace weigh them differently and both move ranking. The first-year target is simple, and I measure it on every project: 25% to 30% of orders arriving through your own channel. That is when your effective commission genuinely starts to fall.
And with AI?
Optimize channels, pricing and unit economics of your dark kitchen. Diego F. Parra is an expert in AI applied to restaurants.
Free tools to apply this now
What brings order to this decision
Costing a cocina oculta by hand, on an improvised spreadsheet, is the fast lane to a business that bills plenty and keeps nothing. Sequence matters: model structure and break-even first, then the menu dish by dish, and only at the end the cash calendar that tells you how many months you can survive before the algorithm recognizes you.
Diego F. Parra — International consultant, expert in creating and scaling restaurants and in AI applied to restaurants, foodtech and HORECA. Methodology applied in 8.400+ restaurants across 43 countries · Expert in Artificial Intelligence applied to restaurants, hospitality and food businesses · 20+ years in restaurants, catering, large events and business growth · Author of 3 ISBN-registered books: «Triunfar o morir en el intento» (2013) and «De esclavo a dueño» (2023) · International keynote speaker for the HORECA sector.
Questions that land every week
How much does it cost to open a dark kitchen from scratch in 2026?
How much does it cost to open a dark kitchen from scratch in 2026?
Between 8,000 and 45,000 USD in startup investment, depending on the setup. A rented corner inside a working kitchen starts at 8,000-14,000 with used equipment, licenses and product photography; a specialized hub with extraction and cold storage asks 18,000-28,000; and your own converted warehouse or container reaches 45,000. Monthly rent runs separately: 600 to 3,500 USD by city and square meters.
What is the real commission for selling on Rappi or iFood?
What is the real commission for selling on Rappi or iFood?
Nominal contract commission runs 18% to 25% by country, category and whether you use platform couriers. The effective rate, which is the one that matters, reaches 28-33% once you add co-funded promos, weekend discounts and the boost that buys position. Always cost at 30% and treat any lower figure as a pleasant surprise, never as your starting assumption.
Is a cocina oculta better than a physical restaurant?
Is a cocina oculta better than a physical restaurant?
Depends what you are buying. The dark kitchen buys speed and low entry risk: break-even in 4-8 months on a tenth of the investment. The physical restaurant buys customer ownership and double the average ticket, in exchange for 14-26 months to break even. If your budget stops at 20,000 USD and you already cook for volume, start dark and build the brand from there.
How do you increase sales on Rappi without giving away margin?
How do you increase sales on Rappi without giving away margin?
Work three levers before touching discounts: the hero dish photo, which moves conversion more than any promotion; dispatch time, which the algorithm rewards above price; and a rating held above 4.7. Then, if you need a push, buy position with in-app placement measured weekly instead of cutting prices, because discounting trains your customer to wait for the discount.
Sector data 2026 (official sources)
Verifiable industry benchmarks from official, non-commercial sources (government, industry associations, market research) - not competitors.
| Metric | Benchmark 2026 | Source |
|---|---|---|
| Cuota conjunta de Serve, Starship y Nuro en flotas globales 2024 | 18% | Mordor Intelligence — Autonomous Delivery Robots Market 2024 |
| Mercado de entrega de paquetes por dron en 2023 | USD 585,9 millones | Grand View Research — Drone Package Delivery Market 2023 |
| Proyección de entrega de paquetes por dron a 2030 | USD 5.238,8 millones (CAGR 38,7%) | Grand View Research — Drone Package Delivery Market 2030 |
| Entregas comerciales por dron de Zipline (abril 2024) | 1 millón (primera empresa en lograrlo) | Grand View Research — Drone Package Delivery Market |
| Unidades de drones de reparto proyectadas 2024 a 2030 | de 32.456 a 275.703 unidades | Grand View Research — Drone Package Delivery Market |
| Cuota del delivery de comida en el mercado de drones 2024 | 36,87% | Grand View Research — Drone Package Delivery Market 2024 |
Related content
Put numbers on your cocina oculta before you sign anything
Whether you are evaluating a dark kitchen or already running one that bills without profit, the work order is identical: model, menu, cash. Start by costing the digital menu against a 30% effective commission and in one afternoon you will see which dishes carry the business and which ones are draining it.