A dark kitchen — a second, delivery-only brand run out of your existing kitchen, with no dining room and no sign on the street — sounds like the easiest revenue you'll ever make. You already pay the rent. The cook is already on shift. Whatever comes in on top is pure profit. That feeling is half right, and the other half comes down to three numbers most owners never put side by side: what the delivery platform actually takes off every order, how much genuinely idle kitchen time you have, and what happens the moment that second brand gets more popular than your idle time can carry.
A dark kitchen is nothing more complicated than a kitchen that cooks for delivery only — its own name, its own menu, its own profile on Deliveroo, Uber Eats or Just Eat, but no premises of its own, no dining room and no staff of its own if it's run out of a kitchen you already have. For anyone who already runs a restaurant, that's exactly the appeal: you already have the kitchen, the cooks and the permits. All you're adding is a second brand that uses burners between lunch and dinner, or on a quiet Tuesday night, that would otherwise just sit there.
The problem is that almost every dark-kitchen guide online is written for the other situation: someone renting an empty commissary unit and starting from zero, with their own rent, their own full crew, and a complete P&L that has to work from day one. That maths doesn't apply to you. Rent and base labour get paid whether the second brand exists or not — so the question isn't 'can this carry a whole restaurant', it's a much smaller, more honest one: what's left per order once commission, food cost and packaging come off, and how many of those orders can you run before your idle hours run out?
This guide works that out in seven steps: how much idle time your kitchen genuinely has, what a delivery platform actually takes, what food cost and packaging cost together, the margin left after that, the ceiling where 'idle' stops meaning idle, how many orders you need to break even, and the reputation risk that shows up on no P&L at all. At the bottom, plug in your own numbers and find out whether your idle hours can carry a second brand — or whether they're already being eaten by it.
Everything runs in your own browser: nothing is sent anywhere and nothing is stored. The figures in this article are a realistic guideline for an independently run European restaurant — your own menu, platform and kitchen capacity are the final judge.
Why the standard dark-kitchen maths doesn't apply here
Every guide, calculator and video about dark kitchens you'll find online is written for someone starting from zero in a rented commissary kitchen: their own rent, their own full crew, their own equipment still to be bought. For that person, every euro of rent and every hour of labour is a real, new cost, and it's a fair question whether the whole thing is worth it as a standalone business.
For a restaurant that already exists, that's the wrong sum. Your rent, your base staff, your insurance and your equipment get paid whether that second brand exists or not. What a dark kitchen out of an existing restaurant actually costs isn't that full P&L — it's only what genuinely gets added: the commission the platform holds back, the extra ingredients, the packaging, and — if orders fall outside your idle hours — the extra minutes of labour that would otherwise have gone somewhere else.
That confusion between 'full cost' and 'extra cost' is exactly why so many owners either dismiss a second brand too fast ('that commission alone kills it') or launch it too casually ('the kitchen's already there, so it's all profit') — both wrong, for the same reason: they're running the wrong kind of cost through the sum.
The ultimate guide Restaurant Finance: 6 Numbers That Decide Your Profit Prime cost, cash flow, break-even and RevPASH: the complete financial system, in plain language. Open the guideThe 7 numbers behind a dark kitchen
They build on each other: from the idle time you already have, through what a platform and your ingredients take off it, to the ceiling where that idle time runs out and the risk that shows up on no invoice at all.
1. Idle kitchen time — the raw material a dark kitchen actually spends
Before you calculate a single euro of platform commission or food cost, there's only one question that matters: how many hours does your kitchen genuinely sit idle, and how much of that is actually usable? Almost every restaurant has a gap between lunch and dinner service, or a couple of evenings a week that are structurally quieter than the rest — and that gap is the only raw material a dark kitchen really consumes.
Make the distinction between the raw gap and what you can cautiously run on top of it straight away. A kitchen with no dinner guests between 2pm and 6pm has four free hours on paper — but in practice, part of that is cleanup, mise en place for dinner, or staff on their break. Plan a first, cautious trial period around a fraction of the raw gap, not the whole thing — the graphic below shows both side by side.
It's exactly this distinction that most online dark-kitchen calculators skip: they only ask 'how many hours does your kitchen sit idle' and run straight ahead with that, as if every idle hour is automatically a usable one too.
For a kitchen open 11:00–23:00: dinner service, the idle gap, and what's actually usable from it.
- Dine-in service (lunch + dinner) — 50%
- Idle gap (raw estimate) — 42%
- Cleanup, mise en place, breaks — 8%
The raw gap on the clock is bigger than what you can actually run on it in practice. The tool at the bottom of this article deliberately starts from a fraction of the raw gap, not the whole thing — exactly the mistake most online calculators make.
2. Platform commission — what actually comes off every order
Deliveroo, Uber Eats and Just Eat earn on every order through a commission on revenue, not on profit — and for a restaurant not using its own drivers, that commission usually sits between 20% and 30% of the order value, before any extra marketing or listing fees. That's not a rounding error; it's nearly a third of every euro a guest pays, gone before you've even reached the ingredients.
Our guide to cutting delivery costs goes deeper into negotiating or avoiding that percentage on your existing menu; here it simply counts as a fixed deduction on the second brand, exactly as the platform itself reports it.
3. Food cost and packaging — the two costs on every single order
After commission come the two costs that hit every individual order: the ingredients, and the packaging a dine-in dish never needs. Food cost is calculated the same way as on your regular menu — as a percentage of the order value — but packaging is a fixed cost per order that's easy to forget: a box, a lid, cutlery, a bag, often €1 to €2 together for an average main course.
Both belong in the same sum, and both should be kept deliberately low on a dark-kitchen menu: pick dishes that survive the ride in a delivery bag without losing quality, and packaging that doesn't copy the plate but does protect it — a dish that arrives soggy, cold or collapsed costs you more in a bad review than the cheaper packaging ever saved.
4. The contribution margin per order — the number that's left
Subtract commission, food cost and packaging from the order value, and what's left is the margin this second brand actually earns you — not the gross margin on your dine-in menu, which doesn't apply here because commission and packaging aren't part of it. Use the tool at the bottom of this article to see what figure that produces for your own menu.
The graphic below breaks that down: of every euro a guest pays on the platform, a slice goes to the platform, a slice to the ingredients, a small slice to packaging — and what's left is what you genuinely keep from those idle hours.
Of every euro a guest pays on the platform for the second brand: commission, ingredients, packaging, and what's left.
- Platform commission — €7 (28%)
- Food cost — €8 (30%)
- Packaging — €1.25 (5%)
- What's left — €9 (37%)
Compare this deliberately with your dine-in margin on the same dish: that one has no commission and no packaging, and is therefore always higher. A dark-kitchen menu priced the same as your dine-in card is almost never profitable.
5. The ceiling — the point where 'idle' stops meaning idle
This is the number most owners forget, and it's the most important of the seven: your idle hours aren't infinite. The moment the second brand gets popular enough to pull in more orders than your idle capacity can carry, the exact reason it looked 'free' in the first place disappears — every extra order now competes with a dine-in guest for the same stove, the same pass and the same cooks, at a moment when a dine-in guest pays full margin and a delivery order doesn't.
So work out your idle capacity just as concretely as everything else: how many hours a week are genuinely idle, and how many orders can one cook realistically turn out in that hour without quality or the dinner service suffering? Multiply those two numbers and you have your ceiling in orders per week — the line above which this stops being extra revenue and becomes shifted revenue at a lower margin.
Once you're approaching that ceiling, there are only two honest options: schedule extra staff for those hours — which changes the sum from step 4, because that's now a genuine extra cost — or deliberately throttle order volume through the platform, so the second brand stays small and profitable instead of large and loss-making.
6. Break-even — how many orders cover the fixed costs
A second brand also carries costs that don't fall per order but per week or per month: professional food photography for the platform profile, a listing fee, the occasional boost or ad spend on the platform, and the time someone spends keeping the menu up to date. Add those up to a weekly figure and divide by the contribution margin per order from step 4, and you get the number of orders you need to break even.
Compare that break-even figure straight away with the ceiling from step 5. If break-even sits comfortably below the ceiling, there's enough room to cover the fixed costs without eating into your idle hours. If break-even sits above the ceiling, this second brand simply cannot become profitable without either extra staff cost or a pricier menu — and that's exactly the kind of sum that needs to happen before launch, not three months after.
7. The reputation risk that shows up on no invoice at all
There's one cost that doesn't appear in any of the six numbers above, and it weighs at least as much: a badly run second brand gets bad reviews on a platform you can't put right with a friendly word at the table. A dish that arrives cold, an order that took too long because dinner service took priority, or packaging that leaked — that becomes a star rating the platform itself uses to decide how often you're shown again, without you ever getting to speak to that guest.
Worse still, platforms and guests are increasingly quick to connect a new, unfamiliar brand back to the address or the kitchen it comes from. A second brand that consistently underperforms is then no longer an isolated risk — it can reflect back onto your main restaurant's name, the exact brand you least want to put at risk.
The practical lesson: only launch the second brand once the kitchen can genuinely handle it calmly, test internally before going live, and treat a slipping platform rating with the same urgency as a bad Google review on your main restaurant — with our review-reply generator where the platform allows it, and otherwise by adjusting the menu or the hours before the problem builds up.
Run the numbers on your own dark kitchen
Enter the average order value, your platform's commission, your food cost and packaging cost, and how much idle time your kitchen genuinely has. The figures are pre-filled with this article's example, so you can see straight away how it reads — overwrite them with your own.
You'll get the margin per order, the ceiling in orders per week your idle time allows, and what that ceiling earns you a year at full use — plus a verdict that says whether you're staying inside your idle hours or already going past them.
Dark-kitchen scan
Order value, commission, food cost, packaging and idle capacity — in one margin-and-ceiling figure.
—
Idle capacity is rarely known exactly; start with a cautious estimate rather than the raw gap on the clock. Everything runs in your browser; nothing is sent or stored.
Two things worth remembering when you read this. A dark kitchen's margin is never your dine-in margin — commission and packaging always belong in it, and forgetting that prices the second brand wrong from day one. And the ceiling isn't a suggestion: it's the point where idle time runs out and every extra order starts displacing a dine-in guest instead of filling empty hours.
This number resolves the same way as the other invisible costs and revenues on this site: you have to see it before you can act on it. Work it out before you set up a platform profile, not after the first orders are already coming in.
What to do this week, this month and this quarter
Nobody works out a dark kitchen fully in one sitting. This order works, because each step makes the next conversation more concrete.
This week — run the scan above with your own numbers
- Count your genuinely idle hours per week, and be cautious: use a fraction of the raw gap, not the whole gap on the clock.
- Ask the platform you're considering for its commission structure — it varies by contract and by region, and the published percentage isn't always the final one.
- Put the margin per order next to your prime cost on your dine-in menu — those two numbers belong in the same conversation.
This month — design a menu that survives the ride
- Pick three to six dishes that are cheap and quick to package and that survive delivery time without losing quality.
- Test internally: leave a dish in its packaging for thirty minutes and taste it yourself before it reaches one paying guest.
- Check whether catering out of your own kitchen is a better-fitting second revenue stream than delivery — our guide to starting catering covers exactly that alternative.
This quarter — watch the ceiling, not just the revenue
- Track weekly order volume against your calculated ceiling — not just revenue, which hides the moment you start displacing dine-in guests.
- Revisit the scan as soon as the platform changes its commission or you notice orders structurally falling outside your idle hours.
- Answer every mediocre or bad platform rating as fast as you'd answer a Google review — the second brand needs the same reputation as the first.
Idle hours are a raw material, not a free gift
A dark kitchen out of your own kitchen can be a genuine, profitable second income stream — but only if you treat it as what it is: a way of selling a limited, countable raw material, namely your idle hours. The moment that raw material runs out, the whole sum changes, and anyone who doesn't work that moment out in advance discovers it only once dinner service starts feeling it.
The sum itself isn't complicated: subtract commission, food cost and packaging, set that against the ceiling of your idle time, and check it against the fixed costs. What most owners skip isn't the difficulty of the sum — it's ever doing it at all.
Then do the same with the rest of your P&L. Our guide to cutting delivery costs unpacks the commission side of the same question further, and together with your prime cost and your catering strategy, they give you the full picture of what your kitchen can genuinely earn outside regular service.