Is a dark kitchen actually profitable?
A dark kitchen can be profitable, but the margin is thin and fragile: in most realistic setups a delivery-only kitchen nets somewhere in the low single digits to low teens as a percentage of revenue — and the two variables that decide which end you land on are aggregator commission and delivery cost. Strip rent and dining-room labor out of the P&L and a ghost kitchen looks cheap to run, but those savings are easily eaten by 15-35% platform commissions and per-order delivery fees. Profitability is therefore less about the kitchen itself and more about your channel mix and how you move the food the last mile.
The dark kitchen P&L, line by line
Let's build the P&L the way a controller would — top to bottom, every line as a percentage of revenue. The numbers below are an illustrative example for a hypothetical kitchen, not real data or a cited study. They're chosen to show how the lines interact, and the ranges are general industry observations you can sanity-check against your own books.
Revenue and channel mix
Start with the most important split in the whole model: where the order comes from. An order placed on a third-party aggregator and the same order placed on your own app generate identical food, identical labor, and identical packaging — but wildly different contribution. Say our hypothetical kitchen does an average order value (AOV) of €22 and 1,000 orders a month, for €22,000 in gross sales. The mix is 70% aggregator, 30% direct. That single ratio will swing the bottom line more than any cost-cutting you do downstream.
Aggregator commission drag
Aggregator commissions commonly run 15-35% of order value depending on market and whether delivery is included. On 70% of €22,000 (€15,400) at, say, 28%, that's roughly €4,312 of commission — about 20% of total revenue gone before you've bought a single ingredient. The 30% direct slice (€6,600) pays no commission. This is why operators obsess over reducing aggregator commission fees: every point of mix you shift from platform to direct drops almost straight to contribution.
Food cost (COGS)
Food cost often lands at 25-35% of revenue. Take 30%: €6,600. Dark kitchens have a structural advantage here if they run tight portioning and recipe discipline — there's no walk-in traffic, so production is more predictable and waste is easier to control. But delivery-only also punishes inconsistency: a soggy or wrong item triggers a refund and a bad review you can't recover with table-side service.
Packaging
Easy to overlook, surprisingly material. Quality delivery packaging — leak-proof containers, insulation, tamper seals — typically runs 3-6% of revenue. Call it 4%: €880. Cutting corners here shows up as cold food and 1-star reviews, so this is rarely the line to squeeze.
Labor
Kitchen labor in a delivery-only operation is leaner than a full-service restaurant — no front-of-house, no servers — but you still need cooks and prep. Figure 20-30% of revenue. At 25% that's €5,500. This is also where multi-brand utilization pays off: running two or three virtual brands out of one kitchen spreads the same labor and equipment across more orders without proportionally more staff.
Rent and utilities
This is the headline dark-kitchen advantage. No prime high-street location, no dining room, no premium frontage — just a production unit in a cheaper zone. Where a restaurant might pay 8-12% of revenue in rent, a ghost kitchen can run 5-8% all-in including utilities. Say 6%: €1,320. If you're weighing the model end to end, our guides on how to start a ghost kitchen and the dark kitchen market in 2026 go deeper on siting and demand.
Delivery cost
Here's the second swing factor. On aggregator orders, delivery is usually bundled into the commission — convenient, but you don't control the cost or the courier experience. On direct orders, you pay for the last mile. Third-party on-demand couriers can cost €4-7 per drop; an own courier fleet can be cheaper per order at density but carries fixed cost when volume dips. For our 300 direct orders, an own fleet at an effective €3.50/order is €1,050; the same orders via a per-drop third-party at €5.50 would be €1,650. That €600 gap is pure contribution — and it's why controlling your own delivery is a lever, not just a logistics preference.
Marketing and CAC
Aggregators bring "free" demand — except you pay for it in commission forever. Direct channels need acquisition spend up front but the customer is then yours to re-market to at near-zero cost. Budget 5-10% of revenue while you're building a direct base; here 7%: €1,540. The right way to read this line is alongside food delivery analytics metrics like LTV-to-CAC, not as a standalone cost.
Payment and fiscal fees
Card processing and fiscal/compliance fees are small but real — roughly 1.5-3% of revenue. At 2.5% that's €550.
Adding it up: does this kitchen make money?
For the illustrative example: revenue €22,000, minus commission €4,312, food €6,600, packaging €880, labor €5,500, rent €1,320, own-fleet delivery €1,050, marketing €1,540, fees €550 — total costs about €21,752. Net: roughly €248, or about 1% margin. Razor-thin.
Now flip two levers. Move the mix from 70/30 aggregator-direct to 40/60. Commission on the smaller aggregator slice (40% of €22,000 = €8,800 at 28%) falls to ~€2,464 — a €1,848 saving. Delivery cost rises a little because more direct orders mean more last-mile you pay for (600 direct orders × €3.50 = €2,100, up €1,050), and marketing ticks up to support direct. Net the two big moves and contribution still improves by roughly €800-1,000, pushing margin toward 5-6% on the same sales. Add a second virtual brand to lift utilization on the existing labor and rent, or nudge AOV from €22 to €25 with smarter upsells, and the same kitchen can reach low-teens margin.
The biggest levers, ranked
- Direct-channel share. Every point shifted off aggregators is a point of commission saved. This is the single highest-leverage number in the model — see own ordering channel vs aggregators.
- Delivery cost control. Owning the fleet turns a variable per-drop fee into a managed cost you can optimize with batching and zone density.
- AOV. Fixed costs per order (packaging, delivery, processing) don't scale with basket size, so a higher AOV dilutes them.
- Prep efficiency and multi-brand utilization. More orders through the same labor and rent footprint is how dark kitchens earn their structural cost advantage.
The thread connecting all four is measurement. You can't manage commission drag, delivery cost per order, or LTV by channel if your P&L only shows monthly totals. Platforms like Toster are built around exactly this — unit-economics and P&L analytics down to the order, RFM and LTV segmentation to see which channel actually retains customers, and an own-fleet dispatcher so delivery cost is a number you steer rather than a line you absorb. Run the numbers per brand and per channel and the profitability question stops being "is a ghost kitchen profitable" in the abstract and becomes "which of my orders are profitable" — which is the only version of the question you can actually act on.
Frequently asked questions
What is a good profit margin for a ghost kitchen?
It varies widely, but a healthy delivery-only kitchen often targets a high-single-digit to low-teens net margin once it has shifted meaningful volume to direct channels. Heavy aggregator reliance commonly compresses that to low single digits or worse, because 15-35% commissions sit on top of normal food and labor costs.
Why are aggregator commissions such a big deal in dark kitchen economics?
Because they're charged on gross order value, not profit. A 28% commission on a kitchen running 30% food cost and 25% labor can be larger than the food itself — so it often determines whether the order contributes anything at all. Shifting orders to a direct channel removes that line entirely.
Is an own courier fleet cheaper than third-party delivery?
At sufficient order density, usually yes — an own fleet converts a per-drop fee into a managed cost you can optimize through batching and tight delivery zones. The trade-off is fixed cost: at low volume, idle couriers can make third-party on-demand the cheaper option. The break-even depends on your orders per hour per zone.
How does running multiple brands from one kitchen help?
Multi-brand (virtual brand) operation spreads the same rent, equipment, and core labor across more orders. Since rent and base staffing are largely fixed, additional revenue from a second or third brand drops a higher share to contribution — provided menus share enough ingredients to avoid inflating food cost and prep complexity.