A ghost restaurant makes money in one narrow way: it strips out the cost of a dining room and keeps more of every order than a dine-in restaurant can. That is the whole thesis. Delivery-only kitchens report margins in the 10% to 30% range, averaging around 15%, against roughly 3% to 5% for a traditional restaurant, according to OysterLink’s ghost kitchen statistics.
But that spread is not automatic, and it is not stable. Every order a ghost restaurant sells moves through somebody’s delivery network, and that network charges 15% to 30% of the ticket. A dine-in restaurant pays that fee on a slice of its sales. A delivery-only kitchen pays it on all of them. So the same structure that creates the margin also contains the one line item capable of erasing it.
This post is about that math, start to finish: what the model actually returns, where each dollar goes, and which levers move the number. If you want the other side of the story first, our breakdown of why ghost kitchens fail covers the structural traps that close these businesses, which is a different question from how the profitable ones stay open.
What this guide adds: Most articles on this topic quote a margin range and stop. This one walks the full P&L of a delivery-only kitchen at a realistic revenue level, shows what the commission tier you are on does to the bottom line, and separates the levers that actually move margin from the ones that just sound efficient.
Save 80% of delivery management time
We handle everything:
- Dedicated operations manager
- Real-time tracking dashboard
- Automated customer notifications
- Urgent issue resolution
The Bottom Line
Ghost restaurants report 10% to 30% profit margins, averaging about 15%, compared with 3% to 5% for a dine-in restaurant.
The margin comes from eliminating front-of-house cost: no dining room rent, no servers, no decor, no prime-location premium.
Delivery commission is the largest single threat to that margin. Published rates run 15% to 30%, and the real, all-in cost per order often lands between 30% and 40%.
At a 25% commission rate a well-run kitchen can hold 5% to 7% on a delivery order. At 30%, that falls to roughly 0% to 2%.
The profitable ghost restaurants are the ones that pull orders off the marketplaces and onto channels they own, keep the menu tight, and treat delivery-only as a tactic rather than the whole business.
What a ghost restaurant’s profit margin actually looks like
Two numbers matter, and people routinely confuse them.
The first is contribution margin on an order: what is left after food cost, packaging, the labor that made it, and the commission on that sale. The second is net profit margin: what is left after rent, insurance, software, marketing, and the owner’s time. Ghost restaurants look spectacular on the first and merely good on the second.
Here is how the two models compare on the cost lines that actually differ:
| Cost line | Dine-in restaurant | Ghost restaurant |
|---|---|---|
| Food cost | 28-35% of sales | 28-35% of sales |
| Labor | 25-35% of sales | 20-25% of sales |
| Rent and occupancy | Prime retail rate, full square footage | Kitchen-only rate, often shared or subleased |
| Front-of-house cost | Servers, hosts, decor, furniture, utilities | None |
| Delivery commission | Paid on delivery orders only | Paid on essentially every order |
| Typical net margin | 3-5% | 10-30%, averaging ~15% |
Food cost barely moves between the two models. A chicken sandwich costs what it costs. The savings are concentrated in labor and occupancy, and the new cost is commission. That is the entire trade, and whether it works depends on which of those two is bigger for your specific operation.
How much ghost restaurants make per month
Revenue for a single delivery-only kitchen in the US commonly runs between $50,000 and $150,000 a month, which nets somewhere between $5,000 and $45,000 in profit depending on how tightly the operation is run. Startup investment typically falls between $75,000 and $200,000 once you account for labor, supplies, insurance, and marketing, per OysterLink’s figures. Most operators who reach profitability get there inside 10 to 18 months, which is fast by restaurant standards and directly attributable to the lower build-out.
The market these businesses operate in is also less rosy than the headline forecasts suggest. Research firms project the global category growing at double digits toward $177 billion by 2032. Meanwhile GrowthFactor’s analysis cites IBISWorld putting the US market at $2.9 billion in 2025, with revenue down 5.2% in 2024. Global market forecasts and the reality of a US operator’s order volume are not the same number, and planning off the former is how operators end up with a kitchen they cannot fill.
The demand side has grown. The National Restaurant Association reported in April 2025 that nearly 75% of restaurant traffic is now off-premise, and DoorDash order volume rose 23% to 3.2 billion orders in 2025. There is no shortage of delivery demand. The question is how much of each order you keep.
Where a ghost kitchen’s overhead savings come from, and where they stop
The savings are real and they are specific:
No dining room. You are not paying retail rent on square footage that holds tables, and you are not paying to heat, light, clean, or decorate it.
No front-of-house labor. No servers, hosts, bussers, or bartenders. This is what pulls labor from the 25-35% typical of full-service down toward 20-25%.
Location stops mattering. A dine-in restaurant pays a premium for foot traffic. A delivery-only kitchen needs to sit inside a delivery radius, not on a high street, so it can take an industrial-park rate.
Smaller footprint. Less space to rent, equip, insure, and maintain.
Now the part most guides skip. Those savings get clawed back in ways that only show up once you are operating:
Shared-kitchen premiums. Commissary and ghost kitchen operators charge a higher rate per square foot than a standard commercial lease, and many charge a percentage of sales on top. Restaurant Dive’s teardown of the category calls this cost layering: platform commission, marked-up menu prices, sublease premium, and percentage rent all stacked on the same order.
Marketing you used to get free. A storefront generates walk-past awareness every day at no cost. A ghost restaurant has no sign anybody sees, so discovery has to be bought, either through platform promotion tiers or through your own advertising.
Packaging. Every order ships. Containers, bags, seals, and labels become a real line item instead of an afterthought.
The honest version: a ghost restaurant trades fixed cost for variable cost. That is a good trade when volume is uncertain and a bad one when volume is high, because variable costs never stop scaling.
Delivery commission is the largest line item on the P&L
Published marketplace rates for 2026 cluster in the same bands. According to Rezku’s breakdown of third-party fees, DoorDash charges roughly 15% for self-delivery, 25% for partner delivery, and 30% for premium placement; Uber Eats runs 15%, 25%, and 30% across its tiers; Grubhub starts near 15% and climbs toward 30% with marketing add-ons.
Those are the sticker prices. OPA’s cost analysis puts the all-in figure at 30% to 40% per order once processing, promotion, and add-on fees are counted, and notes commissions reaching 43% in heavily regulated markets like New York City.
What that does to an order is not subtle. Labrador AI’s margin analysis puts it plainly: at 25% commission a well-run kitchen can hold 5% to 7% margin on a delivery order, and at 30% that falls to 0% to 2%. You are covering costs and buying volume.
Here is the same idea as a full P&L. QSR Pro’s 2026 review models a delivery-only kitchen doing $50,000 a month:
| Line item | Monthly cost |
|---|---|
| Revenue | $50,000 |
| Platform commission at 25% | $12,500 |
| Food cost at 30% | $15,000 |
| Kitchen rent | $5,000-$8,000 |
| Labor | $10,000-$12,000 |
| Marketing | $2,000-$5,000 |
| Net result | Negative to 7% |
That bottom row is the central fact of this business. A ghost restaurant running entirely on marketplace orders at a mid-tier commission does not reliably clear the 15% average. The operators who hit 15% and above are doing something structurally different, and it is almost always the same something.
The four levers that widen a ghost restaurant’s margin
Move orders onto channels you own
This is the lever. Every order that arrives through your own website or app instead of a marketplace keeps the 15% to 30% you would otherwise hand over. Sauce reports one operator that now takes 51% of its online orders directly, generating over $863,000 in direct sales and avoiding close to $130,000 in commission in a single year.
Stay on the marketplaces anyway. They are a discovery channel, and for a brand nobody has heard of they are the only discovery channel. The move is to treat them as customer acquisition and then work to make the second order a direct one, using insert cards, loyalty offers, and direct-channel pricing set a few percent below the app menu.
Keep the menu to 8-12 items
A short menu is not a compromise, it is a margin tool. Fewer SKUs means less inventory sitting idle, less waste, faster ticket times, and fewer chances to make an expensive mistake during a rush. The most disciplined operators in 2026 run a focused list of roughly 8 to 12 items per brand. Complexity is the enemy in a kitchen where multiple brands may share one line.
Control the delivery itself on direct orders
Once a customer orders from you directly, the delivery is yours to arrange, and that is where the commission saving actually gets banked. Some operators hire drivers outright; that improves per-order economics but adds vehicles, insurance, scheduling, and management. Others run a hybrid model, handling planned and recurring direct orders through their own logistics while leaving marketplace orders on the marketplace couriers.
This is the part of a ghost restaurant’s cost structure that food businesses most often leave unexamined. A delivery platform built for local food businesses, like Metrobi, handles the routing and driver side of planned direct orders with multi-stop route optimization and the ability to keep working with drivers who already know your operation, which removes most of the reason operators default back to paying commission on everything.
Price the delivery menu as its own menu
Delivery prices do not have to match what a walk-in would pay, and for a delivery-only kitchen there is no walk-in to be inconsistent with. A 15% to 20% markup on marketplace menus is standard practice and exists precisely to offset commission. Charging the same price on a channel that costs you 25% more to sell through is a choice to lose money.
Ghost kitchen profit margin by operating setup
Where you cook changes the cost structure, and the three common setups behave differently:
Hosted commissary or shared ghost kitchen facility. Lowest upfront cost, fastest to open, highest ongoing rate per square foot. Good for testing a concept, expensive as a permanent home.
Your own leased kitchen. Higher build-out and a longer lease commitment, but you control the rate and there is no percentage-of-sales rent skimming the top line.
A kitchen you already own. The strongest economics by a wide margin. A restaurant, bakery, or caterer launching a delivery-only brand out of existing capacity during slow hours adds revenue against rent and equipment that are already paid for.
That last one is where most of the durable profit in this category now sits, which is the opposite of how ghost kitchens were originally sold.
What separates the profitable ghost restaurants in 2026
The pure-play version of this model, an independent operator launching a delivery-only brand from a rented kitchen with no existing customers, has largely not worked. Funding for the category fell 95% year over year in Q4 2025, from $210 million to $10.5 million, and several of the biggest facility operators have contracted sharply.
What survived breaks into three recognizable groups, per QSR Pro’s review:
Established brands using delivery-only sites tactically. Chains like Chipotle and Sweetgreen open delivery-only locations to reach dense urban areas, carrying brand recognition and marketing spend they already have.
Multi-brand operators running several concepts from one kitchen. This group captured roughly 48% of global ghost kitchen revenue, and it works because the fixed cost of the kitchen gets spread across several revenue streams.
Hybrid facilities that added pickup. Food halls, kiosks, and counters bolted onto delivery kitchens, which is an admission that some customer-facing surface is worth having.
The pattern across all three: a ghost restaurant works as a tool, not as a whole business. It is an efficient way for a food business that already has a kitchen, a brand, or a customer base to add delivery revenue. It is a hard way to build one of those three from nothing, because the commission that funds your discovery also caps your margin, and nothing about a delivery-only kitchen fixes that on its own.
So if you are running the numbers on a delivery-only concept, the question is not whether ghost restaurants can be profitable. They demonstrably can. The question is what you already own, a kitchen with idle hours, a brand people recognize, a customer list, or a way to deliver your own orders, because that asset is where the 15% comes from. Without one of them, you are buying revenue at 25% and hoping the rest of the P&L is generous.
