Ghost Kitchen vs Brick & Mortar (The Honest Cost Comparison)

Ghost kitchens (sometimes called virtual brands or cloud kitchens) get pitched as the future of restaurants. Lower startup cost, lower labor, fatter margin. Some of that holds up. The part the pitch deck skips is the failure rate, which runs up to 60% in some markets according to recent reporting (Restaurant Business Online, CloudKitchens disclosures). So here’s the honest comparison, line by line.

Startup cost #

Ghost kitchen: $75,000-200,000

Most ghost kitchens lease space in a shared facility like CloudKitchens, Kitchen United Mix, or Reef Technology. The buildout is bare bones: equipment, hood, storage, and that’s about it. No dining room, no bar, nothing street-facing for a customer to ever see.

Brick and mortar: $750,000-1,000,000+

Full buildout means dining room, restrooms, a bar if you have one, exterior signage, parking improvements, ADA compliance, full HVAC, and every finish a guest is going to look at. It swings hard by market too. Coastal urban can hit $1.5M+, while suburban runs $500K-$800K.

The difference: Ghost kitchen startup is 75-85% lower than full brick-and-mortar. That’s a real number, and it’s the whole reason the pitch sounds so good. The catch shows up later, on the costs you don’t see in a buildout quote.

Labor cost #

Ghost kitchen: 20-25% of revenue

There’s no front-of-house at all. No servers, no host, nobody behind a bar. Just kitchen staff and a couple of people assembling and packing orders. That thin crew is the second-biggest savings against brick-and-mortar.

Brick and mortar: 28-35% of revenue (NRA 2024 medians)

Full service runs about 1 server per 12-30 covers, and then you’re stacking a host, bussers, runners, the kitchen, and management on top of that. The payroll adds up fast.

The difference: Ghost kitchen labor runs 8-12 points lower. Real money on the labor line.

Rent and occupancy cost #

Ghost kitchen: 8-15% of revenue (shared kitchen space)

You’re renting less square footage and you don’t need a spot anybody can find. Think industrial parks and edge-of-town buildings, sometimes a second floor. Some operators even run a commission structure where you pay a percentage of revenue instead of fixed rent.

Brick and mortar: 6-12% of revenue

A dining room needs real estate people actually pass by. Foot traffic, visibility, and parking all matter, because nobody’s driving out to a second-floor industrial park for date night.

The difference: Ghost kitchen rent can run higher OR lower than brick-and-mortar depending on the deal you sign. I’ve seen shared-kitchen operators charge percentage-based rent that ends up costing more than fixed rent once your sales are healthy. So the “lower rent” line is conditional, and you should treat it that way before you sign anything.

The hidden cost: third-party delivery commission #

This is the line item that wrecks the ghost kitchen pitch for a lot of operators.

Ghost kitchens are 60-95% delivery-driven, and delivery means platform commission. DoorDash, Uber Eats, and GrubHub all-in cost runs 30-40% of order value once you add up commission, processing, promotions, and refunds. See Third-Party Delivery Real Cost for the full math.

A ghost kitchen pushing 90% of revenue through third-party platforms is handing over a 27-36% take rate off the top before it pays for a single thing. A brick-and-mortar usually runs 10-30% of revenue through third-party and keeps the rest as dine-in or first-party pickup, so its blended platform cost lands much lower, around 3-12% of total revenue.

That commission is what eats the labor and rent savings ghost kitchen brags about. Here’s the math.

Ghost kitchen P&L (typical):

• Food cost: 30%
• Labor: 22%
• Rent: 10%
• Platform commission (effective): 27%
• Packaging: 2%
• Operating overhead: 5%
• Total cost: 96%
• Net margin: 4%

Brick-and-mortar P&L (typical):

• Food cost: 32%
• Labor: 30%
• Rent: 8%
• Platform commission (effective, 25% mix): 8%
• Packaging: 0.5%
• Operating overhead: 6%
• Total cost: 84.5%
• Net margin: 5.5%

So the ghost kitchen margin that gets sold as 10-15% on paper usually lands at 3-5% once you’re actually running it. Brick-and-mortar at 4-7% sits right in the same range, sometimes better. Not the gap you were promised.

Profit margin: claimed vs actual #

Ghost kitchen claimed margin: 10-15% (CloudKitchens marketing, industry promotion)

Ghost kitchen actual margin (post-commission reality): 3-8% depending on platform mix

Brick-and-mortar margin: 3-9% (NRA 2024)

Put the platform commission back in and the claimed advantage is gone. A ghost kitchen that builds its own direct ordering, meaning its own website plus delivery, ChowNow at 10-15% all-in, and off-platform marketing, can actually grab the on-paper advantage. The ones leaning on DoorDash and Uber Eats never get there.

The 60% ghost kitchen failure rate #

Restaurant Business Online reporting (2023-2024) and CloudKitchens disclosures put the ghost kitchen failure rate at up to 60% in some markets. Brick-and-mortar is no picnic either, but its first-three-years failure rate sits closer to 30-50%. Ghost kitchen still loses that race.

Why ghost kitchens fail at higher rates:

1. No customer-direct relationship. The platform owns the customer, full stop. The ghost kitchen can’t remarket to them, can’t build a loyalty program, can’t get them back next week on its own terms.

2. Platform dependency. When DoorDash changes its algorithm or bumps your commission tier, the ghost kitchen has no fallback. A brick-and-mortar at least has dine-in customers walking through the door no matter what the app does that month.

3. Quality is harder to control. Food sits in a bag for 25 minutes between the kitchen and the doorstep, and even a great plate degrades in there. Serve that same plate at a table and it’s still hot. That’s a real edge, not a marketing point.

4. Marketing is harder. The customer’s relationship is with the platform brand, not yours. Building a name through somebody else’s app feed is a lot harder than building it on a corner people drive past every day.

5. Volume isn’t there. Most ghost kitchens forecast off optimistic platform reach, and then actual delivery volume in the market comes in under projection. No volume, and the fixed costs never get absorbed. That’s how you bleed out slow.

The break-even reality #

A ghost kitchen needs roughly $650,000-$800,000 a year in revenue to be viable, depending on the market and concept. Drop below that and the fixed costs, meaning rent plus kitchen lease plus insurance plus minimum labor, eat the whole contribution margin.

Brick-and-mortar needs more, around $900,000-$1,200,000 a year. So yes, the ghost kitchen breaks even sooner. The trade is that its ceiling is lower too, because platform reach caps how big you can get.

Same break-even formula drives both of them (see Restaurant Break-Even Formula). The only thing that changes is the cost structure you feed into it.

When ghost kitchen makes sense #

1. Idle kitchen capacity in an existing restaurant. This is the version that actually prints money. You run a virtual brand out of your existing kitchen during the slow hours, your fixed costs are already covered, and the extra revenue comes in at marginal cost. If you already have a kitchen sitting half-used, do this.

2. Test a concept before brick-and-mortar. A ghost kitchen lets you prove demand for a new idea at 10-20% of the capital you’d risk on a full build. If the virtual brand pops, you graduate it to brick-and-mortar with real numbers behind you instead of a hunch.

3. Delivery-optimized concepts. Wings, pizza, sushi rolls, poke bowls, sandwiches, salads. Food that survives the drive, packs clean, and prices into a delivery-friendly ticket around $25-40 average.

4. Markets with strong direct-order infrastructure. Some markets have a higher first-party order share and lean on the platforms less. Check yours before you commit, because this one swings the whole math.

When brick-and-mortar wins #

1. Concepts with high beverage mix. This is the one I’d put money on. Bar programs carry margin, and a cocktail does not travel through a delivery app. I ran bars for five years, and a 35% beverage mix restaurant has structural advantages a ghost kitchen can’t touch.

2. Experience-driven concepts. Anything where the dining room IS the product. Steakhouse, sushi bar, wine bar, fine dining, the special-occasion stuff. Box that up for delivery and you’ve thrown away the whole point.

3. Average ticket above $35. Bigger tickets clear the platform commission math easier, and they get the dine-in margin structure on top of it.

4. Strong local brand potential. Brick-and-mortar builds a name people own in their heads. Ghost kitchen rents that recognition from a platform feed, and the day the feed changes, it’s gone. One compounds. The other resets.

What this looks like in the calculator #

The break-even calculator will model either concept once you plug in fixed costs, variable cost structure, and average check. Run both, then put the break-even revenue and contribution margins side by side.

For the commission piece, run the delivery profit calculator on a typical ticket and see what actually lands in your pocket.

What to do today #

Thinking about a ghost kitchen? Run the break-even math with a realistic platform commission of 30-35% all-in, not the headline 15% somebody sold you. Leaning toward brick-and-mortar? Run it with real labor at 28-35%, the NRA-typical band, not the optimistic 24% that makes the spreadsheet smile. The answer is almost never “this one’s just better.” It’s “this concept, in this market, at this scale, pencils out better one way.”

The pitch that ghost kitchen wins every time is wrong. The on-paper math is wrong too, because it leaves out the commission that’s eating you alive. The honest number sits in the middle, and it depends entirely on your specifics. Run your own.

Sources: NRA 2024 Industry Factbook, Toast Industry Reports, CloudKitchens, Restaurant Business Online, Lunchbox third-party commission analysis.