A $12 burger combo sold through a delivery app can net as little as $8.25 after platform commissions, promotional discounts, and processing fees — before a single ingredient or minute of labor is counted. Now multiply that math by three virtual brands cooking out of the same 900-square-foot kitchen, sharing the same fryer, the same walk-in cooler, and the same two line cooks. If your books can't tell you which of those three brands actually made money this month, you're not running three businesses. You're running one very confusing black box that happens to answer to three different names.
That's the defining accounting problem of the ghost kitchen model, and it's only getting more common. The delivery-only kitchen market has grown into a business worth well over $90 billion globally in 2026, expanding at a double-digit annual clip as more operators launch second and third virtual concepts to squeeze more revenue out of a single lease. The operational playbook — one kitchen, multiple brands, all revenue routed through third-party apps — has matured fast. Most operators' bookkeeping has not kept up with it.
Why Ghost Kitchen Accounting Breaks Traditional Restaurant Bookkeeping
A single-location, single-brand restaurant has it comparatively easy: one P&L, one point-of-sale feed, one rent line, one payroll run. A ghost kitchen operator running two or three virtual brands out of one commissary has to answer a much harder question for every dollar that comes in or goes out — which brand does this belong to?
Three things make that question genuinely hard to answer:
Multi-platform, uneven revenue. Orders arrive from DoorDash, Uber Eats, and Grubhub on different payout schedules, each taking its own cut before the deposit ever hits your bank account. If you only look at what lands in the bank, you're accounting for net cash, not gross sales — and you lose the ability to see how much each platform is actually costing you.
Shared everything. Rent, utilities, the walk-in cooler, the fryer station, and often the labor itself are shared across brands. Unlike a multi-location restaurant group where each site has its own four walls, a ghost kitchen's cost structure doesn't naturally split along brand lines. Someone has to draw that line manually — and if no one does, every brand's "profit" is really just a guess.
Commingled inventory. A commissary running a burger brand and a chicken-wing brand off the same fryer oil and the same walk-in is buying overlapping ingredients (buns, sauces, packaging) that get pulled for either concept depending on the night's order mix. Without a system that tracks which ingredient went to which ticket, food cost becomes a kitchen-wide number instead of a per-brand one — which means you can't tell if Brand A is subsidizing Brand B's losses.
Layer on the fact that many ghost kitchens use accrual-adjacent hybrid tracking (deposits hit the bank on a delay, but the sale happened days earlier) and it's easy to see why so many virtual-restaurant operators find out a brand was unprofitable only after months of running it.
The Real Cost of Third-Party Delivery Platforms
Before you can allocate costs across brands, you need an honest number for what delivery platforms actually take. The advertised commission rates — typically 15% to 30% depending on the plan tier — understate the real hit. Once you add payment processing fees, required promotional discounts, and refund/chargeback absorption, the effective cost commonly lands at 30% to 40% of the order total.
That gap matters enormously for multi-brand accounting, because it means gross revenue and net deposits are two completely different numbers, and your books need to capture both. A ghost kitchen brand that looks fine on "money in the bank" can be quietly bleeding out because the true commission rate crept up when the platform pushed a mandatory promotion.
The fix is to categorize every platform payout into its component parts rather than booking one lump "delivery deposit" line:
- Gross order revenue — what the customer paid, before any deductions
- Platform commission — the base percentage fee
- Promotional discounts — often funded partly or fully by the restaurant, not the platform
- Payment processing / service fees
- Refunds and chargebacks
- Net deposit — what actually lands in your account
Most delivery platforms provide this breakdown in a downloadable settlement report. If your bookkeeping only records the net deposit, you're structurally unable to see which brand's promotions are eating its margin — which is exactly the visibility you need most when you're running more than one concept.
Three Ways to Allocate Shared Rent, Utilities, and Labor
Once revenue is properly categorized per brand, the harder problem is splitting the shared costs — rent, utilities, equipment depreciation, and often labor — across the brands that share the kitchen. There's no single "correct" method; commissary and shared-kitchen operators generally use one of three approaches, and the right one depends on how your brands actually use the space.
1. Square-footage allocation
If each brand has a dedicated prep station or storage zone within the shared kitchen, you can allocate rent, utilities, and depreciation proportionally to the square footage each brand occupies. This is simple and defensible, but it breaks down when brands share the exact same equipment (one fryer, one flat-top) rather than having separate zones — in that case, no brand has more "square footage" than another, even though usage varies wildly by order volume.
2. Revenue-based allocation
Shared costs get split in proportion to each brand's share of total kitchen revenue for the period. A brand doing 60% of total order volume absorbs 60% of shared rent and utilities. This scales naturally with actual usage intensity — a high-volume brand is presumably also using more gas, more water, and more walk-in space — and it's the method most commissary operators default to for indirect costs like supervisor wages, insurance, and general overhead.
3. Labor-hour or machine-hour allocation
For labor specifically, revenue-based splitting can misrepresent reality if one brand's menu is far more labor-intensive per order than another's (think: a made-to-order burrito concept vs. a pre-portioned salad concept). Tracking actual prep and cook time per brand — even via a simple shift log noting which tickets belonged to which brand — gives a more accurate labor allocation than assuming labor scales with revenue.
In practice, most multi-brand operators blend methods: square footage or machine-hours for the physical kitchen and equipment, revenue-share for general overhead like insurance and admin costs, and tracked labor-hours for payroll. Whatever combination you choose, the critical discipline is picking a defensible allocation base for each cost category before the month closes, and applying it consistently — not backfilling percentages after the fact to make the numbers look better.
Tracking Inventory and Food Cost Per Brand
The formula multi-brand kitchen operators increasingly use to isolate profitability by concept is straightforward in principle:
Brand food cost = Σ (recipe ingredient cost × items sold for that brand)
The catch is that it only works if your recipes are built with allocated ingredient costs at the unit level, and your point-of-sale (or aggregator) data is tagged by brand at the item level — not just at the kitchen level. If your burger brand and your wing brand both use the same case of buns, you need a system that knows a bun pulled for a burger order counts against the burger brand's food cost, not a shared "kitchen supplies" bucket.
This is where a lot of ghost kitchens quietly lose the thread: it's easy to track total food cost as a percentage of total kitchen revenue, and much harder — but far more useful — to track food cost as a percentage of each brand's revenue. A kitchen-wide food cost percentage that looks healthy can mask one brand running an unsustainable 45% food cost while another runs 22%. Only per-brand tracking surfaces that.
Practically, this means:
- Recipe cards with locked ingredient costs, updated when supplier prices change
- POS or aggregator exports tagged by brand/menu, not just by order
- A monthly reconciliation between theoretical food cost (what the recipes say you should have used) and actual food cost (what you bought), by brand — the gap tells you about waste, portioning drift, or theft, and it's invisible if you only look kitchen-wide
A Simple Monthly Close Checklist for Multi-Brand Ghost Kitchens
- Pull settlement reports from every delivery platform and break each one into gross revenue, commission, promotions, processing fees, and net deposit — tagged by brand.
- Reconcile net deposits to your bank statement. The delay between order date and deposit date means this month's deposits won't perfectly match this month's orders — that's expected under accrual accounting, but it needs to be tracked, not ignored.
- Apply your chosen allocation method (square footage, revenue-share, or labor-hours) to rent, utilities, equipment depreciation, and shared payroll, and post the split to each brand's cost center.
- Compute food cost per brand using per-brand recipe costs and sales mix, then compare theoretical vs. actual usage.
- Produce a P&L per brand, not just one for the kitchen as a whole. This is the number that tells you whether to keep, fix, or kill a concept.
- Review platform commission rates for creep. A brand's effective commission percentage rising month over month — even without a stated rate change — usually means promotional spend or refunds are climbing.
Why Plain-Text Records Make Multi-Brand Allocation Easier to Audit
Cost allocation only holds up if you can show your work. When rent, utilities, and labor get split across three brands using three different formulas, a black-box spreadsheet that nobody remembers the logic behind six months later is a liability at tax time — and a headache the moment you want to bring in an accountant or investor to check the numbers.
Beancount.io takes a plain-text approach to accounting: every allocation, every journal entry, and every brand's cost center lives in version-controlled text files you can read, diff, and audit line by line. That means the formula you used to split July's rent between your burger brand and your wing brand isn't buried in a spreadsheet macro — it's a transparent, reviewable entry in your ledger, exactly as it was recorded. Get started for free and see how plain-text books make multi-brand cost allocation something you can actually explain, not just something you hope adds up.