Plate costing in restaurants: myth vs reality (profile matrix, 2026)

What is the best costing method for an independent operator with fewer than 15 tables?

For the independent operator with fewer than 15 tables and mixed-channel sales, the best method is CONTRIBUTION MARGIN IN DOLLARS by channel, with the marketplace commission subtracted inside the dish rather than buried in overhead.

The arithmetic is visible to anyone who opens their own cash drawer: a dish at 22% food cost that leaves $3 and turns 4 units a day contributes $12, while one at 34% that leaves $9 and turns 30 contributes $270, twenty-two times more, and the free percentage template tells you to promote the first one. When 80-90% ground beef went from $4.56 to $5.63 a pound according to USDA data in 2026, operators costing by percentage raised prices blind; those costing in dollars knew exactly how many plates they had to sell to cover the gap. If your dining room carries 70% or more of sales, you want single-channel costing where contribution margin is calculated on the full check and the tip stays OUT of the plate cost.

Best for operations with 70% dine-in sales: single-channel costing with a service load

Canadá shows why that split matters: full service moved 43,1% of foodservice sales in 2024 against 46,4% for limited service, per Statistics Canadá, and those cost structures look nothing alike. At a table you pay a server, linen, dishwashing and breakage, but you sell beverage at a 70-80% margin that rarely travels through delivery. One detail almost nobody adjusts: in the United States tips make up 58,5% of hourly income for tipped staff according to Clockify, so loading that onto the plate inflates your theoretical cost and pushes you into price increases your neighbor never made. A group running a ghost kitchen with three virtual brands at one address must cost each brand as a separate business, with the 15% to 30% commission subtracted line by line before margin. Traditional accounting lies here: pushing commission into overhead spreads a cost that exists only when that specific dish sells through that specific app.

Best for ghost kitchens and virtual brands: per-channel costing with commission inside the dish

Run the counterfactual. A $12 dish with $4.20 of ingredients leaves $7.80 in the dining room; the same dish on a marketplace at 28% loses $3.36 in commission and $0.90 in packaging, dropping to $3.54. Grow that channel 40% while the dining room shrinks and your revenue climbs while your cash falls. It happens to operations that celebrate a sales record the same month they cannot cover payroll. The free percentage template is still the right call in three concrete scenarios, and saying otherwise would be selling method. First, if you run a menu under twelve items with one channel and one price: percentage and dollars give you the same order, and the extra work does not pay. Second, if today's problem is not mix but THEFT OR WASTE — the average restaurant wastes between 4% and 10% of what it buys according to The Restaurant HQ, and full service accounts for more than 43% of foodservice surplus per ReFED 2024; there the theoretical-versus-actual percentage is the detector, not margin.

When NOT to choose the popular option: three scenarios where percentage food cost wins?

Third, if you negotiate volume purchasing and need to compare suppliers: percentage is the natural unit for that. And if you just opened, start with percentage and move to dollars in month three.

Four signals from the trade tell you the template you are evaluating does not work in 2026. One: no column per channel means it was written before delivery carried a third of the ticket, and no patch saves it. Two: if it loads payroll, rent and utilities onto the plate, you get a unit cost that rises when sales fall, an accounting absurdity that pushes price increases at the worst possible moment — those expenses belong to break-even. Three: if selling price comes from multiplying ingredient cost by a fixed factor of 3 or 3.5, it ignores elasticity, and with menu inflation hitting 8,8% in March 2023 per the National Restaurant Association, that fixed factor ate customers.

Red flags when comparing costing methods and templates

Four: if it never asks for yield or waste per recipe, it is not costing. If most of your demand arrives through local search, open a costing line for the direct order and treat it as the most profitable channel on your menu, because it is. The paradox stings: the more visible you become, the faster sales grow and the faster margin drains, since the channel bringing volume charges the most. Resolving it does not mean shutting off the marketplace, it means measuring what each share point costs you. A $30 ticket captured by your Google profile arrives whole; that same ticket through an app arrives at $21.60 after a 28% commission. Ten daily orders shifted from app to direct are $2,520 a month of pure margin. Diego F. Parra insists at Masterestaurant that this shift is bought with repeat visits, not discounts: a magnet in the bag beats a 15% off that destroys your dish.

Best for menus with volatile protein: costing with monthly review and prices in bands

When your menu depends on an ingredient that moves on its own, cost it monthly and set the price as a band, not a number. Arabica coffee touched $4.41 a pound in February 2025, an all-time high according to Bellwether Coffee, and farm-level eggs rose 43,1% in 2024 per the USDA, with retail up 8,5% that year and 21,9% in 2025. A breakfast concept that recosts once a year lost two quarters of margin before finding out. The band works like this: you define the minimum contribution margin in dollars that dish must leave, and the price moves inside a pre-approved range whenever the ingredient crosses a threshold. I got this wrong for years recommending quarterly recosting; with this volatility, quarterly arrives late. In a three-unit group, the right matrix combines both methods: dollars to decide the menu, percentage to control the kitchen.

What the decision looks like in a three-unit operation?

Each location gets its target contribution margin by channel and its theoretical percentage per recipe, and the manager answers for the gap between theoretical and actual, which is where waste lives.

Mexico shows the scale of the problem: restaurants make up 12,2% of the country's economic units per INEGI and CANIRAC, nearly all small operations costing with whatever template they found for free. The value shows up in the asset: the median sale price of a small restaurant in the United States reached $773,000 in 2025, 24% above 2021 according to BizBuySell, and that multiple is paid on documented margin. Start this week with your eight best sellers and calculate margin in dollars by channel. Percentage is a purchasing signal; dollars are a selling signal. A dish at 22% food cost leaving $3 of margin on 4 daily units contributes less cash than one at 34% leaving $9 on 30 units — yet the percentage template tells you to promote the first one.

Where the two approaches genuinely split?

Channel costing acknowledges what traditional accounting hides: marketplace commission is a direct variable cost of the dish, as direct as the protein, because it exists only when that dish sells there.

Filing it under general expenses was a convenient bookkeeping habit that now costs you money. Cost structure changes shape by channel. In the dining room you pay a server and linen; in delivery you pay packaging, commission, and a review that can sink your rating because of a courier who does not work for you. Packaging alone eats 4% to 9% of price on low-ticket dishes. Percentage costing blurs CapEx and OpEx, which is why so many owners load the new oven payment onto the plate. Never do that: equipment investment comes back through break-even and cash flow, never by inflating a recipe's food cost. Payroll, rent and utilities stay out too. There is also a governance gap: percentage costing gets done once by the chef, while contribution costing forces owner and kitchen to sit down quarterly with the management P&L open.

Where the two approaches genuinely split — in practice

It is less comfortable. It is also the only version that survives an 18% jump in cooking oil.