Pricing: one price for four different channels is the leak

1. A single price is a channel decision dressed up as arithmetic

Charging the same price in the dining room and on third-party delivery is not a pricing policy, it is a silent transfer of margin to the platform. The effective cost of third-party delivery, once you add commissions, co-funded promotions and refunds for incidents, reaches 30% to 40% of the total order according to OPA! (True Cost of Third-Party Delivery, 2026), while the food cost of a healthy dish rarely goes above 32%. There sits the leak: the channel costs more than the food. When an owner sends three out of every ten units of the signature dish through an app and keeps the dining-room rate, that dish stops funding the operation and starts consuming it. Fixing this does not require leaving the platforms, it requires accepting that price is a per-channel variable and not a number inherited from the recipe book. Govern with contribution margin in dollars and use food cost percentage only as a control alarm.

2. Which metric should govern the price list: percentage or dollars?

Two dishes at 28% food cost can leave four dollars and eleven dollars per unit, and in a service with tight table turns you want the server selling the eleven-dollar one even though the percentage is identical.

The percentage measures purchasing discipline; the dollars per unit pay the payroll, which in full service hit a median of 36.5% of sales in 2024 and 31.7% in limited service, according to the National Restaurant Association Restaurant Operations Data Abstract 2025. Under that structure, a price list built on percentages can be technically flawless and still fail to cover the roster. The right question is not how much the plate weighs, but how many dollars each seat leaves per shift. Below 500 thousand dollars a year, the only pricing decision that moves cash is applying a channel surcharge of 15% to 25% on the delivery menu, and doing it this week.

3. Under 500 thousand in annual revenue: price by channel, nothing else

In this band there is no analyst, no category manager, and the owner cooks or works the register; any pricing architecture that demands weekly upkeep collapses for lack of hands. The operating threshold is simple: if a dish's contribution margin after the platform commission falls under three dollars, that dish leaves the digital menu. With utilities eating between 2% and 5% of total revenue according to Toast (Average Restaurant Electricity Bill, 2025), there is no cushion left to subsidize someone else's orders. Diego F. Parra insists at Masterestaurant that this band stays in the analysis: it is where a pricing mistake kills first. Between 500 thousand and one million dollars a year, the decision is to build two menus with different prices and start measuring margin by time slot. A manager now exists who can sustain the discipline, and the gap between lunch and dinner is usually worth six to ten margin points on the very same dish.

4. From 500 thousand to 1 million: time to split menus and measure by daypart

The threshold I set is 8%: if the differential between the dining-room price and the app price does not clear 8%, the digital menu is working for free. Staffing risk shifts too, because replacing a general manager costs 16,770 dollars in hard costs according to Black Box Intelligence (State of the Restaurant Workforce, 2024), and a price list that squeezes the team to rescue margin ends up paying that bill twice a year. Once past a million in annual revenue, pricing stops happening dish by dish and moves to a popularity-and-margin matrix, reviewed every ninety days with POS data. Volume in this band turns any twenty-cent error into tens of thousands of dollars a year, and waste already weighs: average food waste cost per restaurant runs near 72,000 dollars annually according to The Restaurant HQ (Food Waste Statistics, 2025). The action threshold is a minimum contribution margin of nine dollars across the ten best-selling dishes.

5. Above 1 million: menu engineering from POS data, not from instinct

And here it pays to accept a paradox of the trade: raising the price of the most popular dish almost never sinks its demand, while raising it on the niche dish does sink it, because the guest who orders that one comparison-shops and the loyal guest does not. Above five million a year, the territory of the large-format themed concept or the restaurant signed by a media figure, price no longer chases cost coverage, it sustains a promise, and margin gets defended through experience design rather than surcharges. A venue of that profile that caves to third-party delivery pressure and absorbs the 30% to 40% of the ticket documented by OPA! (2026) loses more than money: it erodes the price anchor it spent years building. My recommendation is firm, and I know it stings: in this band, third-party delivery is limited to a reduced menu of four to six items engineered to travel, with its own price, or it is not done at all.

6. Above 5 million: price becomes positioning and mistakes get paid in brand

Kitchen equipment for a mid-size venue already costs 50,000 to 150,000 dollars according to Rezku (2025); at this scale, that investment demands protection. From ten million upward, or inside a group structure, the decision turns institutional: you create a pricing committee with a fixed calendar, bands by market and veto power over platform promotions. Without that governance, each manager negotiates a private discount and the group discovers the leak six months into the bleeding. The threshold I use is three points: no unit may drift more than three percentage points from its market price band without central approval. The backdrop offers little help, since real projected sales growth for the U.S. sector in 2026 sits at just 1.3% according to the National Restaurant Association, and with expansion that thin, margin is earned inside the four walls. A group growing through new units rather than disciplined pricing is buying expensive revenue.

7. The scenario worth simulating before you move a single price

Before touching the menu, run this exercise: assume you raise delivery prices 12% and lose 8% of that channel's orders. What happens then? With effective commissions of 30% to 40% of the order according to OPA! (2026), the surviving orders pay considerably more margin than the lost ones contributed, and the kitchen frees capacity during the peak, which trims overtime on a payroll already representing 25% to 35% of revenue according to the U.S. Bureau of Labor Statistics. The outcome is almost always more cash on less volume. For years I recommended abandoning the platforms and I was wrong: the problem was never the channel, it was pricing inside it as if it were the dining room. Open the POS tomorrow, filter ninety days, and rank your dishes by margin dollars per unit.

8. What actually changes when price stops being a number and becomes a system

The real difference is not how much you charge but WHICH cost structure you compute against: a uniform markup starts from food cost, which in a healthy operation tops out around 32% of the plate, and ignores that the delivery channel adds 30% to 40% of the ticket in commissions, promotions and refunds (OPA!, 2026) — a variable cost larger than the food itself, invisible to kitchen arithmetic. The second change is the unit of measure. Food cost percentage is a control metric; contribution margin in currency is a decision metric. Two dishes at 28% food cost may leave four dollars and eleven dollars respectively, and in a service with tight table turns you want to sell the eleven-dollar one, identical percentage notwithstanding. The third one is geographic, and inside a local digital engine it is the piece almost nobody models: delivery radius defines the true cost to serve.

9. What actually changes when price stops being a number and becomes a system — in practice

An order at the edge of the polygon burns more courier minutes, more complaint risk and more chance of a bad review, so far-zone pricing should differ from near-zone pricing — something platforms allow through differentiated polygons and most operators never use. The fourth is corporate governance. When the price list lives in the chef's spreadsheet, every increase depends on availability and mood; when it lives in the management P&L with an owner, a threshold and a date, pricing becomes auditable. That is exactly where a buyer's operational due diligence looks first, because an ungoverned price list is an implicit discount on the exit multiple.