Delivery vs dine-in menu: how to cost without giving away margin

Charging the same price for delivery and dine-in is costing you margin

Giving away margin doesn't feel like a serious mistake, which is exactly why nobody fixes it in time: all it takes is uploading the dining room menu to the app without adjusting a dollar. A $10 dish with $3 food cost, 30%, right at the Masterestaurant ceiling, hands over $7 of contribution margin at the table. That same dish, on the app at the identical price, first pays the platform commission, between 18% and 30% depending on the deal, meaning $1.80 to $3 leaves your pocket. Add packaging, $0.40 to $1.20 per order, and the original $7 shrinks to $2.80 or less: food cost stays at 30%, untouched, but the channel's entire profit is gone. Circana measures off-premise at close to 75% of industry traffic, so three of every four orders could be going out at the wrong price. Here's the calculation almost no owner runs before signing with a platform: separating what belongs to the recipe from what belongs to the channel.

Channel costing vs. dish costing: the distinction that changes everything

Commission, packaging, and delivery (18% to 30%, $0.40 to $1.20, depending on brand and format) never get charged to the dish; they get charged to the channel, the same way payroll and rent never get prorated onto a recipe. Food cost stays at ≤32% in both channels because the recipe card doesn't change, the selling price does. You calculate it backward: define how much margin you need to keep, add projected commission and packaging, and that number, not the dine-in one, is your channel price. A dish that costs $10 in the dining room can justify $12.50 to $13.50 on the app without touching the recipe or sounding unreasonable to a customer paying for convenience. Think first about how the dining room works, because that's the yardstick everything else gets measured against: raw materials at ≤32% of the selling price, minimal packaging, and waste already controlled by the recipe card.

Dine-in cost structure: the baseline against which you measure delivery

Payroll, rent, and utilities go to the location's break-even, never to the dish. Under that logic, a $10 dish with $3 food cost generates $7 of gross margin that then absorbs the business's fixed costs; the table is already paid for, so is the server. Diego F. Parra, through years of work with dozens of operators, has seen that 68% of owners who miscalculate delivery carry this exact dine-in logic over, unadjusted, to a channel that drags three extra cost layers behind it. Dine-in and delivery break-even are two different equations. Three cost layers show up the moment an order walks out the door, and none of them exist in the dining room. First, platform commission, 18% to 30% on the price the customer pays, not on what it costs you to make the dish. Second, packaging, $0.40 to $1.20 depending on whether you use simple bags or sealed, temperature-controlled containers.

Delivery cost structure: three layers dine-in never sees

Third, delivery, $0 to $2.50 per order depending on whether you run your own fleet, outsource riders, or use the app's own couriers. Stacked together those three layers eat $2.60 to $5.70 out of a $10 order, so if your dine-in margin was $7, delivery can leave you with as little as $1.30, before covering the channel's own fixed costs: community manager, product photography, order tablet. The Masterestaurant method requires calculating these three layers dish by dish, never as a blanket average. Start from the margin you want to keep per order, not from the price you already charge dine-in: that single change in starting point decides whether the channel leaves you profit or hands it over to the commission. Masterestaurant works with a minimum 60% contribution margin on what you actually receive after commission: if the platform charges 25%, you keep 75% of the selling price.

How to set the delivery price without losing margin or scaring off customers?

To hold onto $6 of margin on a dish with $3 food cost and $0.80 packaging, the minimum channel price comes out to ($3 + $0.80 + $6) ÷ 0.75, or $13.07, rounded to $13.50.

The customer sees that number on screen (35% more than dine-in) and accepts it, paying for the convenience. If your local competitor sells the same dish at $11 on delivery, that's a positioning problem, not a pricing one: improve the packaging before you cut margin. A flat 25% commission is easy to calculate and expensive to live with, because the real contract shifts by app, by city, and by the order volume you've actually negotiated, and averaging it costs you dearly by month's end. Uber Eats and Rappi charge anywhere from 20% to 32%; iFood, 18% to 28% depending on the market.

The mistake of averaging commissions: why negotiating with the app matters so much

A four-location operator who negotiated on volume might be paying 18%, while a single-location owner pays 30%: a 12-point gap that equals $1.20 on a $10 order and, multiplied across 80 daily orders, adds up to $96 a day, $2,880 a month. Masterestaurant recommends auditing each platform's real commission every quarter and recalculating channel prices the moment commission moves more than 2 points either way. AI automates that full recalculation in under 30 seconds per menu, a speed no manual spreadsheet keeps up with. Very few restaurants keep the delivery break-even separate from the dine-in one, and that's exactly the gap: the channel adds its own fixed costs (menu photography, in-app optimization, the order tablet, packing staff where it applies) on top of variable costs per order, commission, packaging, delivery. With a 25% commission, $0.80 packaging, and outsourced delivery at $1.50, the variable cost on a $12 order reaches $4.30.

Delivery channel break-even: the metric missing from 80% of restaurants

If the channel's fixed costs run $800 a month and net contribution margin per order sits at $3.70, you need 216 monthly orders just to clear the channel's break-even, before a single dollar reaches restaurant profit. Masterestaurant models this number for every client before any platform goes live, because opening delivery without knowing it is switching on a leak nobody notices until it already hurts. Line the two scenarios up side by side and there's not much left to debate: with channel costing, net margin holds between $5.70 and $6.50 per order, while flat dine-in pricing drops it to $1.30 to $2.80. The channel price, 25% to 35% above dine-in, is justified to the customer through convenience, and when it's communicated well, dish description, careful photography, packaging that matches, it doesn't drive people away.

Verdict: channel costing or keep giving away margin on three out of four orders

Flat pricing sustains volume short-term but destroys profit: at 100 daily orders and $1.50 net margin, the channel brings in $4,500 a month; with channel costing, those same orders at $6 margin bring in $18,000. That $13,500 monthly gap funds operations or turns into real profit: channel costing wins, and AI works it out dish by dish in minutes. What separates a delivery channel that makes money from one that bleeds it isn't a better spreadsheet: it's understanding that the moment your restaurant goes live on an app, you're running two businesses with two separate P&Ls, even though both cook off the same line. The commission, 18% to 30% depending on brand, city, and negotiated volume, doesn't behave like an occasional discount that hits sometimes. It's a structural cost the platform collects on every single order, no exceptions, and that's why it has to be priced for from the start.

Why costing the channel changes the result?

Here's the rule almost no operator applies: neither that commission, nor packaging, nor delivery ever gets charged to the dish, the same way you'd never prorate payroll or rent inside a recipe.

They are, without argument, channel costs, and as such they get covered by the contribution margin that channel produces and measured against its own break-even. The recipe's food cost, ≤32%, doesn't move in either dine-in or delivery; the only thing that shifts from one channel to the other is the selling price. There's a second mistake, and it's an expensive one because it shows up in your app rating before it shows up in your books: uploading all 60 dine-in dishes to the platform and assuming every one of them survives the trip equally well, which simply isn't true. A creamy risotto, a 63-degree egg, or fries lose the fight against the packaging in about 25 minutes, and the customer notices it in the texture before the first bite.

Why costing the channel changes the result — in practice?

Bloated delivery menus end up stretching the courier thin (too many line items, too much assembly time) and lengthen kitchen time until the app rating drops for the whole location.

The menu that works under the MR method is shorter and more deliberate: only dishes that survive transit make the cut, with packaging designed for them and a channel price already worked out. As I tell my clients, you don't win delivery by matching the dine-in price; you win it by costing the whole channel and cutting from the menu whatever doesn't travel.