Delivery menu vs dine-in: costing each channel without giving away the margin

Step 1: split the P&L by channel before you touch a single price

The first deliverable is a channel-level income statement, dining-room gross sales in one column and net platform revenue in another, because as long as those two live added together you are reading an average that exists nowhere in your kitchen. Pull three months of app settlements, subtract the real commission —somewhere between 15% and 30% depending on your agreement— and write down what actually landed in the bank, not what the POS sales report shows. With median full-service food cost at 32,0% of sales (National Restaurant Association, Restaurant Operations Report 2025), a channel that withholds 27% produces a gross margin that no longer resembles the room's. Verify it this way: if you cannot say within ten seconds how many net USD per order the app left you last month, this step is not finished. Packaging is raw material and belongs in every dish's digital spec sheet with its unit cost, exactly like oil or protein.

Step 2: charge packaging to the standard recipe, never to overhead

A compartmented tray with a tamper lid, the thermal bag and the security seal add between 0,40 and 1,80 USD per order, and that spread is enormous: the dish that ships in one tray and the bowl needing three separate containers so the sauce doesn't drown the crunch live in different financial realities, ones that general accounting hides. Duplicate each spec sheet in the system —dining-room version and delivery version— and add packaging only to the second. How you check the step: open five sheets at random and confirm the delivery version shows a HIGHER food cost percentage than its dining-room twin; if they match, somebody skipped the line. The formula that decides everything is short: menu price minus commission, minus food cost, minus packaging, equals the channel's real contribution margin. An 18 USD dish carrying 5,40 in food cost leaves 70% in the room; once the platform withholds 27%, net revenue drops to 13,14 and margin falls to 58,9% before packaging enters.

Step 3: recompute contribution margin dish by dish with commission inside

Add the 1,20 tray and that dish is already working near 49%. Multiply the gap by monthly volume and you will see the thousands of USD evaporating without showing up in any accounting line, because technically they were never revenue. The deliverable is a table ranking dishes from highest to lowest digital-channel contribution margin. The order rarely matches the dining room's, and that mismatch is precisely the information you did the exercise for. Time to commit: dishes in the bottom quartile of that table leave the digital menu, even when the room loves them and the owner defends them. A well-built delivery menu usually carries between 40% and 60% of the room's items, not for kitchen convenience but because every dish that travels badly generates complaints, remakes and low ratings that later cost visibility inside the app's algorithm. Give priority to what survives twenty minutes in a thermal bag and to what runs cheap by design: pizza works at 15% to 20% food cost against menu price (Sauce, 2025), which is why the digital channel rewards it.

Step 4: cut the dishes that subtract and build a shorter digital menu

Verification: the digital menu fits on a phone screen without endless scrolling, and no bottom-quartile item survived the cut. Digital prices go up, and they go up with judgment, never as a flat percentage smeared across the whole menu. Work out how much each dish needs to recover the dining room's target margin and raise only where the differential hurts: a dish that lost eleven margin points deserves different treatment from one that lost three. The rule holding all of this together is that per-dish food cost stays under 32% in every channel, and the healthy industry band runs from 28% to 35% (National Restaurant Association). Diego F. Parra keeps insisting at Masterestaurant that the increase should not hide: your app guest already pays service and delivery fees, and compares mostly inside the application, not against your printed menu. Verifiable deliverable: every digital dish sits back inside the target margin defined in step 3.

The four mistakes that ruin this exercise

The most expensive mistake is celebrating a rising digital average ticket without reading the settlement, because the ticket rises with the service fee and that money is not yours. Second comes treating commission as a commercial discount when it is a variable cost of the sale belonging to the channel's P&L. Third, leaving packaging in office supplies, a decision that guarantees you will never learn which bowl is costing you money. And fourth, the quietest one, switching suppliers to paper over the problem: 40% of operators named that their main strategy against rising costs (TouchBistro 2024, via Apicbase), and it fixes nothing when the hole sits in the channel's structure. If cash flow tightens, there you have the most frequent cause of closure among small businesses (Inc.), and it rarely comes from the supplier. Something worse than losing money happens: you grow toward insolvency without noticing for months.

What happens if you never separate the two menus?

Picture a location adding 1.400 monthly app orders with the same menu and the same prices as the room;

reported sales climb, the owner hires another cook, negotiates the neighboring space to expand, and aggregate contribution margin slides two or three points per quarter while no P&L line shouts about it. With sector spending up only 2% in 2024 against flat traffic (Circana), that cushion does not exist. The dining room ends up financing the digital channel, Tuesday-through-Thursday cash thins out, and the diagnosis arrives the day payroll is due. That is delivery's paradox: the channel that fills a kitchen fastest also drains the bank fastest when it enters without spec sheets of its own. Your digital menu is ready when you can answer five questions without opening a file. One: what is the app channel's net contribution margin, commission already deducted? Two: what does packaging cost on your best seller, priced to the cent between 0,40 and 1,80 USD?

Closing checklist: how to know everything landed right

Three: which dishes left the digital menu, and why? Four: did every digital item land under 32% food cost, inside the healthy 28% to 35% band the National Restaurant Association reports? Five: how much net money per order does the app leave today versus a dining-room cover? Should any answer begin with «more or less», go back to the step where measurement broke. Revisit this table whenever the platform changes its commission, because the day it climbs from 25% to 30% your digital menu expired without telling you. Commission is not a trade discount, it is a variable cost of the sale. When a platform withholds 27% on an 18 USD dish, you did not sell 18: you sold 13.14. If that dish carries 5.40 USD of food cost, your contribution margin slid from 70% to 58.9% before packaging enters the picture, and that single point across 1,400 monthly orders is 2,100 USD gone from EBITDA without a trace in any ledger account.

The four differences that actually move the till

Packaging is raw material, not stationery. A compartment tray with a sealed lid, the thermal bag and the tamper seal run between 0.40 and 1.80 USD depending on the dish, and that figure belongs in the digital standard recipe exactly like oil or protein. Booking it under overheads is the most elegant way to never learn that your 12 USD bowl barely leaves 4.10. Travel time destroys product, and dishes do not degrade at the same rate. A stew, a rice dish or a braise hold thirty minutes without losing anything; breaded fish, thin fries or a salad dressed with an emulsified vinaigrette lose texture before minute fifteen. Pulling the dishes that drain profitability through complaints rather than cost brought refunds under 1.5% wherever the travel-resistance filter was applied. The purchase decision happens on a different plane. At the table a guest reads for eight minutes, asks the server and lets themselves be guided; on the app they decide in 40 to 90 seconds, with a thumb, comparing photographs.

The four differences that actually move the till — in practice

Digital menu engineering puts the highest marginal profitability per dish in the first six scroll slots and drops the long-section logic that works on paper and buries margin on screen.