Channel Margin Architecture: Dine-In, Delivery and Take-Away with a Separate P&L

Chapter 1 — Why a blended margin is blind accounting

A P&L that lumps dine-in, delivery and take-away into a single line lies to you by design. A blended prime cost of 58% sounds under control, yet it can hide a delivery channel at 68% dragging down the whole while dine-in runs at 51%. I have seen it in dozens of groups: the operator celebrates that sales climb 22% year over year and never notices EBITDA eroding three points, because the channel that grows is the one that leaves the least margin. A dish with 30% food cost in the dining room hits 41% in delivery once you load the platform commission and packaging. That 11-point gap never shows in the mixed account; the average absorbs it. Before signing for the second location, split the result by channel: each euro performs differently depending on how it enters the business. The delivery platform commission is a direct variable cost of that channel, not a general expense diluted across everyone.

Chapter 2 — The platform commission is a direct variable cost, not overhead

When you bury it in the "other expenses" line, you lose the most common capital leak Diego F. Parra finds while auditing groups of 3 to 10 locations. A 28-32% commission on the delivery ticket, plus 1.20-1.80 euros of packaging per order, plus a dedicated runner's time: that is 9 to 14 points of prime cost belonging to that channel alone. Treating it as shared fixed cost spreads the blame onto dine-in, which never caused it. The Masterestaurant rule is blunt: every cost is charged to the channel that triggers it. Only then do you see that a 24-euro delivery order leaves the same absolute margin as a 15-euro dine-in check. Contribution margin per order decides how many orders you need for the same profit, and between channels the difference is brutal. A delivery order contributing 21% versus a dine-in check at 46% means you need more than double the orders —2.19 times exactly— to leave the same money in the till.

Chapter 3 — Contribution margin per order changes the whole scaling math

Without that figure, delivery volume reads as success: a thousand orders a month sounds like expansion. But if each one delivers half the margin, you are running the kitchen, staff and capital just to drain liquidity. I have seen restaurants with delivery at 40% of sales and 18% of real profit. The number that matters is not the average ticket or the order count, but how many cents of each euro survive after commission. That thermometer separates growing from merely billing. Standardizing the channel P&L before opening the second location turns each opening into the replica of a model with a known margin, not a bet. Once you know your dining room runs at 46% contribution, delivery at 21% and take-away at 38%, you can set the target mix of every new site instead of discovering it twelve months late. Diego F. Parra puts it this way in every Masterestaurant audit: scaling without channel accounting is photocopying an error three times.

Chapter 4 — Standardize channel accounting before you scale

The group that standardizes first can rule that no location exceeds 35% of sales in delivery unless its kitchen is designed for that margin. The gap between a group that makes money with ten locations and one that only stacks revenue with ten almost always sits in whether it disaggregated the channel before or after the third opening. A channel P&L is built with three columns —dine-in, delivery, take-away— that share the sales line but split each variable cost at its source. On top, the net sales of each channel. Below, that channel's real food cost, which is not identical: delivery usually weighs 2-4 points more because portions are adjusted for transport. Then packaging, charged in full to the channel that consumes it; dine-in barely uses any. Next the platform commission, zero in the dining room, 28-32% in delivery. The result is three contribution margins you never average again.

Chapter 5 — How to build a three-column P&L

Fixed costs —rent, base payroll, utilities— go at the bottom, to the location's break-even, not to the plate. This format takes an afternoon to build in a spreadsheet and saves five-figure losses a year in a mid-size group, because it exposes which channel finances which. Once you see the margin split by channel, decisions that looked obvious flip. The operator about to raise delivery advertising spend discovers each new order costs margin, and redirects that budget to filling dine-in tables, where contribution doubles. The one about to close take-away for being "too quiet" sees it runs at 38% with no commission and pushes it harder. Diego F. Parra has reordered the mix of dozens of locations with this single change of lens: in a group of six restaurants, moving from 42% to 28% of sales in delivery lifted aggregate EBITDA 2.8 points without touching a single recipe.

Chapter 6 — What decisions change once you see the margin split

Channel accounting is not a month-end exercise; it is the tool that decides where you put the next marketing euro, which dish you promote and through which channel you want your growth to enter in 2026. The consolidated P&L averages; channel accounting disaggregates. A single 58% prime cost can hide a delivery channel at 68% dragging the whole down: the average lies by design. Platform commission is not overhead, it is a direct variable cost of the delivery channel. Burying it in 'other expenses' is the most common capital leak Diego F. Parra finds in groups of 3 to 10 locations. Contribution margin per order changes radically between channels: a delivery at 21% versus a dine-in at 46% means you need more than double the orders for the same profit. Without that figure, delivery volume reads as success while it drains cash. Standardizing channel accounting before scaling turns every opening into a replica of a model with a known margin; scaling without splitting replicates an average nobody has audited.