Delivery commissions: cost each channel vs selling blind
Verdict: Selling through apps without costing the channel is the quietest way to give away profit in 2026. Platforms charge between 15% and 30% per order; on a $25 order at 25%, that's $6.25 gone before touching food or labor. The traditional method uses the same price in the dining room and on the app, and margin evaporates unnoticed. The Masterestaurant method costs each channel separately: app price with commission and packaging already subtracted, dining-room price apart, and an owned channel that captures 15-20% of volume with no commission. Across 8,400 restaurants in our network in 43 countries, those that cost by separate channel hold 5 to 8 points more net delivery margin. The hard rule stands: 32% food cost is the MAXIMUM per dish, and payroll, rent and utilities are never charged to the dish — they go to the break-even point.
Delivery stopped being an add-on and became a structural channel: in many urban restaurants it already accounts for 25% to 45% of sales. The problem is almost no one costs it as its own channel. The dining-room menu price gets pushed onto the app as-is, and the owner assumes the commission is 'already included.' It isn't. The 15-30% commission, the 3-5% packaging, and the payment lag of up to 15 days all coexist with a food cost calculated for table service, not for a dish traveling 6 kilometers inside a bag.
Diego F. Parra, consultant at Masterestaurant, puts it plainly: 'I have seen restaurants with delivery growing 40% month over month and profit falling at the same time, because they never separated app price from dining-room price.' Costing by channel is not a big-chain luxury; it's the difference between delivery adding or subtracting. The 2026 trend across the Masterestaurant network is clear: operators who win at delivery don't lower the app commission — they can't control that — they control their unit economics per channel, dish by dish.
Side-by-side comparison
| Traditional (same price, app and dining room) | Masterestaurant (cost by channel) | |
|---|---|---|
| App commission | ✕Assumed 'included' in the menu price | ✓Explicitly subtracted: 15-30% per order before pricing |
| Packaging | ✕Not counted as a dish cost | ✓3-5% of the ticket charged to the channel's variable cost |
| Target app food cost | ✕The same 32% maximum as the dining room | ✓26-28% max to absorb commission and packaging |
| Dish price | ✕Identical in dining room and app | ✓App price 12-18% higher to protect margin |
| Owned channel with no commission | ✕0% of volume, everything goes through apps | ✓15-20% of volume captured at 4-6 pts higher margin |
| Payment lag | ✕Not projected in cash flow | ✓Up to 15 days modeled in the channel's cash flow |
| Net delivery margin | ✕Falls to 4-8% unnoticed by the owner | ✓Holds at 10-14% with per-channel price and food cost |
How much does selling through a delivery app really cost you?
Selling through an app costs between 15% and 30% commission per order, plus 3-5% packaging: on a $25 order at 25% that's $6.25 gone before touching food or labor.
That's the number the traditional method ignores when it pushes the dining-room price straight onto the app. Diego F. Parra repeats it in every Masterestaurant audit: commission is not an abstract tax, it's a variable cost that must be subtracted before pricing. When you add commission, packaging and a food cost built for the dining room, net delivery margin falls to 4-8% unnoticed. The 2026 trend is to stop treating delivery as an extension of the menu and start costing it as a channel with its own cost structure, dish by dish. The most frequent mistake I see across the Masterestaurant network is one price for two channels with different costs. In the dining room you pay for service and experience; on the app you pay commission, packaging and cash lag.
Dining-room price and app price cannot be the same number
Matching prices gives away 5 to 8 points of margin. The fix is a delivery price 12% to 18% above the dining-room price, a range platforms allow and app customers tolerate because they already assume delivery costs more. But a flat surcharge won't do: high-food-cost dishes need more adjustment than low-food-cost ones. Diego F. Parra recommends recalculating dish by dish, because a uniform markup leaves some dishes at a loss and others overpriced, and both extremes either scare off orders or burn profit. The delivery channel's target food cost must drop to 26-28%, four to six points below the 32% that is the MAXIMUM per dish in the dining room. The reason is arithmetic: the platform commission eats that extra margin, so the dish has to cost less in ingredients to leave room. It's worth restating Masterestaurant's hard rule: payroll, rent and utilities are NOT charged to the dish in any channel; those go to the whole business's break-even point.
The delivery food cost is not the same as the dining-room food cost
What changes between channels is the direct variable cost: food cost, packaging and commission. In 2026, network restaurants that win at delivery build their app menu in reverse, starting with low-food-cost, high-rotation dishes and leaving off the app those that can't absorb the commission. Capturing 15-20% of volume outside the apps is the most profitable operational trend in delivery in 2026, and also the most wasted. That owned channel — WhatsApp Business, ordering page, in-store pickup — eliminates the full commission on that share of sales and raises net margin by 4 to 6 points in the first semester. In the Masterestaurant network we've seen operations go from 8% to 13% net margin in four months by moving just 18% of orders to a direct channel, with an investment under $200. The classic mistake is waiting for the customer to migrate on their own: they don't.
The owned channel: the lever almost no one uses well
They need an explicit incentive, an 8-10% discount for ordering outside the app, and differentiated packaging that makes the direct channel's value obvious. Platforms don't pay instantly: the lag reaches up to 15 days in 2026, and that gap sinks operations that look healthy in the spreadsheet. Costing by channel isn't only knowing how much you earn per order, it's knowing when that money lands. A restaurant selling hard on apps but paying suppliers in cash needs 4 to 6 weeks of working capital just to cover the collection gap. Diego F. Parra has seen businesses with delivery growing 40% monthly enter a liquidity crisis precisely for not projecting the lag. Masterestaurant's cash-flow tool models that gap platform by platform, so delivery growth doesn't become an urgent call to the bank at month-end. The trend that most separates delivery winners in 2026 is using AI to recalculate per-channel margin in real time.
AI applied to per-channel costing: from quarterly review to daily control
When a platform raises its commission, or the protein price moves, manual costing reacts three months later; a well-configured AI reacts the same day. At Masterestaurant we integrate models that watch food cost, commission and packaging per channel, and fire an alert when a dish crosses the profitability threshold. It's not tech for fashion's sake: it turns a quarterly audit into continuous control. Diego F. Parra insists AI doesn't replace the owner's judgment, it amplifies it: it shows which app dishes are losing money today, not next quarter, once thousands of negative-margin orders have already piled up. Costing by channel doesn't look the same across formats, and confusing them is expensive. A dark kitchen depends almost 100% on apps, so its delivery food cost must be the most aggressive, 24-27%, and its top priority is opening an owned channel before month six.
What changes by format: dark kitchen, dining-room restaurant and food truck?
A dining-room restaurant uses delivery as a complement: it can tolerate a 27-28% app food cost because the room carries most of the margin.
A food truck or small format plays differently: low volume, commission that weighs more per order, and a critical need for commission-free pickup. In the Masterestaurant network, crossing format with channel mix is what defines the right price. Diego F. Parra sums it up: the same dish has three different prices depending on where it exits, and whoever uses one number loses on two of the three. A well-costed delivery in 2026 leaves a net margin of 10-14% per channel, not the 4-8% of whoever sells blind. The difference isn't in negotiating a lower commission — the owner rarely controls that — it's in the per-channel price, the tighter food cost and the share of volume running through an owned channel.
What a healthy delivery looks like in the month-end cash register?
Across 8,400 restaurants in the Masterestaurant network in 43 countries, those that cost by separate channel hold on average 5 to 8 points more net delivery margin than those that match prices.
The proprietary datum matters: it's not theory, it's the pattern that repeats in the real cash register. Diego F. Parra closes with a concrete action: open a per-channel sheet today, subtract commission and packaging from your ten best-selling app dishes, and in one afternoon you'll see which ones you were giving away. Commission: the traditional method assumes it's 'included'; Masterestaurant subtracts it explicitly, 15-30% per order, before setting any app price. Packaging: goes from invisible expense to a declared variable cost of 3-5% of the ticket on every dish that travels. Food cost: drops from the 32% dining-room maximum to a 26-28% maximum in the delivery channel, to absorb commission and packaging.
6 differences between selling blind and costing by channel
Price: the app dish sits 12-18% above the dining-room price instead of identical, protecting 5 to 8 points of margin. Owned channel: goes from 0% to 15-20% of volume with no commission, raising net margin by 4 to 6 points in the first semester. Cash: the payment lag of up to 15 days is modeled in the flow instead of surprising the owner at month-end.
How selling blind looks (traditional method)Blind
- The dining-room price is pushed onto the app without subtracting the 15-30% commission.
- Packaging is treated as a minor expense, never as a variable cost of the delivery dish.
- Food cost is measured for table service, not for a dish that travels and cools down.
- All volume depends on two or three apps; there is no owned channel free of commission.
- The payment lag of up to 15 days never shows up in the cash projection.
How costing by channel looks (Masterestaurant method)Masterestaurant
- Commission and packaging are subtracted first, and only then is the price set.
- The delivery channel's food cost drops to 26-28% to absorb the platform commission.
- The app price sits 12-18% above the dining-room price without hurting customer perception.
- An owned channel via WhatsApp and pickup captures 15-20% of volume with no commission.
- Cash flow projects the payment lag and the real cost of each platform.
Side-by-side comparison
| Traditional (same price, app and dining room) | Masterestaurant (cost by channel) | |
|---|---|---|
| App commission | ✕Assumed 'included' in the menu price | ✓Explicitly subtracted: 15-30% per order before pricing |
| Packaging | ✕Not counted as a dish cost | ✓3-5% of the ticket charged to the channel's variable cost |
| Target app food cost | ✕The same 32% maximum as the dining room | ✓26-28% max to absorb commission and packaging |
| Dish price | ✕Identical in dining room and app | ✓App price 12-18% higher to protect margin |
| Owned channel with no commission | ✕0% of volume, everything goes through apps | ✓15-20% of volume captured at 4-6 pts higher margin |
| Payment lag | ✕Not projected in cash flow | ✓Up to 15 days modeled in the channel's cash flow |
| Net delivery margin | ✕Falls to 4-8% unnoticed by the owner | ✓Holds at 10-14% with per-channel price and food cost |
Delivery commissions by the numbers: what the 2026 cash register confirms
“We sold the same in the dining room and on the app, and I couldn't understand why delivery was growing but profit wasn't. When we separated the price and dropped the channel's food cost to 27%, six points of margin I'd been giving away every month appeared. The hardest part was accepting that 3 of my 10 star dishes were losing money on the app, because I never subtracted the commission before pricing them.”
How to cost your delivery channel in 2026, step by step
The first move is to stop using a single price for two different channels. In the dining room you pay for service and experience; on the app you pay commission, packaging and cash lag. Diego F. Parra recommends setting the delivery price 12% to 18% above the dining-room price, a range most platforms allow and app customers tolerate because they already assume delivery costs more. Across the Masterestaurant network, restaurants that separate price recover 5 to 8 points of net channel margin without measurable volume loss in the first 90 days. The key is to do it dish by dish, not with a flat surcharge: high-food-cost dishes need more adjustment than low-food-cost ones, and a uniform markup leaves some at a loss and others overpriced.
Before deciding what to charge for an app dish, first subtract the platform commission — 15% to 30% in 2026 —, then packaging, usually 3% to 5% of the ticket, and only then look at the dish's food cost. If those three items exceed 60% of the final price, the dish isn't viable for delivery even if it flies in the dining room. Masterestaurant does this subtraction on a per-channel sheet, never as a menu-wide average, because one dish with negative margin drags down the whole operation's profit. Diego F. Parra has found menus with 3 of every 10 dishes losing money on the app without the owner knowing. Remember the hard rule: 32% food cost is the maximum per dish, and in delivery it must drop to 26-28% to leave room for the commission.
Depending 100% on two or three apps is the biggest structural risk in delivery in 2026, because a platform can change its commission from one month to the next. An owned channel via WhatsApp Business, an ordering page or in-store pickup that captures 15% to 20% of volume eliminates the full commission on that share of sales and raises real net margin by 4 to 6 points in the first semester. In the Masterestaurant network, operations that moved 18% of their orders to a direct channel went from 8% to 13% net margin in four months, with an investment under $200 in setup. The classic mistake is waiting for the customer to reach the owned channel on their own: they won't. They need an explicit incentive, an 8-10% discount for ordering outside the app.
Platforms don't pay you instantly: the lag reaches up to 15 days in 2026, and that gap breaks operations that are profitable on paper but cash-poor. Costing by channel means projecting when the money from each app actually lands, not just how much you sell. Diego F. Parra insists that payroll, rent and utilities are never charged to the dish — they go to the whole business's break-even point — but they must line up with delivery's real collection calendar. A restaurant billing hard on apps but paying suppliers in cash needs 4 to 6 weeks of working capital just to cover the lag. Masterestaurant's cash-flow tool models that gap platform by platform, so delivery growth doesn't turn into a liquidity crisis at month-end.
And with AI?
Optimize channels, pricing and unit economics of your dark kitchen. Diego F. Parra is an expert in AI applied to restaurants.
Free tools to apply this now
Tools to cost your delivery by channel before you scale
Before adding a single dish to the apps, validate the numbers with three free Masterestaurant tools: each one tackles a different delivery-margin leak.
The first checks whether your per-channel model works, the second calculates how much you can scale without breaking margin, and the third projects real cash flow including the up-to-15-day platform payment lag.
Diego F. Parra — International consultant, expert in creating and scaling restaurants and in AI applied to restaurants, foodtech and HORECA. Methodology applied in 8.400+ restaurants across 43 countries · Expert in Artificial Intelligence applied to restaurants, hospitality and food businesses · 20+ years in restaurants, catering, large events and business growth · Author of the book «From Slave to Owner» (Amazon) · International keynote speaker for the HORECA sector.
Frequently asked questions about delivery commissions in 2026
How much do delivery apps really charge per order in 2026?
Should I set the same price in the dining room and on the app?
What food cost should I run in the delivery channel?
Is it worth opening an owned channel if I already sell on apps?
Sector data 2026 (official sources)
Verifiable industry benchmarks from official, non-commercial sources (government, industry associations, market research) - not competitors.
| Metric | Benchmark 2026 | Source |
|---|---|---|
| Comisiones de delivery | 15–30% nominal · 30–45% efectivo | Nation's Restaurant News |
| Mercado global de ghost kitchens | ~$83.5 B en 2026 (CAGR ~10–15%) | Statista |
| Operación fuera del local | ~75% del tráfico | Circana |
| Tráfico de foodservice | delivery como driver de crecimiento | National Restaurant Association |
Related content
Grow your restaurant with the Masterestaurant method
Applied in +8.400 restaurants across 43 countries.