Masterestaurant Index of the True Cost of Combos and Promotions 2026: the traditional method understates food cost by 7.4 points

Finding 1 — Why traditional combo costing lies about its margin

Traditional costing lies because it divides theoretical food cost by list price and books a profit that the live real cost disputes. Three leaks hit at once and almost never get seen together: the bundle's cross-discount, the delivery commission at 15%–30% (Rezku 2026), and the card swipe fee, 2.35% per transaction (Texas Restaurant Association 2025). A combo priced at 30% food cost on the list SEEMS healthy until it sells through an app carrying that same commission plus a 12% discount off the sticker price — that's where the pretty number breaks. In limited service, the segment where combos get promoted hardest, median food cost already weighs 32.4% of sales (National Restaurant Association 2024), so there's no cushion left to absorb three blind leaks at once, let alone one at a time. Over the NET price —what's left after the bundle discount— is where the contribution margin's survival gets decided, not over the LIST price the traditional recipe cost still uses.

Finding 2 — List price versus net price: where the margin really contracts

Price a combo at $15 with a 12% cross-discount and what actually lands in the register is $13.20; if the dollar food cost was set against the original $15, the real percentage climbs the moment someone does the subtraction. The rule I apply at Masterestaurant fits in one line: cost against what hits the register, never against what the menu says. With limited-service median food cost already at 32.4% of sales (NRA 2024), a 12% discount pushes that ratio up before delivery or card even enter the math. List price is marketing; net price is accounting, and only the second one covers payroll, which in full service already runs 36.5% of sales (NRA 2024). DoorDash and Uber Eats charge 15% to 30% per order —30% is standard— and Grubhub runs 15%–25% (Rezku 2026): that is where the combo's biggest leak lives, exactly in the channel that costs most to promote.

Finding 3 — The delivery commission: the combo's biggest and most ignored leak

What would happen if that same combo sold only dine-in? Real food cost would fall back to the theoretical line, near 30%, and the margin would survive intact; but it sells mostly through the app, so the commission isn't a hypothesis, it's half the real story, and static recipe costing NEVER charges it to the channel. Run the math on a $15 combo at the standard 30% commission: the marketplace keeps $4.50 before the operator covers a single gram of food, and that hole stacks on top of the limited-service food cost of 32.4% of sales (NRA 2024) until the margin is gone. At Masterestaurant we measure this channel by channel, never on average: profitable dine-in, bleeding cash in the app. Two line items explain most of the 7.4 points static recipe costing never books: the card swipe fee, 2.35% per transaction (Texas Restaurant Association 2025), and the combo's waste.

Finding 4 — Swipe fee and waste: the 7.4 points the operator never sees

The card fee alone costs the industry roughly $187 billion a year (National Restaurant Association), and food waste adds another $162 billion annually to the US restaurant sector (The Restaurant HQ 2025) — a figure that grows in combos because they're produced by volume and unsold portions get tossed regardless. Each leak, seen alone, SEEMS minor: 2.35% here, a handful of grams there. Added to the delivery commission of 15%–30% (Rezku 2026), they explain why a combo theoretical at 30% ends up living at 37% or higher. We deduct both BEFORE approving any promotion, not after launching it, because after launch there's no margin left to fix. A combo at 45% food cost can beat one at 28%, and the reason fits in one word: dollars. The traditional method reads percentage and tosses out anything above 32% on sight; Masterestaurant reads what actually reaches the register.

Finding 5 — Percentage of sales versus dollars of margin per transaction

If that 45% combo leaves $6.50 of margin and sells 40 units a day, it contributes $260; a 28% dish that leaves $4.00 and sells 15 units contributes just $60 — the pretty percentage loses, the DOLLAR wins. With full-service payroll at 36.5% of sales (NRA 2024) and healthy occupancy under 10% (Toast), what actually pays those fixed costs is the margin dollar, never the ratio. The bank doesn't charge in percentages, it charges in dollars, and that's why I insist on costing and deciding in dollars per transaction, even when the percentage looks uncomfortable at first. Live real cost gets calculated by starting from the NET price per channel and subtracting the three leaks BEFORE looking at food cost, not after, when it's too late to fix the price. Subtract the bundle discount from the list price first, to find real net revenue.

Finding 6 — How to cost a combo with the live real-cost method, step by step

Apply the channel commission next: 15%–30% for delivery (Rezku 2026) or the 2.35% card fee for dine-in (Texas Restaurant Association 2025). Add a waste cushion to theoretical food cost after that, since the sector loses $162 billion a year to spoilage (The Restaurant HQ 2025) and combos are no exception. Express the result, finally, in dollars of contribution margin per transaction, NEVER in a bare percentage. That's the framework we apply at Masterestaurant before approving any promotion, with limited-service median food cost —32.4% (NRA 2024)— as the baseline, and the ecosystem's costing tool automates these four steps by channel. In a market that has already contracted, a badly costed combo no longer gets forgiven: full service is now roughly 18% smaller than in 2019 (Technomic 2024), and reopening costs between $75,000 and $150,000 for a small takeout spot (Rezku 2025). There's no spare margin to give away on blind promotions, and even less when replacing one hourly employee costs US$2,305 (Black Box Intelligence 2024).

Finding 7 — What this means for the owner in a market that has already contracted

Every dollar a badly costed combo leaks is a dollar less to retain the team that actually delivers. Masterestaurant's synthesis of real sources (NRA, Toast, Rezku, Black Box, Texas Restaurant Association) points to one stubborn discipline: cost on the net price per channel, subtract the three leaks, decide in dollars of margin. Review your best-selling combo today with the live real cost and adjust the price per channel before the next campaign launches; theoretical food cost already lied to you once. Over the LIST price is where the traditional method costs the combo; over the NET price, after the bundle discount, is where the live real cost costs it, and that net is exactly where the contribution margin truly contracts. Theoretical cost takes for granted that everything sells dine-in. Live real cost doesn't: it loads the delivery commission, 15%–30% per Rezku 2026, onto the channel where the combo gets promoted most in QSR.

Finding 8 — Why the traditional method and the live real cost diverge by 7.4 points

Card swipe fee at 2.35% per transaction (Texas Restaurant Association 2025) and waste: two lines the static recipe cost leaves out and the live cost deducts, the exact reason behind those 7.4 points the operator never sees. A combo at 45% food cost can be worth running if its absolute margin lifts the average ticket and speeds table turnover, something the traditional method never catches because it only reads percentage of sales; Masterestaurant reads dollars of margin per transaction, and that dollar is what pays the bills.