Real cost of a combo or promotion: the costing alternatives, unvarnished

The real cost of a combo or promotion equals recipe cost PLUS the implicit gift inside the discount, PLUS cannibalization of the dishes the guest would have bought at full price, PLUS the channel commission on wherever that sale lands. Add up only the recipe card and you are looking at somewhere between 55 % and 70 % of the true number.
For most operators running one or two locations, the right alternative in 2026 is a well-built spreadsheet tracking contribution margin per combo, reviewed every 30 days. Your POS promotions module exists to MEASURE what happened, never to set the price. Specialized costing software pays for itself from the third location onward, or with inventories above 250 SKUs. And the Masterestaurant menu engineering framework is what turns the number into a decision once the combo is already selling and you cannot tell whether to keep it, reprice it or kill it.
An owner in Bogotá sent me his flagship combo: burger, fries and a soda at 34,900 pesos. His recipe card read 11,200 in ingredients, a 32 % food cost, right at the ceiling I consider the MAXIMUM tolerable per dish. The combo moved 1,400 units a month and he was delighted. When we opened the real books, the number came back at 47 %, and the gap sat in no ingredient at all: 62 % of those combos arrived through delivery at 27 % commission, and the included soda displaced a fresh lemonade that carried 4,100 pesos more contribution margin per ticket.
That is the pattern. Theoretical cost lives on the recipe card; actual cost lives in the P&L, and four leaks open up between them that almost nobody books: waste on components that exist only for the combo, cannibalization of the full-price dish, channel commission, and the opportunity cost of a table held 52 minutes by a ticket with less absolute margin than your average. None of the four shows up in the food cost calculator bundled with your POS.
What follows compares the four real routes to that number — spreadsheet, POS module, specialized software and the menu engineering framework — with cost, learning curve and the operator profile each one serves. There is a firm recommendation at the end, plus a four-question decision tree.
Side-by-side comparison
| BEFORE: cost from the recipe card | AFTER: real cost via contribution margin | |
|---|---|---|
| Combo cost, burger + fries + drink (menu price 34,900 COP) | ✕11,200 COP · 32.1 % food cost | ✓16,400 COP · 47.0 % real cost including channel and cannibalization |
| Contribution margin per unit sold | ✕23,700 COP estimated (67.9 %) | ✓18,500 COP actual (53.0 %) after 27 % delivery commission |
| Cost components the method captures | ✕1 of 5: recipe ingredients only | ✓5 of 5: ingredients, 4.8 % waste, commission, cannibalization, table cost |
| Units/month needed to cover 4.2 M COP of fixed cost | ✕177 combos (optimistic math) | ✓227 real combos · 28.2 % more volume for the same break-even point |
| Monthly EBITDA effect at 1,400 units | ✕+7.3 M COP projected | ✓+1.1 M COP actual · 6.2 M COP annual variance from one combo |
| Calculation time per combo and review cadence | ✕12 minutes, once at launch | ✓38 minutes first pass, 9 minutes per 30-day review |
| Decision it unlocks | ✕«The combo is profitable» | ✓«Profitable in the dining room, margin-destroying on delivery: price by channel» |
The 34,900-peso combo that really cost 47 %, not 32 %
A combo costs what its recipe card says PLUS three line items that card never sees: channel commission, cannibalization of the full-price dish, and waste from components that exist only for that promotion. In Bogotá, a burger with fries and a soda at 34,900 pesos showed 11,200 in ingredients, meaning 32 % food cost, exactly the CEILING I consider tolerable per plate. It moved 1,400 units a month and the owner slept fine. Once we opened the books the real number was 47 %: 62 % of those combos came in through delivery at 27 % commission, and the soda displaced a fresh lemonade that carried 4,100 pesos more contribution margin per ticket. Fifteen points of cost that sat in no ingredient at all. Your POS food-cost calculator measures one component of cost —the ingredient— and that is why it understates the real cost of a promotion by ten to fifteen points.
Why does the POS calculator fall short here?
It was built to answer what the plate is worth, not what it is worth to SELL that plate through a given channel to a guest who would have ordered something else.
The giveaway is straightforward: when theoretical food cost reads 30 % and the P&L reads 42 %, that gap is not a typing error, it is the promotion working against you. With a median prime cost of 65 cents on every sales dollar in limited service, per the National Restaurant Association Restaurant Operations Data Abstract 2025, no amount of extra volume absorbs twelve points of leakage. My default recommendation is still a properly built spreadsheet: it costs zero pesos and six to ten hours of your own time. Who it fits: the one- or two-location operator who already runs monthly inventory and wants to set prices on their own judgment, not a software vendor's. The learning curve sits in the middle, because you have to understand the line between variable and fixed cost, and never load payroll or rent onto the plate —those belong to break-even, never to the recipe card.
Spreadsheet with contribution margin per combo
Switching cost is your time, not your cash. And there is a reason beyond price: a spreadsheet FORCES you to think the business through, column by column, while software hands you a conclusion you never audited. Diego F. Parra uses one with clients billing seven figures. The promotions module already bundled in your POS runs between 0 and 45 dollars a month —usually included in the plan you pay for— and switches on in an afternoon. It serves the operator who needs to MEASURE: redemption rate, time-of-day consumption, product mix inside the combo, average ticket with and without the offer. Its limit deserves naming precisely: it records what happened at the point of sale, not what happened to the margin. Aggregator commission and displacement of the full-price dish stay outside its field of view. Treat it as a demand thermometer and carry those numbers elsewhere for the profitability math.
POS promotions module: cheap, fast, blind to commission
As a measurement layer it is excellent; as a judge of whether the combo pays, it simply lacks the information. Specialized costing software pays for itself from three locations up, or with an inventory above 400 SKUs, because at that point reconciling recipes, purchase prices and waste by hand eats a full headcount. It runs 90 to 400 dollars monthly depending on modules, ingests supplier invoices and recalculates every recipe cost the moment an input price moves. That matters more in 2026 than in any recent year: USDA ERS forecasts beef up 7.5 % with the cattle herd at a 75-year low, and non-alcoholic beverages and coffee up 5.7 %. A combo built on beef and soda takes both waves. The curve is steep, four to eight weeks of implementation, and it demands inventory discipline many teams do not have. Menu engineering does not compete with the other three options, it governs them: this is the framework that crosses popularity with absolute contribution margin and tells you what to do with the answer.
Menu engineering: the framework that decides, not the tool that calculates
A combo can carry 47 % cost and still earn its place, if it pulls traffic into the dead three-o'clock window and its absolute margin per ticket beats a 28 %-cost dish selling twelve units a week. There sits the bridge between two ideas that look like enemies: percentage protects cash per unit, absolute margin pays the rent. Who needs it: the owner who already holds the numbers and now has to decide what gets cut. Without clean data from the other tools this framework does nothing; with them, it is the only thing that turns arithmetic into a decision. Say the Bogotá combo takes an 8 % price increase and loses 15 % of its units. It sold 1,400 a month; now it sells 1,190, at 37,700 pesos. With the same 11,200 in ingredients, food cost drops from 32 % to 29.7 %, and the 210 units that walked away do not vanish from the dining room: part of them return to the regular menu, where fresh lemonade contributes that extra 4,100 pesos of margin per ticket.
The scenario almost nobody runs before signing off on a promotion
The operating question is not whether you lose sales, but what absolute margin survives once traffic reshuffles. Run that exercise across three elasticity scenarios before touching price and you will know whether the combo is a traffic engine or a machine for buying guests at a loss. Skip the simulation and the price move is a bet with rent money. Stay where you are if more than 80 % of your combo comes through the dining room, channel commission is zero, and your closed prime cost holds below 60 % of sales, the healthy target Toast and Nation's Restaurant News both place in the 55–65 % band. Theoretical and real cost nearly converge there, and standing up 300-dollar-a-month software to discover it means paying for certainty you already own. Do not switch during peak season either: rolling out costing while the kitchen runs at capacity produces dirty data nobody goes back to fix.
When NOT to switch tools or touch the combo?
And one case makes staying the brave call: when the combo loses three points against the regular menu yet sustains the volume that lets you buy beef at a scale discount.
That combo does not give you margin; it gives you purchasing power, and that is money too. ALTERNATIVE 1 — Spreadsheet tracking contribution margin per combo. Cost: zero, plus six to ten hours of your own build time. Learning curve: medium, since you must separate variable from fixed cost and resist loading payroll or rent onto the dish — those belong to the break-even point, not the recipe card. Best for: one or two locations, monthly inventory already running, an owner who wants to price with his own judgment. This is my default recommendation and I will not soften it: a spreadsheet forces you to think about the business, software does not. ALTERNATIVE 2 — POS promotions module. Cost: 0 to 45 USD monthly, usually bundled into the plan you already pay for.
Four alternatives, with cost and learning curve
Curve: low, live in an afternoon. Best for: operators who need to MEASURE redemption, daypart and product mix. Its limit is serious. The POS tells you what sold, never what it truly cost, because it knows nothing about cannibalization or seat-hour cost. Treat it as a sensor, never as a pricing calculator. ALTERNATIVE 3 — Specialized costing and inventory software. Cost: 89 to 350 USD per location per month, with 20 to 60 implementation hours. Curve: steep through the first quarter, because the system returns truth only if recipes and waste go in clean. Best for: three locations and up, inventories past 250 SKUs, or central production kitchens. Here the math works: one recovered food cost point on 200,000 USD of annual sales is 2,000 USD, and the license costs less than that. ALTERNATIVE 4 — Masterestaurant menu engineering framework. Cost: the method itself, plus the discipline of a 30-day review.
Four alternatives, with cost and learning curve — in practice
Curve: medium-to-steep, since it requires classifying every item by popularity and contribution margin, then moving price, recipe or menu placement according to the quadrant. Best for: the owner sitting on data he cannot act on. This is the step that converts costing into a decision, and where Diego F. Parra concentrates his work with the operators he advises: a number without a decision is bookkeeping, not management. WHAT NO ALTERNATIVE FIXES: if your recipe is not standardized in grams, all four hand you a false number at different levels of sophistication. Standardize first. A badly costed combo inside 300-dollar software is still a badly costed combo. ON PRINTED MENUS AND QR MENUS: when the combo gets communicated, the Masterestaurant recommendation is to ALWAYS keep the printed menu alongside the QR. The printed menu controls service pace, menu narrative and the server's suggestive selling — which is exactly where combo margin rises or falls. The QR complements it for delivery, accessibility, fast price changes and consultation analytics. Never QR alone: both, each with its own role.
Criterion-by-criterion comparison
What your recipe card genuinely solvesThe original option
- It sets the floor: no combo can cost less than the sum of its inputs, and that alone kills suicidal pricing in thirty seconds.
- It is cheap and repeatable — once the recipe is standardized, anyone in the kitchen can recalculate it the day protein prices move.
- It arms you for supplier negotiations, because cost per gram becomes explicit and you walk in with the figure in hand.
- It feeds waste control: if the card says 180 grams of beef and inventory burns 191, you have a 6.1 % leak no promotion can justify.
- HARD LIMIT: it does not see the discount as a cost. A 2-for-1 appears on neither dish's recipe card, yet it doubles ingredient cost per ticket while revenue stays flat.
Where the recipe card runs outMasterestaurant
- Cannibalization: per National Restaurant Association 2026 data, 41 % of combo buyers would have purchased the main dish at full price anyway, so the discount is cash handed to demand you already owned.
- Channel commission: an aggregator sale at 27 % turns a 32 % food cost into a 59 % combined variable cost, and contribution margin flips negative the moment a discount rides along.
- Combo-specific waste: components existing only for the promo rotate worse and spoil faster; we measure 4 % to 7 % above general inventory.
- Table cost: in a 40-seat room running two turns, every seat-hour carries an assigned cost, and a combo occupying that table as long as a 30 % larger ticket is buying revenue with capacity.
- Knock-on effect: a combo that drags down average ticket drags down tips too, and with tips goes server retention, which costs between 0.8 and 1.4 monthly salaries to replace.
Side-by-side comparison
| BEFORE: cost from the recipe card | AFTER: real cost via contribution margin | |
|---|---|---|
| Combo cost, burger + fries + drink (menu price 34,900 COP) | ✕11,200 COP · 32.1 % food cost | ✓16,400 COP · 47.0 % real cost including channel and cannibalization |
| Contribution margin per unit sold | ✕23,700 COP estimated (67.9 %) | ✓18,500 COP actual (53.0 %) after 27 % delivery commission |
| Cost components the method captures | ✕1 of 5: recipe ingredients only | ✓5 of 5: ingredients, 4.8 % waste, commission, cannibalization, table cost |
| Units/month needed to cover 4.2 M COP of fixed cost | ✕177 combos (optimistic math) | ✓227 real combos · 28.2 % more volume for the same break-even point |
| Monthly EBITDA effect at 1,400 units | ✕+7.3 M COP projected | ✓+1.1 M COP actual · 6.2 M COP annual variance from one combo |
| Calculation time per combo and review cadence | ✕12 minutes, once at launch | ✓38 minutes first pass, 9 minutes per 30-day review |
| Decision it unlocks | ✕«The combo is profitable» | ✓«Profitable in the dining room, margin-destroying on delivery: price by channel» |
The figures that settle the decision
“I defended that combo like it was my kid: 1,400 units a month, the card said 32 % food cost, and to me that was winning. Once we split dining room from delivery, the real 47 % showed up along with the 6.2 million pesos a year bleeding out through commission and through the soda that was killing my lemonade. I did not pull the combo. I priced it by channel, 39,900 on the apps and 34,900 in the room, swapped the included drink, and by month two contribution margin per unit went from 18,500 to 24,100 pesos. Same volume, 7.8 million more profit in the quarter.”
How to calculate a combo's real cost in four steps
Weigh combo components as they actually leave the kitchen, not as the written recipe claims. Add measured waste from your last inventory, which on promotions typically runs 4 % to 7 % above general inventory. That figure is your floor: if it already exceeds 32 % of menu price, the problem is neither the channel nor cannibalization, it is the recipe, and no costing method will rescue it. Log cost per gram so you can recalculate in two minutes when your supplier raises protein.
Break combo sales into dining room, takeout and aggregator, then apply each channel's commission and packaging cost. A combo can hold 53 % contribution margin at the table and under 20 % on the apps, and the weighted average hides precisely the channel draining your cash flow. This step changes the decision more often than any other: you rarely kill the combo, you price it differently by channel, which is legitimate and which the aggregators themselves permit.
Compare standalone units of the main dish across the 60 days before the combo and the 60 days after. If standalone sales dropped 300 units while the combo sold 1,400, you gifted a discount to 300 buyers you already had and won 1,100 new ones: good trade. If standalone dropped 1,200 against 1,400 combos sold, you just cut your average ticket in exchange for 200 incremental guests, and that destroys contribution margin beyond argument.
Plot popularity against real contribution margin and place the combo in one of four quadrants. High margin, high rotation: protect it and raise its visibility on the printed menu. High rotation, thin margin: reformulate components or lift price 6 % to 9 %, the band where elasticity rarely punishes. Low on both: pull it this month. High margin, low rotation: that is a communication problem, not a cost problem, and servers fix it with suggestive selling. Calendar the review every 30 days, because input costs move and the decision expires.
And with AI?
Project your food cost, spot margin leaks and simulate pricing scenarios in minutes. Diego F. Parra is an expert in AI applied to restaurants.
Free tools to apply this now
Ecosystem tools that keep the decision alive
You run the math once; the hard part is sustaining the 30-day review when service gets heavy. These Masterestaurant ecosystem tools exist so your combo costing does not end up as a file nobody opens again.
Frequently asked questions about a combo's real cost
What is the difference between theoretical and real combo cost?
What is the difference between theoretical and real combo cost?
Theoretical cost is the recipe card sum: inputs per gram. Real cost adds measured waste, channel commission, cannibalization of the full-price dish and the cost of the seat-hour occupied. In a casual restaurant combo that gap typically runs from 32 % to 47 % of menu price, which explains why EBITDA stays flat even as the combo sells.
Can a 2-for-1 be profitable, or does it always destroy margin?
Can a 2-for-1 be profitable, or does it always destroy margin?
It can work when the second dish carries low food cost and pulls guests who were not coming. It stops working once 41 % of redeemers would have bought anyway, because you double ingredient cost with no added revenue. Practical rule: a 2-for-1 is justified only in dayparts with proven idle capacity, never during peak service.
Should payroll and rent be loaded onto the combo cost?
Should payroll and rent be loaded onto the combo cost?
No. Payroll, rent and utilities are fixed costs and belong to the break-even point, not the dish recipe card. Loading them onto the combo inflates unit cost and pushes prices the market will not accept. The dish carries ingredients and waste; fixed structure gets covered by the aggregate contribution margin of the whole menu.
How often should I recalculate the real cost of my promotions?
How often should I recalculate the real cost of my promotions?
Every 30 days minimum, and immediately whenever a key input rises more than 8 %. Protein prices, delivery packaging and aggregator commissions move several times a year; a costing exercise from six months ago is an expired snapshot pushing you to price against data that no longer exists.
Sector data 2026 (official sources)
Verifiable industry benchmarks from official, non-commercial sources (government, industry associations, market research) - not competitors.
| Metric | Benchmark 2026 | Source |
|---|---|---|
| Facturación anual de la hostelería en el Reino Unido | £144.000 millones al año (2024) | UKHospitality / House of Commons Library 2024 |
| Número de negocios de hostelería en el Reino Unido | 176.685 negocios (marzo 2025) | House of Commons Library 2026 |
| Ventas de servicios de comida y bebida en Canadá | CAD 96.500 millones en 2024 (+4,0% vs 2023) | Statistics Canada 2024 |
| Participación por segmento en ventas de foodservice (Canadá) | servicio limitado 46,4% / servicio completo 43,1% (2024) | Statistics Canada 2024 |
| Peso de la industria restaurantera en los negocios de México | 12,2% de las unidades económicas del país | INEGI–CANIRAC 2024 |
| Pronóstico de precios de carne de res (EE. UU.) | +7,5% en 2026 (hato ganadero en mínimo de 75 años) | USDA ERS (Food Price Outlook) 2026 |
Related content
Grow your restaurant with the Masterestaurant method
Applied in +8.400 restaurants across 43 countries.
