The Silent Commission: What Each Aggregator Order Really Costs You

1. What does each aggregator order really cost?

The real cost of an aggregator order runs 46%–58% of the ticket, not the 28%–35% printed in your contract. The nominal commission is only 50%–65% of the channel's total cost;

the rest lives off the page. Across +8,400 units we audited at Masterestaurant in 43 countries, the owner reads the contract percentage and closes the analysis. It's the most expensive decision-architecture error of the foodtech decade. Add packaging (4%–8% of the ticket), platform shrinkage from errors and disputes (1.5%–3%), the hidden CAC of promos to rank higher (6%–12%) and the cost of capital tied up for 14–30 days. A dish with 30% food cost in the dining room yields positive contribution; that same dish cloned to delivery without its own spec sheet yields negative contribution. The order isn't the one shown in the app: it's the one that survives settlement.

2. Nominal commission versus total channel cost

The visible commission is only 50%–65% of the true cost of running delivery; treating it as the final number disguises the loss. A contract at 30% looks harmless until you stack the layers you didn't sign: premium packaging so the food arrives presentable (4%–8% of the ticket), the in-platform electronic payment fee (1.5%–2.5%), the logistics fee for far-zone delivery and the customer-refund adjustments the platform deducts without asking. I've seen in dozens of restaurants how a P&L celebrating 'only 30% commission' hides a channel eating 52% of the ticket. The boardroom rule is simple: anything eroding price between the customer and your bank account is channel cost, whether or not it's in the clause. Reconcile the deposit, not the order. Between the price the customer sees and what the platform deposits there's a 4%–9% gap the owner rarely reconciles.

3. Gross ticket versus net settlement: the gap no one reconciles

The app charges 20 dollars; between tax withholdings, adjustments for 'platform-funded' promotions, refunds for orders the customer flagged as incomplete and disputes resolved for the buyer by default, the real deposit drops to 17.80–19.20. Multiply that leak by 1,200 orders a month and it's 1,500–2,600 dollars vanishing with no invoice to explain them. At Masterestaurant we enforce a monthly ritual: cross the platform report against the bank statement, line by line, for a full quarter. Nine of ten operators discover charges they never authorized. The gross order is marketing; the net settlement is accounting. Only the second pays payroll. Promotions to rank higher in the app are paid acquisition, and without assigning them to each order the margin is disguised by 6%–12%. The owner runs a 2-for-1 or a 20% discount 'to gain visibility' and books it as generic marketing expense, blurred across the month.

4. Promos aren't diffuse marketing: they're CAC per order

It's a textbook costing error. That discount acquired a specific order from a specific customer; it's CAC, and it must be subtracted from THAT order's contribution. When we assign it correctly in the audit, the promoted ticket usually shows negative contribution of 1.50–4 dollars: the restaurant pays to sell. The discipline we demand: every campaign is modeled as variable cost per order, not as a fixed line. If the promoted order doesn't cover food cost plus packaging plus commission, it isn't growth, it's subsidizing the aggregator with your capital. Cloning the dining-room price to delivery destroys 8 to 15 margin points per order; the delivery dish demands its own spec sheet, its own packaging and its own price. In the room, that dish absorbs a 30% food cost and yields healthy contribution because the guest pays for the experience, the drink and the dessert that lift the ticket.

5. The delivery dish needs its own cost sheet

In delivery it travels alone, with no high-margin beverage, in packaging costing 0.80–1.60 dollars and a commission taking a third. Diego F. Parra repeats it in every engagement: the error I see again and again is the mirror menu, identical price in both channels. The Masterestaurant method fixes it with a separately designed delivery menu, dishes that travel well, pricing set to the channel's real cost and combos that rebuild the ticket. The same dish in two channels is two different products. Platforms settle in 14–30 days, and that tied-up capital costs between 1.5% and 4% of GMV when the business is financed with credit at 25%–40% a year. The owner celebrates volume without seeing that every dollar sold takes weeks to return, while he pays suppliers at 7 days and payroll every fifteen. That cash-flow mismatch forces expensive credit lines or stretched payments, and the financial cost is real even if no one signs it.

6. Tied-up capital: the cost in no clause

In a dark kitchen with 40,000 dollars in monthly GMV and 21-day settlement, we're talking about 2,300 dollars permanently locked. This is where a dark kitchen stops being a growth lever and becomes a liability with a nice logo: when the channel grows faster than the cash that sustains it. Raising volume in a channel with negative contribution accelerates the loss instead of fixing it: each extra order deepens the hole, it doesn't fill it. It's the most common mental trap we correct in the boardroom: the operator sees aggregator GMV grow 20% month over month and reads it as success, when in reality he's scaling a business losing 2–4 dollars per transaction. Doubling 500 orders to 1,000 doesn't double profit; it doubles the bleed. The metric that matters isn't GMV or order count: it's contribution per order after food cost, packaging, total commission, promo CAC and cost of capital.

7. GMV that grows versus contribution that falls

If that number is negative, the correct order is to brake, not accelerate. At Masterestaurant we rebuild the per-order sheet first; only when contribution is positive and stable do we authorize growing volume. Growing on a loss is the most expensive way to go bankrupt with rising sales. The concrete action is one: rebuild the real contribution per aggregator order before approving one more dollar of growth. Take your last 200 platform orders and subtract from each the real food cost, packaging (0.80–1.60), total commission reconciled against the bank, the CAC of the promo that generated it and the financial cost of 14–30 days of waiting. The number left is your truth, not the app's optimistic report. If it runs 46%–58% total cost like most of the 8,400 audited units, you have three levers: redesign the delivery menu with its own price, push your direct channel with lower food cost and zero commission, and renegotiate the fee by volume.

8. What to do Monday: rebuild the order before scaling

Well-costed delivery is profitable; delivery cloned from the dining room is a monthly donation to the aggregator. Decide with the sheet, not the GMV. Nominal commission vs total cost: the visible commission is only 50%–65% of the channel's real cost; the rest lives outside the contract. Gross vs net-settled ticket: between the customer's price and the platform's deposit there are holdbacks, adjustments and disputes owners rarely reconcile. Fuzzy marketing vs CAC per order: promos to rank on top are paid acquisition; unassigned, they cosmetically inflate margin. Dining room vs delivery: the delivery dish needs its own cost card, packaging and price; cloning it from the dining room destroys margin. GMV growth vs contribution: scaling volume on a negative-contribution channel accelerates the loss, it does not fix it.