Cut delivery commissions: the traditional method negotiates the rate, the Masterestaurant method changes where the order comes from

1. Commissions aren't negotiated, they're diluted: shift 18% to 30% of volume to your own channel

Cutting delivery commissions doesn't come from negotiating with the aggregator, it comes from moving between 18% and 30% of your digital volume onto a channel you control, until the platform stops owning the customer. The arithmetic is uncomfortable but clear: pulling two points off a 28% commission across 100% of volume gives you two points of a channel that still belongs to someone else, while shifting a quarter of that volume to a direct order with 6% processing cost returns more than five points of contribution on that slice, and hands you the guest's email and phone. The global online delivery market reaches 1.51 trillion dollars in 2026 and grows at 6.24% compound annually through 2031 according to Statista (2026): the channel is staying, the question is who owns the customer inside it.

2. Why does the commission sit on the wrong line of the P&L?

An aggregator commission is a revenue reduction, not a marketing expense, and booking it wrong is what makes a board celebrate growth that never happened.

When that 28% lands below, among promotional costs, the channel's gross sale enters the top line whole and the dashboard shows a 20% revenue jump while real contribution falls. Booked above, as a deduction from revenue, the same order shows what it is: a 12 dollar ticket that reaches the register as 8.64 before food cost touches it. Diego F. Parra pushes the boards Masterestaurant works with to break the P&L out by channel, with its own net revenue line and its own contribution margin, because until delivery has a column of its own nobody can make a decision about it. A channel without a column is a channel without an owner.

3. Under 500 thousand dollars: differentiated pricing and a 25-order weekly threshold

Below 500 thousand dollars in annual revenue there is one decision worth making: separate the channel price from the dining room price before attempting anything else. A 15% to 18% uplift on the delivery menu absorbs most of the commission without disturbing how the in-house guest perceives value, and it's the only lever an operator this size can execute on Monday morning. Direct ordering at this level starts with an order-and-pay link, not an app: under 25 direct orders per week the infrastructure costs more than it saves. Start with price.

4. From 500 thousand to 1 million: the band where you buy your first owned customer

Between 500 thousand and 1 million dollars a year the goal shifts from price to ownership, and the number governing the decision is 18% of digital volume originating in your own channel within twelve months. At this scale the business already generates 40 to 90 digital orders per week, enough for a customer database to start being worth money: bringing back 12% of those guests one extra time per month through a direct channel pays for the whole toolset and leaves change. Operators who fail in this band are the ones who build direct ordering and then leave it without demand, waiting for the guest to find it. He doesn't find it. You have to buy him inside the aggregator's own bag, with a physical insert and a concrete reason to come back, which is the one contact surface the platform still can't take away from you.

5. Above 1 million: direct ordering stops being a project and becomes a business line

Past a million dollars in annual revenue the direct channel needs a budget, an owner and a quarterly target, or it dies of neglect no matter how well the technology works. The operating threshold is 25% of digital volume direct and an acquisition cost recovered inside three orders: if the second direct order still hasn't paid for bringing that customer in, your problem isn't the commission, it's the offer. Negotiating with the aggregator makes sense at this scale, but from a different chair, because a restaurant already routing a quarter of its orders elsewhere negotiates with a real alternative on the table. Uber Eats moved 74.6 billion dollars in gross bookings during 2024 according to Statista (2024): nobody hands points back out of goodwill, they hand them back when your volume can walk.

6. Over 5 million: satellite kitchens, virtual brands and the high-end case

Beyond 5 million a year the lever stops being commercial and turns structural: owned satellite kitchens that shorten the delivery radius, and virtual brands that load the idle capacity of a kitchen already paying rent. The global dark kitchen market is projected at 171.3 billion dollars by 2033 according to Global Growth Insights, and that figure doesn't describe a fashion, it describes a change in where the order gets produced. Here appears the high-end case worth naming by profile: the celebrity-chef restaurant or the large-format themed venue, with a brand capable of generating demand on its own, that still hands 90% of its digital volume to a third party out of operational convenience. That profile leaves the most money on the table, because it pays commission on demand it created itself.

7. Over 10 million: a group negotiates with architecture, not with emails

From 10 million dollars a year upward, a group or chain stops buying rates and starts buying position: a master agreement with tiered pricing on aggregate volume, direct interface integration that removes the duplicate terminal, and order data returned to the group as a contract condition. The threshold here is 30% of digital volume in the owned channel, held for two quarters, because that's the figure that turns a discount conversation into a mutual-dependency conversation. Swiggy operated with 196,000 partner restaurants across 653 Indian cities in fiscal 2023-24 according to its annual report (Swiggy, 2024): against that scale a standalone group has no leverage, but a group with a consolidated direct channel and data on its own guests does. You build that difference before you sit down.

8. What would happen if the aggregator raised the commission three points tomorrow

Assume the platform lifts your commission from 28% to 31% next quarter, without warning and without room to argue, exactly as it has happened in several markets. A restaurant with 95% of its digital volume there loses three clean points across nearly all its digital sales and has no available answer inside six months. A restaurant with 30% direct loses three points on the remaining 70%, a bit over two effective points, and can push that owned base with a win-back campaign the same week. The tension in this trade is real: the aggregator brings volume you wouldn't bring alone, and that volume is useful. I resolve it firmly and without a comfortable middle: use it as a paid acquisition channel, never as a retention channel. This week, print an insert with one concrete reason to order direct and drop it into every bag that leaves.

9. Where the equation breaks

Commission is not a marketing expense, it is a revenue reduction. Booked on the wrong P&L line, leadership reads growth that does not exist and celebrates volume while contribution erodes. The aggregator does not sell you delivery, it sells you demand. And the demand you rent every month is the same demand you could buy once by building the asset. Cutting delivery commissions starts with an accounting decision rather than a commercial one: separate channel price from dining-room price. A dish priced at USD 12 in the room and USD 12 in the app is not the same product, because one carries up to 30% commission and the other carries none.

10. Where the equation breaks — in practice

The dark kitchen vs brick and mortar restaurant debate gets framed badly when it collapses into square footage. A hidden kitchen lowers occupancy cost, certainly, but inherits the commission problem intact and loses the foot traffic that amortizes acquisition. The global dark kitchen market projects USD 171.3 billion by 2033 according to Global Growth Insights, which confirms the format works; it does not confirm it works without an owned channel. Increasing sales on Rappi and lowering blended commission look like opposing goals and are not: the marketplace is an excellent discovery engine and a terrible repeat-purchase engine. Use each for what it does.