Aggregator Commission or Your Own Delivery: Run the Numbers

01What is left from a single order

Average check 90 000 UZS, food cost 32%, so the dishes cost 28 800. Packaging 4 000. That leaves 57 200 UZS in hand before the order goes out. Now subtract the channel cost.

At 25% commission the aggregator takes 22 500 UZS. That leaves 34 700. Rent, wages and utilities come out of it. The aggregator supplies the courier, so logistics are not your problem.

On your own site the platform cost comes from the package: Light is 1 200 000 UZS a month for up to 1 200 orders, roughly 1 000 UZS per order. The payment provider (Payme/Click) takes 1–1.5%, about 1 100. The courier is yours: your own rider or a TaxiMillenium call, on average 15 000–20 000 UZS. Around 37 000 is left.

At first glance the gap is small. But when the guest pays the delivery fee (10 000–25 000 depending on the zone), your own channel beats the aggregator by 15 000–20 000 UZS per order. Over 1 000 orders a month that is 15–20 million UZS.

02The aggregator's real value is visibility, not the customer

Someone who orders through an aggregator is not your customer. Their number, address and order history stay with the aggregator. Tomorrow they open the app and pick whoever sits higher in the list. You cannot even send them an SMS.

So it is more accurate to treat the aggregator as an advertising channel. It finds you a new person — expensive work, and you pay 25% for it. Your job is to move that person's second order into your own channel.

The move is simple and honest: put a card with a QR in the box — "10% cashback on your next order, on our own site". The QR opens your site, the guest signs in by SMS, and now they are in your database. After the third order they stop opening the aggregator.

03Running both channels from one menu

The main pain of two channels is maintaining the menu twice. A price changes, the stop-list changes, a new dish appears: it has to be repeated everywhere. Within two weeks they drift apart.

In laCafe there is one menu: Yandex Eats and Uzum Tezkor connect to it, availability and prices flow out automatically, and order status moves both ways. An aggregator order lands on the till and in the kitchen exactly like a site order.

One technical requirement: the aggregators need MXIK/IKPU fiscal codes. They live on each dish card. Without them the menu will not upload — the slowest part of the integration, so prepare it in advance.

• Aggregator order: menu, availability and status sync automatically (the aggregator's own commission is separate)
• Site order: no commission, the customer stays in your database; the site comes with Start
• Telegram AI agent order: placed in conversation, from the Standart package
• Delivery zones on the map: each with its own fee, with the nearest branch chosen automatically

04When the aggregator is enough and when you need your own site

At 5–10 delivery orders a day there is little point in fussing with your own channel. Stay on the aggregator, build recognition, teach the kitchen to handle delivery. At that stage the commission is tuition.

Past 30 orders a day the maths flips. 30 × 30 days × 20 000 UZS of difference is 18 million a month. That is more than enough for your own site, a courier and a cashback programme. From that point the aggregator becomes the second channel.

Leaving the aggregator entirely is rarely right. It supplies a flow of new guests and finds you on holidays. The correct model is both: the aggregator introduces, your own channel retains.