ONDC for restaurants, without the hype

1What is ONDC, in restaurant terms?

ONDC, the Open Network for Digital Commerce, is a network, not an app. On Zomato or Swiggy one company owns the app guests browse, your listing, the riders and the payment flow; that control is what the commission pays for. On ONDC those pieces belong to separate companies that speak one shared protocol.

• Buyer apps are where guests order: consumer apps that have joined the network.
• Seller apps are where you list: they hold your menu and send it to the buyer apps.
• Logistics providers deliver the food, unless you use your own riders.

Two things follow. No single company controls the whole order, so the fees are split and small. And there is no single “ONDC app” to check: your orders arrive through your seller app, from guests using buyer apps you may never have opened yourself.

2How does an ONDC order travel?

1. A guest finds your restaurant in a buyer app and places the order.
2. The buyer app passes it over the network to your seller app.
3. You accept it in the seller app, against the preparation time you set.
4. With network delivery, a logistics provider assigns a rider.
5. The rider collects the order and delivers it.
6. The guest pays in the buyer app, and your money reaches you through your seller app, after fees, on the settlement cycle in its terms.

Every step is a different company, which is the point of the network and also the source of most of its rough edges. When something goes wrong, the question “whose step was it?” has an answer, but you may have to ask it twice.

3What does ONDC cost a restaurant?

The cost comes in parts: the buyer app’s fee, the seller app’s fee, payment charges, and delivery, which is usually priced as a separate, visible charge rather than folded into a commission. Together the platform fees are commonly reported at around 3–5% of order value. Seller apps price their own service in different ways: a percentage, a flat fee per order, or a monthly subscription.

Compare that with the big two, where base commission alone runs well into double digits before GST on fees, ads and co-funded discounts. The commission guide lists every line. As an example, a 4% fee on a ₹500 order is ₹20.

Two honest footnotes. Delivery still has to be paid for: on ONDC it is often shown to the guest as its own charge, or paid by you as a promotion, so watch what it does to the price the guest sees. And the apps’ fees carry 18% GST, which you can’t claim back at the 5% restaurant rate. To compare channels, run your own figures through the commission calculator with ONDC’s fees in place of a commission.

The GST on the food follows the same section 9(5) rule as on the big apps: tax on restaurant service supplied through an e-commerce operator is paid by the operator, not by you. On a network with several apps in one order, ask your seller app which of them pays it, and have your CA check how those sales appear in your returns.

The pricing difference is easy to see. To net ₹200 from a dish on a channel that takes 5%, you list it at ₹200 ÷ 0.95, about ₹211. On a channel with a 30% effective take, the same dish needs ₹286. Whether you pass that gap on to ONDC guests or keep it as margin is your call.

4How do you register a restaurant on ONDC?

You don’t register with ONDC directly. You join through a seller app, which handles the connection to the network.

1. Choose a seller app. This is the decision that matters. Compare the fee schedule, the catalogue tools, how orders reach you, the delivery options and how quickly support answers. Ask to speak to a restaurant like yours that already uses it.
2. Submit documents. Your FSSAI registration or licence, PAN, GST details or the declaration that applies, bank details, the outlet address and the area you deliver to.
3. Build the catalogue. Dishes, prices, photos, veg or non-veg marks, preparation times and opening hours. Every buyer app shows what you put here, so a thin listing converts poorly everywhere at once.
4. Choose delivery. The network’s logistics providers, your own riders, or both.
5. Go live and place a test order. Order your own food through a buyer app. One test order teaches you more about fees, delivery estimates and how your photos look than any sales call.

The cost of joining is low, and the documents are ones a restaurant already holds.

5What makes an ONDC listing work?

The same things that work on any delivery app, with one difference: several buyer apps show your catalogue, each in its own layout, so it has to read well without help.

• A photograph of the real dish, in daylight, for every dish.
• Names a guest would search for, with a line on what is in each dish.
• Veg and non-veg marks on every item, and allergens where they matter.
• Preparation times you can hold at your busiest hour.
• Opening hours that match the kitchen’s, so orders don’t arrive after the stoves are off.

6What should you ask a seller app before you sign?

• What exactly do you charge, and what does the buyer app charge on top? Show me one example order with every deduction.
• How often do you settle, and where do I see the deductions for each order?
• Who handles a refund or a complaint, and how quickly?
• What happens to an order when no rider can be found?
• Can my prices here differ from my other channels?
• How do I mark a dish out of stock, and how fast does it change across the buyer apps?
• How do I leave, and how much notice do you need?
• Can I speak to a restaurant like mine that uses you today?

A seller app that answers all eight in writing, without hedging, is a good sign before you have taken a single order.

7What order volume should you expect?

Be careful with headline numbers. Network-wide figures count every category, not only food, and orders divide unevenly between sellers. Volumes have also followed incentives: Business Standard reported in March 2025 that orders in ONDC’s retail segment were falling as the network cut the incentives it funded, from a monthly cap of ₹3 crore in July 2024 to about ₹30 lakh. The same report noted that Paytm, an early buyer app, had taken the network’s shopping icon off its home screen.

So plan for a modest channel. Whatever ONDC brings in arrives at a fraction of the big apps’ deductions, which makes even a few orders a day worth having, if the catalogue and stock are kept accurate. Measure a full quarter before you decide.

What to measure in that quarter:

• orders a day, and the average order value
• what reaches your bank per order, after every fee and any delivery you paid for
• cancellations, and which side cancelled
• delivery times, and the complaints about them
• how many guests come back for a second order

At the end, compare the money that reached your bank per order with the same figure for your other channels. That one comparison tells you how much effort ONDC deserves next quarter.

8What are the menu, pricing and delivery gotchas?

• A third menu to keep true. Your ONDC catalogue drifts like any menu. A dish that runs out in your kitchen must go out of stock on the network too, or you cook cancellations. Someone owns the update, by name, every shift.
• Pricing. With lower fees you can price below your Zomato and Swiggy menus and still keep more. Whether you should depends on any price-parity terms in your app contracts, so read them first. The menu price calculator prices each channel.
• Delivery. With network riders, the person delivering doesn’t carry your name, and the guest will still blame you for a cold meal. Track delivery times from the first week. With your own riders, keep the delivery area honest. If you already have riders for phone orders, ONDC orders can spread their cost over more deliveries, but only within a radius they can cover while the food is still hot.
• Disputes. Refunds and complaints run through the network’s issue and grievance process, across several companies, which owners report can be slower than an app’s in-house support. Keep your own order records.
• Reconciliation. A third channel means a third payout statement. It is small at first; build the habit before the volume arrives.

9ONDC, the big apps or your own orders?

They do different jobs.

• Zomato and Swiggy bring the most new guests, at the highest deduction. They are paid discovery.
• ONDC brings fewer guests, at a far lower fee. It is cheap discovery, where it brings any.
• Your own orders, regulars ordering by phone for collection or at your counter, carry no commission at all, and the relationship stays yours.

If you have to choose an order, build your own orders first, because they earn from guests you already have. Then add ONDC as a low-cost extra.

CountStand covers dine-in and takeaway, so a regular’s phone order for collection goes on the same till as the dining room, with a kitchen ticket, a GST bill and a dynamic UPI QR for the exact amount, paid straight to your own account. ONDC and other delivery-app orders are not open at launch: they come through your seller app’s own tools, so keep those records next to your till’s when you reconcile.

10The verdict, format by format

Cloud kitchens: list. Delivery-only kitchens feel commission the most, so every order that moves to a lower fee helps, and each brand can be listed with its own catalogue.

QSR: list, with stock discipline. Standard, photographed, fast menus suit buyer apps. The risk is cancellations from a stale catalogue, so keep it in step with the kitchen.

Casual dining: list if it’s easy, but put your own orders first. Your dining room is full of guests who could order from you directly. ONDC is a modest extra on top.

ONDC isn’t a replacement for the big apps, and it doesn’t need to be one to earn a place. Register through a seller app, keep the catalogue honest, measure a full quarter, and keep building the orders you own, because the lowest fee on any network is still more than the fee on an order a guest places with you directly.

11Questions owners ask