ONDC for restaurants

What is ONDC, in restaurant terms?

ONDC, the Open Network for Digital Commerce, is a government-backed open network, not another app. The unbundling is the whole idea, so here it is in restaurant terms. On Zomato or Swiggy, one company owns everything: the app customers browse, your listing, the delivery fleet and the payment flow, and that vertical control is what a 25–35% effective take pays for. On ONDC, those pieces are split across independent participants speaking a common protocol. Customers order through buyer apps, consumer apps many of them already use for payments and shopping that have plugged into the network. You list through a seller app, your agent on the network, which hosts your catalog and pushes it to every buyer app at once. Delivery is fulfilled either by logistics providers on the network or by your own riders.

The consequence that matters to your P&L: no single participant has the market power to charge 25%, so nobody does. Commissions are unbundled and small. The consequence that matters to your operations: there is no single "ONDC app" to check, your orders arrive via your seller app from customers on buyer apps you may never have heard of. List once, appear everywhere the network reaches: that is the promise, and in 600+ cities as of March 2026, it is no longer theoretical.

What commission does ONDC charge in 2026?

The number that makes owners sit up: 3–5% of order value, combining the buyer-app and seller-app charges, versus the aggregator stack. Side by side:

Per 2026 marketplace fee breakdowns, the aggregators' effective take lands at 25–35% once the ₹17.58 platform fee, gateway charges, GST on services and co-funded discounts are counted, the full anatomy is in the commission reduction playbook, and the two incumbents are compared line by line in Zomato vs Swiggy for restaurant owners. Against that, ONDC's arithmetic on a ₹500 order is ₹15–25 of deductions versus ₹125–175. Same food, same packaging, a different network, and roughly ₹100–150 more reaching your account per order.

Two honest footnotes. Delivery has to be paid for somewhere: on ONDC it is typically priced as a visible logistics charge (borne by the customer, or subsidised by you as a promotion) rather than buried inside a fat commission, better transparency, but watch what it does to your delivered price. And seller apps price their own services differently, some flat, some percentage, some subscription. The 3–5% band holds broadly, but get your shortlisted seller app's exact schedule in writing. Then run both channel structures through the aggregator commission calculator with your own average order value.

How do you register a restaurant on ONDC?

You do not register with ONDC directly, you join through a seller app, which handles the network integration. The process, step by step:

1. Choose a seller app. This is the decision that matters; everything else is paperwork. Compare on fee structure (flat vs percentage vs subscription), catalog tooling, how orders reach you (their dashboard, or integration into your POS), logistics options and support responsiveness. Ask each candidate for restaurants like yours already live on the network, and talk to one.
2. Submit documents. The set mirrors aggregator onboarding because most of it is law: FSSAI licence, PAN, GST registration or the applicable declaration, bank details, and your outlet address and serviceability details.
3. Build the catalog. Menu items, prices, photos, veg/non-veg flags, preparation times and serviceable hours. This is worth doing carefully, your catalog is what every buyer app on the network renders, and thin listings convert poorly everywhere at once.
4. Choose fulfilment. Network logistics providers, your own riders, or hybrid. If you already run delivery for a direct channel, your riders can serve ONDC orders too.
5. Go live and place a test order. Order your own food through a buyer app. You will learn more about the customer experience, fees shown, delivery estimates, how your photos render, in one test order than in any sales deck.

Cost of entry is genuinely low, seller apps typically charge little or nothing upfront, with the economics sitting in per-order fees, and the document set is one you already hold. Budget the effort in days, not weeks, as a rule of thumb, with catalog quality the main variable.

What order volume should you honestly expect?

This is where most ONDC coverage gets dishonest, so let us not be. Expect 5–15 food orders a day initially. Not fifty. The network-wide numbers are genuinely large, 12 lakh+ daily transactions across all categories, 5 lakh+ sellers, 600+ cities per ONDC network figures, March 2026, but "network-wide across all categories" is doing heavy lifting in that sentence. Food is one category among many, and dividing network transactions by network sellers should cure anyone of extrapolating a single outlet's volume from headline numbers.

The trajectory is the better argument than the level: food orders on the network have been growing 15–20% month-on-month per the same figures. Compounding at that rate does interesting things within a year, but a trajectory is a reason to be positioned early, not a reason to forecast aggregator-scale volume into next quarter's cash flow.

So the honest frame: at 5–15 orders a day, ONDC will not replace your aggregator revenue. It does not need to. Those orders arrive at 3–5% commission instead of 25–35%, margin-rich volume that costs you almost nothing to be available for. And early sellers in each city build catalog quality, ratings and operational rhythm on the network before their competitors have filled out the paperwork. That is the entire first-mover case: cheap option now, compounding position if the growth holds.

What are the menu, pricing and fulfilment gotchas?

The traps owners actually hit, learned the mildly painful way:

• Catalog drift. Your ONDC catalog is a third menu to keep true. Items out of stock on a Saturday night must go out of stock on the network too, or you will cook cancellations and eat the ratings damage. If your seller app integrates with your POS, stock-outs sync automatically; if it is a standalone dashboard, someone owns updating it, by name, per shift.
• Pricing strategy. At 3–5% commission you can price below your aggregator menu and still net dramatically more. Whether you should depends on your channel strategy: a visibly cheaper ONDC price rewards customers for using the cheaper channel; matched prices bank the margin quietly. Decide deliberately, and mind any pricing-parity language in your existing aggregator agreements. The menu price calculator makes channel-wise pricing a five-minute exercise.
• Fulfilment ownership. With network logistics, the delivery experience is a third party wearing no brand, track SLAs from day one, because the customer will blame the restaurant regardless. With your own riders, serviceability radius is your promise to keep honest.
• Issue resolution. Refunds and disputes flow through the network's grievance process across multiple participants, and owners report it can run slower than an aggregator's in-house support. Keep your own order-level records rather than relying on anyone's dashboard.
• Reconciliation, again. A third channel is a third payout statement to tie back to bills. Small at 5–15 orders a day, but build the habit before the volume arrives.

ONDC or your own WhatsApp channel, which should you build first?

It is a false choice, they solve different problems, and the strongest 2026 posture runs both. ONDC is discovery you do not own: new customers finding you through buyer apps, at 3–5% instead of 25–35%. A WhatsApp/QR storefront is retention you fully own: your repeat customers ordering at 0% commission, on a channel no network participant can reprice, with the customer relationship, name, number, order history, in your hands rather than a platform's. If you must sequence them, build the direct channel first: it monetises customers you already have, this month. Then add ONDC as the cheap third lane for incremental discovery. The end state every commission-conscious operator is converging on: aggregators for acquisition, ONDC for low-cost incremental volume, direct for repeat business.

The operational objection, "three channels means three systems", is only true with fragmented tooling. CountStand is an AI-native restaurant operating system for India, offline-first billing, KDS, inventory, GST & compliance, and an autonomous AI manager, in one platform, from ₹999/mo per outlet. Its WhatsApp/QR storefront runs the 0% direct lane natively, every channel's orders land in the same billing and kitchen display, offline-first billing keeps orders flowing when the internet drops mid-rush, and payout reconciliation ties aggregator and seller-app statements alike back to your bills, so adding a channel adds margin, not another dashboard. That is the difference between a channel strategy and a channel mess; see it against your own menu in a demo, and check pricing against one week of recovered commission.

What is the 2026 verdict, format by format?

Cloud kitchens: register this month. Delivery-only economics are maximally exposed to commission, so every order shifted from 25–35% to 3–5% is pure margin, and multi-brand kitchens can list each brand's catalog on the network (see cloud kitchen software for running per-brand costing underneath). At cloud-kitchen margins, even 5–15 daily ONDC orders is real money.

QSR: register this quarter. The document set is one you hold, the effort is days, and QSR menus, standardised, photographed, fast, are exactly what converts on buyer apps. The main discipline is stock-sync so speed does not turn into cancellations; a POS-integrated seller-app setup removes that failure mode.

Casual dine-in: register, but prioritise the direct channel. Your dine-in guests are a captive audience for the 0% lane, every bill and table can carry the QR. ONDC is worth the low-effort listing for incremental delivery, but for this format the repeat-customer channel compounds faster.

The honest verdict: ONDC in 2026 is not a Zomato-killer, and it does not need to be one to deserve a place in your stack. It is a 3–5% commission lane in 600+ cities, growing 15–20% month-on-month in food per ONDC network figures, that costs a restaurant days of effort and almost no money to join. Register through a seller app, keep the catalog honest, expect 5–15 orders a day at first rather than aggregator volume, and keep building the direct ordering channel you actually own, because the cheapest commission on the network is still not 0%.