He Joined ONDC Expecting Orders. What He Got First Was a Reckoning.

ONDC lets a kirana shop sell through apps it does not own. The listing is the easy part — the hard part is that the network can see whether your stock record is true.

Devendra runs a 700-product grocery in Indore. He signed up to sell through ONDC in March, expecting the hard part to be the paperwork. The paperwork took an afternoon. The hard part arrived three weeks later, in the form of a cancellation rate of 31%.

Nearly a third of the orders he accepted, he could not fulfil. Not because he was disorganised — by the standards of his street he was the organised one. Because for the first time, something other than Devendra was reading his stock numbers and making promises based on them.

What ONDC Is, Without the Acronyms

The Open Network for Digital Commerce is not an app and it is not a marketplace. It is a set of rules that lets a buyer app and a seller app talk to each other even though different companies built them. Practically, for a shopkeeper, it means this: a customer using some other company’s app can buy from your shop, and you do not have to be on that company’s platform to appear there.

That is genuinely different from the platform model, where a marketplace owns the customer, sets the commission, and can delist you on a Tuesday. On the network, you are listed through a seller app of your choosing and your catalogue travels.

All of which is real. None of which helps if your catalogue is wrong.

The Reckoning Is Always the Same

Selling across a counter is forgiving in a way that selling across a network is not. A customer asks for Surf Excel 1kg, you glance at the shelf, it is not there, you offer Ariel. Ninety seconds, no harm, no record.

On a network, your catalogue said you had it. The buyer app promised it. The customer paid. And now you are cancelling — which is recorded, aggregated, and eventually used to decide how often you are shown at all.

The network did not create my stock problem. It just started telling other people about it.

Devendra’s 31% cancellation rate was not an ONDC problem. It was a six-year-old habit of keeping stock levels approximately, in his head, which had never cost him anything measurable until something automated started quoting them.

What You Need Before You List, Not After

Four things, in order. Doing them after you go live means learning each lesson in public, at the cost of your fulfilment rating.

The third point is where most shopkeepers over-reach. Devendra relisted in May with 180 products instead of 700 — his fast movers, his reliable restocks, nothing perishable, nothing he could only get sometimes. Cancellations fell to under 4%. His order volume went up, not down, because the algorithm started trusting him.

A shop listing 150 products it can always supply will out-earn the same shop listing 700 it can usually supply. Fulfilment reliability is the ranking input you control most directly, and it compounds — every cancelled order suppresses the next ten impressions.

The Money Question

Network orders are not free money. Devendra’s average network basket was ₹340 against ₹190 at the counter, which sounds excellent until you account for what the order actually costs to serve.

Run those against your actual margin per product rather than against an average. Devendra found that on staples at 6% margin, a network order was close to break-even; on personal care and packaged snacks at 18% to 22%, it was clearly profitable. So he listed the second group and not the first — which also happened to make his basket sizes look better.

What It Is Actually For

A kirana shop is not going to beat a ten-minute delivery operation on speed, and it does not need to. What the network offers is a way to be found by the customer three streets away who does not know your shop exists, and to serve the regular who wants a repeat order without a phone call.

Devendra’s honest summary after six months: about 14% of his revenue, higher margin than his counter average, and one genuinely useful side effect — because his catalogue now had to be true, he finally knew what he was carrying.

Mapis keeps the catalogue and the stock count in one place and updates them as you bill, which is the piece most shops are missing when they list. The listing is an afternoon; the stock discipline behind it is what decides whether the listing earns anything.

Frequently asked questions

What is ONDC and how does it help a kirana store?

ONDC is an open network that lets buyer apps and seller apps from different companies transact with each other. For a kirana shop it means customers using apps you never signed up for can order from you, at your prices, with lower commissions than a traditional marketplace, and without one company controlling your listing or your customer relationship.

Do I need a GST number to sell on ONDC?

For most packaged goods categories a GST registration is required by the seller app you onboard through, though rules differ by category and by seller app. Check the specific requirement with your chosen seller app before registering, because the onboarding will stop at that step otherwise.

Why do ONDC orders get cancelled by the seller?

Almost always because the listed stock number was wrong. Across a counter you can offer a substitute in seconds, but on a network your catalogue has already made a promise and taken payment. Shops that list only products they can reliably supply, and whose stock updates automatically as they bill, see cancellation rates fall from double digits to low single digits.

Is selling on ONDC profitable for a small shop?

It depends entirely on which products you list. Staples at 5-7% margin barely cover network fees, payment charges, packing and delivery. Packaged snacks, personal care and household goods at 15-22% margin leave real profit. Calculate against your per-product margin, not a shop-wide average, and list only the categories that clear the cost.