When you go to buy something on Amazon, the same company facilitates nearly every step in the process. You open the Amazon website or app, search its catalog, select an item sold by an Amazon-designated seller and pay for it using Amazon's payment system. A logistics partner then delivers the product to you, all within a system controlled by Amazon.
The marketplace model binds them together.
India's Open Network for Digital Commerce (ONDC) has an equally ambitious goal, but one that seeks to disrupt Amazon's control of the market. As of July 2026, ONDC has facilitated more than 500 million cumulative transactions. More than two lakh (200,000) retail merchants and over ten lakh (1,000,000) mobility and logistics providers are connected to the network.
But ONDC does not seek to rival Amazon; it wants to make Amazon optional. Think email, not marketplace.
Email Works Because Gmail and Outlook Can Talk to Each Other
Email works because Gmail users can send emails to Outlook users and vice versa. Both services adhere to a common set of standards, or protocols, that allow messages to be sent and received across different platforms.
Now imagine a similar set of protocols for e-commerce. A buyer could search for products on one application while a seller uses another. A separate logistics company could deliver the product, and yet another company could handle the payment. The transaction could still work, so long as the different applications used by the buyer and seller adhere to a common set of standards. This is the fundamental insight behind ONDC, it is an open network, not a marketplace.
Amazon Bundles Everything Together
Marketplaces solve a critical set of problems. Take e-commerce: The buyer needs to find the product, get it delivered, pay for it, leave a review and get customer support. The seller needs to list the product, get discovered, handle logistics and returns, and collect payment. A marketplace bundles together all these services, and in doing so, captures value from both the buyer and the seller. It is easier for the customer to find what they are looking for by limiting their choice of where to buy. And it is easier for the seller to get customers by limiting their choice of where to sell.
But those are costs imposed on the marketplace ecosystem by the bundling of services. ONDC seeks to unbundle them. The marketplace model concentrates power in the hands of the platform, but an open network decentralizes that power.
The Same Seller Can Show Up in Different Storefronts
Take a small-time restaurant or retailer. In the current system, that seller can list their products on a marketplace so that their customers can discover them. But if another marketplace wants to compete, it would have to convince the same seller to list their products on their platform as well. That is a high bar to clear. But on an interoperable network, that same seller could show up on multiple buyer applications, so long as they adhered to the same set of standards. In this sense, ONDC is similar to digital public infrastructure (DPI) initiatives such as the Unified Payments Interface (UPI). By creating a common standard, ONDC and UPI have allowed different buyer and seller applications to co-exist.
The relationship between the buyer and seller is no longer bound to a single marketplace.
The Network Has Gone Beyond Just Shopping
The rise of ONDC has gone beyond just facilitating retail transactions. As of FY2025-26, 218 million transactions have taken place on ONDC across retail, logistics, mobility and financial services.
As of mid-2026, the ONDC network is processing more than 370,000 public-transport trips every day across multiple buyer applications, with most metro ticket sales in India being ONDC-connected. It has also begun facilitating financial services.
The same principle applies to other services as well: unbundling them allows competition to take place at the level of the individual service, not at the level of the bundled product.
Opening Up the Network Creates a New Problem: Who Is Responsible?
A marketplace model has one critical advantage: If something goes wrong, there is always a company to blame. The delivery driver did not show up? You can complain to Amazon. The Uber driver overcharged you? You can complain to Uber. The food you ordered was not spicier than mild? You can complain to the delivery app.
An open network has no single entity that can be held responsible. The delivery service might blame the buyer application, the buyer application might blame the delivery service, the seller might blame the payment gateway, and so on. Interoperability reduces the power of monopolies, but it also creates new challenges when it comes to consumer protection. The network is only as good as the consumer experience it enables, and that experience is ultimately shaped by the companies that comprise the network.
Having Said That, Open Does Not Always Mean Cheap
One common criticism of the marketplace model is that it captures value from both consumers and producers. The same criticism can be levelled at ONDC. Take buyer applications: They have to spend money to acquire customers. Take seller applications: They have to spend money to get listed on the network. Take logistics providers: They have to be paid to deliver goods. Take payment gateways: They have to be compensated for facilitating payments. Advertisements might still show up.
But in a competitive marketplace, those costs might be lower than what a single company would charge for a bundled set of services.
Opening up the network reduces the degree to which any one company can capture value from the ecosystem, but it does not eliminate value capture entirely.
Network Effects Work Against ONDC
Marketplaces have network effects: The more buyers, the more sellers, and the more sellers, the more buyers. That is why Amazon works so well. A large number of buyers means a large number of sellers, which in turn means more choice for the buyers. It is a self-reinforcing loop, and it creates a tremendous amount of value for both buyers and sellers. But it also makes it very difficult for new marketplaces to enter the fray.
ONDC is attempting to overcome this challenge by creating a network of networks. It is not trying to create a new marketplace that competes directly with Amazon, it is trying to create a shared infrastructure upon which different marketplaces can be built.
The network effect works in favour of ONDC, so long as the consumer experience is good enough.
Five Hundred Million Transactions Is Not a Victory
ONDC has had a significant impact on the Indian digital commerce ecosystem. But five hundred million transactions is not a number to crow about, not when it comes to competing with Amazon, Flipkart, Zomato, Swiggy and other large-scale marketplaces. These companies have tremendous brand equity, deep supply chains, strong logistics networks and access to massive amounts of consumer data.
ONDC should not measure its success in terms of how many people are using an ONDC-specific application. Ideally, most people would not even be aware that they are using ONDC. That is how email works, most people are not aware of the protocols that allow Gmail and Outlook to communicate with each other. The early internet was built on a similar philosophy: by standardizing protocols, the internet allowed different websites and applications to interoperate.
ONDC seeks to do the same for commerce. It is an incredibly ambitious goal. Amazon asks the question "How can we build the best possible marketplace?" while ONDC asks the more radical question "What if there does not need to be a marketplace at all?"