By Sidharth Wagle

Abstract

The Government of India (GoI) is planning to implement the Open Network for Digital Commerce (ONDC), a hypothesized protocol to unbundle the current platform-centric model for digital commerce and allow for a freer flow of value within the ecosystem. This article explains the protocol and analyzes its potential impacts on Indian digital commerce. 

Introduction

The platform model and its discontents

Under the current model of digital commerce, a prospective buyer has to register at some platform to purchase a product, say Amazon.com, search for his desired product, and buy from one of the sellers who are also present on Amazon.com. However, this approach has a few main flaws, some of which have been outlined in the ONDC Strategy Paper published in January of 2022. 

Competition between platforms occurs mainly on the features of delivery speed, security, and reach. Such competition on the basis of the platform’s delivery network shapes a market where suppliers face significant economies of scale. This is because, as a platform’s delivery network expands, it becomes easier to reach remote areas, leading to greater revenues, which can further fund increased expansion, creating a flywheel effect. As a result of these leviathan economies, new entrants have virtually no chance of competing with the delivery network of an established platform. Consequently, once a platform scales up, it faces little competition. This inevitably leads to market inefficiencies, since these platforms are no longer incentivized to improve their services. In addition to miscellaneous inefficiencies, there exists another source of welfare loss on the seller’s side caused by this barrier to entry.  

Before exploring this welfare loss, however, it is important to understand the role of trust in a consumer transaction. Trust is an important intangible asset for all sellers. Most transactions are conducted on a common understanding of trust and the improbability of fraud. In a normal transaction, then, as the buyer conducts repeated purchases from a seller, over time, as the seller proves their trustworthiness in delivery or quality, the consumer begins trusting the seller, leading to an increased number of future purchases. 

However, under a platform-centric model, a consumer associates their transaction with the platform and not the seller, since the visibility of the seller is negligible when the consumer is making the purchase. Instead of the seller, this trust accumulates to the platform. As such, sellers are unable to build a significant relationship with the consumer, forcing them to effectively “borrow” the credentials of the platform and conduct the transaction under its aegis. Notably, this locks the seller into the ecosystem of that particular platform and renders them unable to leave, because if they were to leave, they would lose consumer trust and essentially have to start over completely. This “lock-in” effect gives the platform significant and often arbitrary power in dictating the status of the sellers and their products. 

For example, Amazon has been known to copy well-performing third-party products under their in-house brands and manipulate search results to ensure their copy is more visible to consumers than the original product. This tactic aggressively diminishes the original sellers’ revenues. Moreover, as explained above, the sellers have little recourse than simply accepting the situation, since they are locked into the Amazon ecosystem. 

Hence, under the platform model, platforms are given huge amounts of discretionary power, which they can and do abuse. 

Another significant problem with the platform-centric model is that all sellers must first comply with the terms and conditions of the platform they are using. This compliance (including but not limited to photography of goods, English descriptions of the product, and Amazon verification) can be expensive, an undeniable obstacle to small, “hyperlocal” stores like kiranas and other roadside shops.

ONDC’s proposed solution

The ONDC proposes to eliminate these problems through the implementation of a decentralized, federated protocol through which users on different “buyer apps” and sellers on different “seller apps” can all find each other, regardless of the particular app they are using. The technical details will not be covered in this article, but the protocol works as shown in the following figure. 

Figure 1– ONDC representation

The buyer platform nodes forward the search request of the buyer to the routing “layer”, which will multicast (or forward to all seller nodes) that particular request, along with certain metadata like geolocation, price range, and various other preferences. The seller nodes have access to a register of sellers, their goods, locations, and prices, based on the sellers in each seller app. The seller nodes filter through this register to find a matching listing, and then connect the buyer and seller, allowing them to complete the transaction. 

Under this model, there is no singular app or platform in which discretionary power can be concentrated, instead many apps which can specialize and compete in different sections of the protocol. Additionally, Micro, Small, and Medium Enterprises (MSMEs, including kiranas) can be onboarded into the digital commerce space through the use of a single seller app on the owner’s mobile phone. This will boost their revenues and reduce consumer delivery times, since this kind of decentralized network of sellers already penetrates the majority of India’s urban space, without the need for additional investment into large-scale delivery infrastructure. Finally, an official online sales channel could legitimize much of the informal economy, allowing for greater job security and more accurate data collection for future policies. 

Although the ONDC, therefore, solves the major problems with India’s digital commerce space, its deficiencies must be accounted for before implementation. 

Deficiencies and Solutions

Privacy

Naturally, a protocol that links buyers to sellers will produce copious amounts of customer data as a by-product, raising the question of data privacy under the new protocol. According to the previously-mentioned Strategy Paper, “transaction data will reside only with the buyer and seller applications” and the ONDC will allow “platforms to emit anonymized performance metrics”. Although the paper claims that the network will protect any sensitive information, multiple studies have shown that the so-called “anonymized” data can easily be de-anonymized and associated with the correct user, calling into question the actual protection afforded to consumer data. Additionally, users will no longer be able to restrict their data-sharing to one platform. They could now have to, inadvertently, through the federated nature of the protocol, share data with the entire network and all apps on it. This represents a massive breach of privacy, especially for unsuspecting consumers assured of “anonymisation” of their data. 

Private ownership

The non-profit company registered to head the ONDC initiative is “majority-owned by private sector institutions”. Although motivated to ensure alignment with markets and avoid their impairment, one must ask whether a majorly private-owned company will indeed develop the protocol while keeping the interests of the consumer in mind as opposed to a pure profit motive. Additionally, the ownership by capital-heavy firms also creates a dearth of representation of the hyperlocal kirana stores the ONDC claims to help. Adequate consultation and representation with all stakeholders while keeping in mind the good of the market must be a necessary feature of the protocol’s development. 

Trust

Finally, while the ONDC in some measure solves the concentration of trust problem of the platform model, the lack of any singular “accumulator” of trust raises many questions about accountability within the network. For example, a federated protocol could permit abuses of product return policies, which a consumer could find difficult to navigate, given the decentralized nature of the transaction. Another potential area of weakness is the payment process. In a platform model, the user can trust that the platform will enforce prompt delivery upon payment or force the seller to return the money. Under the ONDC, however, the legislation of payment disputes could be a problem since different sections of the protocol, such as delivery, seller-side applications, and buyer-side applications, could use varied payment interfaces and enforcement policies, causing confusion and limiting interoperability.

Conclusion

The proposed Open Network for Digital Commerce, despite its flaws, is a refreshingly modern and novel solution to the problems plaguing platform-centric commerce, not just in India, but all over the world. Its federated nature allows the redistribution of bargaining power from conglomerate platforms to MSMEs while maintaining flexibility due to the minimal oversight needed to regulate it. Additionally, the onboarding of informal businesses into regulated online marketplaces promises a network of ripple effects, from increased tax revenue to greater compliance with other government regulations, further leading to increased standardization and choice for both consumers and businesses. 

Nevertheless, despite its obvious advantages over conventional public policy, it remains to be seen whether the ONDC will be able to effectively deal with the issues raised previously and rise above these drawbacks to become the new standard of digital commerce. 

About the Author
Sidharth Wagle is a first-year undergraduate student at Ashoka University, and his interests include economics, technology, and public policy.