Competition Is an Architecture to Maintain
More payment apps do not automatically create more competition because an app is only one layer of a retail payment. The subject of this series is the competitive architecture linking user-facing interfaces, payment rails, banks, merchants, and public rules. The decisive question is whether a person can choose and switch an app without losing access to counterparties, while providers can still enter and improve. That question gives the reader a practical way to judge a payment ecosystem by its behavior rather than by its number of logos.
Why app counts mislead
An app count answers a narrow question: how many providers offer a visible way to initiate or receive a payment? It does not answer whether those providers are usable substitutes. A substitute must reach the same counterparties, operate with comparable reliability, and remain available on terms that let users and merchants act on their preferences. It must also have a plausible route to scale. If a rival app cannot connect to the people a user needs to pay, its technical presence is not much competition at the moment of purchase.
The distinction follows from the structure of retail payments. The user-facing front end includes the app, wallet, card, or other interface where a payment begins. Beneath it, a retail payment system routes, authorises, and clears the transaction. Wholesale settlement then links payment-service providers and financial institutions. These layers can have different owners, incentives, costs, and concentrations. A market can have many front ends on one dominant card network, or many apps sharing one fast-payment system. Counting the front ends alone hides the relationship that makes each one useful.
The first test is reach. In a closed-loop system, the payer and payee must use the same provider. The provider’s user base is therefore part of its product: more users create more possible counterparties, which attracts still more users. A person may prefer another app’s language, design, or reliability, but choosing it can mean losing the ability to transact. The largest provider can win not because its service is best, but because other people are already there. A rival faces a coordination problem before it can demonstrate its own quality.
Interoperability changes that constraint. When different providers can connect through a common rail, a payer can use one app to reach a payee who uses another. The user’s choice of interface no longer has to match the counterparty’s choice. Network effects do not vanish; they move toward the shared platform, where participation by banks, providers, consumers, and merchants increases the number of possible connections for everyone. This can turn provider-level lock-in into platform-level growth.
That shift has three competitive consequences. First, users can choose by personal fit: trust, language, accessibility, or features. Second, users can reward quality, such as fewer failed transactions, without abandoning their existing payment relationships. Third, an entrant can initially target a subgroup rather than attracting both sides of a payment market at the same time. An app’s early customers can still pay merchants on another provider, giving the entrant a way to build a position before it has its own complete network.
The third consequence matters especially in person-to-merchant payments. Consumers value an app when many merchants accept it; merchants value it when many consumers use it. This two-sided relationship can create a winner-take-most dynamic even when several providers are allowed to operate. Interoperability lowers the coordination barrier by allowing one side to use a new provider while the other side remains elsewhere. It does not eliminate the fixed costs of operating a payment system, the economies of scale in processing, or the need to earn trust and provide reliable service.
The underlying rail can therefore remain concentrated for legitimate structural reasons. Payment systems have high fixed costs and strong network effects among banks, non-banks, consumers, and merchants. Processing at scale can lower average costs. Those features can make a retail payment system tend toward a natural monopoly. Interoperability may make the rail more contestable by providing an alternative, but it does not require several duplicate rails to survive. The competition question is whether the concentrated infrastructure is accessible and whether the front end remains open to meaningful challenge.
This is why visible variety and effective competition must be kept separate. Ten siloed apps can give users less practical choice than three interoperable apps. Conversely, ten interoperable apps can still produce little competition if a device gatekeeper, a dominant merchant interface, or a bundled service makes one of them the unavoidable gateway. The relevant unit is not the logo; it is the user’s ability to obtain a different outcome by choosing a different provider.
The broader evidence supports this caution. The Bank for International Settlements describes rapid digitalisation, new fintech and big-tech entrants, and expanded user choice, while also noting that incumbent banks and card networks retain dominant positions in important markets. In major jurisdictions excluding China, Visa and Mastercard together account for around 95 percent of card transactions in the cited measure. That statistic does not measure all retail payments or UPI, but it shows how scale, acceptance, and network effects can preserve incumbent power while the visible interface changes.
The same logic applies within an open platform. The IMF reports that the largest three UPI apps eventually facilitated more than 95 percent of payer-side transactions. The BIS separately reports that Google Pay and PhonePe account for 80 percent of India’s user-facing front-end market.