Popular on Ex Nihilo Magazine

Global Trends

How India Built the Payment System the World Is Now Copying

Brazil studied it. Singapore linked up with it. The US Federal Reserve built something like it. The IMF called

How India Built the Payment System the World Is Now Copying

Brazil studied it. Singapore linked up with it. The US Federal Reserve built something like it. The IMF called it the world’s largest retail payment system. The World Bank recommends it as a template for developing economies.

India did not set out to build an export. It set out to solve a domestic problem: how do you move money digitally in a country where most people have no credit cards, limited internet, fragmented bank systems, and hundreds of millions without formal financial identity?

The UPI story does not start with a startup or a Silicon Valley idea. It starts with a government committee, a biometric database, and a payments crisis large enough that a half-measure was not going to work. By 2025, UPI processed 228.3 billion transactions worth approximately $3.5 trillion, handled 49% of all real-time payment transactions on earth, and was processing 660 million transactions a day.

The Problem Behind the UPI Story

India in 2012 had a payments problem with several layers.

The obvious layer was cash dependency. Despite a growing economy and a rapidly expanding middle class, the vast majority of Indian transactions happened in cash. Credit card penetration was low. Internet banking was available to a small urban minority. Mobile wallets were fragmented and could not talk to each other. If you had a Paytm wallet and your vendor used MobiKwik, the transaction could not happen.

The less obvious layer was identity. To open a bank account, make a digital transfer, or access most financial services, you needed documentation that hundreds of millions of Indians did not have. No permanent address. No formal employment record. No credit history.

The third layer was infrastructure. Banks ran on different systems with different standards. Interoperability between them was slow, expensive, and unreliable. An NEFT transfer between two banks could take hours. RTGS was restricted to large-value transactions.

Nandan Nilekani, who had built Aadhaar, the biometric identification system that eventually enrolled over a billion Indians, was the person the government turned to in 2012 to chair a committee on financial services for small businesses and low-income households. His committee submitted its report in December 2013. The core recommendation was a single open digital rail that all banks, wallets, and apps could plug into, a common interface owned by nobody and available to everyone.

That report became UPI.

How It Actually Works

Most payment systems work as closed loops. Visa processes Visa transactions. PayPal processes PayPal transactions. WeChat Pay processes WeChat Pay transactions. If you are not in the loop, you cannot participate.

UPI works as open infrastructure. The National Payments Corporation of India, a non-profit set up by the Reserve Bank of India and the Indian Banks Association, built and operates the underlying rails. Any bank, any fintech company, any app can plug into those rails and process payments. PhonePe, Google Pay, Paytm, Amazon Pay, and hundreds of other applications all run on the same UPI infrastructure. A vendor with a QR code does not care which app the customer uses. The money moves the same way.

This interoperability is the design choice that made everything else possible. It removed the network effect advantage that normally lets one payment platform dominate and exclude competitors. It meant adoption could happen across every app simultaneously rather than waiting for one platform to achieve scale.

The identity layer came from Aadhaar. Rather than requiring bank account numbers or IFSC codes, which most Indians did not have memorised and many did not have at all, UPI introduced the Virtual Payment Address. A VPA looks like an email address: name@bank. You share it instead of your bank details. The underlying routing happens invisibly.

The security layer built on this. Every UPI transaction requires a PIN set by the user. The PIN never leaves the device. It is not stored by the app or transmitted to a server. This was a deliberate design choice that removed the fraud vector that had plagued mobile payments in other markets.

Demonetisation Did What the Pilot Could Not

The UPI pilot launched quietly on 11 April 2016 with 21 member banks. Raghuram Rajan, then RBI Governor, did the official launch. For the first three months, not a single transaction went through it. In the fourth month, total value processed was Rs 38 lakh, roughly $45,000. The banks showed little enthusiasm. The fintech industry was not paying attention.

On 8 November 2016, Prime Minister Narendra Modi announced that Rs 500 and Rs 1,000 notes, which together accounted for 86% of all currency in circulation, would be demonetised overnight. Indians woke up the next morning with wallets full of paper that could no longer be spent at most places. Bank queues stretched around city blocks. ATMs ran dry.

The NPCI moved within weeks. On 30 December 2016, it launched BHIM, a reference UPI app built specifically as a proof of concept, simple enough that anyone with a basic smartphone could use it within minutes. The desperation created by demonetisation provided a use case that no marketing campaign could have engineered. Vendors who had never considered digital payments accepted them because they had no alternative. Customers who had always paid cash downloaded apps because cash had become impossible.

UPI transaction volume went from negligible in November 2016 to 1.7 million in December 2016. A year later it was processing 105 million transactions a month. The demonetisation crisis compressed what would have taken five years of consumer behaviour change into approximately eight weeks.

The Numbers at Scale

The growth from that point is difficult to contextualise without specific figures.

Annual transaction volume expanded from 2 crore transactions in FY 2016-17 to 24,162 crore in FY 2025-26, a 12,000-fold increase in nine years. Transaction value over the same period rose from Rs 0.07 lakh crore to Rs 314 lakh crore, a 4,000-fold increase.

UPI now accounts for 83% of India’s total digital payment volume, up from 34% in 2019. The daily average in 2025 was 660 million transactions. In December 2025 alone, UPI processed transactions worth Rs 21.63 trillion. The system handles 3,729 transactions per second, up from 2,348 in 2022, outpacing China’s Alipay, PayPal, and Brazil’s PIX in transaction velocity.

India’s 129.3 billion real-time transactions in the comparison period used by the IMF exceeded Brazil, Thailand, China, and South Korea combined. The gap is not close.

There are now 491 million UPI users and 65 million merchants accepting it. The merchant base includes roadside tea stalls, vegetable vendors, auto-rickshaw drivers, and temple donation boxes. A QR code printout stuck to a chai cart in rural Bihar processes the same UPI transaction as a luxury hotel in Mumbai.

Who Built It and Who Runs It

NPCI is the answer, but it requires some explanation.

The National Payments Corporation of India is a not-for-profit entity set up under the Payment and Settlement Systems Act of 2007. Its shareholders are the major Indian banks. It operates UPI and several other national payment systems including RuPay, India’s domestic card network, and IMPS, the immediate interbank transfer system that predated UPI.

NPCI does not charge transaction fees for UPI payments. This was a deliberate policy to drive adoption. The apps built on top of UPI, PhonePe, Google Pay, Paytm, can charge for premium services but the underlying rail is free. The business model for these apps depends on other revenue: financial services referrals, lending, insurance, merchant services, and advertising.

PhonePe dominates the app layer with around 48% of UPI transaction volume as of early 2025. Google Pay holds approximately 37%. Paytm, which was once the dominant player before a regulatory crisis in 2024 forced it to rebuild its banking partnerships, has fallen to a smaller share. The top two players being a Walmart-owned Indian company and a Google product creates its own interesting dynamic in what was designed as Indian public infrastructure.

The World Is Watching

The UPI story has become a reference point for how developing economies think about financial infrastructure. Singapore’s PayNow and India’s UPI are now interoperable, allowing cross-border real-time transfers between the two countries. Similar linkages have been established with UAE, Mauritius, Nepal, Bhutan, and France.

The US Federal Reserve launched FedNow in 2023 as its own instant payment system after years of watching India demonstrate what was possible. Brazil’s PIX, launched in 2020, was explicitly modelled on UPI’s open-rail architecture and has achieved comparable penetration in a fraction of the time.

The World Bank and IMF have both cited India’s digital public infrastructure model, Aadhaar plus UPI plus a data consent layer, as a template for financial inclusion in developing economies. The idea is simple enough: build the rails publicly, keep them free, and let private players compete on top of them rather than fighting to own exclusive networks.

India did not invent instant payments. Several countries had faster payment systems before UPI launched. What India invented was the combination of open architecture, zero-cost access, biometric identity, and mass-market design that made instant digital payments work for a construction worker in Rajasthan and a tech executive in Bengaluru simultaneously.

What It Has Not Fixed

UPI’s scale is real. The frictions that remain are also real.

Digital fraud has grown alongside adoption. The volume of UPI-related fraud cases reported to banks rose significantly through 2023 and 2024 as the user base expanded into demographics with less digital literacy. The government set up the 1930 cybercrime helpline partly to manage the volume of UPI fraud complaints.

Internet connectivity in rural India remains patchy. UPI requires a smartphone and a data connection. The 491 million users are not uniformly distributed. The wealthier, more urban, better-connected segments of the population account for a disproportionate share of transaction volume.

The zero-fee model creates a sustainability question for the app ecosystem. If no one charges for the rail, the apps built on top of it need to make money elsewhere. That has pushed PhonePe and Google Pay toward financial services cross-selling that may not always align with user interests.

And the concentration at the app layer contradicts the open-rail vision to some degree. A system designed so that no single entity owns the payments infrastructure has ended up with two players, one American and one American-owned Indian company, controlling 85% of the transaction volume.

None of those are fatal problems. But for a system described as the template for global financial inclusion, they are worth naming.

Sources


Ex Nihilo magazine is for entrepreneurs and startups, connecting them with investors and fueling the global entrepreneur movement

About Author

Malvin Simpson

Malvin Christopher Simpson is a Content Specialist at Tokyo Design Studio Australia and contributor to Ex Nihilo Magazine.

Leave a Reply

Your email address will not be published. Required fields are marked *