What’s ailing UPI and how to fix it10 min read . Updated: 07 Dec 2016, 05:29 AM IST
Since multisided platforms run into the chicken-and-egg problem from the get-go, they have to adopt different strategies to attract different groups
Conspicuous by their absence—an ubiquitous phrase that is attributed to the ancient Roman writer Tacitus, who coined it in reference to the relatives of the Roman figure Junia Tertius missing her funeral. He could well have been describing the Unified Payments Interface (UPI) in the aftermath of the momentous demonetization announcement by Prime Minister Narendra Modi on the evening of 8 November.
Piloted in April and launched in August, the brilliant UPI was meant to be the answer to all of India’s small payment travails. The National Payments Corporation of India (NPCI) had been working on the platform for a while, with the promise of a simple mobile app that makes both peer-to-peer and merchant payments easier.
UPI is designed to work 24x7—a user can have a single mobile number and app linked to multiple bank accounts, creating a virtual payment address (VPA) to help send and receive money. A UPI user can link a bank account to any app—even that of a different bank, or of a third party.
There are facilities to schedule payments, make payment requests and set up bill payments. And a user can do all this without sharing any credentials—not even a phone number. The VPA identifies the user and gives away no other information to the counterparty.
The UPI system has been live for a quarter now, so it is natural that as the government pushes for a “less cash" economy after demonetization, there should have been a push for greater UPI adoption.
Instead, we see an advertising blitz by banks and third parties talking about their own wallets and e-payment services—the very services that UPI was supposed to kill.
Yet, Paytm, the leading Indian digital wallet, reported a jump from 115 million users to 150 million users in the first week after demonetization.
So, what’s not working for UPI?
The answer to that question may lie in internalizing the essence of the 2016 book Matchmakers: The New Economics of Multisided Platforms by David Evans and Richard Schmalensee.
The NPCI has so far focused on creating a wonderful platform addressing a big area of friction, viz. small payments, and enrolling banks and third parties into using the platform. But it stops there, or rather, further efforts continue in doing more of the same.
Evans and Schmalensee in their book explain the concept of multisided businesses—platforms where there are no linear value chains, but rather several groups of distinct users who interact on the platform. UPI is a classic example of a multisided platform but it not being run like one.
Multisided businesses aren’t new—newspapers, classified advertisements and credit cards are all examples of offline multisided business, where the business had to enrol different types of users and match their requirements.
The advent of the Internet morphed these businesses into online platforms with certain signature characteristics: low or no inventory, low employee base, low resource, high scalability.
Multisided platforms involve different user groups interacting with each other in a bid to substantially reduce the high interaction friction outside the platform. They bring together different types of customers, but each side is to be treated like a customer and not a user.
Multisided platforms benefit from both the direct network effects as well as indirect network effects. Direct network effects mean that the greater the number of people connected to a network, the higher the value of the network to each person who is part of it.
However, what differentiates multisided platforms from traditional businesses is the presence (or absence) of indirect network effects—the value of the platform for a group of customers depends on how many members of other customer groups participate. UPI has benefited neither from the direct nor from the indirect network effects thus far.
The NPCI approach to UPI has been to just leave the platform adoption and scaling to chance. Yes, the banks are being encouraged to participate in the platform, and various app providers—banks or third parties—have launched their offerings on the Google Play Store. But UPI does not treat the various sides of the platform—the users sending money, the users receiving money, the app developers and the banks—as distinct entities to be enticed and treated differently.
The approach has been to put out the platform and then let each side of the platform figure out for itself what to do with it. Also, in general, the NPCI has left it to the banks to educate their respective customers on UPI and let them drive customer enrolment.
The two distinct roles of sending and receiving—where the same person or business can perform a different role in a different transaction—have also not been viewed differently by the NPCI. And banks have no incentives to push UPI other than their moral commitment to the NPCI.
When transactions originate from the UPI platform, banks don’t get fees or commissions on their products like debit and credit cards. Users don’t log in to their Internet banking platforms. And they lose track of transactions—valuable data on customer spending habits, which helps them cross-sell and upsell their products.
So, while banks are going live on UPI, it is not a high-priority project for them. And when they do go live, they are conveniently burying the UPI functionality inside their larger wallet or Internet banking apps.
In such cases, those interested in UPI may not be able to easily find out how to use the functionality. Banks are forcing “multi-homing" on their app users, not promoting UPI exclusively and leaving open the option of switching to an in-house platform.
Hence, a quarter after the product launch, UPI has failed to create a thick market. There just aren’t enough participants on each side with whom other participants may want to interact.
In fact, UPI is struggling right now even for direct network effects—if enough people don’t use it, there is little incentive for more people to use it. Much like a social network. The indirect network effects aren’t even in play, because the NPCI isn’t talking to different sides differently.
UPI means different things to different participants. And therefore it needs custom messaging for different user groups. This necessitates, first of all, identifying these user groups.
One quick win could be identifying low-margin, high-volume trading businesses that dot the business landscape of India—ideal for UPI adoption.
Small businesses have to pay anywhere from 1.5% to 4% of transaction value for participating in other digital payments platforms. For trading margins which tend to be in the range of 5% to 8%, such costs are unsustainable.
With cheques considered unreliable, these businesses naturally gravitate to cash payments. Such businesses could have easily been an initial focus for UPI propagation and uptake.
Since multisided platforms run into the chicken-and-egg problem right at the inception, they have to adopt different strategies to attract different customer types. The platforms have to incentivize each customer type differently, and this can include different pricing strategies.
It is quite common for such platforms to lose money on one or a few sides, so as to create a thick market overall, and make their money on other sides. UPI unfortunately has not tried driving this critical mass on any one side, relying solely on organic adoption.
The NPCI could have run campaigns to enrol traders or small merchants or rickshaw drivers or cash-and-carry buyers or agriculture produce market committees—India is full of payment friction.
Many of these customers would have potentially adopted a low ongoing transaction cost system for an initial upfront fixed fee—it was a question of trying out different pricing and customer acquisition strategies.
The NPCI could have also incentivized banks to integrate their cash management solutions for small businesses with UPI—this could have helped get banks interested in pushing it as a fee business even as they lose commissions on payment transactions.
On a multisided platform, demand from one group of customers does not depend only on the price the group is paying for the access. It also depends on the demand from other customer groups with whom they have to interact.
While UPI remains a uniformly low-cost solution—effectively free as of now—it does not automatically attract new customer groups on the back of this pricing.
Traditionally, multisided platforms have always invested in self-supply of content or activity in other business areas, when they aren’t attracting enough customer groups. Given that UPI does not provide any specific incentive to banks, this option is more or less ruled out—banks and financial institutions were the best placed in the larger UPI ecosystem to get the initial set of transactions going.
Successful multisided platforms also invest in governance. Even when the larger ecosystem is open, the platform imposes basic rules of engagement for the participants. UPI lags on this count.
While any member of the UPI platform can launch an app, the currently available apps are very different from each other. Not all apps cover all the functionalities that UPI offers. Users have reported transactions failing between some specific app pairs.
Some apps show counterparty name and account number in full on pre- or post-transaction confirmation screens—using the VPA was supposed to mask such information.
These teething troubles don’t help the UPI cause—the guidelines on how apps should be designed and information presented should have been more stringent.
Multisided platforms almost always run into trouble if they do not create the right set of laws and rules, and impose penalties on those who break them.
When these rules work well, it is possible to generate direct as well as indirect network externalities as per Evans and Schmalensee.
When these rules do not work well, the platform can easily stagnate or collapse under the weight of user abuse. The NPCI should go deeper into some of these initial problems and solve them before they deter the initial set of users from repeat transactions.
If UPI has to perform a big role in the post-demonetization world of payments in India, it needs to immediately up its marketing and management ante. The NPCI should specifically focus on these five things to ignite interest in the platform.
First, the NPCI should take an ecosystem view and find ways to compensate banks for loss of control, commissions and data. UPI has to be at the heart of various transaction banking offerings and the NPCI should facilitate this process.
Second, a few large customer groups need to be identified and specifically onboarded to the UPI platform. This is standard customer acquisition work, but the NPCI may have to define industry-specific use cases to drive adoption and usage.
Third, there may be a need to tailor pricing strategies for various customer groups. Rather than saying that every transaction is free, the NPCI could look at charging a participation fee from specific customer groups. This will be feasible if the overall transaction costs for these groups reduce despite paying an UPI access fee.
Fourth, UPI apps should be more controlled and standardized. The process of certifying these apps by the NPCI against 100% feature compliance before they go up on Google Play Store should be made more stringent. There should be a central mechanism of recording any deviations and such instances should attract penal intervention from the NPCI.
Fifth, the NPCI should move from being a platform provider to a platform administrator. It should be an active participant in running a business, not just creating the initial infrastructure for others to drive transactions.
Of course none of this is possible without the central government and the Reserve Bank of India actively supporting such behaviour on the part of the NPCI.
Beyond just regulatory and financial support, the NPCI should be allowed to seek the services of professionals who have run multisided platforms and understand their complexity and unique characteristics.
Tacitus once remarked that the desire for safety stands against every great and noble enterprise. If the central government wants UPI to be a serious player in the inevitable “less cash" future of India, it should empower the NPCI to take enabling decisions. Else, banks and wallet providers will continue to run roughshod over a brilliant technology idea.
Aashish Chandorkar is a management consultant with an interest in how digital technology is changing businesses.
Comments are welcome at firstname.lastname@example.org