A guest post arguing the chicken and egg challenges of payments mechanisms.
Note: This is a guest post by Karthik Shashidhar, a reader of the Platform Thinking blog and a Bangalore-based quant consultant. Karthik is a Resident Quant at the Takshashila Institution and is fascinated by the concept of liquidity, outside the financial markets, and particularly applied to platform markets.
Kenya’s hugely successful payments platform “M-Pesa” was brought to India a couple of years back by Vodafone, in association with ICICI Bank. This allows an “M-Pesa” account holder (who is a customer of both Vodafone and ICICI Bank) to transfer money to another “M-Pesa” account holder. So far, the scheme hasn’t taken off anywhere the way it has in Kenya, where “M-Pesa” transfers reportedly account for over 60% of the country’s GDP.
In a similar effort, Apple, in its latest version of the iPhone, introduced the concept of Apple Pay, which allows users to maintain an “E-wallet” in their iPhone 6, and then use this wallet (which is tied to a credit card at the back end) to make payments to merchants who have machines that can accept payments using Apple Pay.
Recently, the Reserve Bank of India (India’s Central Bank) issued a notification that enforces a second factor of authentication on all credit card payments in the country. This means that cab companies such as Uber cannot operate in India the way they do elsewhere – where your credit card gets automatically charged at the end of a journey. This has led to Uber’s competitors in India, such as Ola Cabs or Meru Cabs, introducing their own “wallets” where people can store money and use it for that particular platform.
None of the three payment mechanisms mentioned above are likely to take off. This is a fallout of the fact that none of them are interoperable, and that constrains both the reach and the incentive to adopt these mechanisms.
(Sangeet’s note: I discussed the issue of backward compatibility a couple of years back in this post here, which complements this current one perfectly.)
Any payment mechanism can be seen as a platform, which needs to be adopted by two sides – the payers and the receivers. Unless there are enough payers using a particular mechanism, there is no incentive for a receiver to invest in that particular mechanism. And vice versa.
(Sangeet’s note: The chicken and egg problem is an issue all platforms struggle with but because of the nature of payment mechanisms as being only value transfer platforms rather than value creation/addition platforms (e.g. YouTube), simultaneous adoption is, even more, important. Value creation/addition platforms can potentially stage the platform on one side first and then move to the other.)
Take Apple Pay, for example. Reports say that this mechanism is going to be exclusive to Apple phones and that too only to the iPhone 6. Apple’s global market share is only about 15%, and even in the US it is around 40%. Even in the most developed markets, the smartphone market share of Apple is less than 30%, and the market share of the iPhone 6 is likely to be much lower than that. Considering that only a small proportion of all payers are likely to be able to use this mechanism, there isn’t adequate incentive for a merchant to invest in accepting money using Apple Pay.
While both Vodafone and ICICI Bank have a reasonable market share in India, the intersection of their market shares (the market that can use “M-Pesa”) is not particularly large. As a result, even if you have “M-Pesa” enabled, there is only a small set of other people with whom you can transact using this mechanism, thanks to the low penetration of the intersection of user bases (Vodafone and ICICI Bank) that is necessary for transacting using this mechanism. Thus, you have no real incentive to make the effort to install and start using the “M-Pesa” platform.
Ola Cabs or Meru Cabs’ closed loop mobile wallets may suffer from the same fate. Why would one want to lock up your money in a wallet that can only be used with one or two vendors? There is no real incentive for users to move their money to such vendor-specific wallets, thanks to which these vendors have started offering massive discounts for moving money to such wallets (Ola, for example, is giving a 100% “cash back” for moving money to its wallet, which is a rather steep incentive).
(Sangeet’s note: One might argue that this worked fine for Starbucks’ payment card, but Starbucks transactions presumably have higher liquidity than booking cabs in Indian cities.)
The key to making a payment mechanism successful is ubiquity, leading to the network effect. Visa, MasterCard, and PayPal are excellent examples of payment mechanisms that have been massive successes as platforms. Visa and MasterCard have achieved it by being ubiquitous and interoperable – most merchants in developed countries and a significant proportion of merchants in other countries accept Visa and MasterCard, which makes it useful for a customer (or a “Payer”) to hold one of the two cards. The issue of interoperability needs to be stressed here – Visa and MasterCard cards and machines are interoperable, which means vendors don’t need to invest in two machines. If this interoperability didn’t exist one of the two providers could have got “Betamaxed” (Betamax was Sony’s videotape format, which was technically superior to VHS, which it competed with, but lost out on account of network effects).
PayPal, on the other hand, became a successful payments platform by being ubiquitous and a virtual monopoly. In the late 1990s, when the mechanism gained traction, there were few other competing mechanisms for moving money on the internet. Thus, before any competitor could react, PayPal had managed to get for itself a large enough market share that gave it the necessary network effects to be successful.
(Sangeet’s note: PayPal had crazy burn rates and took all the risk of fraud upon itself in the initial days to drive adoption of the platform. That coupled with a sign-up fee for every new user and its ability to solve a key pain for eBay users, led to massive adoption.)
So how can a new payment mechanism (such as M-Pesa or Apple Pay) gain traction? There are essentially two ways – one is the PayPal route, where you enter with so big a bang that you quickly have a large chunk of the market, and network effects make it necessary for the rest of the market to adapt to you. Given the plethora of payment options that are present now, it is unlikely that any player will be able to establish this kind of domination without significant investment.
(Sangeet’s note: Sangeet’s note: One should also note that most payment plays today are less about payments and more about using payments as a method to get data and then build new services on that data. e.g. Square. The problem is that transaction margins are razor thin and often negative, and one needs a lot of venture capital to survive that.)
The other option is to make it interoperable. Apple Pay, for example, could introduce an Android app (which might cannibalize on Apple iPhone sales, but increase traction of Apple Pay itself). This could potentially increase the number of devices that can pay using this mechanism, and it thus gives incentive for merchants to install the mechanism that allows them to accept payments using Apple Pay. There is a parallel to this within Apple itself – when the iconic iPod was first introduced, it was only interoperable with Apple computers. After much internal debate, Apple finally introduced iTunes for Windows, and made the iPod interoperable with Windows, in 2003, and that year iPods saw a 235% growth in sales (http://www.forbes.com/2004/01/14/cx_ah_0114aapl.html).
(Sangeet’s note: One does need to bear in mind that the key reason Apple Pay has all those credit cards stored is the app store. Apple Pay is a consequence of that. An Android app will need to have a compelling incentive, apart from just being a new payment mechanism, for users to enter their credit card details.)
What of M-Pesa? The RBI is in the process of introducing new classes of banking licenses, one of which is the “payment bank” – banks that are to be used solely for payments and making deposits and which are not allowed to lend. Mobile phone service providers such as Airtel and Vodafone are expected to apply for these licenses.
Once this license is granted, Vodafone will not need the partnership of ICICI Bank anymore in order to implement M-Pesa. This will automatically increase the network on which M-Pesa can operate significantly. That, however, is not going to be the big increase.
Once payment banking licenses have been issued, and Vodafone and Airtel and other prominent carriers have their own payment mechanisms, it will make sense for them to become interoperable. This interoperability will allow users of “Vodafone Money” (just made up this name) to transfer money to users of “Airtel Money” (and vice versa). And at that point, mobile money transfers may start seeing massive network effects and take off (Vodafone and Airtel can monetize their respective customer bases by levying interoperability charges on cross-network mobile payments, like banks do with ATMs).
(Sangeet’s note: There are, of course, other issues involved with interoperable wallets that are more regulatory in nature and will need to be addressed as well.)
In order to succeed, it is imperative for companies coming up with payment mechanisms to realize that payment mechanisms are inherently platforms, and that interoperability is going to be the key to the success of such mechanisms.
When everyone is building new social products, how do you get users to choose yours over the others?
Why offline network structures are important to online network effects.