[From MicroSave´s Financial Inclusion in Action blog, 5 August 2014]
We have seen in a number of countries how, when they work well, branchless banking and especially mobile money systems can reach millions of people. But beyond the headline numbers on customers reached, the record of such systems as a vehicle for financial inclusion is still mixed: we can hardly talk about a globally-proven solution.
Let me draw some stylized facts from the international experience:
Branchless banking systems have only tended to work at large scale. There does not appear to be an easy, gradual incremental path for providers wishing to deploy branchless banking solutions. There seems to be a chasm between the large numbers of institutions that have run sub-scale pilots and the much smaller set that have succeeded in establishing commercially sustainable branchless banking operations. As a result, there are very few examples of smaller entities –whether banks, mobile operators, microfinance institutions, or other third parties— successfully incorporating branchless banking solutions in a sustainable, impactful way.
The space is still dominated by mobile operators. Few banks in the world seem to have made sizable bets to develop agent networks, and most of those who have built agent networks have tended to see them as an add-on for specific services (e.g. utility bill or credit collections, social welfare payouts) or for specific segments (e.g. poor, rural people) rather than as an extension of their core business. Non-financial companies with a retail or distribution background have been reticent to jump into the space. Therefore, the space has been left largely to mobile operators, who have an easier time conceiving of a transactional, high-volume, low-touch approach.
Customers tend to use branchless banking systems relatively infrequently, and only for a limited range of applications. The median active user is likely to make a transaction only once or twice a month – typically a remote person-to-person or bill payment, and some mobile airtime purchases. It is not common to see branchless banking being a “stepping stone” or “gateway” into the use of a fuller range of financial services. In fact, where mobile money has flourished, it is far more common to see the opposite: fully-banked people adopting mobile money as “liquidity extension” to their banking service.
Branchless banking is not fundamentally reducing people´s reliance on cash. Most mobile money transactions start and end in cash. We may refer to it as a mobile or electronic transaction, but most customers would understand it as a cash-to-cash money transfer, akin to what Western Union has always done. The payment may be electronified, and as a result the distance that cash needs to move is much reduced. But the underlying money is not electronified, since the value is largely held in cash before and after the transaction. Branchless banking systems have generally failed to position the store-of-value function of customer accounts among the previously un- or under-banked, and the result is that the majority of accounts are actually or practically empty.
Branchless banking systems tend to exhibit relatively low levels of service innovation. Branchless banking –and in particular mobile money— systems are about exposing financial service platform functionalities directly to the customer by digital means. But this has not brought on the kind of constant innovation that has been the hallmark of internet business models. Of course, the need to work on basic phones has hampered the ability to innovate, but the fact remains that most branchless banking providers have brought on new services or optimized their user interfaces not more frequently than annually, if at all.
There are of course counterexamples to each point, but they are few. Zoona in Zambia is a small, independent organization growing a purely mobile-based money system incrementally by exploiting specific niche opportunities. The much larger bKash in Bangladesh operates largely as an independent entity, even though it is backed by BRAC Bank that is part of one of the most influential organizations in the country. Equity Bank in Kenya is making a big push into the mobile space with its acquisition of a mobile virtual network operator (MVNO) license.
The above factors are all inter-related, like distinct symptoms of a broader malaise. The pattern of starting and ending in cash most transactions in cash raises costs and presents a brutal business challenge of having to ensure sufficient density of liquid agents in each locality served. Higher transaction costs make the system less compelling for lower customer-value-adding transactions, such as savings or face-to-face merchant payments, which on the other hand, offer the highest potential pool of transactions. In the face of low usage levels per customer and the inherent network effects of payment businesses, the economics can only work for those able to aggregate the largest number of customers, and in particular mobile operators with a mass-market transactional business model. Other big players such as banks may not see a positive business case, or if they do, may fear that the new branchless banking activity may cannibalize their core business or be margin dilutive. As a result, few players in each market enter the business, and when they do they tend to underinvest in IT platforms, staffing and marketing spend. With such shoestring resources, they become easily overwhelmed by day-to-day operational issues and do not devote much attention to the service roadmap. With lack of effective competition, innovation falters.Let´s not concede that branchless banking must push the unbanked into the arms of the larger banks and telcos in the country. Now that we have a good decade of experience with mobile financial services, it behooves us to look back on the trajectory and see what course-corrections can be made to spur more competition and innovation for the benefit of the world´s poor. This should start with regulation, which needs to shift from being merely enabling to being pro-competitive, as I argue in this paper.