This is an excerpt from a recent CGAP paper, The Early Experience with Branchless Banking. The paper synthesizes the observations and research of the CGAP Technology Program. Gautam Ivatury and Ignacio Mas wrote the paper, with substantial input from the entire program team. This blog series will cover seven observations, four uncertainties and four predictions for branchless banking - what we call mobile banking and other technology-enabled banking solutions.
Two thorny problems for bankers and regulators considering branchless banking have been to ensure that customers are not defrauded by agents and that agents have sufficient cash on hand when customers want to make withdrawals. The concern is that customers will mistrust the financial institution and even lose confidence in the banking system if they are victims of fraud or if they cannot get money out of the agent.

But preliminary unpublished ethnographic research in Kenya on M-Pesa suggests that customers will do neither: in several instances, M-Pesa customers continued to use agents for cash withdrawals that earlier had insufficient cash to dispense. Anecdotal evidence suggests that customers’ trust of Safaricom, the entity ultimately holding customers’ funds, is what is leading them to continue using these agents.

Although the evidence on how customers respond to cash shortfalls at agents is limited, by and large customers seem to appreciate there is no guarantee of cash availability. Indeed, the agent’s key role is less about maintaining large cash balances to meet all eventualities, as much as undertaking trips to the bank on behalf of customers when liquidity runs out. Customers will understand that when cash runs out at an agent, all it requires is a trip by the agent to the bank to get more. And now only one person need make that trip rather than each customer of the bank. The open questions are how many trips to the branch will be required, and will agents be paid enough through commissions to make those trips. Also, how can cash be balanced to reduce the time between these trips in places far away from bank branches? In the end, branchless banking through agents may not be a solution for very remote locations until the predominance of cash is replaced by a predominance of electronic payments and transfers.

We are still looking into how much customers save by making branchless banking transactions. But overall, poor and unbanked customers, in particular, have been accustomed to skipping work and traveling hours to open a bank account or make a withdrawal, and receiving altogether abysmal service from many of the formal financial services poor people use. In this context, local banking agents are well-known community members bringing low-cost, hitherto unavailable services to places where no services—utilities, mobile phone coverage, government services—work reliably.