How Is the Pandemic Affecting Agents? Here’s What Providers Tell Us

Business has not been easy for Pak Bu since the COVID-19 (coronavirus) pandemic struck. The 46-year-old BNI Bank agent and owner of a small grocery store in Jakarta, Indonesia, says that his usual customer base of local residents and food vendors has dried up in recent months. Sales have dropped 75 percent. And the revenue he earns as a digital financial services (DFS) agent has taken a hit. Some customers still come to buy staples with digital government subsidies, but the fees he used to earn on regular transactions like remittances, airtime top-offs and small purchases of cigarettes and other goods have dwindled.

"I am just happy to still be able to eat daily,” he said.

He’s far from alone among small business owners who serve as DFS agents. Over the past few months, CGAP has been asking mobile money operators, digital banks, fintechs and other DFS providers across Asia and Africa how the COVID-19 pandemic is impacting their agent networks. Two overarching challenges stand out for providers as they try to maintain and grow agent networks to serve larger numbers of low-income customers.

COVID-19 is changing DFS transaction types, which could force providers to revisit pricing and fee strategies

An agent helps a customer in Bangladesh
An agent helps a customer in Bangladesh. Photo: Ma-Moni Telecom, Brammonbaria, Bangladesh

According to many of the DFS providers we’ve spoken with, lockdowns have precipitated sharp declines in provider and agent revenue over the past three months – up to 60 percent in some cases. This is true for dedicated agents (those who facilitate only financial services) and nondedicated agents (those who facilitate financial and nonfinancial services) alike. Despite lockdowns being relaxed in some markets, many providers report that the percentage of active agents in their networks has still not returned to precrisis levels.

These declines in revenue come at a time when providers in most regions report an overall increase in the number of DFS transactions, mainly as a result of governments asking providers to waive fees on many of these transactions.

Interestingly, the overall increase in DFS transactions is led by a rise in the number and value of digital bill payments and, for some markets, government-to-person (G2P) payments. In contrast, the number and value of person-to-person (P2P) transactions vary more across countries but show a tendency to decline. We are told these changes are accompanied by a significant decrease in the number and value of cash-out requests within agent networks, while cash-in requests have remained steady.

The apparent increase in high-value, low-frequency transactions like bill payments and G2P transfers, accompanied by a decrease in low-value, high-frequency P2P transactions mostly facilitated by agents (such as cash-in/cash-out for remittances, airtime purchases and merchant payments), seems to be corroborated by other market consultations from MSC/Caribou and the Washington Post.

These changes in the structure of transaction types, due to changes in customer and agent behavior, may be temporary. If they persist over a long period of time, however, fees that DFS providers pay to agents based only on the number, not the value, of the transactions they facilitate most likely will need to be revisited, along with the pricing of services to the end-customer. A more adequate compensation of agents may be necessary to ensure agent network sustainability. DFS providers should conduct rigorous monitoring and analysis of disaggregated DFS transactions, along with customer interviews, to assess how customer and agent transactional behavior may be changing and how long lasting these changes will be.

The financial inclusion community also has a stake in analyzing disaggregated DFS transactions to understand how lower-income segments are being affected. It may be that while wealthier customers are conducting more DFS transactions than ever, lower-income customers – who rely on struggling agents for smaller value and more frequent transactions – are dropping out. If this is the case, financial inclusion policies can focus on ways to reverse the trend.

Agents who offer only financial services are struggling more than others

Another challenge emerging from our conversations with DFS providers is that dedicated agents are having a harder time recovering than nondedicated agents. There are some exceptions in countries where high volumes of COVID-related government subsidies are being disbursed in rural areas via dedicated agents, like in India. However, in other markets we are told that nondedicated agents are improving their activity rates faster than dedicated agents.

The apparent edge for nondedicated agents over dedicated agents appears to be the fact that they offer a wider range of services than dedicated agents. This diversification enables them to reallocate scarce working capital to DFS transactions associated with services less affected by the crisis, reducing their exposure to the hardest hit services.

For example, we hear of shops in India and Indonesia that are stocking up on basic staple foods and e-floats that allow them to honor cash-in/cash-out requests associated with restocking or selling these goods. At the same time, they have reduced e-float associated with the trade of goods and services that have seen demand drop, like remittances, credit transactions and purchases of nonessential goods. There also are reports that nondedicated agents are becoming more mobile as they begin to explore food and goods delivery, while they facilitate related digital payments. Pak Bu, the entrepreneur we spoke with in Jarkarta, had tried this approach.

“I try to sell my stuff online, accepting orders through my phone and take what they buy to their house. But it has not been easy because their income now goes down,” he said.

These changes may be temporary, too, but they reveal insights into what makes agent networks more resilient during a crisis. Agent networks are weathering the COVID-19 pandemic better in countries where agents are able to facilitate DFS payments for a wider range of goods and services. This should motivate providers, policy makers and regulators to explore with greater urgency if and how they can integrate agent networks into the wider digital economy.

Looking ahead: Building more resilient agent networks

Our initial conversations with DFS providers suggest that countries with the most resilient agent networks are those in which these networks were developed as part of a broader strategy to build a robust DFS ecosystem that reaches vast segments of lower-income people. In these strategies, DFS agent networks facilitate an increasing array of valued financial and nonfinancial services. A single agent e-float can be used for diverse products and services supported by a fully interoperable banking system connected to mobile wallets. DFS providers in China, Colombia, India, Indonesia and Kenya are implementing this principle in different ways, as documented in CGAP’s “Agent Networks at the Last Mile” (2019).

Strategies to build more resilient agent networks tend to be supported by G2P programs that aim to reach areas where vulnerable people live. As part of these programs, some policy makers and regulators have collected geopositioning data to identify which types of agents have an edge in different parts of a country or among vulnerable customer segments. Others have developed incentives to expand the range of providers allowed to enter payment markets and manage DFS agent networks, giving G2P beneficiaries more choices of where to collect subsidies. Efforts to expand G2P distribution have been most effective when done in collaboration with private DFS providers.

As customers, agents and DFS providers adjust to survive the current pandemic, they also are revealing which DFS distribution strategies are more resilient. We should take note to prepare for the next crisis.



Most digital financial services users require agents to help them switch between worlds of cash and digital currency. Global evidence suggests six principles for building viable agent networks in rural areas home to poor, financially excluded populations.

Add new comment