Bringing digital finance agents to the last mile in Indonesia

At first glance, Indonesians appear to have excellent access to banking agents and cash-in/cash-out (CICO) services. There are over 500,000 bank agents across the country helping customers open accounts and deposit or withdraw their money. Yet over 50 percent of adults in Indonesia do not own a financial account, and one-third cite the distance they have to travel to a bank branch as the reason they are unbanked. Additionally, Indonesians who do own accounts use them infrequently for a very limited number of use cases, mostly to receive government payments and send remittances.

Clearly, there is a need to expand both the quality and reach of CICO agent networks that offer valuable financial services to deepen financial inclusion. But how?


As is the case in many countries, Indonesia’s agent network challenge is concentrated in rural areas, where high operational costs coupled with low transaction volumes can be problematic for digital financial services providers and their agents. According to a sample of agents assessed by MicroSave, 26 percent of Indonesia’s agents operate in such areas. Alarmingly, the same report shows that 26 percent of agents nationwide manage to only break even or operate at a loss.


Global evidence gathered by CGAP and the Bill & Melinda Gates Foundation shows that a variety of emerging CICO agent models are improving the economics of last-mile agents. Our research shows that the financial services providers, policy makers and regulators leading emerging CICO models have generally followed six principles and tailored them to their local contexts.


Below, we offer some reflections on where Indonesia stands against these principles.


Principle 1: Enable rural CICO agents to generate more revenue streams


In many countries, bank- and telecom-led agent network models have struggled with profitability in rural areas due to their costly business practices and heavy reliance on revenue from CICO transactions. Meanwhile, e-commerce companies in countries like China are reaching deeper into rural areas because they enable an array of revenue streams for agents. For example, Alibaba established its Rural Taobao network serving 30,000 villages in just three years, surpassing the rural banking agent reach of some of China’s biggest banks.


Similarly, Indonesia’s largest fintech industry association, Asosiasi FinTech Indonesia (AFTECH), has documented how some e-commerce companies are coming up with their own approaches to service aggregation. AFTECH estimates that there are a whopping 5 million e-commerce or fintech agents in the country who are already facilitating digital financial services. This e-commerce ecosystem is showing greater potential than established agent networks to viably serve rural areas. Specifically, unlike banking agents, e-commerce agents are associating CICO transactions with an ever-growing number of use cases, which can provide more value to customers and generate greater fee revenue for agents.

Photo: Vered Konijnendijk, CGAP

This agent in rural Indonesia uses the same e-commerce platform and e-float to stock her shelves that she uses to sell airtime, conduct bill payments and facilitate money transfers. E-commerce companies are coming up with some of the most innovative approaches to building viable agent networks in remote areas.


However, current regulations prevent e-commerce agents from offering a full suite of financial services. These agents are not allowed to provide cash-out services. This limits the ability of e-commerce agent networks to support use cases that require a cash out, such as remittances.


Additionally, current regulations limit the revenue streams of banking agents who are currently authorized to do CICO by prohibiting them from partnering with multiple financial services providers. Especially in rural areas, this restricts agents’ ability to aggregate revenue-generating transactions from different providers.


Principle 2: Make CICO agents more accessible to rural customers


When customers live close to their CICO agents and personally know them, they are more likely to sign up for and trust digital financial services. In Indonesia, banking agents tend to cluster around bank branches. Agent Network Accelerator’s interviews with dozens of agents across the country in 2017 revealed that almost all operated within 15 minutes of the nearest branch. One of the reasons for this is that being a short walk from a bank branch makes it easier for agents to manage liquidity.

The emerging CICO agent models in e-commerce make it easier to manage liquidity over greater distances by enabling DFS providers to partner with third-party agent or merchant networks that offer a wider range of services. This diversity of services offered can help agents better balance CICO requests.


Principle 3: Expand the range of people who can serve as CICO agents