Mobile money: Getting to scale in emerging markets

| Article

More than a billion people in emerging and developing markets have cell phones but no bank accounts. Many low-income people store and transfer money using informal networks, but these have high transaction costs and are prone to theft. Mobile money is beginning to fill this gap by offering financial services over mobile phones, from simple person-to-person transfers to more complex banking services. To date, there have been more than 100 mobile-money deployments in emerging markets; at least 84 of them originated in the past three years.1

Only a handful of these deployments have reached a sustainable scale; some notable examples include M-Pesa in Kenya, MTN Uganda, Vodacom Tanzania, FNB in South Africa, and GCASH and Smart Money in the Philippines. Even these players have not gained much traction for financial services beyond simple transfers and payments. We sought to find out what drives on-the-ground success and to develop a preliminary set of prioritized, actionable recommendations. We interviewed and conducted workshops with more than 40 leading mobile-money providers (primarily mobile-network operators and banks) and industry experts, which we supplemented with a survey of about a dozen providers. Among the experts we consulted was Michael Joseph, the former CEO of Safaricom in Kenya and “father” of the M-Pesa money-transfer service, which has inspired many recent deployments around the world (see sidebar “Getting people to send money home: Six questions for Michael Joseph”). We have also begun compiling a benchmarking database that includes performance data for about 20 percent of existing deployments (see sidebar “Some preliminary mobile-money benchmarks”).

Our research revealed the three most critical success factors to implement after a provider has launched its deployment and starts to face execution problems: (1) pay close attention to managing the agent network; (2) create a compelling product offering; and (3) maintain corporate commitment (Exhibit 1). As the exhibit shows, we identified four additional success factors, but they are more important when deciding whether to enter a market and design a new deployment than they are during the immediate postlaunch phase.2 We are now in the process of testing these recommendations with providers. We hope to spur both debate and, more important, real breakthroughs in achieving success at scale in mobile money.

Exhibit 1
Seven success factors make mobile money work.
MoSociety_Mobile Money_ex1

Managing the agent network

Our findings suggest that managing the agent network is the most critical postlaunch success factor. Agents conduct the cash-in and cash-out functions, enabling customers to convert cash into electronic money and back again in convenient locations; in the eyes of the customer, the agent is the face of the company. This means the agent can either build or destroy trust and credibility.

Many providers focus on building their agent networks as fast as possible, without careful attention to the agents’ business case and profitability. According to our research, this is a mistake. We propose three key tenets in managing an agent network: (1) grow the customer base and the network in tandem; (2) understand agent economics and risk—the business case for agents is not that simple; and (3) only enroll agents who have the right skills and dedication, and be prepared to train and retrain.

Grow the network with the customer base

Getting the agent network rollout right is one of the most complicated aspects of launching mobile money. If a provider enlists too few agents, customers perceive the system as difficult to use, or even useless. On the other hand, if there are too many agents, many of them cannot generate enough business to cover the cost of managing liquidity. As a result, they stop maintaining their electronic money float and cash balances. This is bad for the agents and damages the system’s credibility with customers. It is best not to let such an imbalance continue. For example, when one African provider realized that it had too many agents relative to transaction volume, it cut its agent base in half.

To ensure ease of use for customers and the appropriate level of business for agents, a provider must take a considered approach to network growth. As an example, one of the keys to Safaricom’s continued success has been its decision to match network growth to customer-base growth, ensuring a steady 1,000 transactions per agent per month (Exhibit 2).

Exhibit 2
M-Pesa’s customer-base growth drove agent-network growth.
mobile money

The practical implication is that the agent network can start out smaller than commonly expected. The initial network will likely number in the hundreds, not thousands, and it does not have to cover the entire country. Safaricom launched M-Pesa with just 400 agents in a country of almost 37 million people. For larger countries, some experts urge a regional launch, accompanied by later rollouts.

Develop a deep understanding of the agent business case

This advice may sound simple, but many providers have failed to ensure that their agents have a manageable path to profitability. Experts note that it takes a year or more for agents to see profits from mobile money. The Gates Foundation maintains that agents must process around 30 to 50 transactions per day for their business to be viable. Understanding the business case for agents means analyzing their day-to-day operations and developing solutions to their biggest pain points, as well as understanding the trade-offs they face.

For example, managing liquidity is the largest cost for agents, representing one-third of their total costs. In this context, there are two sides to liquidity: managing physical cash on hand and managing the amount of e-money the agent has bought as “e-float.” Agents must invest significant amounts of money up front in e-float and other costs, which in many countries can be as much as three times per capita GDP. In the early days of a deployment, there are too few transactions to offset these costs. Even when transaction volume increases, agents often see their revenue only once a month, when commissions are paid. Another cost of liquidity management is the risk of robbery for agents holding large amounts of cash for payouts, which has been a problem for some Brazilian banking correspondents.

Liquidity is also an agent’s most vexing logistical challenge. Most agents struggle to maintain enough cash or e-money to meet customer demand on peak days, like government paydays. And even on normal days, most M-Pesa agents must typically plan to send someone to the nearest “superagent” (that is, an intermediary agent—larger than others in the network—that helps the provider manage parts of the network and ensures that smaller agents can conveniently manage liquidity) or bank branch to rebalance cash and e-money.

Some of these timing issues have relatively simple solutions. For example, paying agents to sign up customers can help the agent earn some income before transaction volumes really kick in. However, to ensure this does not become an incentive to sign up a large number of dormant customers, these commissions should be staggered, with parts becoming due, for example, at initial registration, first deposit, first financial transaction, and continued use. Providers should also consider up-front loans to high-potential agents. The key is to understand the full business case and design incentives, loans, and other financial aspects of the relationship appropriately.

Invest in maintaining agent quality

Agent quality is critical for maintaining customers’ faith in the system. The best agents maintain liquidity consistently, market mobile money effectively, educate customers in mobile-money use, and remember their training instructions. Such agents are an essential asset in mobile money, but acquiring them requires careful recruitment, training, and monitoring, with a particular emphasis on protecting customers from fraud.

Mobile-network operators (MNOs) interested in launching mobile money have often tried to convert their prepaid airtime distribution network to a mobile-money network, but this strategy has not always worked. Airtime dealers are rightfully wary of cannibalization of their business by mobile money—once users adopt mobile money, what is to stop them from buying airtime over their handsets? This could kill the agents’ airtime sales. This discrepancy in incentives between airtime sales and mobile money has been a major barrier to mobile-money growth for an MNO in Southeast Asia.

And the switch to becoming a mobile-money agent does not initially appear lucrative, since commissions on mobile money are typically lower than commissions on airtime. Therefore, providers seeking to build a network may need to be choosy with respect to existing airtime sellers—recruiting only high-potential early adopters—and then branch out to different sources of high-quality talent, such as retail chains, or aggregators, such as superagents.

In addition to recruiting effectively, providers must offer agents ongoing training services, especially at launch and when introducing new features. Because airtime dealers have high employee turnover, retraining agents by visiting each shop up to twice a year may be necessary. Safaricom has outsourced this successfully to the independent contractor Top Image. Other providers, including EKO in India, have managed training well in-house, choosing to outsource only the more day-to-day functions of agent management (liquidity balancing, for example).3 Consistency, continuous follow-up, and in-shop training are all important. Experts told us that agents prefer not to send their employees off-site for training and that employees who receive off-site training typically do not retain what they have learned as well as those trained on-site.

Monitoring agents for fraud is also critical. Many customers do not understand how to use the technology and will hand their phones to agents to perform the tasks involved in a transaction for them. Dishonest agents can easily send customers’ money to themselves or to friends. Safaricom took a strong stand on this kind of behavior, firing 80 of its first 200 or so agents and taking away their airtime business. This sent a clear signal that Safaricom would not tolerate theft, which the company wisely coupled with generous customer compensation for all fraudulent transactions.

Creating a compelling product offering

Most experts consider a product (or suite of products) that generates two to two and a half transactions per month per customer to be a commercial success. So, which products are most likely to achieve this level of transaction volume?

Recent mobile-money deployments have generally taken one of two approaches to product design. The first has been simply to copy M-Pesa’s successful product in Kenya, focusing on person-to-person (P2P) transfers and echoing M-Pesa’s value proposition: “Send money home.” The second has been to launch a complete financial- product line with the hope that one of the products or the package as a whole will generate enough demand to achieve sustainability. Neither of these strategies has met with great success.

A more promising approach would be to focus first on identifying a “killer app” that would drive a high volume of transactions onto the platform, and then promoting this product through both above-the-line (ATL) marketing (which includes mass-market advertising channels such as television, radio, newspapers, magazines, and outdoor advertising) and below-the-line (BTL) marketing (which includes nonmedia marketing techniques such as direct mail or face-to-face sales). We believe that the most successful deployments will expand into other products only when a killer app gains traction.

Choose a narrow killer-app product to go to market

The first step in designing a killer app is to conduct a detailed customer-needs assessment that maps the current behavior and pain points of the provider’s most important customer segments, and then implement a service to address the needs identified. M-Pesa did this effectively when it offered Kenyans something that addressed a real pain point: a simple, safe, low-cost way to transfer money. In other countries, customers may have a completely different pain point. “Send money home” is not a magic formula that will resonate universally. Some providers have instead focused on utilities payments for their core urban-based customers, as alternatives pose significant cost and time challenges. Other possible killer apps could be point-of-sale payments, business-to-business transfers, mobile bill pay, mobile payroll, international remittances, agricultural insurance, savings, or microloans.

Promote the killer app through significant and ongoing marketing

Experts frequently cite poor marketing as one of the key barriers to adoption. Our research validates this observation. For example, a customer-acquisition funnel for the Philippines reveals ample room for improvement at each stage of the funnel. Our research also suggests that the most successful mobile-money providers have combined both ATL and BTL marketing to drive registration and transactions.

For mobile money to gain traction, people must be aware of the deployment and its value proposition (Exhibit 3). ATL marketing that utilizes a variety of channels can quickly and effectively increase widespread general awareness. ATL marketing messages, however, must be careful to balance aspirational messages that depict higher-income people with the sense that the product is for “average people.”

Exhibit 3
Considerable marketing efforts are needed to build a base of frequent users.
MoSociety_Mobile Money_ex3

One expert told us that simply removing the gold jewelry from the models in marketing images had a major, positive impact on customer acquisition. When the provider used models wearing gold jewelry, most people assumed that mobile money wasn’t for “people like them.”

BTL, face-to-face interaction is critical to drive trust. Potential customers must receive quite a bit of education about how the new service works and what it could mean for them before they will trust the system. Individual agents and marketing representatives will be most effective in building this customer trust, especially early in the deployment. Experts estimate that most customers need 10 to 15 minutes of face time with an agent or other mobile-money representative in order to feel comfortable using the product.

All marketing activities should drive customer adoption and usage. A common mistake is to focus marketing efforts on customer acquisition at the expense of use. Messaging should communicate the benefit of frequent use and educate customers about the variety of situations in which using mobile money could benefit them.

Our research suggests that companies may need to revise their expectations regarding the level of marketing investment needed. For example, once Vodacom refined its marketing campaign in Tanzania so that it would resonate with customers, the company wound up spending about twice the amount per customer that its affiliate Safaricom had spent in Kenya—primarily because there were fewer potential customers in Tanzania and they were harder to reach.

Wait to achieve minimum scale with initial product before diversifying

After attracting a loyal customer base, the most successful mobile-money deployments will diversify the product offering to attract new customers and/or increase the transaction volume per customer. When M-Pesa started to catch on in Kenya, grocers and other merchants began accepting M-Pesa payments. In 2010, Safaricom began offering crop-insurance products to M-Pesa users. A similar path of diversification will likely characterize other successful deployments, as diversification both draws new customers and allows deployments to capture more of the spend of existing customers. Diversifying too quickly or starting with a diverse platform of products can confuse customers and dilute messaging.

Set yourself up for future success

Early in the deployment, the provider should have a vision of what it hopes to be known for. This vision should drive the ultimate suite of products the provider will offer (Exhibit 4). For instance, will this eventually be a banking platform, or will it be a trading and payments platform? Based on successful models, it is clear that becoming a payments platform is a promising strategy in many cases. The sheer number of transactions in the business-to-business or government-to-people spaces that still occur in cash indicate that some effective business models have yet to be built. For instance, in a recent payments map of India, we estimated a total current volume of $9 billion in person-to-person transactions versus $6.8 trillion in business-to-business transactions.

Exhibit 4
A company’s aspirations should drive its product offerings.
MoSociety_Mobile Money_ex4

Sustaining a strong corporate commitment

Our research suggests that even successful deployments require three to five years to reach profitability, which can strain corporate commitment. An enthusiastic, committed corporate culture that is willing to make significant up-front investments and patiently wait for impact at scale is crucial to mobile-money success.

Mobile money has great potential to be profitable for MNOs in the long term because of the combination of direct fee revenues and indirect benefits (churn reduction and lower airtime-sales-channel costs being the most significant). Our conservative estimates indicate that both MNOs and banks can generate significantly more profits once mobile money reaches scale. From what we have seen so far, though, mobile money tends to become profitable only after it begins to “go viral” and enjoy the benefits of network effects. Because new technologies and people’s money are involved, it takes time for people to become comfortable and trust the system. Therefore, strong corporate commitment and faith in mobile money’s future profitability is crucial for success in the launch and scale-up phases.

Companies often have to commit more funds than expected to reach scale. Most successful mobile-money launches have had the backing of millions of dollars. Safaricom poured over $30 million into M-Pesa. Other successful deployments received similar levels of investment: Vodacom spent $25 million, and MTN Uganda put up $10 million in initial investment. Numerous other deployments in the last few years have attempted to get off the ground with less than $1 million, only to find themselves unable to gain traction. Mobile banking is not simple; it requires a thorough, careful understanding of customers and agents, and a willingness to invest in the buy-in of both these groups. If companies are not willing to invest and stay committed, the deployments will not reach scale.

The potential for mobile money remains promising, and the next generation of success stories will help shape the innovations that enable hundreds of millions of the world’s citizens to participate more effectively in the global economy. The biggest barriers to mobile money’s success arise from common misperceptions or mistakes in execution. Providers could vastly improve upon the low success rates of mobile-money deployments to date by better understanding their customers and agents and investing more confidently and patiently in mobile money.

Explore a career with us