Agent Network Expansion: What Can We Learn from Cote d’Ivoire?

For mobile money to reach scale there are many aspects providers need to take into consideration. One of them is deploying a high-quality distribution network with extensive coverage, especially in rural and remote areas where access to financial services is often limited. In Cote d’Ivoire, not only has the market boasted impressive acceleration in terms of registered customers but the number of transaction points in the distribution networks has also seen tremendous growth. Data show that between 2012 and 2013 the number of agents grew from more than 3,000 to 12,000.

In 2013, CGAP provided funding, technical assistance and knowledge sharing to one mobile money player in Côte d’Ivoire. The objectives were twofold: (1) to expand the reach of mobile money services and improve the quality of the agent network, and (2) extract lessons learned. The project spanned a period of 18 months.

Now that the project ended, what are the results?

More than 1,500 new agents were deployed during the project. Starting from a pool of about 4,000 airtime resellers and small shops, the agent selection process was designed to recruit those demonstrating the greatest potential for mobile money activity.  Two categories of agents were targeted because of their proximity and their close daily interaction with the population:

• 1,456 agents, consisting of neighborhood shops;
• 72 agents, consisting of managers of cellphone booths including 31 exclusively located in rural areas.

Faster deployment of agents. The mobile money provider tested and implemented adjustments to processes and arrangements. Supervisors were introduced to train and support agents with e-money management, customer experience enhancement, and fraud prevention.  This helped reduce operating costs and considerably shorten deployment time (less than three weeks, instead of more than one month).  Today, 45 agents a week on average can be deployed.

Agent performance improved. The mobile money provider implemented a new approach to agent selection, with the introduction of tools such as a scorecard for qualitative appraisal, GPS to optimize geographical deployment, and a system of personalized supervision and monitoring. This led to better quality in the overall agent portfolio. Newly recruited agents conducted almost 10 million transactions over a 10-month period, averaging between 400 and 600 customers a month per agent.

Personalized supervision improved agent satisfaction. Two hundred seventy neighborhood shops received personalized supervision and monitoring.  On average, five agents were visited per day and per supervisor, with one or two visits per week.  A survey performed on 150 agents shows high level of agent satisfaction.

Beyond these positive results, what can the project teach us when it comes to deploying mobile money agents?

Overall, it is essential that providers ensure a clear vision and buy-in across the organization to deploy agent networks at scale. A dedicated team and resources, suitable procedures, logistical support and ongoing supervision and monitoring of agents at the onset are also important factors to ensure success.

Conclusions can also be drawn from the three phases of deployment:

Selection/Validation: With the right skillsets and tools, agents will contribute to ensure the viability of the distribution network.  So, the selection of agents is an essential step and if not done properly it can lead to additional costs and delays. Overall, the simple appraisal scorecard has helped the provider expedite the hiring process while optimizing the financial capacity of its agents’ portfolio and ensuring that people with an adequate profile are recruited. But two specific challenges also came up during this phase. First, some agents needed to be converted from the informal to the formal sector. This was particularly the case for cellphone booths managers who needed to obtain a business license to distribute electronic money.  A partnership was established with a government agency to find suitable licensing requirements for this particular category of agents. Second, some agents need to borrow money from financial institutions to meet new working capital requirements and this requires supervision.

Training/Deployment: Agents who met their supervisors during their initial training period proved to be more efficient and results-oriented. Their performance also improved when supervisors monitored agents’ activity using simple indicators, such as level of transactions, stock of electronic value unit, etc. That said, several factors should be taken into consideration as they can delay the deployment of the agents: the paperwork needed to be able to exercise the activity of agent as well as the fact that not all applications from potential agents come completed. Moreover, human resources at the provider level must be in place to activate the SIM cards of agents to match the growth in agent deployment; if not, this can create an important bottleneck.

Monitoring/Supervision: Qualified staff is essential to supervise and motivate agents so that their performance is optimized. Supervision of agents has also helped improve the ability to deal with periodic fluctuations in deposits/withdrawals and manage the demand for units of electronic value. Finally, financial incentives for agents to open new customer accounts are an additional driver to improve performance.

The spread of agents across Cote d’Ivoire is promising but it also raises a number of questions regarding their activity level and health. We hope that providers in Cote d’Ivoire and across WAEMU can capitalize on these lessons to ensure healthy and sustainable growth of mobile money distribution networks.