Safaricom M-PESA: Using Data Analytics to Improve Customer Service and Products

Providing access to financial resources remains one of the main current goals of global international development. The simple provision of access to finance does not however necessarily change the market outlook for providers, nor does it resolve the constraints on customers. Service providers have to constantly improve the quality of the services they offer. Firms that manage by the three pillars of Responsible Finance should pay careful attention to fair and respectful treatment of clients, constantly updating transparent mechanisms for complaint resolution. This allows not only the creation of formal financial opportunities for the unbanked population, but also fully satisfy the needs of clients in terms of responsible customer service. The case of M-PESA in Kenya demonstrates how a data driven approach can successfully identify the most crucial gaps in services. Market players can develop necessary strategies to answer these challenges, significantly improving the quality of such services.

M-PESA was the pioneer of DFS at scale, and by 2016 had 20.7 million registered customers, a thirty-day active customer base of 16.6 million, and reported revenue of $450 million. When Safaricom launched the service in 2007, there were no templates or best practices; everything was designed from scratch. Continuous operational improvement was essential as the service grew. Uptake was unexpectedly high from the start, with over 2 million customers in the first year, beating forecasts by 500 percent. This huge demand forced the team to tackle capacity issues well before they had expected to do so. At this early stage in the product lifecycle, a bad customer experience could quickly erode customer confidence, so the operations team had to proactively anticipate scaling problems in both the technology and business processes. Data-driven metrics helped the team plan and guide operations appropriately. As uptake was unexpectedly high from the start, the number of calls to the customer service center was correspondingly much higher than anticipated, resulting in a high volume of unanswered calls. To improve call response levels and achieve their key performance indicators, the customer care team needed to make some changes. The problem was first tackled by recruiting additional staff, but recruitment alone could not keep pace with the increase in customer numbers. To identify bottlenecks and prioritize solutions, the team analyzed their data. PABX call data and issue resolution records were examined and some key findings were:

  • Length of call time: the average call was taking 4.5 minutes, around double the length of time budgeted.
  • Key issues for quick resolution: the two key call topics were forgotten PINs, and customers sending money to the wrong phone number; this covered 85 to 90 percent of the longer calls coming into the call center.

The analysis allowed bottlenecks to be identified, passing key insights into operations. It also highlighted the unexpectedly high incidence of some difficulties that customers were experiencing, namely erroneously sending money and forgetting their PINs. Managing against the Unanswered Calls KPI therefore delivered broader operational benefits. Using the analytic results, operations implemented a resolution strategy. Firstly, by understanding lengthy versus short problem types, difficult issues could quickly be identified and passed to a back-office team for resolution. This reduced customer waiting times and freed up the call center representatives, allowing more customers to be processed per day. Secondly, operations and product development teams worked to reduce times across all call types. This was achieved by improving technical infrastructure and the M-PESA user interface, mitigating the problems that caused lengthy calls. The combination of initiatives reduced the Call Length KPI and number of Unanswered Calls KPI, shifting both to acceptable levels despite customer numbers continuing to grow beyond forecast levels. Thirdly, this assisted in prioritization of interventions and refining consumer education activities to manage recurring issues such as pin activations.

The M-PESA case is an excellent example how data analytics solutions can be successfully used to provide significant improvements in customer service and create an opportunity for service providers to enhance the quality of their offerings and address the most urgent needs of customers. Developing a better complaint resolution system, the company both creates a responsible approach to finance by treating the customers efficiently and respectfully, and builds its reputation by providing a better service in an increasingly competitive market. Both the DFS provider and the customer benefit from such approach in a long-term perspective.

Adapted from a case study presented in the Data Analytics and Digital Financial Services Handbook (June, 2017), this post was co-authored by Susie Lonie and Oleksiy Anokhin, IFC-Mastercard Foundation Partnership for Financial Inclusion, for the Responsible Finance Forum Blog.