Paper: Predictors of Downstream Retention of Mobile Financial Services

“Mobile Financial Services in Kenya, Uganda and Tanzania: Downstream Prediction for Lasting Economic Retention”. Cameron Turner, Aman Ahuja, and David Gutelius. [Access via academia.edu]

In this paper we sought to provide directional frameworks for mobile network operators (MNOs) and non-government organizations to better deliver digital financial services in East Africa.

While “upstream” adoption – awareness and sign-up for services – is relatively well understood, there is a dearth of understanding about “downstream” retention – utilization and activity beyond the initial 90 days. We surveyed available data assets that enabled meaningful longitudinal analysis of downstream retention, and proposed future work to improve data availability.

Using statistical modeling techniques we identified predictors of retention and differentiated between usage of services such as payments and loans on mobile devices. We investigated retention predictors in mobile financial services using statistical modeling, challenging common beliefs about demographic factors like gender and emphasizing the need for nuanced data analysis.

While initial findings are provided with caveats around data availability, we found age and education to be the primary demographic factors leading to retention. Conversely, gender was a poor predictor contrary to popular belief. Further analyses separates basic vs. sophisticated use of mobile financial services.

A follow-up phase (including primary data collection) was proposed to confirm these findings and to move beyond basic demographics into more nuanced prediction and development of modeling tools useful to practitioners.

This work was under context of the Financial Servies for the Poor (FSP) group at the Gates Foundation. Grant was provided by the Bill and Melinda Gates Foundation and administered by Rockefeller Philanthropy Advisors.