Rural Markets Detection for Financial Inclusion

Uganda, Tanzania & Madagascar

Client Brief

  • Financial inclusion is often very low in rural areas in large part because of low densities, challenging topography, low income levels, etc.

  • The usual "static mobile money agent" model is unadapted to those geographies: volumes of transactions are too low for agents to be profitable

  • Identifying when and where rural populations gather (for market days, religious events, social events, etc.) on a regular & recurrent basis can help MNOs deploy roving agents


  • Using mobile data only, Masae Analytics developed an algorithm to detect where and when people tend to gather periodically (every Tuesday, every two weeks...) by detecting unusual peaks in mobile usage. It also allowed a definition of the "catchment area" of those gatherings (where are the people who gathered living)

  • This then allows MNO to build a network of roving agents to meet with rural people on the right days in the right place


  • Online secure platform with maps displaying where and when markets occur (week day, frequency...)

  • Other complementary information is provided like "size" of the market (number of mobile users detected on market days), intensity of GSM usage (minutes, data usage, SMS...), mobile money usage, mobile money penetration in GSM users...

  • Field visits are conducted on detected places to ensure the performance of the algorithm