Crowdsourcing at Work: Mapping Financial Access in Uganda

This post is the first in a series designed to help donors working on financial inclusion explore and better understand the role of information in spurring growth and innovation in financial markets for the poor. 

Access to digital financial services is fundamental to enabling poor people to become more economically stable, prosperous, and resilient. These services – payments, credit, savings and insurance offered through mobile phones or other technology – are reaching millions of people around the world who had not previously been included in the financial system. Historically, being included in the system meant living close enough to access a physical bank branch, credit union, or ATM location. With the advent of mobile money, location still matters, but the proximity of the nearest mobile money agent is now more important than the nearest bank branch or ATM. In Kenya, for example, banking through the mobile-phone based savings and loan product M-Shwari has generated Sh153 billion in deposits from over 10 million customers.

In order to increase digital financial inclusion, service providers must have the capability to analyze gaps in coverage and strategically fill those gaps. Dynamic, rapidly changing agent networks translates to a need for sophisticated, real-time location data and analysis tools. While much has been accomplished to collect and geolocate data on mobile money agents (see, for example fspmaps.com), often the methodologies used in prior collection efforts revolve around periodic national surveys that are cost-prohibitive and pose sustainability challenges. Data on provider networks can only be updated every 1-2 years and do not keep pace as locations open and close. In addition, previous processes miss out on local knowledge and insight as communities have not to-date been involved in the solution. Bringing in communities also results in savings as the cost of data collection by private-sector firms is much higher than community/participatory approaches. In addition, data becomes out of date and is not self-sustaining without further paid collection activities.

The Humanitarian OpenStreetMap Team (HOT) is an international NGO which provides training and support to communities to use open data for humanitarian response and economic development. HOT, with support from the Bill and Melinda Gates Foundation, will initiate a six-month pilot program in Uganda starting in December 2015. The pilot program will address the need for a data collection methodology that is inexpensive, accurate, allows for frequent collection, and integrates seamlessly into existing processes. The pilot will examine the cost and effectiveness of involving university students and community members in collecting data on financial services available in the areas where they live. The program will train and equip participants with the necessary skills, equipment, and other resources to collect and map this data. While students and community members are not paid, they will receive a small stipend to cover their food and transportation expenses while working on the project.

OpenStreetMap (OSM) will be used as a technology enabler to convene government agencies, local NGOs, academic partners, existing financial services technology startup communities, and the local OpenStreetMap community in Uganda around a single open data platform. Data in OpenStreetMap can be freely exported and analyzed by all – including financial serve providers and telcos – to improve coverage of their provider networks. It can also be used to facilitate dialogue among providers by groups such as Financial Sector Deepening (FSD) Uganda, or by tech entrepreneurs to develop apps focused on increasing mobile money utilization.

Data reliability/accuracy
Involving students and community leaders in the data collection process can result in more meaningful and reliable data. For example, Uganda’s Local Councils often maintain lists of financial service locations in their zones. Having a trusted community member present during survey of a mobile money agent can yield more truthful answers to questions. HOT’s methodology during the initial training and survey includes a two-step collection validation process, where data collected by one team is validated by a supervisor and then spot-checked. A sample is also re-surveyed by a second team to corroborate results. Over time, the formal two-step process is replaced by the “crowd” of over two million OSM contributors who continually work to correct and improve data quality.

Incentives for providers to contribute
While the community mapping approach does not rely on direct provision of data from financial service providers, it does aim to benefit providers and over the long-term incentivize their uptake of open data platforms. In the future, providers themselves may also more openly release data which can then be verified by local mappers. It comes down to business gains: providers will contribute data to an open platform when they can derive benefit from the platform. Contributions of data to a platform such as OpenStreetMap might initially appear risky: after all, competitors will have open access to each other’s agent network data. At the same time, OpenStreetMap has the potential to drive new business and new agent network expansion approaches. By contributing to such a platform, providers will have a comprehensive data foundation upon which new analytics tools can be developed to provide sophisticated network analysis and expansion scenario planning.

Long-term sustainability 
The sustainability of accurate, up-to-date financial service location data boils down to incentives and motivation. Providers and communities will be incentivized to keep data up-to-date when they receive value from the data. This value could come in the form of mobile apps or businesses that rely on accurate OSM data to fill a critical need. Communities will also have other incentives to continue to update the data. Free training and skills building is one motivator: community-led mapping will include training in the latest open data collection and open source GIS tools. By forming long-term partnerships with universities, rather than a one-time training, open data can be integrated into university GIS curriculum and faculty OSM champions identified. Over time, these steps contribute to building a highly-functional local infrastructure and community capacity for mapping while shifting funding required for additional training from international sources to local sources.

The availability of data in openly accessible formats may help to foster the growth of local technology entrepreneurship and small businesses that create mobile apps or services using the data. The more apps and businesses that rely on OSM data mean the more shared benefit is derived from keeping the data up to date. In the long-term, the resulting apps or SMS-based services can drive behavior change toward increased availability, usage, and access to financial services.