Globally, there is a scramble to get access to call detail records (CDRs). Sometimes it feels like the next gold rush in the Wild West (and South, East and Center) of Africa. MNOs in Africa get constant requests from consultants and firms to analyze their customers’ data. During the Ebola crisis in 2014, one MNO in Sierra Leone received on average five requests per week to analyze the data for the purpose of fighting the epidemic. Why all the excitement?
These enormous datasets, easily comprising of a few billion data points, hold all the information on how, when and where customers use their mobile phones. Only a few years back, the amount of data was too massive to be analyzed. Even though data generated is growing exponentially, the advent of cloud computing and new programming languages in computer science, originally invented for artificial intelligence in gaming, make it possible to find patterns and specific user profiles not visible just a few years ago. Quite literally, it is now possible to find the needle in the haystack.
All over the world, start-ups, established software giants and computer scientists at universities are developing algorithms to use CDRs to predict churn, customer adoption of new products and increasingly the creditworthiness of customers. This work has led to interesting findings. For instance, did you know that people who move around more and have more contacts are more likely to use new products like mobile money? Or that clients who call mostly after 10 PM have better repayment rates?
MNOs rarely have the internal capacity to do this kind of analysis on their own and are often willing to pay hefty sums to specialized firms because they hope Big Data will be the magic solution for deeper and more stable customer relations. At the same time, the profit margins for Big Data assignments are attractive since cloud computing power can be rented for a few thousand dollars per assignment and small teams of one or two computer scientists will do most of the programming. Sounds like a win-win situation where African firms are among the most advanced in the industry, right? So, why should international organizations get involved?
There are three good reasons for an honest broker to enter the market: quality assurance, consumer protection and public benefit. Let us look at them in turn.
First, Big Data analytics remains a new and relatively unchartered territory. We still do not know all the possibilities that the analysis of these large datasets offer. At the same time, there are no universal quality standards (like the ISO 9000 quality management standards) for Big Data work. This means that commercial firms sometimes promise results that turn out to be much harder to achieve in practice. Through its own body of work, the World Bank Group is working towards best practice examples and is already able to assist partners in developing countries to judge the merits of technical proposals and assess the quality of results produced. Equally, the UN Foundation’s Mobile Hub is a new alliance to address gaps and issues around coordination within the mobile for development (M4D) sector. The goal of this five year initiative is to equip public and private sector stakeholders in the digital development ecosystem in Africa and Asia with the upstream systems, policies and knowledge necessary to design and deploy large scale, high-impact M4D services for the poor in the developing world.
Second, CDRs are generated by the users of mobile phones, who usually do not know nor have they agreed explicitly that their data will be shared with other firms. The World Bank Group is working with regulators to strengthen privacy protection while continuing to allow the analysis of CDRs. At the same time, its own Big Data projects only use anonymized data and include awareness raising components for MNOs and consumers to ensure that consumer rights are not violated.
Thirdly, and perhaps most importantly, these dataset are generated by hundreds of thousands of people who should benefit from the analysis. Or, as Jaron Lanier has argued in “Who owns the future”, people deserve to be compensated for the data they produce. Of course, the analysis of private firms leading to more precise targeting of consumers will have benefits for consumers similar to smarter ads on websites. However, with the support of donors such as the MasterCard Foundation or Bill and Melinda Gates Foundation, the WBG is able to provide this analysis to MNOs while also using the data for the public good. More precisely, CDRs are being used to build new, more cost-effective poverty maps, to be better prepared to respond effectively to natural disasters and other emergencies, understand more precisely migration patterns or to try predicting (un-)employment. Similarly, the UN’s Global Pulse works towards harnessing big data safely and responsibly as a public good. Its mission is to accelerate discovery, development and scaled adoption of big data innovation for sustainable development and humanitarian action. Digital data offers the opportunity to gain a better understanding of changes in human well-being, and to get real-time feedback on how well policy responses are working.
While it is important for the WBG to stimulate private sector activity, including African tech firms and start-ups, the potential of Big Data analysis to support shared prosperity and jobs in developing countries calls for international organizations to be involved in this space. Big data is generated by all of us and hence should be a public good to be used also by impartial institutions for the benefit of all, but especially for those most in need.
INFORMATION DRIVES INNOVATION. THE OVERLOAD OF DATA TO BE PROCESSED IN INFORMATION AT EVERY LOCAL LEVEL - WILL NOT ONLY FINANCIAL INCLUSION BUT SHARED SUSTAINABLE PROSPERITY. NEEDS TO BE DEBATED FURTHER BY ENGAGING THE ACTUAL STAKEHOLDERS'.