Deciphering the Data Deluge: Making Data More Useful for Financial Inclusion
Editor’s note: This post is part of NextBillion’s series, “High Tech Buzzwords: Hype or Real Impact,” — one of several topic areas we’ll be covering through special series this year. Click here for more details on our 2018 series.
The excitement around Big Data in the financial inclusion space is understandable. Imagine a world where an emerging markets consumer can look at Google Maps and see financial access points throughout their local community, with enough real-time data to choose the financial services that best suit their needs. This scenario isn’t that far-fetched, says Camilla Nestor, CEO of MIX.
But this enticing potential is somewhat tempered by the challenges of gathering this data, distilling it into a usable format, and presenting it in a way that makes sense – and by the risks it can imply for low-income customers. That’s why MIX manages FINclusion Lab – a data analytics platform launched in 2014, and now supported by MetLife Foundation and other partners – to help market actors make sense of financial inclusion data through interactive dashboards and maps. The platform integrates national and sub-national data points from numerous sources, facilitating insight generation, business planning and financial inclusion measurement in some of the world’s most challenging markets.
We spoke with Nestor about this initiative and the impact it’s having on financial inclusion efforts around the world. Check out our Q&A below.
James Militzer: Give us your “elevator pitch” for the FINclusion Lab: What needs does it address for the financial inclusion community that other data sources don’t?
Camilla Nestor: FINclusion Lab creates single datasets and databases where previously siloes existed. For example, the data – which primarily includes access point location and demand-side data like population density, cellular coverage, poverty rates and the like – is usually found in project documents (PDFs), or separate online locations managed by regulators, or even individual Excel files from financial institutions. Bringing it all together in one place allows users – often regulators, financial institutions or others – to conduct analyses across different types of data including service points (geo-coded data), credit/deposit usage and demographics. It also allows users to visualize the data across geographies and drill down to more specific locales. Because we publish this data in a highly interactive format, users can explore the data based on their specific questions or interests. For example, a user can explore a particular district or type of financial service provider, or pick a reference period to view trends.
JM: Do you have any examples of how the platform is currently being used to improve financial access and health in emerging markets?
CN: Regulators and policy makers, for example, are able to identify geographic areas where there are either no access points or very few access points. They can also examine the socio-economic data in these same areas, which may indicate whether there is an opportunity to create incentives-based policies for financial service providers to expand. Similarly, these same actors can explore the interactive data to understand market concentration of financial service access points and identify opportunity areas. Additionally, bringing this data together allows stakeholders to monitor and analyze progress on financial inclusion targets.
JM: Where does the platform’s data come from, and how do you determine what data sets to include?
CN: This can be the difficult part. Typically, in a particular country, our team at MIX gathers a group of local stakeholders, including funders, financial service providers, regulators, associations and others. We work with them to understand the priority use cases and, based on those use cases, determine exactly what data are needed. For example, if a priority is to understand access to finance for smallholder farmers, it may be especially important to capture data on agricultural markets in that country. The same group of stakeholders also helps our team to identify the available data sets, which often take different formats. These conversations with local stakeholders are key to creating useful information tools. In the end, these same actors (and others like them) are also the users.
JM: Why did the MIX decide to aggregate existing data, rather than generating new data? Are there any downsides to this approach?
CN: The short answer: Creating new data is expensive! There have been a number of experiments with surveying to gather data, for example, collecting data using enumerators or surveying all the agents in a particular country. But this approach is very expensive to undertake and, even if there is funding available for the initial collection, it is difficult to find support to maintain and update those datasets on a continuous basis, even annually.
Instead, we prefer existing data sources. In particular, we see regulators as the preferred source of data because those collection costs are already baked into the system, so to speak. If we can prove the value with those data sources, we are then also more likely to see them improve their data because they are using the data in meaningful analysis.
JM: How does MIX work with other players in the industry, to avoid duplicating efforts or siloing as you aggregate and share this data?
CN: Any public domain data that MIX “normalizes” to ensure its reliability and standardization is available from the “About the Data” tab on the FINclusion Lab website (free registration is required to download the files). We also reuse existing datasets, such as those created from financial services providers’ service point censuses and through actors like the FSD Network in Africa. Sourcing reliable data – as you probably already know – is always challenging. We want to make sure that good, useful data is available to the financial inclusion community so decisions can be better informed.
JM: What are some of the biggest challenges you’ve faced in operating the platform, and how have you overcome them?
CN: Data availability for the full set of priority use cases – as determined by the initial group of local stakeholders – is always a challenge and one that cannot be easily overcome in the near term. Rather, we hope that by demonstrating the value of this analysis, it will spur various actors, including financial service providers and regulators, to act to bridge these data gaps.
One challenge we’ve faced and overcome is that of geo-location. We geo-locate data by aggregating it at the most detailed geographic (or administrative) unit. In the context of the United States, a county is more detailed than a state, and a municipality is more detailed than a county. In many markets, those administrative units change with surprising frequency and datasets do not always line up, especially at the most granular level. In those cases, instead of tossing that data out, we map it to the next highest administrative level. Obviously, it reduces the granularity of the data but it increases the reliability. This is a good example where the underlying data has quality issues and we’ve developed a solution to accommodate those changes. Again, if actors see the insights that can be derived from this information, it could lead to improvements in that underlying data. But that is a long-term challenge.
JM: Why are data aggregation and mapping so important – and how much more effective could “Big Data” be in global development, if we had better ways of synthesizing it?
CN: You can imagine a scenario where you look at Google Maps and are able to see financial access points down to a very local level – including mobile money or banking agents, branches and other types – updated through crowdsourcing rather than through expensive surveys. And taking that further, you might be able to see real-time data on mobile money agent liquidity, what financial products are offered at which access points, cost data, reviews – everything a consumer would need to do educated shopping for financial services. This isn’t so far-fetched and has considerable implications for some of the major financial inclusion data collection efforts underway.
JM: Relatedly, is Big Data all it’s cracked up to be in the financial inclusion space? Does it live up to the hype – or are any unexpected challenges/downsides diluting its impact?
CN: Big data holds enormous potential to create better-designed products, improve the efficiency of delivery channels, and pull excluded consumers into the financial sector. It’s obviously a very exciting space. But of course, there are risks – particularly around data privacy. Depending on how digitally connected they are, people can generate hundreds of data points about their lives each day, raising real issues around who can access that data, and how it’s used. Even in mature markets with comparably strong privacy laws, you have situations that expose personal data about people – who never agreed, or “opted in” to having their data collected and sold. The breach of trust with consumers – in particular, low-income consumers who are often less likely to trust formal financial services – can be enormous. I hope we can apply these lessons to efforts in emerging markets to ensure we’re tapping the upside potential of big data to further financial inclusion without exposing households to risks they’re ill-equipped to handle.
JM: Do you have any thoughts on the challenges facing donor-financed initiatives that serve a public good, but that often lack a clear revenue model that could allow them to exist as for-profit businesses? What is the path forward for these types of initiatives, when/if their donors’ priorities shift?
CN: From the outset, donors and implementing organizations should be designing initiatives with a long-term sustainability model in mind; there needs to be a clear and realistic vision of how such functions will work in the future without ongoing donor support. For functions of a private nature this is pretty straightforward. For functions of a more public nature, it entails identifying actors that have the incentives and capacity to perform or pay for essential functions in the long term, and then using catalytic donor resources to build towards that vision.
But commercial revenue is not the only path to sustainability. In some instances, initiatives can be sustained if governments value the work and have the ability to support it. Either way, it’s necessary to design and adapt the initiative to deliver what market actors are willing to pay for. From the donor perspective, we hope to see a clear vision of how these public or collective functions will be provided and paid for in the future, without further donor support. A long-term view also encourages donors to act as conveners, identifying and uniting multiple actors to ensure that market-building initiatives are complementary.
Main photo: An example of one of the FINClusion Lab’s interactive dashboards, via MIX.
Note: MetLife Foundation is a NextBillion partner.
James Militzer is an editor at NextBillion.