Drowning in Data, Searching for Information: The Role of Funders


As the range and scope of data-related initiatives expand, funders in financial inclusion increasingly need a clear strategy and framework for reviewing the merits of the project proposals they receive. Not only do they need to decide which has the potential to fill market gaps in a sustainable way, they also need to evaluate how their investments in data ultimately can benefit underserved populations.


Since the early 2000s, funders increasingly have invested in data initiatives to address information gaps. (See Box 1 for a review of key terminology, including “data initiative.”) At the global level, funders support large-scale surveys, such as the Global Findex,1 which informs development actors and market stakeholders about financial access, use, outreach, consumer capabilities, preferences, and behaviors.

Box 1. Key terminology

Big data: Often refers to predictive analytics. It also may refer to data sets that are so large or complex that traditional data processing application software programs are inadequate for their analysis.

Data initiative: An umbrella term that refers broadly to different types of data intervention, such as data source, data platform, or technical support (to a market actor, for example).

Data platform: Where data are hosted and accessible for interactive analysis. Examples include Finclusion, Fspmaps, the G20 financial inclusion indicators data portal, and the CGAP Funder Explorer.

Data source: Organized data, disseminated report, and/or dataset. Examples include demand-side studies (like FinScope, the Global Findex, and Financial Inclusion Insights [FII survey]) and supply-side studies (like the IMF Financial Access Survey).

At the country and market level, funders invest in national or local surveys and data platforms that help inform policy decisions. They also support efforts to build financial-sector infrastructure, such as credit bureaus, to capture and share data that can help promote inclusive, competitive, and responsible financial markets.

At the institutional level, funders invest in analytic tools and capacity-building efforts to improve the ability of regulators to monitor markets. They also support financial services providers (FSPs) in effectively leveraging data in product development, such as using big data to predict client behavior and understand consumer preferences.2 Funders also may commission research at all three levels to inform their own strategies and programming.

Over the years some data initiatives have not met expectations and are being supplemented by new efforts (El-Zoghbi 2015). These include more recent big data initiatives where funders facilitate partnerships between FSPs or mobile network operators (MNOs) and data analytic firms to improve data management. Figure 1 highlights 26 data initiatives that are supported by different funders. Each data initiative shown is designed to address specific constraints or opportunities; they cannot easily be mixed or matched with each other.

Funders need to consider the broader data ecosystem and not necessarily one-off, often unsustainable, data solutions when they fund data initiatives. They need to address questions such as which data initiatives sufficiently address the underlying reasons for missing or inadequate data, when is collecting more data the answer, and how do we provide support without crowding out sustainable approaches from the private sector. This Brief recommends the following approaches that funders should consider:

•  Clearly articulate how the data initiative will improve financial inclusion.

•  Define the sustainability plan from the start.

•  Use a “Will-Skill” framework to identify partners and the support they need.

•  Find the right balance between publicly shared and privately kept data.

•  Collaborate with stakeholders to generate useful data—and avoid duplicating efforts.

Clearly articulate how the data initiative will improve financial inclusion. Funders need to be clear about what change they expect to see in the financial system because of their data initiative, what behavior they are trying to change, and how they plan to measure this change. Is it scale? Diversity of products? Diversity of providers? Better designed products? Competition? Outreach to rural areas, or to specific client profiles?

They also need to be explicit about how the data initiative will help trigger this change. Funders need to collaborate with relevant market actors to ensure that the data will be used—and that there are reasonable expectations of what change the data can lead to. Articulating a theory of change will help to align partners on expected results.3 Sometimes, data that can be collected quickly (e.g., focus group interviews) are more useful than nationally representative information that takes a long time to gather. Funders should establish strong measurement systems that monitor both the intended and unintended results of data initiatives and test assumptions.

Funders need to keep in mind that access to more and better information does not always change incentives or the behavior of the intended users, which are also shaped by social norms and capabilities. Funders should thoroughly understand the motivations and goals of stakeholders before investing in a data solution.

Define the sustainability plan from the start. There are many examples of funders who invest in both demand and supply data sources and platforms. Funders should consider the market context when they seek to subsidize a survey or data platform. For example, extremely undeveloped markets with nonexistent or very weak information functions or data sources may require subsidies, but as markets mature, information solutions should transition toward sustainable models.4

Funders need to strike a balance between one-off data collection for a specific purpose (e.g., to design and monitor their own programs) and longer-term investments that develop data as a permanent supporting function.5 The latter effort could be driven by different players, for example, a bureau of statistics or a network association, that would make data regularly available to providers or governments. The data could be used to inform policies, business practices, and investment decisions.

When funders support market information functions or data sources, they need to take a long-term view of what success looks like in one, three, or even 10 years. They should clearly articulate the use cases of data collection from the outset. This includes decisions on when and whether they should exit from the proposed information function and turn over its funding to the public or private sector. A range of questions should be addressed. If funded by the private sector, are there enough potential customers willing to pay for the service down the road? If funded by the public sector, is there a long-term source of subsidy and commitment to keep the information function operating long enough to be useful for market actors and to meet funder expectations? Is a short-term or a long-term solution needed? These issues need to be considered at the onset of the initiative so that funders’ exit strategies are planned, data solutions for market needs are met, and funder expectations are met (El-Zoghbi 2015).

Funders should use a sustainability analysis framework that is similar to the example shown in Table 1. The sample framework identifies key data functions, such as data identification, collection, housing, analysis, dissemination, and use (The Springfield Centre 2014). Funders can use this type of framework to help them decide whether their activities could lead to long-term behavior change within the information market and to ensure that suppliers actually supply and users actually use information. Interacting with market players will help funders to think through which players are best suited to perform which functions sustainably. The technical requirements of a data platform have to match the capacity of local actors to perform the functions. Long-term sustainability can be achieved by building capacity in national bureaus of statistics, government partners, domestic consultancy companies, and regulators.

Use a “will-skill” framework to identify partners and what support they need. After confirming that there is a realistic vision of success and either a business case for a data investment or a long-term plan for public funding, funders should determine what their interventions will need to do to achieve this vision. The will-skill framework can help funders to identify potential partners and what the partnership would entail (see Figure 2). This framework is applicable only to long-term data initiatives and does not apply to one-off data projects.

Depending on a partner’s incentives (will) and capacity (skill), different kinds of support might be required (The Springfield Centre 2014). For example, for partners with high will but low skill, funders may decide to focus on building capacity (e.g., a funder could provide technical support to a financial regulator on data housing and analysis). Building the capacity of market actors improves the sustainability of the data initiative and optimizes the use cases. Institutions with high skill but low will might not be the right partners, unless funders can influence their incentives or opposition to change. Potential partners with high will and high skill should cause funders to determine how their investment would add value. On the other hand, potential partners with low will and low skill could pose significant risks.

Find the right balance between publicly shared and privately kept data. Publicly shared data can help promote competition and incentivize certain behavior by market actors. In 2006 in Pakistan, Microwatch was created to promote competition by sharing the data of individual MFIs (e.g., the geographical distribution of active borrowers, gross loan portfolio, number of loans distributed, disbursements, average loan size). Keeping certain individual provider-level data anonymous, Microwatch highlights high levels of over-indebtedness and opportunities for providers (e.g., noting underserved districts) and provides benchmarks. The right balance between publicly shared and privately kept data helped to stimulated competition and transparency in Pakistan’s microfinance sector.

However, what promotes competition in one market may hinder it in another. For example, agent registries that require FSPs to share too much data on agents can discourage agent network development, because FSPs may fear that competitors will poach their best agents. In most cases, though, making data public is desirable—as long as it does not compromise privacy concerns for customers (Mazer, Carta, and Kaffenberger 2014). To support data initiatives, market actor incentives, the competitive environment, the trust in the information system, and consumer protection should be analyzed.

Collaborate with stakeholders to generate useful data—and avoid duplicating efforts. Dialogue with relevant market actors and other funders when designing data initiatives can help to improve the quality of data, address market needs, and seize opportunities to align and support ongoing efforts. Collecting data is costly. The cost of conducting a nationally representative demand survey with a sample of 3,000 individuals can be upwards of $500,000, depending on a range of factors.

Over the past 10 years, funders have conducted many large demand surveys and one-off market research on specific segments or themes, and have developed several related data platforms (e.g., Fspmaps, MIX, Finclusion, and Microfinance Transparency). Many of these initiatives were designed to inform funders and to measure the status and progress of access, pricing, and other factors within and across markets, and of their programs and overall portfolio.

Before they invest in data collection, funders need to assess what data and information are available in the market and how these data have been applied. Several similar data collection efforts have been conducted in some of the same markets in the space of a few years. In Uganda, for example, the Finscope survey, the FII survey, and Global Findex were all conducted between 2011 and 2013. While these differ in methodology and there may have been valid reasons why a funder would support each, nuances in the headlines from each can create conflicting messages that can confuse stakeholders (Broens Nielsen and Butkus 2014). Ensuring stakeholder buy-in and/or ownership and considering specific use cases for data are important. Funders should cooperate with each other to make sure that the potential value of the data is well-articulated and benefits as many stakeholders as possible.

When it comes to investing in demand-side data sources, funders should make the data and methodology public. This will help to limit duplication and enable others to explore whether the methodology can be replicated in different markets. Researchers, market actors, and funders need to be able to work with the data to review and validate findings. Making the data and methodology public will also enable deeper analysis and combining of datasets to generate new insights.

Looking Ahead

The cost of smart phones is dropping fast, connectivity is expanding, and the data trail following poor people continues to grow. The expansion of digital technologies will open up exciting new opportunities for reaching more of the financially excluded. But it also raises challenging issues around data privacy and data security. Funders have an important role to play in exploring how data-enabled financial services can benefit poor people and in helping governments to balance the new risks that are emerging with the opportunities provided by a digital world.


Broens Nielsen, Karina, and Abby Butkus. 2014. “Making Sense of Financial Inclusion Data Sources.” Blog post, 3 June. http://www.cgap.org/blog/making-sense-financial-inclusion-data-sources

Burjorjee, Deena, and Barbara Scola. 2015. “New Funder Guidelines: Market Systems Approach to Financial Inclusion.” Washington, D.C.: CGAP. http://www.cgap.org/publications/new-funder-guidelines-market-systems-approach-financial-inclusion

El-Zoghbi, Mayada. 2015. “Why Do Many Data Solutions Fail to Meet Expectations?” Blog post, 18 December. http://www.cgap.org/blog/why-do-many-data-solutions-fail-meet-expectations

Mazer, Rafe, Jessica Carta, and Michelle Kaffenberger. 2014. “Informed Consent: How Do We Make It Work for Mobile Credit Scoring.” Blog post, 2 September. http://www.cgap.org/publications/informed-consent-how-do-we-make-it-work-mobile-credit-scoring

The Springfield Centre. 2014. “The Operational Guide for the Making Markets Work for the Poor (M4P) Approach,” 2nd edition. http://www.springfieldcentre.com/wp-content/uploads/2014/09/2014-09-M4P-Operational-Guide-with-watermark1.pdf

  • 1.    Global Findex is a comprehensive cross-comparative demand-side database on financial inclusion, based on interviews with about 150,000 adults in over 140 countries.
  • 2.    Examples of services providers supporting market actors in big data analytics include Real Impact Analytics, Altai Consulting, and Flowminder.
  • 3.    A theory of change explains how interventions are understood to form pathways to a series of results that contribute to achieving the final intended outcomes. It can be developed for any level of intervention—a project, program, policy, strategy, or organization. It is usually presented as a diagram. Progress along the pathways is based on contribution hypotheses and is governed by assumptions, the most important of which are made explicit.
  • 4.    “Information functions” is a broad term. To make relevant decisions, FSPs, policy makers, regulators, supervisors, etc., need information on the clients (e.g., identity, needs, degree of satisfaction, credit history, and asset quality), the suppliers (e.g., range of services, tariffs, sustainability), market coverage (e.g., current users vs. market potential), relevant regulations, and macroeconomic factors that might affect their operations. Clients should have access to clear and transparent information, especially on financial services, suppliers, and customer protection regulation. Different information functions can be managed by either a public- or a private-sector actor like a credit bureau, for example.
  • 5.    For more information about supporting functions see Burjorjee and Scola (2015, p. 22).

Add a Comment