Why do Many Data Solutions Fail to Meet Expectations?

2015 saw a shift in the data and information solutions built to serve the microfinance industry. Earlier this year, Microfinance Transparency and Moody’s Social Performance Assessment Program called it quits. Other platforms, like the MixMarket, started changing their business models toward a more commercial approach as donor funding to sustain the free services was slowly drying up.

Many of these data or information solutions failed to meet the expectations set by their founders – but why? I argue that what we are seeing today is typical of any solution that is not designed based on the underlying incentives that drive decision making and behavior. Because these incentives are typically not fully understood, projects are conceived with an idealized set of behaviors or assumptions about what might happen in the future.

Many, although not all, of the data and information solutions created to serve the microfinance industry have been donor-driven. Typically, the approaches taken fall into three broad categories. The first approach was to spur a market by subsidizing a data platform that would eventually “crowd in” actors who would be willing to pay for the service down the line. This has happened for some information solutions – like Mix Gold and Mix Silver: services that are built off of the free platform MixMarket, but provide investors a much deeper set of metrics on their investees. But this transition has not been as smooth for other solutions. The second approach – used often by those supporting ratings firms – was designed to be paid services from the onset. But these ratings firms have generally failed to garner the demand to sustain their operations and many of them have either closed or are considering some form of consolidation. Finally, the last approach was to create solutions that are public goods, with the idea that funders would sustain their existence for as long as they were needed.

Data solutions need to address a perceived problem. Before any donor embarks on a new data/information solution, it is important to ask:

  • What is the behavior you are trying to change?
  • Why are these stakeholders not doing what you want them to do in the first place (i.e. what are their incentives)?
  • How would data/information change those behaviors;
  • Who/what would make this information available, both now and in the future?

The best way to illustrate this is to think about one of the entities that has recently closed its doors – Microfinance Transparency (MFT). As a concept, MFT was highly lauded when it was created by Chuck Waterfield in 2009. The mission of MFT was to offer transparency on product pricing by promoting public disclosure and education so that consumers and other stakeholders are able to make more informed decisions. The solution created was a global platform that posted pricing data (interest rates) for various microfinance products and financial service providers in the countries in which MF Transparency collected data. In a meeting hosted by CFI in April 2015, Chuck Waterfield said that MFT used to spend 50% of its time and resources on the convincing side of things rather than on the data analysis side of things. MFT was conceived as a ‘public good’ – something that required public resources to sustain its operations. It was built on the premise that with transparent pricing information, actors such as investors, support organizations such as networks, and MFIs would put downward pressure on MFI pricing. But the expected change in the behavior of these actors, even when the information was publically available, did not materialize. The decision to close MFT was well documented in a blog that Chuck Waterfield published earlier this year.

In hindsight, it is clear that that MFT did not understand the basic incentives driving the actors it hoped to influence. MFT made the assumption that investors would be the main source of pressure on MFI pricing, and this did not pan out as expected. Investors had no incentive to lose out on deals with MFIs by demanding that their MFI investees lower their interest rates to clients. In addition, many of the top tier MFIs had plenty of potential investors - microfinance investment vehicles (MIVs) and development finance institutions (DFIs) – to choose from that were competing for their interest, so they had no incentive to adjust their pricing.

When building a new data or information solution, funders must make difficult decisions early on. For instance, will the product be a public good or a private good? If a private good, are there enough potential customers who would pay for the service down the road? If a public good, is there a long-term source of subsidy to keep the product running for a long enough period of time? Does this information solve an actual problem? Do we need a short-term or a long-term solution? Only by addressing these questions upfront will we be able to envision data and information solutions that can meet expectations while serving the industry in a way that is viable.

Add new comment