This session moderated by SPTF followed a semi-structured interview, during which speakers from three financial service providers (FSPs) discussed what kind of client outcomes data they collect, what lessons they have learned over the years, and how they use data to make strategic and operational decisions. The three organizations represented by the speakers were FINCA International (with 2.7 million clients having a global presence and offering loans, savings, and insurance), Friendship Bridge (active in Guatemala, offering its 23,700 clients group lending, individual loans, and non-financial services) and 4G Capital (active in Kenya and Uganda, serving 200k clients with merchant finance & business training, last-mile retail finance and partner finance).
Amelia GREENBERG from SPTF opened the session by introducing the speakers and by sharing the key definition of outcome versus impact. Outcomes are changes for (end-beneficiary) clients that are plausibly associated with the services offered by the FSP, though these changes might have been influenced by other factors as well. Impact on the other hand attempts to quantify what exact percentage of the overall change was caused by an intervention. To understand what is happening in a client’s life it’s sufficient to measure change at the outcome level.
Greenberg started with the question what type of outcome monitoring the three FSPs are offering to their clients. Corne NAGEL of 4G Capital responded that they do a mix of quantitative and qualitative data collection. For the first, they use call centres through which they reach a sample of 500 clients/week. Scott GRAHAM answered by stressing that everything FINCA measures is part of their Theory of Change. The indicators part of the ToC explain who their customers are and whether they are financially excluded. Another element of their generic framework and indicator of the impact is the uptake and usages of their services by the clients. This indicator bridges well FINCA’s social mission and business strategy. To monitor their overall mission, a standardized short phone survey is done; in 2020 they reached a sample of 20,000 customers with that.
According to Caitlin SCOTT of Friendship Bridge, their ToC leads to 3 main goals: increased resilience, improved empowerment, and business development. The Progress out of Poverty (PPI) index is an important instrument used for every client on an annual basis. Friendship Bridge complements the PPI with additional data collection for indicators related to empowerment and self-esteem through individual interviews with a sample group every two years. They use a variety of qualitative methodologies, for example through focus groups offering them more in-depth or subtle information. Finally, they use 60 Decibels as an external evaluation partner for a survey to triangulate the internally collected information. Data gathered by that survey are being benchmarked.
The second question discussed related to what has changed over time for the three organisations regarding outcome measurement and what are their main learnings? Scott kicked off by saying that the PPI index has proved to be very useful to Friendship Bridge for providing poverty information and for segmenting other information/data. Friendship Bridge has learned over time to complement quantitative information with qualitative information and thus the application of a mixed method. Sometimes the PPI information is not enough to make actionable decisions. Graham response addressed more tactical learnings. One of their biggest mistakes was paying insufficient attention to data quality. Taking advantage of technology has been a big leap forward. Over time they also understood better what needs to be addressed under ‘research’ (addressing deep-rooted issues) and what under regular ‘monitoring’ (measuring expected change). Another change for FINCA was the converging of business logic and social impact in recent years with more focus on ‘use cases’ in the growing digital world. For 4G Capital, the introduction of early warning systems has been a big step forward allowing to differentiate between factors of influence. Nagel also mentioned that they learned not to lose themselves in quantitative data alone. The Covid period has made them realize that there was an increasing gap between what they offer to their clients versus what clients need. For example, they learned that customers were self-regulating, not taking more than needed even when it was possible.
Greenberg continued with the question whether there is a perfect set of indicators and questions to use? Nagel answered that it is difficult to isolate a specific set of questions and that the applied approach is sometimes equally important. For example, using native language, the use of right channels and timings and methodologies to reach a younger population. Scott added that the set of indicators used is driven by your ToC. However, there are a few things that you should be evaluating in this financial business: are your customers satisfied, and are you doing any harm?
Understanding retention behaviour with more in-depth qualitative methods is also important. Finally, Graham mentioned that in FINCA’s standard year-to-year surveys, one fruitful area is financial health with several questions to check vulnerability. Usage is another consistent topic, for example by asking about substitutes and the value of a service for clients.
The final question discussed was what is the business case of outcome measurement/management and is there a clear line between commercial information and social impact information? Graham recommended separating financial health and poverty status. Age for example is a more important factor for financial health than poverty. Another example related to Covid is that early in the pandemic people were hesitant to take more debt, having more concern about restructuring. For FINCA this meant that they didn’t need to rush to get out in the market to offer loans but to focus more on restructuring as a service provider. Nagel continued by explaining that 4G Capital makes use of agents to constantly interact with their clients to better understand the why behind data. For example, the repaying time shifting from 1 to 2 weeks can be an indicator. In her response, Scott advised using data continuously to learn along the way and to make it an actionable cycle. For example, quality of life measurement helps Friendship Bridge to make findings more actionable.
The session was concluded by sharing takeaways and recommendations with the audience:
Graham mentioned having different indicators for monitoring compared to what you would use for more in-depth research to refine your impact thesis.
A member of the audience asked ‘What types of indicators are used for each purpose, and how are the monitoring data used in regular operations? ‘Graham responded that monitoring data are more about who is being served and to identify the type of clients (segmentation). Research is to understand the type of outcomes that are happening as a result of specific product features.
The next question related to: ‘the use of data to be able to move into predictive analytics, so as to anticipate which segments/cohorts will do well or not do well (i.e. using data science techniques); and how to monitor data through any dashboards that keep data updated in real-time?’
According to Nagel, the collected data can help predict adjusting the loan when needed. At the same time, it’s supplemented with information from their early warning system. Scott added that they created profiles of clients and their needs, and this information is used to make strategic decisions. In general, they like to stay away from creating barriers by using such profiles, and do not exclude potential clients.