Using Data to Better Understand Savings Needs of Customers

The use of data to drive management decisions and product design is well known in the financial sector. Today, most commercial banks utilize data analysis to support their decision-making. Grameen Foundation’s Microsavings Initiative uses these same techniques to help advance the mission of our partner microfinance institutions (MFIs) to effectively bring savings products and services to their poor clients.

Before data analytics is possible, we need the right data from our partners on the ground. One of our Microsavings Initiative partners, CARD Bank (a rural bank focused on financial services for low income clients in the Philippines), has been working to improve its data collection methods and IT infrastructure – two key drivers that make implementing data analytics possible.[1]

CARD Bank understood the importance of integrating customer-level financial transaction, demographic, and poverty data but was not fully able to utilize data to inform its business decisions. Its work with our Microsavings Initiative showed how good data could help its management understand, target and tailor its products to better serve clients.

The data it needed was already available: financial transactions are tracked in the databases at the branches, while it collects demographic information and administers the Progress out of Poverty Index® during the application process for loans and savings accounts.

By analyzing customer data, CARD Bank now has a more nuanced understanding of its clients. This allows it to do more effective targeting of their savings products and matching those to the needs of their customers. Here are a few examples of what we are learning about CARD Bank’s savings customers.

Data analytics for branch management

Data analytics can be used at both the client and bank branch levels. By aggregating customer data at a branch level (displayed sample data N=160K), we learn more about the average type of client each branch is serving. Figure 1 shows branch level poverty rates for customers living on less than $2.50 a day/PPP, along with their average savings balances.

Interestingly, the data show that the poverty profiles of CARD Bank’s branches ranges widely, from branches with nearly 70% of customers under the $2.50 a day/PPP to those with only 20% (see Figure 1). In addition, Figure 1 illustrates that the savings balances are not correlated with poverty rates – thus the poorest and the less poor branches do not show high discrepancies in regards to savings balances. The average poverty outreach under $2.50 a day/PPP is 47% and the average total savings balance is around $63. This type of descriptive analysis could help CARD Bank choose branches for pilot testing new products and services.

Which products for which customers?

CARD Bank is also using analytics to understand how to target specific products to specific customers. It is easy to think that customers are selecting products based on their needs. It’s actually more likely that they are simply picking from the products available to them. Data analytics can help an institution create and tailor products to new and existing clients by helping the institution better understand what its customers need. Customer targeting is an important outcome of any analytical effort and this is particularly vital with financial products: knowing which clients are using which products and defining customer profiles can go a long way towards ensuring the products meet the financial needs of the poor.

CARD Bank offers voluntary savings products, as well as compulsory savings accounts that are tied to loans. When we examined data on clients who have both the compulsory savings account and a regular voluntary savings account, we found they had a unique profile. After running a statistical model, we discovered having those two accounts was more popular among customers with four key characteristics: a salaried job, a higher level of formal education, membership in CARD Bank for more than two and a half years, and more than $65 in total savings (Figure 2).


Given the fact that those customers with two accounts (“cross sold customers”) tend to share these four characteristics, we concluded this particular voluntary savings product was not reaching the poorest and most vulnerable customers. This makes sense because it is a purely branch-based product, without proximity to poorer customers who may live further away from the bank branches.As part of this analysis, we also discovered that customers who share those characteristics were four times more likely to have the regular voluntary saving product. This segment represents about 20% of the customer base and should be targeted for a marketing campaign given its size and the fact these customers show a higher probability of opening this saving product than the rest of the portfolio.For the bank to reach a poorer customer, it had to develop a savings product to better meet these clients’ needs. In response, CARD has developed a new voluntary savings product with a lower opening balance, deposit collections in the field and ATM withdrawal access. CARD Bank conducted a pilot which enabled it to learn what opening balance amount was required to attract the poorer clients. This resulted in an increase in the number of accounts opened by those customers living on less than $2.50 a day/PPP from 27% to 33% after dropping the open balance to an amount they could afford.Grameen Foundation and CARD Bank continue to work together to design, tailor, and market savings products to meet the needs of its poor customers. It is clear that CARD Bank is already learning and benefiting from data analysis to support decision making. Whether we want to understand branch level poverty outreach, product usage patterns at different poverty levels, or predict which customers are more likely to open a new account, data analytics is important to help organizations and institutions make smart choices and better serve their clients. A new case study on CARD Bank’s experience provides deeper insight into the business questions CARD Bank faced and how it used the data to strengthen its savings strategy.

[1] CARD Bank maintains proper confidentiality and privacy of its clients in compliance with regulatory requirements and Grameen Foundation worked with CARD to maintain that privacy in this project.



10 August 2012 Submitted by Peter van Dijk (not verified)

Anything that CARD or any other Microfinance institution or purely social micro-credit program does will not contribute to structural, sustained poverty reduction when it does not combine its technical and financial activities with government level changes. Such advocacy work is most important for three fundamental issues: (1) the formal registration of individuals and their businesses, (2) the government level confirmation of real minimum wages (over the absolute poverty minimum) and (3) effective universal access to legal protection.

Add new comment