Using Behavioral Economics to Create New Savings Habits

March 24, 2014

By Kimberly Davies, Program Officer, Financial Services

You’ve finally gotten that new savings product off the ground and into people’s hands. They sign up in droves—and then they go silent.

Dormant accounts can be a huge problem for financial service providers.  It puts a huge dent in their bottom line and could jeopardize their sustainability. Their clients also suffer because they don’t gain the experience needed to better manage their assets.

During Grameen Foundation’s Microsavings work with CARD Bank, we found that 58 percent of its new Matapat savings accounts were never used. People wanted to save, signed up for the account, thinking it would be a good fit for their needs. But, they never made transactions.

Grameen Foundation worked with Behavioral Economics experts, ideas42, to address this issue. Through a successful pilot, we were able to raise the balances in the Matapat accounts by 37 percent and increase the number of initial deposits that were higher than the required minimum by 15 percent. Overall, clients were 73 percent more likely to transact in a new account.

How we did it:

Applying behavioral economics techniques, we used both qualitative and quantitative research methods to determine the causes of low savings balances. The diagnosis focused on four primary barriers: clients became anchored to lower deposit amounts, they opened accounts without a clear intention of how to use them, they didn’t enroll in regular savings collections because it doesn’t seem important at the time, and they had abstract savings goals. With that, ideas42 worked with CARD Bank and Grameen Foundation to develop a small randomized control trail (RCT) to test several specific interventions. These interventions included goal-setting, planning how to reach that goal, the feeling of having made a commitment, and personalization of the experience.

What we learned:

One of our biggest learnings was that while client education is extremely important, it is just one step when providing financial products and services to the poor. Many times clients want to save, but this desire is not necessarily met because of competing financial needs.  Sometimes we can focus on education being the solution to increasing adoption and usage, when it’s actually bridging the “intention-action” gap that will help facilitate easy product uptake and regular usage.

We also found that using a client behavior focus is an important part of user-centered design. Focusing on the client experience from opening an account through regular usage allows us to identify unseen bottlenecks that prevent clients from using formal financial services to help them reach their goals. Additionally, during market research people can sometimes say they want one thing, but their actions show differently, so an additional behavioral quantitative focus can bring in rich client insights.  Incorporating behavioral economics techniques into research can help us better understand our clients.

Lastly, while RCTs can often be seen as incredibly timely and expensive, we found that RCTs or “RCT-lites” can be built to bring in useful elements to track a treatment’s impact, without spending years or millions of dollars. With ideas42 we were able to test the treatment that ideas42 developed as a bundle, which included subtle changes in the sign-up form to open the account, the addition of a savings plan, the addition of a savings calendar, and SMS texts. Bundling all of the test items together made the treatment analysis much faster and cheaper than a typical RCT. While the term RCT can bring specific visions of time and cost commitment, we should consider levels of intensity that an RCT can take, and see components like randomly selecting pilot groups, comparing treatment outcome to control groups, and doing data analysis after a pilot as useful  activities, which can all be tailored depending on resources and time available.

More details about our pilot’s structure, outcomes, and learnings are available in the full case study.