Using psychologically satisfying numbers to pay off loans faster

Advertisment framing, Round up framing
Experiment Type
Field Experiment
Increase program enrollment, Repay loan
Reduce debt
Focus Areas
Marketing & messaging
Behavioral Concepts
Framing Round number bias
Partner Type

What Happened

It worked. reframing the value proposition from “save money on a mortgage” to “earn money back from a mortgage,” we were able to increase click-through-rates on their online ads by 59%.

It worked. Both arms of the study inlfuenced 10% of people to change their behavior, while the "round up" framing added an addition 3.1% of people impacted.

Lessons Learned

The results suggest that an "earnings" framing (instead of a "savings" framing") may make saving on a mortgage payment more appealing.

These results suggest that while a simple prompt to increase loan payments may be effective, adopting a "round up" frame may be even more effective.


For a large share of Americans, housing costs are a severe burden. In 2014, the total number of households using more than 30% of their income for housing increased to 40 million, with almost 20% of households using more than 50% of their income to pay for housing costs.

To address this financial burden, we partnered with fin-tech startup EarnUp to help Americans pay off their home mortgages faster and increase long-term savings. EarnUp has built software that intelligently syncs users’ income and expenses. This expense-matching software helps users save thousands of dollars on debt interest reduce loan term by over five years, avoid default, and build net worth faster.

Through this partnership, we became interested in understanding the following questions:

  • What makes saving on a mortgage payment appealing?

  • How can we get people to increase their loan payments?

Key Insights

We conducted a full behavioral diagnosis of EarnUp’s current sign up process, led in-depth interviews with the leadership team, and managed a data analysis of users’ current payment patterns.

Our behavioral diagnosis led us to two key insights:

  • We found that some customers were already paying more than their required mortgage payment.

  • A large percentage of these users were overpaying because they rounded their mortgage payment to the nearest $5 or $10. Behavioral insights have consistently shown the strong tendency of households to use heuristics to simplify financial matters. Rounding is one heuristic found to be useful for LMI households as an easier alternative to complex budgeting.


Experiment I

The team kicked off the partnership by trying to understand what drives people to save money on a mortgage payment. Given that EarnUp offers borrowers a very appealing saving opportunity, why isn’t everyone with a mortgage jumping to sign up?

Our hypothesis was that while “saving money” should be appealing, it is not nearly as appealing as “earning money.” We tested this hypothesis by making a slight language adjustment to EarnUp’s online advertising materials. By reframing the value proposition from “save money on a mortgage” to “earn money back from a mortgage,” we were able to increase click-through rates on their online ads by 59%.

This low-cost intervention, based on behavioral principles, helped drive new users to EarnUp. Based on this research, the company changed its name. Prior to the research study, the company was named APASave. After the research was conducted, the company changed its name to EarnUp.

Experiment II

Our next experiment focused on our second question: How can we get people to increase their loan payments?

Given the combined insights from EarnUp’s data and existing literature on rounding, we wanted to test if prompting people to round to psychologically satisfying numbers could cause people to overpay more than simply telling them to overpay.

We used a simple A/B email test and sent emails to a random selection of 3000 of EarnUp’s active user base. These users were split into two conditions, a rounding condition and a control group.

In both conditions, participants received an email that told them how much the average EarnUp customer saves when they overpay their loan, their next payment amount, and a message asking them to contact EarnUp to start overpaying.

Both emails were identical, except that participants in the rounding condition were prompted to round up their monthly payments, while participants in the control condition were simply prompted to overpay. Participants could either reply to the email or call a phone number to increase their monthly payment.

Increasing the number of people who overpay on their mortgage payments, no matter how small, will have significant impact on their users’ loan term, interest payment, and overall borrowing costs.

To put this into context, according to the Federal Housing Finance Agency, as of November 2016, the average mortgage amount for a new loan was $312,700 at a 30-year-fixed interest rate of 3.76%. If a borrower paid just an extra $25 a month on this loan, they will save over $7,000 dollars and pay off their mortgage a full year sooner.


In reality, our intervention produced a similar result. We compared the number of people who opted to increase their loan payments in the “over pay” and “round up” conditions. Across both conditions, about 10% of people who read our email indicated that they wanted to increase their payment. This represents the high latent demand for debt acceleration. A simple prompt to increase your regular payments influenced 10% of people to make a change.

Comparing the two conditions, the “round up” framing increased the number of people who opted in by 3.1% over the “over pay” condition. There was no difference between the dollar amount that people increased their payment. The average user who increased their monthly payment, did so by approximately $60 per month.

We used data provided by EarnUp to calculate the average size, interest rate, and remaining loan size of their users. Using these numbers to approximate, we estimate that we saved the average EarnUp user about $8,000 over the life of their loan. That is approximately $1,300,000 in interest saved for users who opted to round up or overpay in our experiment. Perhaps most impressive of all - 86% of people who opted to overpay had never previously committed to overpaying on their loans. These consumers are now on a new automated path to savings and financial well-being.