A Case Study: How reframing a feature can make the user and the business better off
eMoneyPool is a sharing community where members take turns borrowing from a common fund to pay for short-term goals.
It’s a centuries old financial practice with roots around the world. When people have trouble accessing capital, they band together and work as a team, not allowing circumstances to get in the way of them achieving their financial goals.
A money pool works like this: Each member agrees to contribute a certain sum (let’s say $100) per month and in a predetermined order, each person in the pool gets to take home all of the cash from one of the meetings. People now have a windfall of $1,000 if the pool has 10 members. It’s basically a zero-sum game, if you will. Members get out exactly what they put in, but they’re able to direct a stream of income on a particular point in time when the funds are most important to them, like to pay their yearly taxes.
The Problem:
Money pools require that a certain number of people have joined before it can officially start. Imagine that the pool has a tipping point. As soon as there are enough people in the pool, the pool can start and the money sharing can begin. Otherwise, the start date keeps getting pushed back, creating a delay for the members of the group and lost revenue for the company.
Logically it makes sense that all new users who come to eMoneyPool for the first time would join one of these ‘tipping point’ money pools in hopes they can trigger it to start as soon as possible.
However, the eMoneyPool product team noticed very quickly that instead of joining an existing pool, new users were creating their own money pools. This meant that eMoneyPool had a lot of incomplete pools on their system that were waiting to start.
Why would people want to start new pools?
Each person in the pool was charged a 5% service fee. A $1,000 money pool would require each person to make 10 payments of $105, regardless of their position in the groups rotation.
All things being equal, if a person was going to join a money pool and pay 5%, they might as well take an early turn and get the money sooner.
Because of this logic, the early spots were usually taken. Instead of choosing a later spot to help fill the pool, they would create a brand new pool and take the early spot from their own pool.
What could eMoneyPool do to solve this?
The existing incentive system was set up to encourage people to take the early positions. If eMoneyPool was going to shift this behavior, they would need to design an incentive that would encourage people to take the later positions.
How should they do this?
To answer this let’s consider Kahneman & Tversky’s 1981 study where they asked people which of these programs would they would adopt?
- Program A: There is a 1/3rd possibility that 600 will be saved. [1]
- Or Program B: there is a 2/3rds probability that 600 people will die.
In both programs, the probabilities of surviving are the same. However an overwhelming amount of people choose program A. Program A is a positive frame and avoids mentioning the possibility of loss.
For eMoneyPool this research suggests that they need to frame the later payouts in a positive frame.
eMoneyPool created a tilted fee structure where early turns would be charged the highest rate of 6% and later turns would be charged the lowest rate of 3%. If you’re willing to wait for your turn, you get charged a much lower service fee.
The behavioral science kicker?
Instead of saying you pay a higher interest rate if you take an earlier position, eMoneyPool says you pay a lower interest rate if you take a later position. This is a positive frame.
After eMoneyPool changed their service fee structure, the problem of incomplete money pools disappeared. Each user now had a different decision to make, do I need the money now, or can I wait and save myself on the service fee? As a result, Instead of the users being spread out over various pools and taking the first spot, everyone was condensing into the open and available pools that needed more people and choosing later spots for a reduced fee.
The changes to the fee structure was recent and statistically significant sample sizes are still in the making, but early results show a significant drop in incomplete pools, roughly 60%, and faster start times for each pool. Which gives the users a better user experience and more active users for the company.
Background:
With a grant from The MetLife Foundation Irrationals Labs hosted 12 financial inclusion startups at their third StartupOnomics conference. At the three day event Irrational Labs co-founders, Dan Ariely and Kristen Berman, as well as PhDs and professors from the nation’s leading behavioral and economics programs helped companies consider how behavioral economics can improve their users’ health, wealth and happiness.