We spend refunds more than other money
Because it’s already earmarked for spending and free from obligation, money refunded from a purchase is more likely to be spent over other income we receive.
Yu, T., Cryder, C., & LeBoeuf, R. A. (2024). Refund psychology. Journal of Consumer Research, 51(2), 238-255.
Impact
Refunds are a major, growing headache for retailers, costing $816bn a year (16.5% of all sales).
Lets turn the refund problem into higher retail sales and groundbreaking financial products.
The study
The studies
600 people were randomly assigned to either a payment or refund group and told that they’d received $20 for their research efforts.
Importantly, those in the refund condition were also told that they’d spent this $20 on a shirt, which they’d then returned and got refunded.
Both were then offered the chance to buy some headphones for $20.
Those in the refund condition were much more likely to buy.
600 people were randomly assigned to either a payment or refund group and told that they’d received $20 for their research efforts.
Importantly, those in the refund condition were also told that they’d spent this $20 on a shirt, which they’d then returned and got refunded.
Both were then offered the chance to buy some headphones for $20.
Those in the refund condition were much more likely to buy.
399 people were either given money from a just-refunded product or simply extra money to spend. They were then asked how likely they'd use this money to buy some headphones.
They were more likely to buy headphones using refund money over new money to spend, despite this second group being better off!
400 people were split into two groups, with one presented with money framed as a refund and the other as a tax windfall.
Those in the refund frame were more likely to buy an iPad than those given a tax windfall.
This is because refunds feel more free from obligations than tax windfalls do, because we've already got plans for the windfall.
500 people were told they’d bought a printer that they’d returned and got refunded.
They were split into two groups with the printer needing replacing or not, and were asked how likely they’d use the refund to buy an air fryer.
People were much less likely to spend refunded money from a printer on an air fryer when the printer was framed as needing to be replaced.
This obligation to replace an essential purchase forms a boundary condition for the Refund Effect.
Yu, T., Cryder, C., & LeBoeuf, R. A. (2024). Refund psychology. Journal of Consumer Research, 51(2), 238-255.
Jerome's Expert View
Key Takeaways
Retailers (simple):
Help customers close the loop on an incomplete purchase.
A refund is an unfinished task, causing initial disappointment followed by confusion about what to do next.
Provide recent refunders with a suggested alternative that's desirable, quick to arrive and pre-validated with strong social proof.
Retailers (advanced):
Provide a unique moment to buy an item from our own Wishlist.
In the week after a refund has been fully processed, highlight any items that are on limited promotion to motivate respend.
Ensure that the offer is attractive and priced similarly to the refund.
Banks:
Protect savers from the Refund Effect to help them reach their goals.
To help us control our spending and make refunds feel less like free, spendable money, banks should provide a feature that auto-saves any refunds we receive.
This would supercharge the sort of savings Monzo Bank customers see from their successful Round-ups feature.
Boundary conditions
Accidental purchases aren't earmarked.
Say you bought two sets of headphones, the second accidental pair wouldn’t be earmarked as spending money and therefore wouldn’t be re-spent.
We're obliged to replace essential products upon refund.
Because of this, the Refund Effect works more for discretionary purchases (e.g. clothes) and less for essential ones (e.g. a work computer).
Future questions
Say you got a refund a month ago.
Are you less likely to spend it than if you got it in the past week?
My intuition is yes.
Why? Because over time, the refund will lose its mental accounting tag and blend back into your general balance, if not actively segmented and made salient somehow on, say, a customer account dashboard.
If so, timing really does matter on when you follow up with customers.
Is there a sweet spot for purchase sizes where spending the refund is more likely, like refunds on smaller items costing $10-100?
And is it less likely if the purchase is a large enough chunk of one's monthly salary like, say, a plane ticket costing $1000?
Given its cost, will we even categorise it as 'spending money' if the flight gets cancelled and we're refunded?
These questions aren't covered in the paper, yet the answers could become important boundary conditions.
However, my hunch is that the larger the purchase size, proportionally, in relation to our general income, the less we're likely to respend that refund.
It'll simply have too many alternative obligations attached.
Say you get a refund for some shoes for $40. Are you more likely to spend that money on another item if it's for the same $40 price?
My hunch is yes: retailers are best suited to promote items of equivalent value ($40), primarily to ease cognitive processing.
However, it would be interesting to see implementations that also promote an upsell to, say, a $60 item, because it'll feel like we're only really spending $20.
I do wonder if there's some discomfort experienced in having to return an item, related to a loss of self-identity?
Check out the Self-Expression Deep-Dive for more on this.
If so, is our greater desire to spend the refunded money on a similar or even a superior substitute a way of cancelling out this grief?
We spend refunds more than other money
Because it’s already earmarked for spending and free from obligation, money refunded from a purchase is more likely to be spent over other income we receive.
Yu, T., Cryder, C., & LeBoeuf, R. A. (2024). Refund psychology. Journal of Consumer Research, 51(2), 238-255.
The study
Impact
Refunds are a major, growing headache for retailers, costing $816bn a year (16.5% of all sales).
Lets turn the refund problem into higher retail sales and groundbreaking financial products.
The study
The studies
600 people were randomly assigned to either a payment or refund group and told that they’d received $20 for their research efforts.
Importantly, those in the refund condition were also told that they’d spent this $20 on a shirt, which they’d then returned and got refunded.
Both were then offered the chance to buy some headphones for $20.
Those in the refund condition were much more likely to buy.
600 people were randomly assigned to either a payment or refund group and told that they’d received $20 for their research efforts.
Importantly, those in the refund condition were also told that they’d spent this $20 on a shirt, which they’d then returned and got refunded.
Both were then offered the chance to buy some headphones for $20.
Those in the refund condition were much more likely to buy.
399 people were either given money from a just-refunded product or simply extra money to spend. They were then asked how likely they'd use this money to buy some headphones.
They were more likely to buy headphones using refund money over new money to spend, despite this second group being better off!
400 people were split into two groups, with one presented with money framed as a refund and the other as a tax windfall.
Those in the refund frame were more likely to buy an iPad than those given a tax windfall.
This is because refunds feel more free from obligations than tax windfalls do, because we've already got plans for the windfall.
500 people were told they’d bought a printer that they’d returned and got refunded.
They were split into two groups with the printer needing replacing or not, and were asked how likely they’d use the refund to buy an air fryer.
People were much less likely to spend refunded money from a printer on an air fryer when the printer was framed as needing to be replaced.
This obligation to replace an essential purchase forms a boundary condition for the Refund Effect.
Jerome's Expert View
Key Takeaways
Retailers (simple):
Help customers close the loop on an incomplete purchase.
A refund is an unfinished task, causing initial disappointment followed by confusion about what to do next.
Provide recent refunders with a suggested alternative that's desirable, quick to arrive and pre-validated with strong social proof.
Retailers (advanced):
Provide a unique moment to buy an item from our own Wishlist.
In the week after a refund has been fully processed, highlight any items that are on limited promotion to motivate respend.
Ensure that the offer is attractive and priced similarly to the refund.
Banks:
Protect savers from the Refund Effect to help them reach their goals.
To help us control our spending and make refunds feel less like free, spendable money, banks should provide a feature that auto-saves any refunds we receive.
This would supercharge the sort of savings Monzo Bank customers see from their successful Round-ups feature.
Boundary conditions
Accidental purchases aren't earmarked.
Say you bought two sets of headphones, the second accidental pair wouldn’t be earmarked as spending money and therefore wouldn’t be re-spent.
We're obliged to replace essential products upon refund.
Because of this, the Refund Effect works more for discretionary purchases (e.g. clothes) and less for essential ones (e.g. a work computer).
Future questions
Say you got a refund a month ago.
Are you less likely to spend it than if you got it in the past week?
My intuition is yes.
Why? Because over time, the refund will lose its mental accounting tag and blend back into your general balance, if not actively segmented and made salient somehow on, say, a customer account dashboard.
If so, timing really does matter on when you follow up with customers.
Is there a sweet spot for purchase sizes where spending the refund is more likely, like refunds on smaller items costing $10-100?
And is it less likely if the purchase is a large enough chunk of one's monthly salary like, say, a plane ticket costing $1000?
Given its cost, will we even categorise it as 'spending money' if the flight gets cancelled and we're refunded?
These questions aren't covered in the paper, yet the answers could become important boundary conditions.
However, my hunch is that the larger the purchase size, proportionally, in relation to our general income, the less we're likely to respend that refund.
It'll simply have too many alternative obligations attached.
Say you get a refund for some shoes for $40. Are you more likely to spend that money on another item if it's for the same $40 price?
My hunch is yes: retailers are best suited to promote items of equivalent value ($40), primarily to ease cognitive processing.
However, it would be interesting to see implementations that also promote an upsell to, say, a $60 item, because it'll feel like we're only really spending $20.
I do wonder if there's some discomfort experienced in having to return an item, related to a loss of self-identity?
Check out the Self-Expression Deep-Dive for more on this.
If so, is our greater desire to spend the refunded money on a similar or even a superior substitute a way of cancelling out this grief?
Yu, T., Cryder, C., & LeBoeuf, R. A. (2024). Refund psychology. Journal of Consumer Research, 51(2), 238-255.
We spend refunds more than other money
Because it’s already earmarked for spending and free from obligation, money refunded from a purchase is more likely to be spent over other income we receive.
The study
600 people were randomly assigned to either a payment or refund group and told that they’d received $20 for their research efforts.
Importantly, those in the refund condition were also told that they’d spent this $20 on a shirt, which they’d then returned and got refunded.
Both were then offered the chance to buy some headphones for $20.
Those in the refund condition were much more likely to buy.
Yu, T., Cryder, C., & LeBoeuf, R. A. (2024). Refund psychology. Journal of Consumer Research, 51(2), 238-255.
Scarcity
We value things more when they’re in limited supply
Social Proof
We copy the behaviors of others, especially in unfamiliar situations
Prospect Theory
A loss hurts more than an equal gain feels good
Reciprocity
We’re hardwired to return kindness received
Framing
We make very different decisions based on how a fact is presented
Loss Aversion
We feel more negative when losing something than positive when we get it
Self-Expression
We constantly seek out ways to communicate our identity to others
Default Effect
We tend to accept the option pre-chosen for us
Priming
Our decisions are shaped by memories recalled from things just seen or heard
Anchoring
What we see first affects our judgement of everything thereafter
Scarcity
We value things more when they’re in limited supply
Social Proof
We copy the behaviors of others, especially in unfamiliar situations
Prospect Theory
A loss hurts more than an equal gain feels good
Reciprocity
We’re hardwired to return kindness received
Framing
We make very different decisions based on how a fact is presented
Loss Aversion
We feel more negative when losing something than positive when we get it
Self-Expression
We constantly seek out ways to communicate our identity to others
Default Effect
We tend to accept the option pre-chosen for us
Priming
Our decisions are shaped by memories recalled from things just seen or heard
Anchoring
What we see first affects our judgement of everything thereafter