| Role | Product Designer (Contract), research + design end to end |
| Team | Me, 1 PM, Engineering |
In a NutshellReshop's instant refund had always been free. The business needed to start charging for it without losing the customers who made the feature successful.
I researched how customers would react to a fee and found that when they discovered it mattered more than the amount itself. That led me to recommend showing the fee upfront, even knowing it could hurt conversion.
It did, but less than expected. Conversion dropped ~2.7% relative to control, the fee generated ~$1,300 in its first week, and most customers who didn't want to pay chose the free bank option instead of leaving.
The challenge
Reshop handles returns for a bunch of DTC brands: Steve Madden, Alo Yoga, Dolce Vita to name a few. Reshop gives customers their refund back to their original payment method almost immediately, instead of making them wait for the merchant to process the return.
Until this point, that experience had been free. Reshop now needed to find out whether customers would pay for the convenience, and what introducing a fee would do to conversion.
I focused on three things:
| Pricing elasticity | Disbursement mix | Messaging |
|---|---|---|
| Pricing elasticityHow much would people pay, and how does demand shift as the price goes up? | Disbursement mixIf OFP costs money, where do people go instead? What does that do to conversion and retention? | MessagingHow do we present a fee without it feeling like a bait-and-switch? |
Research
Before changing the live experience, I ran an unmoderated study with 27 participants through Maze. I wanted to understand what people were willing to pay, how they expected the fee to work, and how much explanation they needed before making a decision.
I went in thinking we could keep "Learn more" focused on the benefits of instant refunds and introduce the fee later. The research changed my mind.
What changed my direction:
Timing mattered more than the amount.
Most participants already expected instant refunds to cost something. A $1–$2 fee on a ~$90 refund felt reasonable. What created friction was discovering it late.
People wanted the fee upfront, with more detail available when they needed it.
I tested three ways of presenting the fee. 72% preferred seeing the amount upfront with the explanation one tap away.
The free option needed context.
Participants didn't understand why card refunds had a fee while bank refunds were free. Combined with low familiarity with Reshop, that made clear communication especially important.
What I tested:
Key findings:
| Timing beats amount | Card vs. bank felt arbitrary | Bank visibility |
|---|---|---|
| Timing beats amountPeople expected a fee going in. What broke trust was finding it late, not the price itself. | Card vs. bank felt arbitraryWithout an explanation, a fee on card but not bank read as random. | Bank visibilityKeep a free alternative easy to find. |
Exploring more directions with AI
Once research pointed me toward "disclose earlier, explain why," I used AI to explore more directions and expand the range of ideas quickly.
I explored different ways to frame the fee, build trust, and decide how much information to show upfront versus on demand. From there, I compared the strongest directions against what I'd learned in research and narrowed them down before testing.
Design
The research reframed this from a pricing problem into a transparency problem. Two decisions shaped the final experience.
Decision 1: Show the fee upfront, even if it hurt conversion
Research showed that people weren't particularly bothered by the $1–$2 fee. The bigger issue was discovering it after they'd already started down the instant refund path.
Showing the fee earlier came with a tradeoff: putting a price in front of customers sooner could reduce how often they chose instant refund. But hiding it until later created the exact problem we'd seen in research.
I recommended keeping the fee upfront, even if it meant accepting some conversion loss.
The tradeoff:
| Gave up | Gained |
|---|---|
| Some customers would reconsider instant refund once they saw the cost. | A more transparent experience that addressed the trust issue we repeatedly saw in research. |
There was also an integration constraint. The refund card itself was rendered through Narvar, so I didn't control its layout. I owned what we told customers about the fee and when we told them.
The fee moved earlier in the flow so customers knew there was a cost before continuing to Reshop.
Decision 2: Keep the free option available without making it the default
We still needed a path for customers who didn't want to pay for an instant refund.
Rather than giving card and bank refunds equal visual weight, I nested bank under a "no-fee option available" expander. This kept the free alternative one tap away without making it compete directly with the experience we were trying to monetize.
The Plaid linking experience itself was third-party, but I designed the steps around it. Adding a new payment method in the middle of a return was already a significant ask, so I kept the path as short as possible.
The free bank option stayed easy to find without competing for the same visual weight as the original payment method.
Fee tooltip copy. The fee treatment also carried through from testing: the amount stayed visible upfront, with an explanation available through the info icon.
Impact
The team ran a controlled A/B test on Steve Madden traffic, with a 2.5% fee on original-payment refunds while bank refunds stayed free.
The conversion drop was smaller than the team expected. More importantly, most customers who didn't want to pay didn't leave Reshop. They moved to the free bank option instead.
Revenue from customers who paid for the instant option more than offset the shift to bank, making the fee profitable from the start.
What the results showed:
| Profitable from day one | Bank absorbed the shift | Upfront disclosure held up |
|---|---|---|
| Profitable from day oneFee revenue more than covered refunds that moved to the free bank option. | Bank absorbed the shiftMost customers who avoided the fee stayed in Reshop and switched to the free bank option instead. | Upfront disclosure held upThe final conversion impact was smaller than the early results suggested. |
The results gave the team enough confidence to move into a second phase, testing the model with more merchants and exploring higher price points.