Item details: sheets → one step at a time
Bottom sheets felt complicated in bright sun. After testing with associates, each question became its own step, and Add item only turns on when the item is complete.










The bigger picture · one Returns program, two channels
Kiosk and curbside weren't separate ideas. Both tested the same question: will members handle a return their own way, if we meet them where they already are? Each started as a scrappy pilot on Express Returns, the app associates already used, so we could learn fast before building anything new.
32%of eligible returns, double the 15% target
Proved members are happy to finish a simple return on their own, without waiting in line for an associate.
10%adoption in 5 weeks, across 3 clubs
Proved members will return without leaving the car, when an associate brings the return to them.
Nextwhat both pilots point to
Pick items, answer questions and see the refund in the app, then just drop off at the kiosk or the curb.
A return still meant parking, walking in and waiting at Member Services. Members already used curbside to get products into their car. Curbside returns started as my idea: could the same curbside take products back?
Prove the behavior before building the technology.
Associates used the Express Returns iPad app from Member Services. One email told pickup members returns had moved to the curb.
I adapted Express Returns into an Android app for the handheld associates already carry for pickup orders.
Tradeoff: Phase 1's iPad app wasn't built for one hand outdoors. We accepted that to prove demand first.
What changed once associates used it at the curb.
Bottom sheets felt complicated in bright sun. After testing with associates, each question became its own step, and Add item only turns on when the item is complete.




Lead with the membership card. One scan pulls up the member's purchases. Receipt and order search stay one tap away.



A refund can't be undone at the curb. The method shows speed upfront, then one explicit confirm before it goes through.



From trunk to refund, without leaving the car.
MemberParks in a Returns spot and checks in.
AssociateHeads out with the pickup handheld.
Pilot: QR, text or call. Future: app check-in.
MemberShows their membership card.
AssociateScans it to pull up purchases.
MemberHands over the item.
AssociateScans and matches it to the purchase.
Future: know what's coming before reaching the car.
MemberPicks where the money goes.
AssociateConfirms the refund at the car.
MemberDrives away. Receipt by email.
Use the arrows or scroll sideways · tap any screen to enlarge
Prototype screens recreate the real experience with dummy data, so some names, prices and item numbers are placeholders.

















A low-tech pilot proved demand and showed what was worth building.
Clear arrival steps mattered as much as a fast refund.
The frustration wasn't only wait time, it was control. Members wanted chores that fit around their day.
Product principleDon't make members come to the return. Bring the return to them.
Designed for the future too: app check-in, automatic eligibility, arrival detection and wait-time visibility.
Right after deploying, we talked to associates and surveyed them on training time, ease of use and time to complete a return.
Next case studyA five-minute return only feels fast when every minute feels predictable.
Some of this work is confidential, so I share the details one-on-one. Happy to walk you through it.