RETAIL KIOSK / UX DIRECTOR + PROJECT LEAD
I redesigned a self-service kiosk so more people could trust it enough to finish.
MinuteKey asks someone to trust unfamiliar hardware with an everyday object they cannot afford to get wrong.
I led the redesign around one idea: reduce uncertainty at every step, from choosing a key to removing the finished copy.

01 / THE TRUST PROBLEM
The interface had to earn confidence before asking for commitment.
The previous experience exposed too much system complexity at once. Customers were being asked to understand the machine while also making an unfamiliar purchase.
Cleaner screens were only part of the answer. Each decision also needed to feel safe, legible, and reversible.

02 / THE STRATEGIC MOVE
Use the system’s wait time to build confidence, not test patience.
Smart and car keys require background processing. I used that unavoidable wait to explain price, quality, accuracy, and the white-glove service without adding another step.
- 01SIGNAL
The wait created doubt.
Customers were asked to trust an expensive service while the kiosk appeared to be doing nothing.
- 02DECISION
Make waiting useful.
Put the most reassuring proof inside the processing time instead of adding more screens.
- 03EXPERIENCE
Confidence without friction.
The product explains the service while continuing to locate and configure the key.

03 / ONE KIOSK, MULTIPLE SERVICES
The redesign had to make every kind of key feel understandable.
Car keys were only one part of the system. I also designed distinct paths for smart keys, RFID fobs, and house keys. Each came with different hardware, fulfillment, pricing, and moments of risk.



04 / LEADERSHIP + COLLABORATION
I pushed us to design for a kiosk instead of putting a website on a bigger screen.
I owned the end-to-end experience and wireframed the full journey, then partnered closely with a design-system engineer on the UI. A UX writer helped shape the service story, while our illustration and animation partners made the physical instructions and feedback easier to understand.
The 16:9 landscape surface was the defining constraint. I advocated for large, obvious touch targets and an interface that used the available space instead of shrinking familiar web patterns into it. We used the client’s research and interactive prototypes to show the behavior, animation, and pacing. That helped people understand the experience in a way static screens could not.
That mattered most for smart keys, which could cost around $600. We had to explain the white-glove programming service and make MinuteKey feel as credible as a dealership before asking for the purchase.
05 / FROM FLOW TO SYSTEM
Turn one complicated kiosk into a sequence of obvious decisions.
The system established a common rhythm across traditional keys, transponders, smart keys, key fobs, and house keys. Each flow could carry different information without changing how customers understood progress.
ORIENT
Make the available key types immediately understandable.
NARROW
Turn vehicle selection into one clear decision at a time.
REASSURE
Use unavoidable wait time to explain quality and value.
CONFIRM
End with an unmistakable receipt, delivery promise, and exit.
Follow the connected kiosk journey from service selection through purchase.


06 / WHAT CHANGED
The core experience improved. Smart-key conversion moved from zero.
Our partner reported key-flow completion of 70% after launch, up from 62%, and car-key completion above its goal. Smart-key conversion reached 3% from zero. It remained below the 10% target, but established that people would purchase a high-ticket service through the kiosk. These are partner-reported results, not a controlled experiment.
- 70%
- KEY-FLOW COMPLETION UP FROM 62%
- 33%
- CAR-KEY COMPLETION VERSUS A 20% GOAL
- 3%
- SMART-KEY CONVERSION UP FROM 0% · 10% TARGET
07 / REFLECTION
I would establish the confidence baseline before redesigning the next flow.
The project showed that clearer sequencing and well-timed reassurance could improve completion. What it did not capture was a strong before-and-after measure of customer confidence.
That is the next layer I would add: measure perceived risk before launch, then connect the behavioral result to what customers actually felt.