Designing an AI nail-sizing experience where 96.6% of first-time users measure accurately.
I designed myEdit, an AI nail-sizing app for Kiss Products. By iterating on the tutorial guidelines, I achieved a 96.6% measurement accuracy in the final beta test.
Traditional Kiss Products kits ship 32 nail tips per package to fit anyone. The trade-off is a time-consuming sizing process, leaving 22+ extra tips wasted per kit.
To solve this, we kicked off myEDIT to allow consumers to scan their nail size.
The original tutorial was text-heavy, failing to clearly deliver how to capture the four fingers and thumb. This resulted in a severely low System Usability Scale (SUS) score of 46.7.
Attempting to scan 4 fingers on the tutorial page
Lifting the thumb off the table during capture
I redesigned the graphics for better visual clarity.
However, data showed that users skipped the guide in under 3 seconds.
“Make users watch, not read.”
Focused on guiding users to start their 4 fingers at the table edge, preventing them from lifting their thumb afterwards.
Visually contrasted ‘Do’ and ‘Don’t’ examples to deliver the correct posture.
96.6% Success Rate
In the beta test with 89 participants, 86 out of 89 users successfully completed the measurement process.
