I wanted to figure out whether adding a physical component to a typically digital interaction would make the experience more engaging or just add friction. Spoiler: it's complicated.
Initial prototype testing the concept
"The gap between 'novel' and 'useful' is real."
what worked
The physical component did increase engagement in testing — people spent more time with it and remembered the interaction better. Gestures felt natural once learned. The 'magic' moment when physical and digital connected worked surprisingly well.
Successful interaction moment during testing
what didn't work
Onboarding was a nightmare. People didn't know what to do without heavy instruction. The physical object felt gimmicky to some users. Technical limitations around gesture recognition caused frustration when it didn't work perfectly every time.
One of many failed attempts
learnings
The gap between 'novel' and 'useful' is real. Physical interactions need to solve an actual problem, not just be interesting. Also learned a lot about the technical constraints of sensor accuracy — what works in a demo doesn't always work in real conditions. Would try this again with a clearer use case in mind from the start.