Research capture
A note-taking app that lets UX researchers tag interviews while they happen.
Role
Product designer and researcher
Team
1 designer, 2 engineers, founder
Timeline
4 months
Platform
macOS and web
Context
Ennisfold was a two-person startup building tools for user researchers. Their first product recorded interviews but did nothing with them.
Note 1
I joined as their first designer, which meant I was also their researcher, interviewing people who interview people for a living.
Problem
Researchers spent about two hours after every one-hour interview finding the moments that mattered.
Note 2
Researchers wanted to find the moment someone said something surprising, and a transcript made that slow.
Decision
One key press during the interview drops a marker. Afterward, each marker becomes a clip with 20 seconds of context on either side.
Note 3
We tried tags, colors and emoji. A single key won because you can press it without looking away from the person talking.
Outcome
In a four-week pilot with nine researchers, time spent reviewing after an interview fell from about two hours to 40 minutes.
Note 4
Nine people is a small pilot. I say so on purpose, because the founder used these numbers to raise money.
What we cut
Automatic highlights. The model picked loud moments, not important ones, and researchers stopped trusting all the markers.
Note 5
Cutting the AI feature was a hard sell in 2025. The pilot diaries made the case for me.
Before and after
Note 6
Both come from the same interview. The clips can go straight into a team meeting; the transcript has to be read first.
Results
40 min
review time per interview, down from about 2 hours
9
researchers in the pilot
4 weeks
pilot length
Lessons
The feature researchers kept was the one they could use without looking away from the person talking. Next time I would test during live interviews from the first week.
Next project