User Interviews

A user interview platform that does the interviewing.

An AI moderator runs the conversation by voice or by text, asks the follow-up you would have asked, and hands back transcripts, themes, and quotes. No calendar invites, no note-taker, no time-zone math.

Free to start · No credit card · 20 free responses every month
AI interview · in progressVoiceText
Moderator

Walk me through the last time you tried to export a report. What were you actually trying to do?

Participant

Honestly I gave up and screenshotted it. It was faster than finding the right format.

Follow-up

“Faster than finding the right format” — where did you look first?

Question 3 of 6Transcribed · coded on submit
The scheduling tax

Six interviews isn’t a sample. It’s what fit in the week you had.

The talking was never the hard part. The hard part is recruiting, finding a slot that works in two time zones, rescheduling the two people who don’t show, taking notes while staying present, and then spending a week turning eleven hours of recordings into four slides.

So teams run six interviews, treat the two loudest quotes as the finding, and move on. Not because six is the right number — because six is what the calendar allowed.

The hard part of user interviews was never the talking. It was the calendar.
How it works

Brief it. Send it. Read it.

You still decide what to ask and what it means. Everything between those two points is handled.

1Step 1

Brief

Write the goal, not the script.

Give the moderator a research goal and the topics it must cover. It builds the guide and adapts the order live.

2Step 2

Field

Send one link.

Participants join whenever suits them, straight from a browser. They pick voice or text — no app, no install.

3Step 3

Read

Themes, with the quotes underneath.

Every session is transcribed and coded the same way. You read themes and open the verbatims behind any of them.

Versus booking calls

Nobody has to be free at the same time.

A scheduled interview needs two calendars to agree. An AI-moderated one needs a link. That single difference is what turns a study from "the six people who could make Thursday" into "everyone we invited".

It also removes moderator drift. The twentieth session gets the same brief as the first, asked the same way, so a difference between two participants is a difference in them and not in how tired you were.

  • Runs around the clock, in every time zone at once
  • No no-shows to chase and rebook
  • The same guide every session — no moderator drift
  • Ten interviews or two hundred, at the same calendar cost
  • Transcribed and coded consistently, not by whoever took notes
Modality

Voice or text — the participant decides.

Voice runs in real time in the browser. People talk instead of typing, so answers come longer, messier, and more useful — which is exactly what you want from a user interview.

Text is for everyone else: the person on a train, in an open-plan office, or who simply would rather write. Same guide, same follow-ups, same coding on the way out.

  • Real-time voice in the browser — nothing to install
  • Text chat with the same adaptive follow-ups
  • Choose the modality per question, or let the respondent choose
  • Fixed-count, AI-decides, or time-based probing
The follow-up

The follow-up is the whole point.

A survey takes the first answer. An interview asks why. The moderator reads what the participant actually said, notices when an answer is thin or contradicts an earlier one, and probes — using the guidance you wrote, not a generic "tell me more".

You can watch it work: every session keeps its full transcript next to the summary, so a theme is always one click from the words it came from.

  • Probes thin answers instead of accepting them
  • Your guidance, applied in every session
  • Attention and quality checks on the way in
  • Full transcript kept beside every summary
What you get back

Findings, not a folder of recordings.

A
B
C
Output · 01

Themes with the quotes underneath

Open-ended answers are coded into themes automatically. Click any theme to read the verbatims it was built from.

Output · 02

Transcripts and recordings

Every session, transcribed and searchable. Keep the recording or set the study to discard it — your call.

Output · 03

A report you can send

The readout your stakeholders were going to ask you to write anyway, with the evidence already attached.

Interview everyone, not just the six who booked.Your first study is free.

Write the goal, send the link, read the themes. Twenty free responses every month, no credit card, no sales call unless you ask for one.

GDPR-compliantNo credit card2-minute setup