AI User Research

Ask AI personas. Get a score you can rank.

Show a panel an icon, a pitch, an art style or a line of store copy. They answer in their own words, and each answer gets a score on your own Likert scale.

Start free3 studies a month free · See pricing
AI User ResearchIllustrative replayReady

What to test

Ask questions to synthetic users

Each persona answers in their own words. The answer is then compared with your reference set, the statements on the right.

The panel

0 of 5 answered

Scored against your reference set

–

Waiting for the first answer.

How close the answer is to each statement
–out of 5

Panel average

How scoring works

Personas don’t pick a number. They tell you what they think, and we turn that into a score. Here’s one answer from start to finish.

STEP 1

A persona answers in their own words

Marcus is one of five on the panel. He sees the Otter icon and the question, and replies the way a person would.

STEP 2

We read the answer for meaning

Two people can say the same thing in different words. What counts is what they meant, not how they put it.

STEP 3

You define the scale

You write five statements, from “never” to “definitely”. That’s your Likert scale, in your words.

STEP 4

We find where the answer belongs

Marcus sounds most like “I’d probably tap this icon,” with a little pull toward the statements on either side.

STEP 5

You see how sure it is

A hesitant yes and a firm yes are different answers. The score keeps that difference instead of rounding it away.

STEP 6

It becomes one score

Every answer ends up as a score out of 5, so you can compare it with everyone else on the panel.

STEP 7

Each option is scored on its own, then ranked

The panel sees each option separately, so one can’t colour how they see the next. They only meet side by side at the end.

Test anything you can show or say

Put a few options in front of the panel, ask them all the same question, and see which one they like best.

Icons

Pick the icon before you commit to it.

“How likely are you to tap this icon?”

One hand-made puzzle a day. No timers, no lives.

Pitches

Find out if an idea lands before anyone builds it.

“Could you see yourself playing this every day?”

Art styles

Show the same scene drawn a few different ways.

“Does this look like a premium game?”

Restore a seaside town, one tile at a time

Store copy

Test the line people read just before they tap.

“How clearly does this tell you what the game is?”

A panel that sounds like your audience

Tell us who you want to hear from: how old they are, what they play or use, and what makes them quit. Keep the same panel for every study so you’re always comparing like with like.

PR
Priya, 29

Plays puzzle games on the train, ten minutes at a time

MA
Marcus, 41

Lapsed console player, games after the kids are asleep

JU
Jun, 19

Tries whatever is climbing the charts, drops it fast

EL
Elena, 56

Daily word-game player, leaves at the first forced ad

SA
Sam, 34

Works in design, notices when something is cheaply made

Start free.
Pay when it pays off.

Every plan scores answers on your own Likert scale. No features locked away halfway through a study.

Free

See what one study can tell you.

$0
per month
No card needed.
  • 3 studies a month
  • AI User Research
  • Ready-made personas
  • Answers scored on your Likert scale
Popular

Pro

For teams testing every week.

$5$3.50
per month
Billed monthly.Billed $42 a year.
  • Unlimited studies
  • AI User Research
  • Store Page Testing
  • Panels built to match your audience
  • Options compared side by side

Team

For studios running research at scale.

Custom
Priced by team size.
Invoicing and contracts.
  • Everything in Pro
  • Seats for your whole team
  • Custom panels built with you
  • Invoicing and a contract
  • A named contact for support

Know what people think before you build it.