More interviews
than classic IDIs
50x
SOUNDS FAMILIAR?
You have the why. You’re asked for how many.
“Twelve interviews. Leadership asks if that’s representative.”
Your insight is solid. Your sample size makes it easy to wave away. curly runs the same depth at n=500+, so every pattern comes with a percentage.
“Three weeks of fieldwork, two weeks of coding, and the deadline was yesterday.”
Recruiting, scheduling, moderating, transcribing, tagging. curly runs the interviews in parallel and codes them as they come in.
“Only the people who have time on a Tuesday at three.”
Scheduled interviews skew toward whoever is available. With curly, people take part whenever it suits them, from home, in almost any language.
LIVE PROBE · AFTER “IT FEELS A BIT CHEAP”
Every answer gets the follow-up you’d have written into your guide. Laddering, clarifying, projective prompts. Always neutral, always the same care.
THE CORE OF THE METHOD
Your guide.
500 conversations.
You design the discussion guide like you always do: topics, stimuli, probing goals. curly moderates every interview along it, follows up on what each person actually says and brings the conversation back when it drifts. What changes is the math. The same interview, 500 times, at the same time.
Deep
Consistent
Scalable
Countable
THE DETAILS YOU’LL BE ASKED ABOUT
Everything below runs automatically. Every step stays in your hands.
Recruiting
THE RIGHT PEOPLE, AND MORE OF THEM
Recruit from any panel, your own CRM contacts, or a shareable interview link. Screeners and quotas as usual. Participants join from their browser with audio and video. No app, no scheduling, no no-shows.
Discussion guide
YOUR STRUCTURE, YOUR PROBES
Open questions, laddering, projective and scenario tasks, plus ratings where you want them. Show concepts, packaging, ads, videos, live websites or clickable Figma prototypes, identically to everyone.
Moderation
A MODERATOR THAT FOLLOWS YOUR RULES
Neutral, non-leading, with defined probe depth. It notices hesitation and contradictions and asks again. You review the guide before launch, watch interviews as they happen and adjust between waves.
Quality control
YOU CAN’T FAKE A CONVERSATION
Every interview is quality-checked automatically and in real time: consistency, engagement, rushing, fake answers. Voice, tone and non-verbal signals raise the authenticity bar far beyond attention checks.
Flagged interviews are excluded before they touch your data.
Analysis
FROM TRANSCRIPT TO THEME
Transcripts, codebook, themes and key quotes are generated live as interviews come in. Because n is large, every theme comes with a count: how many said it, in which segment, and whether the difference is significant.
Traceability
EVERY THEME HAS A FACE
Each finding links back to the clips and verbatims behind it. Put together a showreel for the management meeting in minutes.
QUAL · QUANT · CURLY
Depth
Follow-up questions
Non-verbal signals
Uncovers the 'Why'
Scale
Large Sample Size (n=500+)
Low / no manual effort
Fast turnaround (in days)
Low cost
Output
Qualitative analyses
Quantitative analyses
Statistical confidence
Real quotes
Fraud detection / panel authenticity
METHODS FAQ
Can an AI really probe as well as an experienced moderator?
Is n=500 still qualitative research?
Who owns the discussion guide?
Do I get the raw material?
What about focus groups?
Where does the data live?
We'll run it as a real study. Same guide, hundreds of conversations, full analysis. Then put it next to your last interview project and compare.
Get curly
