Synthetic users vs real users
Pilot with synthetic.
Decide with real.
Synthetic users are quick for drafting a study. Real people are how you decide. Cookiy supports both and always labels which is which.
A model can tell you what people usually say. Only people can tell you what they actually do.
Synthetic users and AI personas are fast and cheap, and good for rehearsing a study. For decisions you need synthetic participants vs real users handled honestly: pilot on synthetic, then run the real study with verified people, which Cookiy makes almost as fast.
Compared
Know what each is for.
| Synthetic users | Real people on Cookiy | |
|---|---|---|
| Speed | Minutes | Hours: 100+ interviews in a day |
| Cost | Very low | Free to start, no credit card |
| Good for | Piloting a guide, rehearsing, generating hypotheses | Decisions: what to build, price, say or fix |
| Emotion and tone | Simulated, averaged | Real hesitation, frustration and delight in the recording |
| Surprises | Rare: models echo what they have read | Frequent: the finding nobody expected |
| Evidence for stakeholders | Weak | Quotes, clips and counts from real people |
How we use both
Synthetic first.
Real where it counts.
1 · Rehearse
Run the discussion guide against synthetic participants to catch confusing questions.
2 · Field
Launch to verified real participants; the AI moderator follows up where it matters.
3 · Keep them apart
Synthetic sessions are labelled and never counted in findings from real people.
Keep exploring
Related: AI user interviews, Participant recruitment and User research tools guide.
About synthetic users.
What are synthetic users?
AI-generated personas that answer research questions as if they were a type of customer. They are built from language models, sometimes grounded in real data, and respond instantly.
Are synthetic users reliable?
They are useful for drafting and piloting, and unreliable for decisions. They tend toward the average answer, miss the unexpected and cannot show real emotion. Treat their answers as hypotheses to test with real people.
Does Cookiy offer synthetic participants?
Yes, for piloting a discussion guide or rehearsing a study before launch. Synthetic sessions are always labelled as synthetic and never mixed into findings from real people.
When should I use real users?
Whenever the answer will change what you build, how you price it or what you say. Cookiy makes real interviews fast enough that you rarely need to substitute.
What is the best alternative to synthetic users?
Real people at similar speed. Cookiy recruits verified participants and runs AI-moderated interviews in parallel, with first findings in about four hours.
Real people,
almost as fast.
Free to start. No credit card. Video on this site is licensed stock footage of real people.