For marketing, comms and research teams
Know which message lands before you spend the media budget.
Put ad copy, headlines, emails or announcements in front of a simulated audience built to resemble yours. Get a ranking, the reasons behind it and how it differs by segment, in minutes rather than weeks.
Join the waitlistHow it works
Bring your options and your audience
Variants of copy or creative, and a description of who you're trying to reach. Add your own survey or campaign data to ground it.
A simulated audience reacts
Each simulated person compares options, says which they prefer and why.
Get a ranking with honest error bars
Which options are clearly ahead, which are too close to call, and how segments disagree.
Where simulation is strongest
Ranking text options is where AI-simulated audiences have the best published track record. In a study of 70 survey experiments, published in Nature in 2026, GPT-4’s predictions of which treatments would work correlated strongly with the real results (Stanford GSB).
The same research shows the limits: simulations overstate how big effects are, and are weaker on real-world behaviour than on survey answers. So we use simulation for what it’s good at, which is narrowing twenty options to three, and we show you how much to trust each result.
Join the waitlist
We're building this now. Tell us what you'd simulate and we'll reach out as early access opens. The most interesting use cases get in first.