Test-screen any film or show on a digital twin of the US population.

A recruited test screening costs up to $35,000 and takes weeks to organize. So most films get fewer test screenings than the team would like. TestScreen runs the screening on a synthetic audience recruited from a digital twin of the US population, with results within 24 hours.

Because testing is rationed. At up to $35,000 a screening, the trailer cut, the title and the alternate ending usually get decided by the people closest to the film. And the room is whoever turned up locally on a Tuesday night, not the viewers the film was made for. With results back within 24 hours, TestScreen makes it practical to test every cut on the audience you're actually selling to.

Traditionally, by renting a theatre, recruiting a room and waiting weeks for one night of reaction. TestScreen screens your cut on an audience recruited from a digital twin of the US population. You get engagement and eight emotions, every 10 seconds, with what viewers said at each moment, within 24 hours.

Yes. Most of a development slate never reaches production, often for financing or timing reasons, before any audience has seen it. Put the pilot, animatic, sizzle or short in front of an audience recruited from a digital twin of the US population, and within 24 hours you see which stalled projects hold their audience, who they hold, and what to change before you put budget behind reviving one.

Yes. Put the same cut in front of two audiences side by side and see exactly where they split, on one timeline, and what each audience said at that moment.

Yes. Each screening is its own session, so the same audience comes to the new cut fresh. You compare your edit, not two different rooms.

We build about 340,000 synthetic Americans from US Census and ACS microdata, checked against the census to within about 1 percent. Each one has a personality from a trait model that extends the Big Five, a job, and their own news sources. We recruit your audience from that population, so when the real audience skews, ours skews with it.

Yes. Each synthetic person reads their own news sources, and their news is updated every day. A trailer lands with an audience that knows this week's headlines.

The population is checked against the census, then benchmarked against real-money prediction markets and published Gallup and Pew results. The prediction-market benchmark maps where the population is reliable and where it is weaker, so you can see that before relying on an answer. Our validation papers set out the method and its known limits.

We recruit every audience from a standing population calibrated to official statistics, the way you would recruit real viewers. When it skews, we tell you rather than smoothing it out.

A chatbot answers as one voice, from training data frozen at its cutoff. TestScreen recruits an audience of individual synthetic people, each with census demographics, a personality, a job and their own news, and reads how each of them reacts to your cut, every 10 seconds.

Ad testing puts a commercial, promo or campaign in front of a sample of the audience before it runs, to see whether it lands and with whom. It has usually meant recruited panels and a wait of several weeks, so most advertisers test the flagship spot and nothing else. TestScreen for advertising is coming next; join the waitlist and tell us what you would test first.

No. Your cut is what the audience watches. It never trains the model your competitors use, or anyone else's. A model calibrated on your own content is yours alone, and only happens with your sign-off.

Your footage and briefs are used to run your screenings inside your private workspace, nothing more. We do not sell your data. Production runs on AWS, with full DPAs and NDAs available.