A researcher, on average, watches 200 clips daily. That’s a thousand videos each week, or 50,000 per year, give or take. The company’s research is aided by proprietary software called Riff, which generates feeds based on niche viral keywords. Maybe you’ve seen 15 or 20 good videos of babies tasting lemons for the first time, or troops surprising mothers at Christmas, or dogs and parakeets becoming unlikely friends. A Jukin researcher has seen these, too, plus all the duds that never made it to your feed.
Of those 200 reviewed on a typical day, perhaps three or four may be good enough to license, generally for a fee between $50 and $5,000, and often a revenue split. Researchers contact clip owners by any means necessary — usually through YouTube or Facebook but occasionally over dating sites, or wherever else they can be found. In all, the company has paid out more than $10 million in royalties to video owners.
Its most successful clips tend to fall into one of three genres: “cute,” “fail” and “win.” “Cute” is dancing toddlers, puppies and clumsy kittens, anything involving that optimal eye-size-to-head-size ratio. “Fail” is whatever tenses the shoulders in empathy — wipeouts, face plants, schemes gone awry. I’m told the best fails are those that fail in unexpected ways, denying Chekhov’s gun — e.g., the skateboarder doesn’t eat it on the ramp, but then a guy in the background is kicked in the nuts. “Win” is whatever affirms faith in humanity, so women giving birth in cars and on airplanes, and also people catching marshmallows in their mouths from far away.