For companies looking to reach consumers online and through social media, the ultimate success is for their ads to “go viral” — that is, to be passed around the Web on a scale that surpasses anything a conventional marketing or ad campaign could hope to achieve, thereby burrowing deep into our collective consciousness with minimal expense.
But of course, if going viral were easy, everyone would do it. In fact, it’s proved all but impossible to manufacture viral content on demand — or even to predict with any degree of accuracy which videos are most likely to be passed around.
But the folks at Unruly Media think they’ve now figured it out.
Founded in 2006, Unruly Media is a U.K.-based video-technology company in the business of predicting virality. Last year, revenue for Unruly tripled — from about $9 million to $27 million, and the company has been expanding its global reach with offices now in Amsterdam, Sydney and New York City, where it recently launched its new “social video lab.” The space is designed to showcase the technology the company uses to predict viral videos.
Unruly uses facial-tracking technology to determine when viewers of a video are feeling a range of emotions, from happy to sad to distressed or angry. That real-time information is used in combination with traditional focus-group questions after a video is viewed. Cat Jones, Unruly’s head of business development, says by using those tools, her team can make predictions with 80% accuracy.
Last week, TIME visited Unruly’s social video lab to check out the process for ourselves. Our conclusion? This isn’t yet a pure science, but predicting virality is getting much more sophisticated. Check out the video above to see for yourself.