And while market research and Google search data can predict opening weekend performance of movie releases, the industry as a whole has not really capitalized on technology to early on predict which investments will fail, break even, or deliver huge financial returns.
Greenlight Essentials, founded by University of Waterloo engineer Jack Zhang, wants to introduce Hollywood to the predictive power of Big Data. Zhang’s proprietary algorithm, which began as a university research project, analyzes the plot components of movie releases and correlates that information with audience preferences mined from social media, providing what Zhang believes is a magic bullet formula for box office success.
“The general consensus in Hollywood is that film is like venture capital. Most studio investments end up losing money, and many fail because they have misread the audience,” says Zhang. “It is very financially risky to invest tens of millions of dollars in a production without some assurance it is going to turn a profit. Studios have very detailed financial data on their films and their marketing efforts, but what they don’t have is the algorithm to break down the film itself. That’s the key.”
Zhang’s Greenlight Essentials platform leverages an extensive database of plot elements, for instance a father and daughter relationship, then combines that data with other film, related elements, such as genre, and matches the information up with demographic audience data such as age, group, gender and geography to determine the ideal match between storyline and audience preference.
“It is amazing how much data we can gather through social media. What we can learn about people would be impossible even a few years ago,” says Zhang. “But today, people post everything all the time. It allows us to take a real time pulse of people’s likes and dislikes and then feed that information back to them through film storylines.”
Getting Hollywood to buy into the power of Big Data has been a bit of an uphill battle, confesses the young founder. His first discussions with studio executives, held in 2014 at the Toronto International Film Festival were met with a healthy dose of scepticism. The same discussions, however, helped Zhang shape his solution. He was able to refine the interface, making the interpretation and presentation of the data more accessible to the industry. “Now you don’t need a math degree to figure it out,” chuckles Zhang.
To firmly prove the value of his technology to the industry Zhang and his team also decided to deliver evidence of its power in terms Hollywood could understand. They created a trailer for their own horror film, complete with a plot built through machine learning. “We used our software to identify 20 evergreen plot elements, and gave those elements to a creative team. Then we hired a high school co-op and spent about $30 dollars shooting the trailer.” The resulting clip, released on social media has attracted 1.4 million views since it was launched, and more than 50,000 requests from fans anxious to see the full feature. “It’s a film that doesn’t exist, has no cast, no named director. All we know is that the audiences want to see it,” says Zhang.
The ploy worked. The trailer has now piqued the curiosity of Hollywood studios. Zhang has been back and forth to LA over the last few months for meetings and is planning to head back again after the film festival season ends. He is also hoping to officially launch the Greenlight Essentials platform at Cannes this year. Mentorship and funding, made available through the Waterloo Accelerator Centre’s AC Jumpstart program, made possible by FedDev Ontario was instrumental in helping Zhang prepare his LA sales pitch, and allowed him to hire a full stack developer to build out the latest release of his product.
And that horror movie? Well that’s a go as well. Jack Zhang, the University of Waterloo math guru who set out to prove Big Data is the next Hollywoosd wave is now going Hollywood himself. His film, “Impossible Things” is now in production with Concourse Media out of LA, financed by Productivity Media and will begin shooting later this year.