Everyone loves Star Wars. That’s why the franchise is worth over 30 billion dollars. So when Lucasfilm, now a part of Walt Disney Studios, decided on whether to begin production on a sequel in 2013, they were confident it would be a money machine. And they weren’t wrong — Star Wars: The Force Awakens grossed over 2 billion dollars worldwide, launching it to third place on the list of highest grossing films of all time.
This was a no-brainer for Disney, but not all movies are instant hits. Hollywood’s production companies need to be sure that their movies will sell a lot of tickets at the box office before investing millions of dollars in them — The Force Awakens cost Disney more than 300 million dollars. A lot of modeling, data analysis, and financial planning come before launching a film into production.
Pilot, a Boston-based startup with a team of three cinephiles, thinks that it can use big data and a proprietary algorithm to predict how well a film will do long before the film even hits theaters.
Pilot takes in data about a proposed film — inputs like cast, director, writer, composer, budget, release date, genre, and many more — and spits out a prediction for how much the film will gross domestically, both in the first weekend and over the film’s lifetime.
And the model is insanely accurate. Pilot’s median error is 2 million dollars for first weekend box office revenue. To put that into perspective, the latest Star Wars movie grossed $248 million its first weekend. Through machine learning and constant data gathering, every iteration of the model gets better and better.
Cofounder and CEO Alan Xie says he first saw potential for the platform while attending film festivals during his time at Harvard. He met with actors, producers, and investors in the film industry. “A lot of film industry folk were interested in the idea of integrating data into their workflow,” says Xie.
Pilot, now a team of three, had interesting beginnings. “It really began when Alan Rozet and I tried to predict who would win the Oscars,” says Xie. Rozet is the other cofounder of Pilot. “We had a data set that allowed a bit of initial exploration.” But the duo decided they could pivot from predicting the Oscars to actually analyzing the box office.
The team applied for the Deans’ Cultural Entrepreneurship Challenge at Harvard with a prototype of Pilot and won 5000 dollars in grant funding. They’ve been iterating on the model ever since.
Pilot is browser-based, so subscribers can access the database from anywhere. “It’s useful not just for producers but for marketers and insurers — basically anyone who needs to solve a valuation problem,” says Xie. Currently, executives in the film industry pay consulting firms or in-house staff thousands of dollars for analysis of a film in pre-production. With Pilot, subscribers can generate models to their heart’s content. “If the movie has Channing Tatum one week and has Jonah Hill the next week, a subscriber makes that adjustment in the model and instantly gets a new report.”
Pilot works for smaller production companies as well. “We have one of the most comprehensive databases in terms of US theatrical release films, producers, composers, and writers,” says Xie. Production companies with budding stars can benefit from Pilot’s data.
Pilot has international ambitions. Within the next few months, they hope to expand their model to predict box office revenue on a country-by-country and weekly basis. The startup recently turned down an investment of 400,000 dollars from a Chinese firm. “We’ll be profitable next month,” Xie proclaimed. Not bad.
With so much data floating around, almost every industry is using technology like machine learning, big data, and artificial intelligence to break down an overflow of information into something comprehensive and useful. “At the last Sundance Film Festival,” Xie says, “guess which studios made the most acquisitions?”
It wasn’t Disney or Warner Brothers, if that’s what you’re thinking.
“It was Amazon and Netflix,” Xie says. Two data giants.
It’s time for the old-fashioned industry of film to take advantage of the ever-growing quantity of data. Pilot is Hollywood’s big step into the future of data analysis.