Making movies the ‘Moneyball’ way
Over the course of April and May, you may see ads for the forthcoming film “World of Warcraft,” based on the popular video game. If you’re the sort of person who plays “Warcraft” or likes movies set in fantastic realms, Matt Marolda will have done his job.
And if you get bombarded by ads for “Warcraft,” but you’re the kind of person who steers clear of movies involving swords, mages, and thrones, Marolda will have failed. His job, as chief analytics officer for the California-based movie studio Legendary Entertainment, is to make sure that when Legendary releases a $100 million film like “Warcraft,” or the now-in-theaters baseball documentary “Fastball,” it gets the most bang for every marketing dollar it spends.
But Marolda’s 50-person “applied analytics” team in Boston, which until now has maintained an extremely low profile, does much more than that. It also advises Legendary on which movies it should make, which actors to cast, and when to release the finished product. (Legendary, headquartered in Burbank, also produces TV shows, comic books, and YouTube content.)
When it comes to building databases about what consumers are interested in and leveraging them to run its business more intelligently, “Legendary is one of the few studios doing a lot,” says Tom Davenport, a Babson College professor who has written extensively about analytics. And while companies like Legendary and Netflix have been setting the pace for how Hollywood uses data to try to make more hits and get more people to see them, this kind of intensive analytical work is something every business is trying to get better at. It’s why “big data” and “predictive analytics” have become such big buzzwords.
Marolda says the Boston outpost dates back to 2013, when he got a phone call from Legendary CEO and founder Thomas Tull. Marolda’s previous venture, StratBridge, had created analytics software for professional sports teams — in part to help them make decisions about which players to recruit and trade. Tull, who is part-owner of the Pittsburgh Steelers, knew of StratBridge from the sports world.
“He called me out of the blue and said, ‘Do you want to do Moneyball for Hollywood?’” Marolda recalls, referring to the book and movie about how the Oakland A’s used statistical analysis to improve their team’s fortunes. “I thought for 30 seconds and said, ‘Yeah.’” Legendary hired Marolda and acquired part of StratBridge’s business.
In the sports world, says Marolda, using analytics “never created a great athlete. But they can help a manager put him in the best position to succeed. It’s the same with entertainment content. You want to put it in the best position to succeed. We’re just trying to increase our odds and be smarter.”
It starts when Legendary is considering whether or not to make a movie. Marolda’s team, which includes people in Burbank and also Beijing, digs through social media posts on services like Twitter and Facebook to determine “is there an audience for it,” he says, and tries to estimate how big that audience is. With the 2015 release “Straight Outta Compton,” about the gangsta rap group NWA, the question was whether people still loved the group’s music from the late 1980s, “and how do people perceive Dr. Dre and Ice Cube,” two of NWA’s founding members. The answers were positive, and the film, with a budget of about $28 million, earned slightly more than $200 million in ticket sales.
The Legendary analytics group also gathers data about historic box office patterns on specific weekends, tries to estimate when other movies will be released, and chooses the optimal date.
But the real focus of Marolda’s team is making sure that the millions of dollars spent on a movie’s marketing campaign are spent wisely. Despite all the progress at creating computer-generated characters and environments, most movie marketing “is a bit archaic still,” says Larry Gleason, formerly a top executive at MGM. Studios spend $50 million promoting a big budget film on television, Gleason says, “because they can’t take the risk of having a major picture not open.”
Marolda is trying to take a more methodical approach — one that involves aggregating vast amounts of data about the entire US population, which Legendary obtains from lenders, retailers, and social networking sites. It’s so detailed, with home addresses, names, e-mail addresses, Twitter handles, and extra-curricular interests, that Marolda acknowledges it can be a bit creepy.
As a movie’s release date nears, the analytics group looks for people who are talking about it enthusiastically on social networks like Twitter and Facebook — those people are likely movie-goers. It then starts making small purchases in online advertising targeted at those people, to see which taglines, trailers, and artwork are most likely to earn a click or a view. Marolda says it amounts to “tens of thousands” of mini ad campaigns to see which groups of consumers respond best to which ads — and who turns out just not to be interested. Then, the analytics group builds bigger campaigns, both online, on TV, and on other media, based on what messages worked in those tests, and what it has learned about the people who responded.
“Even if you haven’t self-declared as a gamer” on social media, Marolda says by way of example, but you share 100 other characteristics with someone who is a gamer, don’t be surprised if you start seeing “Warcraft” ads any day now.
Despite not having a website of its own, or doing any marketing of its services, Marolda says his team has begun doing analytics work for other movie studios, as well as carmakers, sports leagues, and politicians. The Boston office is being expanded; it’ll soon be able to accommodate 100 people, Marolda says.
Of course, there are cases where Legendary’s analytics group discovers that no one is very enthusiastic about a forthcoming film. Marolda says that was the case with “Black Hat,” a 2015 hacker drama that starred Chris Hemsworth. When that’s the case, “you want to batten down the hatches and pull back on the marketing spend,” Marolda says.
Even extreme analytics prowess can’t help a total turkey.