Jumpstart founder Kartik Hosanagar, who holds down a day job as the John C. Hower Professor of Operations, Information and Decisions at the Wharton School, said the algorithm frequently strikes gold. It found YouTube shorts created by actor Anna Hopkins, who’s appeared in Freeform’s “Shadowhunters” and “The Expanse” but was unknown — and unrepresented — in the industry as a writer. “We called her and she said, ‘How did you guys find me, did somebody from NBC send you?,’” Hosanagar recalled. “‘We said no, no, no. An algorithm found you.’”
Hopkins had just sold a spec script to NBC around the time Jumpcut’s technology had identified her as promising talent. Now, Jumpcut is working with her. “It shows that the approach works,” Hosanagar said. Jumpcut is one example of how the industry is turning to data and artificial intelligence to discover talent, identify business opportunities, and navigate a landscape dominated by streamers that keep a tight lid on their own insights. Jumpcut is also using AI to demonstrate that “risky” projects from underrepresented creators and talent have an audience; it’s developing a college-set TV comedy series with the popular Facebook group Subtle Asian Traits. “We have been employed to support those overall deals for our writer clients, our director clients, with the streamers,” said Joseph Kessler, global head of UTA’s data division UTA IQ. Launched in 2017, “when we first started we were kind of the shiny new toy,” he said. “Now we’re part of the mainstream fabric of the agency. At any given time we’re working on somewhere between 20 and 50 projects at a time … The vast majority of agents in the company rely on us to provide the insights they’re looking for.” UTA culls its data from social media, ecommerce, and search sources, building a picture of how consumers’ behavior could impact business opportunities. During the pandemic, UTA IQ started using its technology to identify promising musicians online while venues were closed and is expanding that approach to identify digital influencers and actor/creator talent. The data and analytics are so useful in brand partnerships that UTA IQ has started training agents working in that field to pull their own data.
Kessler said book deals are one of the most frequent uses of his division’s services. “We have these amazing publishing agents who know the industry very well,” he said. “We can supplement that by helping them understand how powerful is [a client’s] audience that they currently have and how difficult or how easy would it be to persuade those people to buy a book by those artists that they follow. We have, dozens of times, supported and elevated the compensation that our clients get for book deals by bringing them from one industry into the book industry.” In January 2020, Warner Bros. signed a deal with AI-project management platform Cinelytic, which crunches numbers to model everything from the value of a particular star to how much a film can make in theaters based on various release dates. Cinelytic has since expanded its digital insights into what works on streaming platforms, which supply almost no data to anyone outside the companies themselves. “If you were selling them a title, before it was very difficult to understand: Is that valuable to them or not?” said Cinelytic co-founder and CEO Tobias Queisser. “In negotiations, you always want to understand what the value of your property is to the other party.” Theatrical release strategies can rely on historical data created by the third-party reporting of box-office numbers; projecting the value of a streaming title is less certain. To help ameliorate that issue, Cinelytic turned to an unlikely source: Queisser said ilegally downloaded TV and movies help project performance on legal platforms, and with striking accuracy. The company tracks around 100 million transactions daily on peer-to-peer streaming sites and torrents for pirated titles from platforms like Netflix, Hulu, and HBO Max as well as illegally uploaded theatrical releases. By comparing titles where streamers reported audience data, or by using third-party data providers, Cinelytic created models that correlate illegal content consumption with legal viewing up to 90 percent, he said. Unlike theatrical box office, streaming and pirating share a common currency: time. Consumers have paid their subscription fees for streaming services, and torrent sites are free. By looking at everything from iTunes downloads to theatrical grosses and streaming market-share, Cinelytic also tries to address questions when a release strategy doesn’t work. In a company blog post last month, Cinelytic detailed why Sony’s mid-budget video game adaptation “Monster Hunter,” which grossed $42 million worldwide in its December 2020 release, might have been a better pick for streaming by comparing it to the market share captured by Netflix’s “Army of the Dead.”
“If we’re selling talent and content out in the marketplace and they’re buying, and they’re using data to make those buying decisions, it would stand to reason that we’d want to have at least equally good, if not better data than they have,” Kessler said. “What we’re doing is we’re using data to establish the true value of talent and content.” Sign Up: Stay on top of the latest breaking film and TV news! Sign up for our Email Newsletters here.
