Media

How Cinelytic is using AI to help Hollywood reboot for the streaming wars


While Hollywood giants have plunged into the streaming wars with massive vaults of content, they still face a yawning consumer data deficit as they try to catch up to industry leader Netflix. Cinelytic wants to help them level the playing field.

The L.A.-based startup has compiled a broad array of data to fuel its platform that helps studios understand in real time how choices ranging from scripts to actors could impact a project’s risk profile and revenue potential. While Hollywood studios have been making bets based on box office data and audience surveys for decades, they still have nowhere near the audience insight that Netflix has at its fingertips.

With subscription-based streaming set to become the primary way consumers discover and experience Hollywood’s content, traditional film and TV producers will eventually be awash in new forms of behavioral data. Studios are starting to turn to AI to help manage and analyze the data in a way that can actually drive more effective and profitable decisions.

“We launched over four years ago with the aim to bring best practices from different industries like finance and tech to build what we call a film business intelligence platform,” Tobias Queisser, CEO and cofounder of Cinelytic, said. “I went into the film industry for two years and I thought there was lack of data and insight to improve how value was assessed.”

VB Transform 2020 Online – July 15-17. Join leading AI executives: Register for the free livestream.

Of course, the creative side of filmmaking has been transformed by the technology and graphics behind special effects and the use of drones for filming. But Queisser said he found that the business side remained stuck in the past. He had previously worked in finance, a world governed by real-time data analytics, before trying his hand at films, where he found himself looking at basic Excel spreadsheets and attending meetings over several weeks to make decisions.

Frustrated by the antiquated methods, he teamed up with cofounder Dev Sen, who had worked at NASA for 15 years building risk assessment models. They eventually brought in Christian Monti, a former studio executive at Paramount and Warner Brothers, to be COO.

Together, they built a platform that offers a handful of different features connected to different parts of the filmmaking process, from greenlighting a project to distribution and marketing.

Underling the platform is data Cinelytic has drawn from different sources. Much of it is historical data that it has licensed from third parties about a film’s performance at the box office, paid TV distribution such as HBO, the rental market, and DVD sales. That data is then matched alongside categories such as the film’s themes, the actors, and the directors.

From there, the platform acts like a fundamental business intelligence dashboard. For instance, sitting in a meeting, an executive can swap in and out the names of different actors for a project to see how that might impact the revenue forecast.

The system also crunches script summaries, character description, and length. Overall, there are 19 attributes the platform analyzes to project its box office performance and other future revenues. In these extremely complex calculations, executives try to estimate how much a film may earn across a wide range of territories, what kind of deals to sign with distributors, and how to create a reasonable budget that maximizes chances of success.

“This tells you what risk you’re taking and what is the marketplace,” Queisser said. “And anything you want to know, you don’t have to wait days for an answer. You just change the parameters and it adjusts automatically. The benefit of the system is that it’s real time.”

Up to this point, studios could likely generate some limited version of this analysis on their own. But the company’s algorithms promise to allow not just for more complex number crunching, but more accurate and faster, Queisser said.

Even if studios could find a way to duplicate some of this analysis, the move into streaming puts them at a disadvantage compared to Netflix and Amazon. These companies have amassed data on users for more than a decade. Typically, studios wouldn’t have access to the underlying data, just the information related to any royalty payments they receive.

As services such as HBO Max, Disney+, and NBCUniversal’s Peacock launch, they face the challenge of paying huge sums to acquire customers. And it will take them years to generate enough data to match that of Netflix and Amazon.

Cinelytic is helping address this issue by gathering its own data on illegal downloads of content on P2P pirate networks. Queisser said the company has found that this offers a highly reliable layer of data that enriches the other sources. Feeding that into the company’s other data on content performance and its algorithms gives producers a window into how a film may perform on their streaming service.

“Nobody really knows what works on the streaming platforms,” he said. “So we monitor illegal downloading of content and found there is a very high correlation between what they watch.”

So far, Cinelytic has raised $2.3 million in venture capital. And in January, it signed a notable deal with Warner Bros. Pictures International.

But to continue making inroads, one of the challenges the company faces is how such data is presented and used by people involved with the creative side of the entertainment business. As studios got bought and rolled up into giant corporations starting in the late 1970s, Hollywood has been increasingly criticized for being too motivated by the bottom line. In recent years, the industry has turned to film franchises, big special-effect tentpoles, and sequels because they offer a measure of predictability.

Queisser said there’s always a risk that AI-driven data analysis will just be used to reinforce the filmmaking culture at studios. But he also argues that better forecasting could liberate producers if they feel more secure that a smaller, more experimental project could make a return with some adjustments in budgets, marketing, or casting.

“If you have more information and can see a broader picture, then the better your decision making can be,” he said. “And that allows you to take more risks because you are more comfortable with the models. But yes, the simplest way to do something is to just replicate something that has been done before.”

Likewise, he emphasized that Cinelytic doesn’t touch the creative side of a project. It doesn’t suggest script or story changes, or how a scene should be filmed. He knows the topic of AI is bound to make actors and directors wary.

“When it’s about algorithms and words like AI, it sounds scary,” Queisser said. “But I think once people understand what the limitations and the benefits really are, then we win them over. This is really about giving them the opportunity to take a deeper dive into the financial side, how to optimize that, see what the risks are, and decide if this is the best way to proceed. We think it’s a very pragmatic use of AI.”



READ SOURCE

This website uses cookies. By continuing to use this site, you accept our use of cookies.