- Netflix is known for relying on data and using analytics to make creative decisions.
- Streamer data is both a powerful weapon and a “flawed science,” former executives say.
- Discussions about show renewals can be “contentious” between creative executives and data analysts.
Netflix famously loves data. The streaming giant she calls herself “a company driven by data since its inception,” one where “analytics arms a company’s decision makers with actionable metrics, insights, predictions, and analytics tools so everyone can be stellar at their job.” And creatives and executives across Hollywood have come to fear and respect the streamer’s vaunted analytics philosophy.
For those who haven’t made creative deals with Netflix, its armies of data scientists and technologists have assumed an almost mythical power, allegedly turning audience data into clear decisions about creative projects.
But the process is much more complicated than that. According to three former Netflix executives who spoke to Insider, as well as others who have done deals there, the streamer’s reliance on data is both a powerful weapon and a “flawed science.” And while the company’s creative decisions stem from collaboration — sometimes “contentious” — between content and data executives and data merchants, the latter have far more influence at Netflix than at legacy Hollywood studios.
How Netflix’s data—better known as the content, strategy, and analytics teams—work
The TV and film executive teams are assigned employees from what Netflix calls its Content, Strategy and Analytics (CS&A) team, a unit “primarily responsible for contract evaluation, programming strategy, performance/audience insights and competitive analysis,” recent Netflix job listing. Essentially, CS&A researchers help creative executives make decisions about whether to buy or produce a particular project, insiders said, and provide an estimate of how much the company should spend on it.
“When you want to buy a project, they’ll do the analysis and calculate it,” one former Netflix executive told Insider.
Compared to similar teams at traditional production houses and Hollywood studios and networks, Netflix’s CS&A executives have more of a front-row seat to creative discussions, insiders say.
At legacy studios, the decision about whether to greenlight a show or buy a movie is “much more instinct,” said another former Netflix executive. At a traditional entertainment company, the person added, market research and valuation of deals would typically fall to a junior creative executive rather than a separate analytics department.
“Sometimes it’s been helpful and sometimes it’s been completely useless,” the executive said of Netflix’s analytics. “If something is really unique or risky, they’re not going to say, ‘Go with God’ or ‘Take that risk.’ They’ll say, ‘Our data doesn’t support it [it].”
He Netflix Tech Blog, the company breaks down how it uses “machine learning and statistical modeling” to help content executives make decisions. The process involves examining similar shows and movies, or comps, that have already been released — a standard practice across the industry — and then plotting those comps on a “similarity map” in which “multiple similar titles appear closer together in terms of spatial distance metrics such as Euclidean distance.”
The company also takes into account how many viewers can tune in and how the title will be presented in other regions of the world – the latter is a key component, given Netflix’s international ambitions.
That can be especially useful when evaluating whether to buy a particular IP, said a third former Netflix executive.
“Some of it is a flawed science,” the first insider said, noting that Netflix has evolved its equation over the years. Once it had more originals under its watch, the company shifted from looking at comps among titles it acquired to comps among titles that were unique to its platform. Streamer has also adjusted the way talent works in creative calculations.
‘Bridgerton’ renewal is a no-brainer, but ‘any other show on the service is a struggle’
The question of whether to extend the series for an additional season is one where discussions can become “contentious,” a third insider said.
Hits like “Bridgerton” produced by Shonda Rhimes or the fantastic “The Witcher” are no brainers. But “every other show on the service is a struggle,” this person said, adding that even shows that have repeatedly made Netflix’s Top 10 list may not garner the watch hours or completion rates considered high enough for CS&A researchers to offer recommendation for renewal. That can sometimes drive a difference of opinion between show hosts and data analysts, especially on shows that have smaller audiences but speak to underrepresented audiences, this person said.
At Netflix, the executive said, “the bar is so, so high.”
The atmosphere is typically collaborative, with CS&A researchers and show execs working together on renewal proposals for, say, Netflix head of US scripted series Peter Friedlander, and sometimes global TV head Bela Bajaria. But sometimes data experts need a little persuasion from showrunners to help a ballooning show get another season.
“It’s, ‘How do we form the image to get buy-in from Peter to get the renewal?’ We work with CS&A to get [Friedlander and Bajaria] to understand why we’re framing the data in a specific way,” explained a third former executive — adding that sometimes an impassioned plea from a prominent show host or actor can make all the difference.
People ultimately hold the reins on whether the project gets built. All that data can be helpful when weighing whether to order a title, but former Netflix executives who spoke to Insider said that each decision — and the risk that comes with it — “is owned by the creative director.”
While Netflix has recruited executives from legacy Hollywood studios like Disney, NBCUniversal and more, some in the industry still see it as a technological distraction, who say it doesn’t give younger content executives the tools to develop their own creative judgment and audience sense.
“These streamers don’t have the proper training,” said another former insider. “Data fills a gap in human knowledge and human experience — the skills you learn when you get at it the traditional way.”