When it came to creating an original show, movie streaming company Netflix reached beyond the usual methods of trying to come up with a concept and cast that people wanted to see.
Because, in truth, “usual methods” means guessing. An educated guess, perhaps, but a guess nonetheless.
Netflix thought they could do better, so they turned to data analytics for guidance. Researching the massive amounts of data they had on movies and television shows people had chosen to watch, they looked at what people watched all the way through and any comments they made about what they watched. With more than 93 million users, they had a mammoth database to work with.
They found this:
- Director David Fincher’s movies – such as “The Social Network” and “Fight Club” – were popular and often watched all the way to the end
- Actor Kevin Spacey remained popular among viewers in a variety of films across many different genres
- The original British series “House of Cards” had been a popular choice for Netflix users.
Netflix got the rights to remake the British political thriller and hired Fincher to direct and Spacey to star. It became the most streamed content in the United States and 40 more countries, according to the New York Times.
That’s a hit. And it all came about because of predictive analytics.
Netflix seems more purely data-driven than many companies, particularly in the entertainment sector. With the large amount of data they have on viewers watching habits, they have an advantage over most companies.
That’s more information than broadcast television network executives have when deciding which television show to green light.
And it’s not just what people watch. It’s also when they watch it, whether or not they complete it, at what point did they stop and a host of other data points. In other words, they can tell you how many people started watching “The Sopranos” and in which season – or even what episode – a significant number of viewers bailed and never finished the series.
But there’s even more. Netflix also knows:
- When users paused and fast forwarded
- When users watch certain kinds of content (TV shows vs. movies, for example)
- What device they use (TV, computer, tablet, phone)
- The ratings viewers give each show
- How fast people binge watch shows. Horror and thrillers are the fastest – a whole season in just four days. However, even the slower ones – such as sophisticated comedy “Arrested Development” – take just six days.
In addition to the “House of Cards” example of using data to pick what property to produce as an original show, Netflix also uses data in other ways.
This is something that other sites (Amazon, for example) do routinely, but the key here is success in making recommendations. It’s already an old 21st century joke about the ridiculous recommendations you used to get from some sites after watching a certain type of movie. However, the algorithms have become more sophisticated, so much so that Netflix reports 75% of viewings on Netflix are driven by recommendations.
Deciding On Movies
Netflix has to pay film companies to show their movies, and they use analytics to determine what movies are worth getting. They use many of the same data points used in the “House of Cards” examples. What stars are popular? What type of television shows do well in the long run?
Netflix also has moved into buying film distribution rights, and most assume they also are using analytics to determine which movies to buy. Part of the reason is the eclectic mix. Last year, that included “Tallulah,” a movie starring Ellen Page; “The Fundamentals of Caring,” starring Paul Rudd and Iranian horror film “Under the Shadow.”
As noted above, the huge collection of data that companies such as Netflix and Amazon possess offers an advantage over traditional media companies.
That has yet to change. Netflix and other digital companies do not share their information with movie companies or the media in the way theaters share box office numbers. Because of the potential use of the data to help them in developing projects, some in Hollywood hope to make the digital numbers public.
However, that change might require the guilds that represent writers and other film workers – who could receive more money based on higher viewership – to fight for transparency in digital revenue numbers.
That doesn’t look like it will happen anytime soon. For the near future, expect Netflix to continue using data analytics to its advantage and for Hollywood and television executives to clamor for more access to the company’s large amounts of data.