Originals Born of Data Mining: Netflix’s Well-known Secret?
By Joshua Wood
Netflix has an ever growing catalog of films and TV shows, and alongside their success in the streaming business they are recognized by their ability to pump out new and original content each year. In 2016 it, has earned them 8.83 billion dollars, 2 billion dollars more than their previous annual income of 6.78 billion dollars in 2015. Netflix has created a countless number of original series, starting with House of Cards back in 2013 which became an instant hit among Netflix’s audience. However, the success of the show wasn’t a shock to Netflix. The show wasn’t released based on it simply being a good show, but based on data that was gathered by users on Netflix.
Businesses use analytics in order to get an accurate reading of their customers and what they want. In using analytics, companies are able to optimize their marketing strategies and deliver better products to their customers. Netflix takes the data that they receive from their viewers and use it in order to make informed decisions on content that they release. This data gives them an accurate prediction on the success of a show they will release, eliminating any chances of failure from the equation. While on the surface, this may seem cheap, thinking that Netflix will only create original content based on if it’s going to be successful, it’s actually not a bad system. Netflix is creating original content based on user data. Meaning they’re tailor making TV shows for their audiences.
Being an internet company, Netflix has this advantage of forecasting their success. Traditional television networks only have their ratings to go off of, which really are only approximations. Networks have to green-light a pilot based only on intuition and what’s been successful before. Netflix has the ability to take out the guesswork and utilize more accurate predictions when creating their original content.
Now you might be asking, what kind of data is Netflix getting from their users? Well, all the data they gather is essentially the user’s behavior on their platform. The easiest type of data to collect being what shows you watch. If Netflix sees that you’re particularly interested in horror, they’ll recommend you more content in that genre. But that’s at its most basic level. Netflix calculates data such as your “completion rate” of a series. They figure out how many of their users watched a series all the way through. Then, they’ll get a percentage of those that have completed that series and once they have that, they’ll look at those who didn’t complete it and see what their cut off point was and why they stopped watching. Netflix takes this into account in order to learn why users may have stopped it the middle of a series. They take this in order to avoid or correct issues in the content they create, so users are less likely to stop watching. If they know a series is doing well in terms of there being a high percentage of those completing it, then they know there’s a good chance of a new season being a success. But the data they collect goes even deeper than that:
· When you pause, rewind, or fast forward
· When you pause and leave content (and if you ever come back)
· Your ratings for shows
· Searching, browsing and scrolling behaviors
· The date you watch
· Where you watch content, etc.
The list goes on. Netflix uses all this information in order to make decisions on their content. Even to the point where they’ll make ten different trailer cuts for the same show, each tailored to different types of customers. Trailers that you view on Netflix are based on your behavior on their site. If you watch a lot of content with action in it, then you’ll probably see a trailer for a show where it mostly shows the action scenes from that show. If you watch a lot of content with drama, then you’ll most likely see a trailer for the same show but one that shows mostly the drama between characters. Either way, it’s the same show, it’s a just a form of marketing that Netflix uses in order to hook you into watching it. Which makes an interesting point; there isn’t a trailer that’s a ‘one size fits all’ type of deal. You either get turned off by a movie trailer because it features too much action, or doesn’t feature enough depth in its content. This is why we see multiple trailers being put out for a single movie. Nowadays, you don’t see only one single trailer, but ones that are tagged at the end with “…Trailer #1”, “Trailer #2”, “Teaser” and so on.
What users watch truly does influence what type of shows Netflix creates. Before House of Cards was green-lighted, Netflix knew:
· A lot of users watched the David Fincher-directed movie The Social Network from beginning to end.
· The British version of “House of Cards” has been well watched.
· Those who watched the British version “House of Cards” also watched Kevin Spacey films and/or films directed by David Fincher.
Each of these factors had a fairly large and consistent size audience. Netflix used this combined data in order to ensure that the $100 million investment that went into making the show would have a large enough audience. This type of data guarantees Netflix’s growing success and there isn’t any sign of them slowing down.