Netflix and Big Data: a Recipe for Success

Published by Engagement Labs on July 31, 2014

Here at engagement labs, we love data and we love when companies understand what people really want. So, we decided to examine a data-driven company, Netflix, and its hugely successful series: House of Cards.
According to Mohammad Sabah, Senior Data Scientist, here is a list of what Netflix tracks (1) : more than 25 million users, about 30 million plays per day (tracked every time users rewind fast forward and pause a movie), about 4 million ratings, about 3 million searches per day, geo-location, time of day and week, where users are watching (zip code), what device they use to watch, when they leave content (and if they ever come back) as well as browsing and scrolling behavior. Thanks to its web-based model, Netflix Analysts are very much aware of what people want, watch and like, in real-time. The difficulty is in predicting what will be a success, according to the data we already have, and what risks they should take. “Data can only tell you what people have liked before, not what they don’t know they are going to like in the future” said John Landgraf, President and General Manager of FX Networks.

According to its analytics, Netflix knew that many of its users watched David Fincher’s The Social Network in its entirety. They also knew that the British version of House of Cards was well received and that those who watched it, also tended to watch Kevin Spacey films and/or films directed by David Fincher. Netflix has a lot of users in all three circles of interest and was able to create a Venn diagram intersection that helped programmers take a bet on user interests (2). Big bets are now being informed by Big Data, and no one knows more about audiences than Netflix.

House of Cards was the first time any company had ever used such data in the creative production process for a T.V. show. Another interesting fact is the recommendation algorithm. It understands what each user would like to see based on ratings and behavior. In fact, after having shot the first season, Netflix released several trailers depending on each individual viewer’s taste. Viewers were assigned his or her own trailer featuring different cast members or behind the scenes looks at the creation of the show. This allowed House of Cards to pique the interest of targeted Netflix users and maximize the viewing potential. That is how the magic happened…!

Although it was a roaring success, the series could have generated more conversations online. People love to communicate with each other during screenings, as demonstrated recently with Breaking Bad. House of Cards missed out on a lot of Twitter conversations by releasing all the episodes at once. The screening of a new episode could have created a number of dedicated online communities. Instead, many users found themselves watching the entire series within a week and wondering when the second season was ever going to start…

Orange is the New Black (3) is a similar series which has just been released and is already proving to be a huge success. Netflix used the same process in the creative production and asked Jodie Foster to direct an episode. The difference in this case is that Netflix, now confident in its process, did not have to promote the TV program as heavily as House of Cards (TV, print and online advertising). Other than a few commercials, they barely marketed it at all, relying solely on their recommendation section.

What do you think about using data in the creative process for launching new products? Is big data stimulating creativity or restraining it? Drop us a line and let us know: contact@engagementlabs.com

Source

  1. Harris, Derrick. 14 June, 2012. “Netflix analyses a lot of data about your viewings habits
  2. Carr, David. 24 February, 2013 “Giving Viewers What They Want
  3. 3. Harris, Derrick. 25 July, 2013. “A Netflix, big data can affect even the littlest things.

We used #elisten™ to facilitate our research and instantly find interesting articles and influencers about Netflix / House of Cards / Orange is the New Black.

Follow Us:


For more information on eValue™ or to request a demo, visit www.evaluesuite.com 

For sales contact us at: sales@engagementlabs.com

For media inquires contact us at: media@engagementlabs.com


About eValue™:

The eValue™ score measures the effectiveness of a brand’s overall social media performance on Facebook, Twitter, YouTube and Instagram and provides a score from 0 – 100. It’s the aggregate of the three subscores including Engagement, Impact, and Responsiveness that work together to create one top-level KPI. Our subscores are calculated by using hundreds of sub metrics which are then benchmarked against a hand-picked database of 75,000 verified brand accounts on each channel.

Engagement: Measures the level of interaction generated by your content and how well your community reacts to it.

Impact: Measures how many unique users have potentially been exposed to a piece of content posted by the channel’s admin through organic, viral and paid reach.

Responsiveness: Measures the rate, speed and quality of your responses to fans.