I found the Alexis Madrigal article, “How Netflix Reverse Engineered Hollywood”, to be the most interesting and relevant reading for this week. The way in which Netflix classifies and categorizes their movies and TV shows is the perfect example of metadata. The article describes how Netflix pays their employees to watch films and tag them with all kinds of metadata and specific genre categories. Then through algorithms that the company has developed, these specific genres and sub genres come together to create extremely specified and tailored search results and suggestions for Netflix subscribers.
This is where it gets interesting… You see, it’s not just like any old search engine where they have some sort of complicated algorithm that helps sort through endless amounts of metadata (in this case, movies and TV shows). Rather, the unique algorithm of Netflix learns what you like to watch based on your recent viewings, automatically searches through a seemingly endless amount of media, and then presents that media to you in an organized form represented by “altgenres” that Netflix has created to classify and sort very specific types of shows and movies.
The genius of Netflix is the great success this feature has in retaining subscribers. Netflix revealed that “members connect with these [genre] rows so well that [they] measured an increase in member retention by placing the most tailored rows higher on the page instead of lower”. The more accurately Netflix can interpret your likes and interests and then present those likes and interests back to you, the more likely you are to continue enjoying watching movies and shows on Netflix. As Madrigal simply put it, “The better Netflix shows that it knows you, the likelier you are to stick around.”
With such an intuitive and revolutionary algorithm, however, it’s easy to forget at what price these luxuries are coming at. What exact kind of information does Netflix pull from us to determine our likes? Is it simply our interests and how many times we’ve watched a certain show or movie? Or does Netflix go further than that, collecting data from us such as location or relationship with other Netflix subscribers? For the most part I’m just hypothesizing—I don’t want anyone to think that I’m being paranoid—but I do think it’s worth thinking about. It’s the same concern we have with search engines and what kind of boundaries they’re setting on our own personal privacy on the web. After doing some research, actually, I did find that Netflix was involved in a couple of lawsuits involving possible violation of user privacy:
On another note, I thought that there was a neat connection between Netflix and the Nathan Yau reading Data Points: Visualization That Means Something. Yau talks about the importance of interpreting and presenting data in a visually pleasing way. I did a quick search to compare the design layout of Netflix from when it was first beginning in the 1990s and what it looks like today. You can see from the two images below that Netflix’s organization of data has improved drastically since the company first began. No doubt this has affected the amount of subscribers that Netflix has retained.