I updated Sergio’s below item with most of this, but figured I should give it a post entirely of its own, given how long it is.

I reached out to Netflix to ask about how they categorize their titles as they come in. I haven’t received a reply yet, but I did a little digging to find past posts on Netflix’s tech blog, as well as interviews Netflix engineer executives have given in which they were asked about categorization of their films, and this is what I learned about their process.

In short… from a Wired magazine interview, Netflix says it has over 40 people "hand-tagging" TV shows and movies for them as they come in. These are typically freelancers who are TV and film buffs, some with experience in the entertainment industry, who are hired for this very specific purpose. Their job as "analysts" is to be objective, and they are trained accordingly.

Adding to the Wired piece, the Consumerist website published an item on one of those so-called "analysts" as the Netflix exec calls them – a woman who makes some part-time income a movie tagger for Netflix. This is a portion of the piece: "She says she gets a list each week from Netflix HQ to let her know which handful of titles (around five at a time) she is supposed to watch and help to categorize. While watching, she uses a spreadsheet to take notes of all the info that goes into the 100+ data points used by Netflix to tag each title. “It covers everything from big picture stuff like storyline, scene and tone, to details of whether there is a lot of smoking in the movie,” the part-time tagger explains. “We’re looking for people who have knowledge of movies and TV shows,” explains Todd Yellin, vice-president of product innovation at Netflix. He explains that the woman… is a good fit because she is not just an independent film maker, but has also worked as a script supervisor — a very detail-oriented gig. Potential taggers must successfully pass a test before being hired… For their efforts, taggers earn a few hundred dollars a week."

Now while the above explanations may not directly address the "race" question, we can assume that, among the list of criteria that these movie taggers hired by Netflix are given to use in their categorizing of movies they watch, the race or ethnicity of the characters in each film, as well as the kind of story it tells, are likely somewhere on that list.

For example, if it’s a "black film" (as in whether the film tells a story about black people), the movie taggers probably check a box on the form Netflix provides them, which is later entered into Netflix’s algorithm, which then ensures that *like* films are recommended together; hence the "More Like This" list of recommendations – at least, in part. Surely there are other criteria other than "race" on Netflix’s list, but it could be that certain criteria carry more weight than others, and "race" could be top-heavy. Again, I’m simply speculating here based on available info. When/if I hear back from Netflix, I’ll write up an entirely new post with what I learn.

Something else you have to consider are your individual viewing habits. If the bulk of your Netflix watching comprises of content that centers specifically on black characters, there’s probably a good chance that your recommendations will also include a lot of films and TV series that tell stories primarily about black people. It’s a combo human-machine strategy that Netflix uses, as I’ve learned.

But, really, this isn’t anything new folks. I’d broaden the conversation to include the ongoing debate about what a black film is, and whether black filmmakers make films specifically for black audiences, or whether non-black people can truly appreciate "black stories" (whatever they are); there’s also the universality of storytelling, no matter who the storyteller is and who the characters are,  etc, etc, etc… It’s one result of a much wider and long-standing industry practice with regards to how content is classified, which, in turn, can affect how a film or TV series is marketed, how much money is invested in its marketing, and even how much funding films or TV series that have received a certain classification, receive. Obviously the goal of most content creators is to have their creations reach as wide an audience as possible, regardless of skin color, and even general preference. But Netflix is just one branch of a very large tree with many branches. We have to get to the root of it to resolve these matters one way another.

Stay tuned… this likely can’t be reduced to a single explanation on algorithms.