It's no secret that Netflix has used its micro-genre movie and TV categories to track each user’s viewing habits with startling accuracy. Not only do they know your favorite actors and directors, but how much you enjoy “African-American Crime Documentaries” and “Romantic Indian Crime Dramas.” They use this data when creating their own shows, like House of Cards and Orange Is the New Black.
What does this mean for screenwriters?
In his Atlantic article “How Netflix Reverse Engineered Hollywood,” journalist Alexis Madrigal has meticulously plotted out the methods that Netflix uses to classify each and every movie in its library. Since Netflix is famously tight-lipped about its usage data, Alexis used an elaborate plan involving a script that copies their genre descriptors--such as “Feel Good” or “From the 1980s.” He found that the rental company sorts movies into 76,897 unique micro-genres.
In short, he reverse engineered Netflix’s reverse engineering.
For those of us in the industry, this means that “Big Data” has already taken up shop in Hollywood. They'd like us to believe that now there's no more guesswork involved in finding out what people enjoy watching. Are surveys and focus groups antiquated? Of course, Netflix’s data only applies to Netflix subscribers, but with over 40 million members, that's a pretty large set.
We might see less trial-and-error in what movies and pilots are greenlit. Netflix’s strategy already involves only greenlighting projects that are surefire hits according to their data. This is opposed to what many consider to be a less scientific approach -- the TV business of shooting dozens of million-dollar pilots, with only a few ever reaching the air. Netflix’s method is far more efficient, which means it might win in the end. But is it actually accurate?
The upside is that you can use this data to your advantage. Alexis’s article graphs out the largest micro-genres in Netflix’s library, meaning they are attached to the largest number of other micro-genres. This does not mean that they have the largest number of movies within them, due to complicated algorithms that Alexis goes into in his article.
The biggest micro-genres are:
...and are overwhelmingly set in Europe. So if you want to show up in the highest number of genre searches, write a movie about kings and queens who are married and have children. On the other hand, you’ll have to compete with hundreds of other films about the same topics. Take a look at Alexis’s other graphs for more.
All this number crunching might push Hollywood in certain directions. Will we see more films about marriage, royalty, and parenthood? Only time will tell.
One thing’s for sure: it’s a brave new world out there, and the lure of constructing a winning formula from mountains of data is gleaming like gold did for Cortez. How do you think big data will change Hollywood? Will it change how you write or what topics you write about? Leave us a comment here or on Twitter!