Netflix (NFLX) gave us a peek into its secret sauce, a.k.a its recommendation algorithm, and it turns out it’s not just random guessing. There’s actually a lot of science behind it!
In crafting its personalized recommendations, the company tracks things like how you interact with the service, what you searched for, rated, played, and even how you navigated the menus. Movies or television shows placed toward the top are more likely to be played.
Now that Netflix is more focused on streaming, it’s grasp on what users like is even better. “Predicted ratings” have fallen by wayside now that Netflix can tell when you stop playing something halfway through. Combine that with a team for 40 freelancers who are constantly tagging its streaming content so it’s sorted appropriately and sensing that viewer behavior changes depending on what time of day or even your location, the algorithm is pretty powerful.
Of course, with Netflix collecting all that data, it also knows when you’re a two-faced liar. Carlos Gomez-Uribe, the VP of product innovation, says they know you’re lying about liking Waiting for Superman:
A lot of people tell us they often watch foreign movies or documentaries. But in practice, that doesn’t happen very much.
Go ahead and openly enjoy your fifth viewing of Charlie’s Angles: Full Throttle — you can’t trick them.
