How Netflix Helps You Decide What to Watch
Two of the streaming service's engineers reveal how it makes recommendations and the extent to which a title's placement on the home page affects whether you'll check it out.
How does Netflix introduce you to movies and TV shows you might be interested in?
An interview with two of the company's engineers responsible for the streaming service's recommendation algorithms provides insight.
Netflix uses a more complicated system that user ratings to tell you what you might want to watch, Carlos Gomez-Uribe, vp product innovation and personalization algorithms, and engineering director Xavier Amatriain tell Wired.
PHOTOS: From 'Arrested Development' to 'House of Cards,' Exclusive Portraits of Netflix's Stars
With instant streaming, as opposed to Netflix's DVD-by-mail service, subscribers don't provide as many recommendations, Amatriain explains, because if they don't like something, they'll just stop watching.
Netflix knows what you watched, searched for, rated and even scrolled past, and they know when and where, including on which device, you saw it. That data is factored into several algorithms to determine what you might want to see next, the engineers explain.
However, Netflix is still working on providing recommendations based on context (the time of day, day of the week, device you're using and location).
They also still recommend two- and three-star movies, knowing that people don't always want to watch artistic masterpieces.
"People rate movies like Schindler’s List high, as opposed to one of the silly comedies I watch, like Hot Tub Time Machine," Gomez-Uribe says. "If you give users recommendations that are all four- or five-star videos, that doesn’t mean they’ll actually want to watch that video on a Wednesday night after a long day at work. Viewing behavior is the most important data we have."
Gomez-Uribe also revealed that the closer to the first spot on a row a title is, and the higher up on the page it is, the more likely it is people will watch it.