Flipping through TV programs has changed a lot in the world of on-demand streaming. There are far more options than there used to be, and they’re all available at once. But one thing hasn’t changed: People are still impatient.
Netflix knows that the average subscriber spends a mere 60 to 90 seconds scanning movies and TV shows on the platform before giving up, according to a recent report (pdf) by Netflix’s Carlos Gomez-Uribe and Neil Hunt. During that time, they’ll review roughly 10 to 20 titles—about three in detail—on one or two screens. And if they don’t find something that interests them, they’ll likely switch to television or another service.
To ensure that potential viewers don’t leave, the company has invested a lot of time and money perfecting its recommendation engine—a task that became significantly more complex after Netflix expanded from 60 countries to 190 last month.
The move forced Netflix to review all of its algorithms to make sure they worked for a global audience—factoring in things like language and cultural nuances. But it also gave the streaming service access to data from members around the world, which helps improve its personalized recommendations.
For example, one way Netflix suggests titles by finding other members with similar movie and TV show tastes, and seeing what they like. Now, those communities are much bigger, with members spanning borders. This allows Netflix to make recommendations that go beyond the country you live in or what’s in your catalog.
As Gomez-Uribe pointed out in recent blog post, it’s challenging to make recommendations for viewers with niche interests, like “foodies” or anime fans, based on the viewing preferences of a single country. “Tapping into global insights makes our personalized recommendations even better because now our members benefit from like-minded viewers no matter where they are in the world,” he said.
Those recommendations need to be on point, every time, if Netflix wants its users to stick around.