Facebook is changing its News Feed algorithm in an attempt to deliver more meaning and less clickbait to its users.
Until now, Facebook has bumped up the visibility of certain stories based on the number of clicks, likes, and comments those stories receive. The more people click on a story, the higher it rises in others’ feeds, making it more likely to receive further clicks.
But that system encourages clickbait, Facebook found, with sites writing intentionally provocative headlines (“You won’t believe what happens next …”) to try to gain more visibility.
Now, Facebook is tweaking its algorithm to include more stories that people enjoy but don’t necessarily click or comment on.
The company surveyed more than 1,000 people about how the content they see on Facebook could be improved, and also asked tens of thousands of users to rate the stories on their News Feed from one to five stars. Facebook discovered that many people like seeing certain kinds of content even if they don’t click the “like” button or leave a comment. In other words, there’s a disconnect between how much people report wanting to see certain kinds of stories, and how much they actually click on stories.
“The actions people take on Facebook—liking, clicking, commenting or sharing a post—are historically some of the main factors considered to determine what to show at the top of your News Feed,” Facebook software engineers Cheng Zhang and Si Chen wrote in a blog post. “But these factors don’t always tell us the whole story of what is most meaningful to you.”
To better make that determination, Facebook may rely on factors such as the amount of time a user spends looking at a story, or whether the story contains keywords that match things the user has posted in the past. (If you post a lot of statuses about dinosaurs, there’s a good chance you’ll be interested in paleontology news.)
Whether something is “meaningful” or not is highly subjective, and no algorithm can perfectly predict what people will want to see online. But Facebook’s new algorithm may do a better job of balancing quantitative measures such as clicks with qualitative judgments about what people find meaningful. In the process, it may also help tamp down on inflammatory headlines and the practice of directly asking users to click on a story.
These practices “will likely only cause temporary spikes in metrics that might then be rebalanced by feed’s ranking over time,” the Facebook engineers wrote.