It's not just about deploying Like-hungry algorithms anymore
Facebook wants to build a News Feed full of stories you actually want to read.
The company announced in a blog post Monday that it will starting putting greater emphasis on human assessments of what constitutes a “good” Facebook post, rather than just relying on straightforward engagement metrics such as Likes and clicks.
The change is a long time coming for Facebook, which began paying a handful of its users to rate the quality of their own News Feeds on a daily basis back in 2014. This group, dubbed the feed quality panel, now numbers more than 1,000 and is selected to be representative of the platform’s U.S. user base.
By analyzing the way these users rate posts on a one-to-five scale, Facebook found out that there were some posts that users wanted to see at the top of their feeds that they didn’t necessarily want to Like—someone mourning the loss of a loved one, for instance. Conversely, there are other posts that attract a lot of engagement but end up frustrating users. Think news stories with deceptive headlines and other clickbait.
Ideally, Facebook wants to populate users’ feeds with stories that both spur engagement and leave users satisfied.
“News Feed will begin to look at both the probability that you would want to see the story at the top of your feed and the probability that you will like, comment on, click or share a story,” the company explained in the blog post. “We will rank stories higher in feed which we think people might take action on, and which people might want to see near the top of their News Feed.”
Facebook is in the midst of an ongoing effort to make the News Feed more “fundamentally human,” in the words of Chief Product Officer Chris Cox. Incorporating the sentiments of actual people into rankings to augment the cold calculus of an algorithm can help achieve that goal. Giving users the opportunity to communicate through a wider variety of human emotions on the platform will also help — in the next few weeks the company is rolling out an expansion of the Like button that will let users express more specific emotions instead, such as anger, sadness or love.