Advertisement

How Facebook Is Using Surveys to Improve Its News Feed and Give Users What They Want

Insights move beyond clicks

Facebook’s algorithm will include probabilities users want to see a given story and the likelihood of engagement. Getty Images

Facebook is taking more of a "survey says" approach to what users will see in their news feeds.

In a blog post published today, Facebook software engineers Cheng Zhang and Si Chen explained that the Menlo Park, Calif.-based company's traditional ways of determining what shows up in feeds—likes, clicks, comments, shares—don't always give a full picture of what users want. And the changes could weed out publishers that rely on clickbait for traffic.

Facebook asked more than 1,000 users to rate their daily experience and provide feedback on how it can be improved. It also surveyed tens of thousands of people around the world to better understand how well feeds are being ranked. Respondents were asked to give stories a rating of up to five stars according to how much they wanted to see each one show up in their news feed.

Facebook uses algorithms to determine what people see based on each user's connections and activity on the platform. Updates to the news feed algorithm will include the probability that users want to see a given story and how likely they are to engage with it.

"We saw through our research that people reported having a better newsfeed experience when the stories they see at the top are stories they are both likely to rate highly if asked and likely to engage with," the engineers wrote.

So, how will this affect Pages, the section brands use to manage their Facebook page? That depends on the scope of the audience, how frequently a brand posts and what it posts. But Facebook said brands asking users to take action on a post won't help the way it used to.

"Pages might see some declines in referral traffic if the rate at which their stories are clicked on does not match how much people report wanting to see those stories near the top of their newsfeed," according to the blog post. "This update helps rebalance those two factors, so people are seeing relevant stories to them."

Advertisement
Advertisement
Adweek Blog Network