On July 17, Tim Herrera spent six hours scrolling through his Facebook News Feed.
What may sound like procrastination or killing time was actually Herrera’s own experiment to understand Facebook’s News Feed algorithm.
Herrera, a digital audience producer for the Washington Post, moved to Washington, D.C. this spring after living in New York for more than six years.
Curious to see how that affected his News Feed, he embarked on an experiment: first, he cataloged every news item on his News Feed until they started to repeat — he found 1,417 unique posts. No event RSVP or like went unnoticed.
“I think most people generally know that Facebook shows only a sliver of what goes on in their networks, so I wanted to see precisely how small that sliver really is,” Herrera told Business Insider.
Then, Herrera went through and noted all of the July 17 status updates and posts made by his 403 Facebook friends and the 157 Pages he likes. After crunching the numbers, Herrera realised he’d only seen 738 of the 2,593 new updates created by his connections that day — roughly 29% of the day’s activity.
“Considering the average U.S. user spends around 40 minutes on Facebook per day — or about one-tenth of the time I spent in my News Feed — it’s easy to imagine that percentage dipping far, far below my 29%,” he said in his analysis of the News Feed experiment.
The further down the News Feed rabbit hole Herrera went, the stranger the posts he was shown became. Herrera saw almost no news from Denver, despite growing up there, liking several Denver-based news Pages, and being connected to friends who live here. And blogs he doesn’t read that often — like Jezebel and The Cut — made up a significant number of the posts he was shown, while websites he likes almost weren’t shown at all.
But toward the end of his experiment, Herrera started seeing stuff his high school friends posted, only because he had checked out their status updates earlier. “That’s a miscue,” he said, in regard to the News Feed algorithm. “And just because I don’t click Like on something from a friend doesn’t mean I don’t want to see their posts in my News Feed.”
Not much is known about how News Feed’s algorithm works. Other journalists have tried gleaning information about how different Facebook features operate.
WIRED’s Mat Honan liked every single thing he saw on Facebook for two days, liking thousands of posts and Pages in the process and unintentionally spamming his friends’ feeds. The Atlantic’s Caleb Garling set out to trick Facebook’s algorithm, crafting a carefully worded status to garner as many likes as possible.
But the outcomes of these experiments don’t tell us much about what makes News Feed elevate certain posts and hide others, or why Tim Herrera is always shown updates from someone he worked with at Pizza Hut when he was 16.
“There’s a long way to go before Facebook learns us well enough that it can predict what we truly want, versus guessing what we want based on past activity and all the other signals it uses,” Herrera told Business Insider. “But over the years we’ve ceded more and more control of News Feed to Facebook’s machines, so eventually it’s possible that it really will know us better than we know ourselves.”
Business Insider Emails & Alerts
Site highlights each day to your inbox.