After weeks of controversy and criticism about the news feed, and reduction in reach for some Pages, Facebook outlined the four main factors it uses to determine if a post shows up in the news feed. It also insists Page reach didn’t increase overall, though an news feed algorithm change did start penalizing Pages that get complaints and that reduced total complaints by a double digit percentage.
Facebook’s news feed product manager Will Cathcart told a room of reporters at Facebook’s headquarters this morning that “The problem we face with news feed is that people come to Facebook everyday, but people don’t have enough time to check out absolutely everything that’s going on.”
So to determine if any given Page post shows up in the news feed, Facebook looks at four main factors:
1. If you interacted with an author’s posts before. If you Like every post by a Page that Facebook shows you, it will show you more from that Page.
2. Other people’s reactions to a specific post. If everyone else Facebook shows a post to ignores it or complains, it’s less likely to show you that post.
3. Your interaction with posts of the same type in the past. If you always Like photos, there’s a better chance you’ll see a photo posted by a Page.
4. If that specific post has received complaints for other users who have seen it, or the Page who posted it has received of lots complaints in the past, you’ll be less likely to see that post. This factor became a lot more prevalent starting in September 2012.
Cathcart used a Star Wars-themed example to explain how these work. Let’s say Darth Vader posts that he and Luke Skywalker have confirmed that they are father and son. To determine if Yoda saw this post in his news feed, Facebook would look at if Yoda had Liked or interacted with posts by Vader in the past, if Leia and Han Solo Liked the relationship post by Vader when Facebook showed it to them, whether Yoda tended to interact with relationship change posts in the past, and whether anyone else had complained about Vader or this particular post by the Sith Lord.
Facebook often makes changes to its news feed sorting algorithm, commonly referred to as EdgeRank, similar to how Google revises PageRank. Facebook says it makes these changes to the weighting of these factors to try to increase engagement and general satisfaction. The impact is that maybe reach goes down but some Pages, but people interact more with the news feed overall, according to Cathcart.
Most recently, it made a major revision in September. Users can click a drop-down arrow next to any news feed post to complain by reporting it as spam or hiding it, saying they don’t want to see more posts similar to it. Cathcart says “we started penalizing things that had an above average rate of complaints, and rewarding things that had a below average rate of complaints. Facebook believes the change was a success because engagement went up, and “complaints went down in the double digit percentage.”
However, Cathcart confirms that while the change “had a varied impact across Pages, he median reach of Page content was relatively unchanged after the algorithm change. But that will impact differnt pages differently depending on what people those pages are connected to, and what other things those people are connected to.
There are conflicting reports about the actual impact on average reach, though. Some like PageLever say it stayed stable, but others like We Are Social and SocialBakers say that average Page reach decreased by as much as 50%, which really hurt some businesses who relied on purchases or ad impressions from traffic driven to their site by their Facebook Page posts.
But here’s the real deal. Those analytics providers are only looking at a relatively small number of Pages, typically less than 1000. Facebook’s product marketing director for ads Matt Idema tells me what those studies show ”all depends on what set of pages you’re looking at and how many pages you’re looking at. We’re looking at all the Pages. The median reach did not decrease.” But what about average reach? Idema shut that down saying that as for the difference between median and average reach, “I’m pretty sure they’re indistinguishable.”
So basically, if an analytics company’s clients had their Page reach decrease significantly, then that company’s clients may be more likely to receive complaints. Just to be sure, I double checked with Facebook and they confirmed that average Page reach did not decrease. So why all the hubbub? Because Pages whose reach increased or stayed the same aren’t making waves, but those that had their reach decrease are being very vocal.
Facebook has never been so up front about the fact that it makes major revisions to the news feed sorting algorithm. Even though it’s seeking to create a better user experience and confirmed it does not make changes to sell more ads, some businesses got hurt. This is the unfortunate reality of relying on a centralized marketing channel like Facebook opposed to an open channel like email.