{"id":645175,"date":"2013-03-04T17:33:10","date_gmt":"2013-03-04T22:33:10","guid":{"rendered":"http:\/\/gigaom.com\/?p=616597"},"modified":"2013-03-04T17:33:10","modified_gmt":"2013-03-04T22:33:10","slug":"facebook-responds-to-criticisms-of-newsfeed-says-its-algorithms-are-designed-to-keep-users-happy","status":"publish","type":"post","link":"https:\/\/mereja.media\/index\/645175","title":{"rendered":"Facebook responds to criticisms of newsfeed, says it\u2019s algorithms are designed to keep users happy"},"content":{"rendered":"

Are you more likely to see posts surfacing in your newsfeed if the author pays to promote them? NYTimes writer Nick Bilton<\/a> wrote this weekend that he’s seen engagement on his posts drop recently, but Facebook has come back to refute some of his arguments<\/a>.<\/p>\n

Bilton’s column pointed to a drop in likes and shares from his subscribers (an option for public figures to allow non-friends to follow their updates<\/a>), and the immediate jump he saw once he paid $7 to promote the post. Bilton questioned how Facebook surfaces items in the newsfeed, asking if advertising and promoted posts will push out posts that users share but don’t pay to promote.<\/p>\n

But Facebook came back Monday with a blog post titled “Fact Check”<\/a> that warns users not to take one person’s anecdotes too seriously, and explain how the algorithims come to be, saying the tweaks the company administers\u00a0to the newsfeed formula are meant to give users a better experience on Facebook:<\/p>\n

\n

“There have been recent claims suggesting that our News Feed algorithm suppresses organic distribution of posts in favor of paid posts in order to increase our revenue. This is not true. We want to clear up any misconceptions by explaining how the News Feed algorithm works.<\/p>\n

First, in aggregate, engagement \u2013 likes, comments, shares \u2013 has gone up for most people who have turned the Follow feature on. In fact, overall engagement on posts from people with followers has gone up 34% year over year.<\/p>\n

Second, a few data points should not be taken as representative of what actually is happening overall. There are numerous factors that may affect distribution, including quality and number of posts.<\/p>\n

News Feed shows the most relevant stories from your friends, people you follow and Pages you are connected to. In fact, the News Feed algorithm is separate from the advertising algorithm in that we don’t replace the most engaging posts in News Feed with sponsored ones.”<\/p>\n<\/blockquote>\n

The post goes on to explain<\/a> that the company has tweaked how much of public figures’ content is shared out to their subscribers, and arguing that you can’t compare interaction with posts from one year to the next, calling it an “apples to oranges” comparison.<\/p>\n

It’s a little unclear why user interaction on a post from year to year would be an unfair comparison, but Facebook has explained the\u00a0algorithms\u00a0behind its newsfeed before<\/a>, noting that it does change the formula quite frequently and relies on negative feedback to figure out what users don’t like. The question is, how does the company’s desire to promote paid posts work its way into the\u00a0algorithms.<\/p>\n

The company explained in November that<\/a> the three most important factors that determine whether you’ll see a post in your feed are how you’ve reacted to the publisher in the past, how other people have reacted to the publisher’s story, and how you’ve reacted to similar stories before. So it’s reasonable to assume that if users stopped engaging with Bilton or other public figures posts, they might start popping up less frequently in the newsfeed.<\/p>\n

\u201cWe make changes to the algorithim all the time, at least weekly,\u201d Cathcart said in November. \u201cWe work all the time to say, \u2018Can we better predict what people are looking at? Can we better predict what people won\u2019t want to see or are less likely to interact with?\u2019\u201d<\/p>\n

\"Facebook<\/a><\/p>\n

\"\"<\/p>\n

<\/a><\/p>\n

Related research and analysis from GigaOM Pro:<\/strong>
Subscriber content. Sign up for a free trial<\/a>.<\/p>\n