Facebook’s news feed is the constantly updating list of stories in the middle of your home page. It includes status updates, photos, videos, links, app activity and likes from people, Pages and groups that you follow on Facebook. With over 2 billion monthly active users, Facebook has to decide which of the possible stories to show each person in their news feed. This article will examine how Facebook’s complex News Feed ranking algorithm works and the factors it takes into account to determine which stories are shown at the top of your news feed.
What Factors Does Facebook Consider for Ranking News Feed Posts?
Facebook uses machine learning algorithms and models to rank and select stories for each person’s news feed. While the exact algorithm is proprietary, Facebook has revealed some of the main factors that influence ranking:
Affinity
How likely is the viewer to interact with the post based on their past interactions? If you frequently like, comment on or share a friend’s posts, you are more likely to see more posts from that friend.
Weight
Not all interactions are equal. For example, commenting on a post signals more value than just liking it. Certain actions like sharing a post or viewing a video multiple times add more weight.
Time decay
Newer posts are weighted more heavily than older posts. The news feed algorithm favors recent activity over older posts.
Frequency
Too many posts from one source in a short time span will signal to the algorithm to rank those posts lower. This prevents any one overactive poster from dominating the news feed.
Relevance
Posts relevant to the viewer’s interests and connections will be ranked higher than unrelated posts. Facebook determines this through factors like pages followed, groups joined, and interactions with friends.
How Facebook’s Algorithm Learns Over Time
Facebook’s news feed algorithm utilizes machine learning to evolve and improve its ranking system based on new data over time. It learns in two main ways:
User Feedback
The algorithm monitors all user activity on Facebook to understand their preferences. This includes likes, comments, shares, clicks, hides, snoozes and surveys. For example, if you consistently hide posts from a certain friend, those types of posts will be shown less in the future.
Deep Learning Models
Facebook trains deep learning models on vast amounts of anonymized usage data to uncover patterns. These machine learning models help the algorithm get better at predicting which posts will be engaging and relevant to each user.
Major Factors Impacting News Feed Ranking
While Facebook considers thousands of signals, these are some of the major factors known to influence which posts appear towards the top of your news feed:
Close Friends and Family
Posts and updates from close connections like best friends and family members are typically ranked very high. Facebook can detect these close relationships based on factors like frequency of interaction, photos tagged, and status updates mentioning each other.
Engaging Content
Posts that get a lot of likes, comments, shares and clicks tend to be ranked higher because they are engaging with more viewers. This signals to Facebook that such content should be shown to more people.
Video
Video posts have been given priority in recent years. Facebook’s algorithm favors native video content as it boosts viewership and time spent on the platform. Live videos also tend to rank very highly.
Posts You Have Interacted With
If you comment on, like, share or otherwise interact with a post, you are much more likely to see similar updates from that person. This signals relevance.
Comment Threads
Being part of an active comment thread also gives a post a higher weight. Facebook wants to keep you engaged in conversations, so will push posts with active discussions.
Sponsored and Advertising Content
Facebook incorporates paid advertising and promoted posts into the news feed. These are marked as “Sponsored” and ranked based on a combination of their ad bid, budget and estimated relevance to you.
Pages and Public Figures You Follow
Having an active like or follow relationship with a Facebook page or public figure will put their page posts higher in your feed. Pages you’ve recently visited or interacted with also get prioritized.
Groups You Have Joined
Posts from groups that you are an active member of tend to be ranked higher in the news feed. Especially if you frequently post or comment in the group yourself.
Birthdays and Events
Facebook looks for upcoming birthday’s, anniversaries and other meaningful events to give you reminders and nudge you to engage with friends around these events.
How Facebook Curates Your News Feed
While the ranking algorithm controls the order of posts, there are also other curation factors that impact the diversity of stories shown:
Link Posts
To promote meaningful interactions, Facebook limits the number of link posts shown from pages and public figures. This pushes up status updates, photos, videos and activity from friends.
Stale Content
Known as “last actor hiding”, Facebook will hide posts in your feed from people you haven’t interacted with in a long time. This prevents stale content from inactive connections.
Overly-Promotional Content
Pages or public figures that excessively post clickbait, sensational or low-quality content are throttled to limit their reach. Facebook wants to show content that sparks meaningful engagement.
Publisher Paywalls
Facebook scans link posts to detect paywalls and limits their appearance in the news feed if they significantly reduce traffic to the publisher site after being clicked.
Related Articles
If you read an article, Facebook will occasionally show you other recommended related articles to keep you engaged.
Balancing Content Types
To provide an experience with diverse content, Facebook balances the proportion of posts from friends, pages, groups and other sources in your feed.
How You Can Influence What’s Shown in Your News Feed
While you don’t control the algorithm, there are ways to influence what appears in your Facebook news feed:
Like and Follow Pages
Be active in liking and following the pages, personalities and organizations that post content you want to see. This signals Facebook to share more of their posts.
Join Relevant Groups
Joining groups around topics and communities you are interested ensures you see more of that content in your feed.
Comment and Interact
Commenting on or reacting to a post with emoji tells Facebook you want to see more like it. Being active with your real-life connections leads to seeing more of their updates.
Share Meaningful Posts
Sharing personally relevant posts to your own Timeline or in Messenger conversations indicates posts you value. Facebook picks up on these sharing signals.
Snooze or Unfollow
Making use of snooze or unfollow options for friends or pages that post irrelevant or negative content reduces their reach and frequency.
Provide Feedback
Facebook occasionally prompts you for feedback on posts, so respond to these surveys to refine your preferences. You can also submit direct feedback to Facebook on undesirable content.
Conclusion
Facebook’s news feed ranking algorithm is extremely complex, relying on deep learning models trained on massive amounts of data to determine relevance and engagement. While the exact details remain a secret, affinity, weight, time decay and frequency are known major factors Facebook considers. Favoring close friends, video content, and active comment threads also help curate a personalized feed for each user. While you don’t control the algorithm, interacting more with the people and content you care about will produce a news feed tailored to your connections and interests over time.