For Google News
This Is Exactly How Social Media Algorithms Work Today
By Christine Warner on May 3, 2018
Each social media network has a unique set of technical elements, intricate logic, and usage analytics that make up the algorithm that delivers content to its users. Seen in the fluctuation of social media algorithms in the past, the biggest consideration-or debate-is balancing personal relevance with post timeliness.
These algorithms are managed by dedicated teams of software engineers, data scientists, content strategists, and more, so nothing you see is an accident. Here's a look inside the criteria and priorities that Facebook, Instagram, Twitter, and LinkedIn use to surface content and updates in users' personal feeds and search results-and what it all means for your social media strategy.
Facebook Prioritizes Meaningful Interactions
Facebook's latest algorithm update, which they announced in January 2018, now puts posts "that spark conversations and meaningful interactions between people" at the top of your News Feed. This means the content that receives the most engagement-reactions, comments, and shares-will be the highest.
Facebook keeps a historical account of your personal engagement with posts by friends and brand pages, and the Facebook News Feed algorithm also predicts what you want to see based on those past interactions with friends or brands. Then, it aims to put similar posts upfront in your News Feed from the same people and pages and from those with relevant posts or profiles.
According to Facebook, it defines engaging content and "meaningful interactions" based on:
- Comment Activity: Back-and-forth comments on posts or a news article or video prompting lots of discussion
- Interactive Posts: Posts that you might want to share and react to, such as:
- A post from a friend looking for advice
- A friend asking for trip recommendations
As a marketer on Facebook, you should focus on engagement as the pathway to reach. Let Facebook's engagement metrics lead your measurement strategy: reactions, comments, and shares.
Although Facebook explicitly states that they will "show less public content, including videos and other posts from publishers or businesses," all hope is not lost for appearing in the Facebook News Feed. The more your page followers engage with your content, the more you will continue to appear in their feeds and the feeds of their friends. Following the criteria above, create content and copy that triggers conversation-or even a healthy amount of controversy. Facebook also calls out live video as a high-engagement format, getting six times as many interactions as normal videos.
Facebook admitted that with these changes, pages will likely start to see their overall reach, video viewing time, and referral traffic dip. However, pages that keep conversation going among friends will see less of a negative effect. Followers can also proactively choose to see page content first through the "See First" option instead of "Default" in your page's "Following" drop-down menu.
In its latest public announcement about Facebook Search in October 2015, Facebook explained its approach to search results. As you type in the search bar, Facebook predicts your complete search query in real-time, then includes posts and conversations shared publicly and from your friends' accounts.
Instagram Features Newest Posts First
After updating its feed algorithm in March 2016 to have "posts you care about most" appear first, Instagram reverted to a time-based feed in March 2018.
Why backtrack? Instagram states the change was the result of user feedback to make its feed "feel more fresh" through timely content. So it seems that while Instagram felt users would appreciate a more personalized feed based on interaction history, users actually care more about the most recent content.
To summarize the differences between the two updates:
- Relevancy Algorithm: Launched in March 2016, it reordered content in feeds based on the user's relationship (a.k.a. interaction history) with the person posting and the post's timeliness
- Recency Algorithm: Launched in March 2018, it places content in feeds based on each post's timeliness
Instagram's current time-based feed doesn't mean that post engagement isn't as important as before. But it does mean that post times are more important to ensure your content appears in your followers' feeds at the ideal time.
While there might not be a silver bullet solution for the perfect distribution time, do pay attention to the peak engagement times (or time frames) of your followers-specifically to make the most of any action-oriented content.
Unlike its feed algorithm, Instagram's Search algorithm operates based on personal activity. The mix of photo, video, and story results depend on the people you follow and the posts you like. As your activity changes, so will your search results.
Twitter Puts Personal Relevance Before Timeliness
The current Twitter timeline algorithm, released in February 2016, places relevant tweets first based on your interaction history with the accounts you follow. When you visit Twitter for the first time in a while, the top recommended tweets are highlighted in a section titled “In case you missed it.”
These recommended tweets show up in reverse chronological order, meaning the most recent tweets appear first. The rest of the tweets that appear beneath them are also in reverse chronological order.
Twitter made this change so that you can "catch up on the best tweets from people you follow." Also, Twitter found that featured tweets get more live interaction to stimulate commentary and conversations.
Not a fan of Twitter's recommended tweets? You can also choose "See less often" right in your timeline or turn it off completely in your Settings by deselecting "Show me the best tweets first." So while your timeline defaults to personalization, you can have control over which you want to see first:
- Recommended Tweets: Tweets surfaced based on your personal interaction history
- Timely Tweets: Tweets surfaced based on post time in reverse chronological order, with the most recent posts appearing first
Of course, marketers aren't able to know the split of their own followers who do and don't use the recommended tweet timeline. So when in doubt, optimize for engagement and time. It can only help to create interactive content and distribute it at the optimal times for engagement.
It's important to note that although Twitter's timeline is presented in reverse chronological order, the results of Twitter search are ordered according to relevance-not time.
Twitter rolled out this search algorithm in December 2016 after much experimentation with balancing tweet and account results, managing the density of video and image results, and optimizing Twitter's machine learning model. All testing had the goal of giving searchers diverse yet hyper-relevant results, using logic to tweak the relevance scoring and tweet positioning. Twitter landed on the current search algorithm since the engineering team found that it yielded higher engagement, with people spending more time on Twitter and becoming more active.
LinkedIn Presents Updates by Individual Relevance
As of March 2018, LinkedIn's Feed algorithm focuses on understanding your individual preferences to give you relevant updates per visit. Upon every visitor session, the algorithm scans and ranks tens of thousands of posts in less than a second.
The thousands of signals used to present this content are grouped into three main categories:
- Personal Identity: Your individual LinkedIn profile, including your workplace, skills, and connections
- Content Engagement: The total views, total likes, update topic, post recency, language, and mentioned people or companies
- Personal Behavior: Your past likes and shares, interaction frequency, and time spent in your Feed
The overall mission of LinkedIn's Feed team is to help users be as productive and successful as possible. Instead of evaluating relevance based on click-through rates, LinkedIn analyzes a variety of signals to measure your interest. These signals include time spent reading, insights from your LinkedIn social graph, and user experience research. Therefore, brand success on LinkedIn is largely determined by time spent with shared content, likes, and comments. Again, it seems that engagement is core to pleasing the algorithm.
Much like Facebook, LinkedIn's Search bar suggests search results as you type, whether for people, companies, or groups. For People searches specifically, LinkedIn's proprietary algorithm returns results through analysis of the searcher's activity and search history, resulting profiles, and other LinkedIn users who have performed similar searches.
What's It All Mean for Marketers?
With all the technical details considered, here's each social media network's default feed algorithm primary criterion (excluding custom user settings):
- Facebook: Relevance
- Instagram: Timeliness
- Twitter: Relevance
- LinkedIn: Relevance
Since relevance is the most common factor, let engagement metrics lead your social media strategy. Even though Instagram is all about post timing, you're more likely to appear in Instagram Search if your posts get a lot of likes and comments.
This is the current state of social media algorithms, but you never know when the next optimization might come and change it all. So follow each network's news and engineering blogs, and stay alert!
For more stories like this, subscribe to the Content Standard newsletter.
Featured image attribution: Erik Lucatero
