What are Social Media Algorithms?

The world of social media is vast and complex – content isn’t one-size-fits-all. Creating quality content is the first step. Next, it’s getting your content seen by your target audience. Social media algorithms play a huge part in this, and here we look at some of the pros and cons of social media algorithms and how to make them work for your business. Each social media platform has its own algorithms for your news feed and other site features. We dig a little deeper into how to improve your content’s chances of “winning” across some of the biggest platforms.

The purpose of social media algorithms

Any new account set up on any platform will see a mix of random content that may not align with their interests. They will know some basics – age, sex, location, language setting, and friends/connections, which will help to begin with. But over time, the content will align with their interests. The platforms “learn” what type of content their users engage with. And when they know what a user is engaging with, they will begin to prioritise more of the same.

Your engagement is hugely important to them, as engaged users stay on the platform. The longer you are on the platform, the more time they must show you advertising in the hope that you click and buy from those adverts. And the longer you are on the platform, the more data they have into your interests and purchases. Further improving their ability to deliver content, they are confident you will engage with them.

Now that you know why they do it let’s look at social media algorithm examples from the most prominent platforms.

Social media algorithm examples

Facebook

With nearly 2 billion daily active users and about 3 billion monthly active users, Facebook continues to be the biggest social media platform. At the start of 2022, Facebook announced that your “News Feed” was now called “Feed”. The latest change in a long line of algorithm developments began in 2009. It started by bumping posts with the most Likes to the top of its users’ feed. From this first basic algorithm, ranking signals have been added (and some removed!), and their significance adjusted.

So, from what started with the “Like” button, today’s Facebook algorithm is by far the most complicated of all social media networks. The introduction of reaction buttons, with weighting, added to stronger reactions. The priority is given to posts that spark conversation and video content that holds a user’s attention for more than one minute. The Facebook algorithm is constantly changing – a huge team is working on AI to improve the algorithms. Add to this a user’s ability to Favourite up to 30 people or pages and their ability to Hide Posts or Hide Ads. Again, it allowed the platform to collect more data on the things that do and don’t interest a user.

Relevancy Score

The Feed algorithms consider three main variables and give the post a relevancy score. This score is based on predictions of if a user will interact (in any way – good or bad) and is specific to you and that post. And when every post that could show up in your Feed has its relevancy score, they are sorted into the order they appear in your Feed. The same for Ads – they are also given a relevancy score. The three main variables are:

Source: Who posted it? You will see posts from sources that you interact with. This includes friends, businesses, public figures and news sources.
Type of content: What is it? Posts are prioritised based on the type of content that a user interacts with most. So, if you watch more videos, you will see more videos. Or maybe it’s photos or links.
Interactions: How are people reacting to it? Positive engagement is seen as proof of quality. So, successful content, particularly with those you interact with, will be prioritised.

Instagram

Part of the Meta family with nearly 1.5 billion monthly active users – a significant player for marketing in any industry. In 2016 the platform started sorting users’ news feed by relevancy rather than recency. Like its sister company Facebook, all posted content is analysed using captions, location data, hashtags, mentions and engagement metrics. It will cross reference this with the information it has on its users and delivers the content they are most interested in seeing. Hugely crucial for marketers if they want their content to be shown to users.

Today it uses different algorithms for each area of the platform. So, the algorithm controlling what you see in your news feed differs from those used for reels, stories and the explore page. Across all four content types, six factors affect how content is ranked:

Follower Relationship: Do you follow each other? Do you regularly message each other? The more you interact, the more you will see their posts.
Interest: Does a user often interact with this content? Instagram tracks a user’s preference for content type and format. They will prioritise content they know you like.
Relevance: How relevant is the content? This includes its “newness” as well as any current trending topics.
Frequency: How often do you use the App? Infrequent users will only see very relevant content.
Following: How many people are you following? The more people you follow, the more people and posts compete for space.
Usage: How long are you spending on the App? If you are only using the App for a short period, you are more likely only to see posts from those you regularly interact with.

The algorithm’s use on Instagram

This algorithm is then applied in various ways depending on what you are looking at – your News Feed, Reels, Stories, or Explore. Perhaps the Explore tab is the most important for accounts trying to gain more followers. Here the algorithm analyses posts that you have previously interacted with. It will then rank and display content from related accounts you don’t follow.

As well as this, like Facebook, Instagram recently introduced the Favourites options. They allow users to “Choose the accounts you can’t miss out on”. And the “Following” option allows the user to view posts in chronological order simply. They are putting the power back in the hands of the user and allowing them to remove (some of!) the influence of the algorithm – in their News Feed, at least!

Twitter

With around 500 million active monthly users, the platform experienced a huge backlash when it first introduced its algorithms. #RIPTwitter was a trending hashtag, and the platform now allows users two options:

The Home timeline known as Top Tweets – or the with algorithms
Latest Tweets – or the without algorithms

Even with the ability to remove the influence of algorithms on your timeline, in reality, Twitter algorithms are inevitable. They will still be influencing Trends, Topics, the Explore tab and the recommended accounts. Because its algorithms are machine learning based, Twitter admitted that it is difficult to completely define the factors used in determining your feed. Following bias concerns that Twitter identified with the machine learning algorithms, it created a new team tasked with identifying and addressing these biases. Its Machine Learning, Ethics, Transparency and Accountability (META) team is working to address algorithm bias, including racial, sexual and political preferences. This review and ongoing development are bound to lead to changes.

Although it can not completely confirm the factors involved in its algorithms across all platform areas, Twitter has given some clues about their workings.

Recency – the algorithm will look for topics that are popular and are happening NOW rather than things that have been popular for a while.
Relevance – this is based on your previous interactions on Twitter. What you Tweet, Tweets that you engage with, topics that you follow and accounts that you interact with the most.
Your location – based on your current location, but this can be set to a specific location in Trends.
Engagement – How many people in your network and the broader platform are Tweeting, Retweeting or commenting on Topics, Trends and individual Tweets?
Media type – adding images, videos, polls and GIFs will improve its ranking.

Topics, Trends and Recommended Accounts

Despite a user’s ability to remove the influence of algorithms on their timeline with the Latest Tweets option, the algorithm will still influence the other areas of Twitter. For Topics, it uses an algorithm to suggest Topics based on what it thinks you will like. These topics are also customisable as you can follow and unfollow topics, following a topic with allowing the algorithm to show you tweets, events and adverts related to this topic. But you can also tell Twitter topics you are not interested in.

There is no escaping “Trends” on Twitter – they appear on your timeline, the Explore tab, search results and on profile pages. Trends are predominantly location-based – using the user’s current location. However, you can choose to see trends from a specific area rather than your current one.

The Twitter algorithm recommends accounts that it thinks you would like to follow. These suggestions are based on data you may or may not have given Twitter access to. Firstly, your contacts and, secondly, your activity on third-party apps and websites. Restricting access will have a heavier influence on the other factors, which are your location, your activity on Twitter and any promoted accounts.

Yes, despite what the platform aspires to be, it is still a social network. And like all social networks, LinkedIn uses algorithms to rank its content for its users. With over 850 million users and more than 58 million companies, the platform has become more than a job search network. The algorithms process billions of posts to make your newsfeed as enjoyable as possible.

Firstly, the platform’s algorithm determines if a post is genuine or spam. It will then categorise it as high-quality or low-quality if it thinks it’s genuine. You can be sure that anything spam will disappear into the ether, and even low-quality content may struggle to see the light of day in some people’s timelines. Creating easy-to-read posts that encourage a response whilst not over hashtagging or tagging others unnecessarily will help you hit the high-quality mark.

If you pass the spam test, the second stage is your post being tested with some of your followers. And depending on the engagement it receives, the algorithm will push it to a broader audience. This process takes place during a short timeframe, so posting on LinkedIn is very time critical. You want your audience to be online, and you need to be available to respond to replies, comments and questions.

With good engagement, your post will start to reach a much wider audience using three ranking criteria:

Connection –firstly, your followers, people that have worked at the same organisation as you, and people you have engaged with on the platform in the past.
Interest/relevance in the topic – users’ interests are based on past activity. This includes the groups they are in, the companies they follow and previous hashtags they have responded to.
Probability of engagement – how likely will the user engage in your post?

Social media algorithm manipulation

Some might call it manipulation – we call it an excellent social media strategy. There is no golden bullet that will work across all your social networks. But here are our essential tips to improve engagement and make your content visible to more people.

Keywords – what are you trying to say, and how will people search for content like yours? Knowing the right words to include is vital across all platforms. Think like your customers and use a free keyword planner such as Google.
Grammar and spelling – bad grammar and poor spelling will ring alarm bells for those spam filters. Make sure you use Microsoft Word’s spell-checker and “Read Aloud” option, or Grammarly is another great free app.
Inappropriate content and language – this is a surefire way to have your content buried or, even worse, removed and your account blocked.
Rich media – using quality images or short, engaging videos will increase engagement.
Include relevant hashtags – overusing hashtags or using generic #like #comment will damage the credibility of your post. Take a look at some free hashtag generators, such as All-Hashtag, or use the search option on each platform.
Location – tag locations where you want local users to find your content.
Mentions – tag relevant accounts and others likely to respond and share your content; likewise, don’t tag irrelevant accounts.
Timing – knowing when to post and being available to respond when your content goes live will increase engagement. Meta provides this information on your accounts, or one of the third-party schedulers can help with this too – check out Hootsuite or Later.
Interact – respond to questions quickly, like comments, answer DMs, but also interact with other accounts and the content they are posting.
Be part of the community – join groups, post consistently (this frequency can vary depending on your business), and understand your audience.
Outbound links – avoid outbound links in the actual post; these can be added to the comments. Social media platforms want to keep people there, so sending people away will negatively affect your hard work!
Use original content – posting the same material across all networks can cause problems, especially on Instagram, where TikTok watermarked content will be deprioritised. Use SnapTik to get rid of any watermarks before posting on Instagram!
Make the most of new features – new features or formats on a platform are usually given a boost, so don’t be afraid to give them a try.
Analysis – it’s important to find out what is working and what isn’t. What has worked well for you in the past is always a good starting point when creating new content.
Find your voice – brand tone of voice and its consistency is essential. You want to create a persona and be instantly recognisable.

Download your FREE Social Media Algorithm Manipulation Tips here….

If all this sounds a little daunting, then working with a team of experts like us will help turn these complex algorithms into basic maths. Get in touch today for more information on how we can help with social media strategy, brand tone of voice and complete marketing solutions.