The conversation about AI has exploded in the last year and a half or so. But the truth is that some form of AI has been guiding your online experience from the very first day you signed into Netscape Navigator (just me?).
Social media algorithms are all powered by AI in the form of machine learning that incorporates various ranking signals to prioritize and personalize content for every user.
It’s never a good idea to try to game an algorithm in social media posts, but understanding the most important ranking signals can give you a strategic advantage over your competitors.
Get our 2024 Social Trends report, and dig into the juicy data that’ll help you reach your goals in 2024—which is shaping up to be social’s best year yet.
A social media algorithm is a set of rules and signals that rank content on a social platform. It organizes content on social feeds based on how likely each individual social media user is to like it and interact with it.
A social algorithm’s purpose is to create a good user experience by making individual user’s feeds interesting and engaging. Algorithms are the reason why no two users will see exactly the same social content, even if they follow all the same accounts.
As you can see from our algorithm definition, social media requires algorithms because they determine what you see out of the incredibly vast potential content pool. For brands, this means algorithms influence how likely your content is to be seen – both by your followers and by people who don’t yet follow you.
And the algorithms can have exponential power. Social media algorithms are designed to serve up content they think a user will like, based on the history of past ranking signals (more on those later). If the user likes, interacts with, or follows any of these posts or accounts, the algorithm quickly learns to serve up and suggest even more of the same.
That’s why following a couple of new accounts, or even liking a few posts from accounts you don’t follow, can create an influx of related content in your feed. After all, what is an algorithm in social media for if not to deliver exactly what you’re looking for to your social feed? Brands, you want to be part of that influx of content.
There’s a reason why the main TikTok user feed is called the For You Page. It’s content specifically selected for you, based on the way you have interacted with the app in the recent past. X (formerly Twitter) has also adopted this terminology, offering users either a For You or a Following timeline.
But, of course, there’s no human being sitting behind a desk shuffling content into the feed of each TikTok or X user. (What a job that would be!) Instead, those recommendations are made by algorithms.
The algorithms of all social media companies are different, but they are all based on machine learning and a set of factors called ranking signals. These are exactly what they sound like: signals used to rank the value of each individual piece of content for each individual user at a specific point in time.
While the science behind them is extremely advanced, the overall function is simple: They scan the entire pool of content available, then score and rank it to determine what appears in a user’s feed.
For example, here’s a visual representation of what’s happening under the hood of the Facebook algorithm:
Source: Meta
Ranking signals are individualized because they are often based on your previous interactions with the platform. And, like we said above, they’re self-reinforcing with an exponential effect. The algorithm’s behavior is entirely focused on keeping you on the platform by giving you content that you just keep watching, liking, or tapping on.
Sp that’s how the algorithms work based on user behavior (in a nutshell). In the next sections, we’ll talk about how content creators and brands can “communicate” with the algorithms that power social media (and help algorithms surface their content to more users). That requires an examination of the specific ranking signals for each social media platform.
We can never know all the details of a platform’s algorithm – social media networks guard that recipe like a secret sauce. But social platforms have become more open in recent years about their ranking signals, and how their algorithms work in general.
That information can guide meaningful adjustments to your content strategy so the algorithms work for you, rather than against you.
Here are the most important known ranking signals for each social platform.
1. Instagram algorithm(s)
Instagram revealed last year that each part of the platform is powered by a different algorithm. Here are the known ranking signals for each.
Instagram Feed algorithm
- Your activity: We’ve talked a lot about this above – it’s the content you like, share, save, or comment on.
- Post info: Post popularity is first here, including not just how many people interact with the content but how quickly they do so. Other factors like date, time, and location come into play here, too.
- Poster info: This signal aims to determine how interested you might be in the account that posted a particular piece of content, using signals like how often people have recently interacted with their account.
- Your history with the poster: You’re more likely to see content from people you follow, message with, or otherwise engage with. For brands, this means it’s critical to encourage and respond to follower engagement.
Instagram Stories algorithm
Note that Stories don’t include posts from accounts you don’t follow, except for ads. So for Instagram Stories, the algorithm social media context is all about sorting the priority of content posted by accounts you follow.
For brands, that means you won’t get new eyeballs here, but you still need to understand the algorithm to extend your organic reach.
- Viewing history: If you always watch an account’s Stories, it will move to the front of your Stories list. If you tend to skip that account’s Stories, it will move to the back of the line.
- Engagement history: Watching is good, but engaging is better.
- Closeness: This is similar to your history with the poster for the Feed algorithm, except in this case the algorithm is specifically evaluating how likely you are to be friends or family. This is a tough one for brands, so focus on the first two.
Instagram Explore algorithm
Instagram says the Explore algorithm is specifically designed to show you the types of content you are likely to “do something with.” The most important actions here are liking, saving, and sharing.
Your own personal Explore Page is practically a personality statement in one screen.
The ranking signals are:
- Post info: The level and speed of interaction with a post. This is more important for Explore than for any other Instagram surface.
- Your activity in Explore: Similar to the equivalent ranking signal for Feed, except that this one more heavily weighs content type.
- Your history with the poster: This is not as relevant to Explore, but if a post is ranking well on the other factors, this can give it an added boost.
- Poster info: Same as for the Feed algorithm, here specifically looking for content from a wide range of accounts.
Instagram Reels algorithm
Brands, take note. Instagram says Reels have “an emphasis on entertainment.” Here, they want to show content that users are most likely to reshare, watch through to the end, and like. (Hint: Entertainment is also an important driver of ROI.)
Interestingly, they also want to show you content where you are likely to visit the audio page, since they use this to get a sense of whether you’re inspired to make a similar Reel of your own.
Here are the ranking signals:
- Your activity: You get how this one works by now…
- Your history with the poster: Again, this is less relevant here than in Feed, but it can be an added bonus.
- Reel info: Basic details like audio track and visual, as well as that ever-important level and speed of engagement.
- Poster info: Here this includes number of followers and level of engagement.
Tip: Eileen Kwok, Hootsuite’s Social Marketing Specialist, recommends keeping up with Head of Instagram Adam Mosseri’s weekly video updates to stay on top of new features and hints at what the Instagram algorithm favors.
For more key insights, check out our full blog post on how to work with the Instagram algorithm.
2. TikTok algorithm
TikTok’s algorithm is especially important to understand because unlike most social platforms, TikTok is designed to surface new content rather than showing content from people you already follow. Here are the known ranking signals:
- Previous interactions. This includes signals like accounts followed and hidden or content you’ve engaged with or marked not interesting.
- Behavior on the Discover tab. This factor analyzes content characteristics like captions, sounds, effects, and trending topics.
- Location and language. Content from your own country or in your own language may be preferenced.
- Trends. Using trending sounds and effects can help make your content more discoverable.
- Follower count does NOT matter. TikTok’s real distinction is that follower count is NOT a ranking signal.
“A good way to find out whether the TikTok algorithm favors your content is checking the percentage of users that saw your videos on their For You page,” says Eileen, who runs Hootsuite’s TikTok account. “A high number means that the algorithm is helping you get discovered by a wider audience by placing your videos in their home feed.”
For more details, check out our full blog post on everything you need to know about the TikTok algorithm.
3. Facebook algorithm
Known Facebook ranking signals:
- Facebook connections. Your Feed will primarily be filled with content from people and Pages you follow and interact with. All the more reason to encourage your followers to engage with your content!
- Content type. Users who watch videos get more videos. Users who interact with photos get more photos, and so on.
- Likelihood of user engagement. This is a prediction of your likelihood to comment on, like, or share a post, based on your past activity.
- Relevancy score: This is a guess at how “meaningful” you’ll find a post. It includes factors like whether a post appears to be clickbait or link to a low-quality webpage, both of which reduce exposure through the algorithm,
Find more details in our post on how the Facebook algorithm works.
4. YouTube algorithm
Like TikTok, YouTube is less about who you follow and more about what the algorithm serves up for you to watch. And take note! While there are certainly ranking signals involved, the “North Star” should be the long-term value you provide to your audience. That’s according to Todd Beaupré, YouTube’s Director of Growth and Discovery
Known ranking signals:
- Your activity: YouTube recommends new videos based on which videos you have watched in the past.
- Videos often watched together: This helps YouTube identify videos you’ll like based on other users’ behavior.
- Channel and topic watch time: If you spend a lot of time watching a particular channel or topic, YouTube will likely give you more of that.
- Video performance: This includes factors like average view duration and average percentage viewed, as well as likes/dislikes and feedback from post-watch surveys.
Source: YouTube Creators Insider
- Trending songs and sampled audio: For Shorts only, YouTube also uses these factors to understand trends a viewer might be interested in and show them more relevant content.
Learn more in our post on how to increase views with the YouTube algorithm.
5. LinkedIn algorithm
Heads up! The LinkedIn algorithm underwent some significant changes in June 2023. (Get all the details in our full post on the LinkedIn algorithm.)
Here’s a quick visual of how it works:
Source: LinkedIn Engineering Blog
And here’s one important thing brands should keep in mind. You’re not going to go viral on LinkedIn. The algorithm is specifically designed to prevent that. This is a platform where understanding your audience and creating top-notch content that serves their specific needs is the best way to get noticed.
Known LinkedIn ranking signals:
- Identity: How closely you’re related to a follower or someone in your extended network. Your skills and profile location are also factors.
- Content: Factors like content popularity, what and whom is mentioned, and how meaningful the discussion in the comments section is.
- Member activity: Which groups, hashtags, pages, and people someone follows and their social interactions with posts on a similar topic.
We get into much greater detail in our post breaking down the intricacies of the LinkedIn algorithm.
6. X (formerly Twitter) algorithm
X allows users to choose between a For You timeline and a Following timeline.
Source: X
Here are the known X ranking signals for For You:
- User connections: Which accounts and topics the user already follows.
- Content quality: whether a post is “relevant, credible, and safe.” Abusive and spammy content will not be recommended.
- Popularity. Both in general, and specifically how people in your network are interacting with the post.
Get the full scoop in our post on the X algorithm.
7. Pinterest algorithm
Pinterest works a little differently from the other social platforms. It’s much more focused on search.
That said, even the ranking of search results requires an algorithm. Pinterest has specifically said that, “Quality and relevance matter more than frequency.” Here are the known Pinterest ranking signals:
- Active engagements: Pinterest is specifically looking for “high signals of meaningful engagement.” The biggest metric here is saves.
- Metadata: This applies to brands’ product catalogs rather than their pins. Be sure your metadata is detailed and includes keywords.
Learn more in our post on Pinterest SEO. It shares lots of juicy details you can use to make your Pins more discoverable.
In case that brings up even more questions, we’ve also got a blog post all about social SEO and how it’s different from social media algorithms.
8. Threads algorithm
Threads is still a baby of a social platform, but it’s up to 130 million monthly active users as of Q4 2023. For brands wanting to get in during these relatively early days, here are the known ranking signals:
- User interactions: How many times you’ve liked a user’s posts (both as a straight number and as a percentage of their posts that you’ve seen), and how often you’ve viewed their profile on Instagram.
- Popularity: How many other people have viewed or replied to a particular post, and how many times other users have clicked on the author’s profile.
Learn more in our post on all things Threads.
You now know why social media algorithms exist and how they differ across platforms. Here are some overarching tips for scoring points with social media algorithms in general.
1. Understand social SEO
Wait, aren’t we talking about algorithms here? Now you have to understand search engine optimization too?
Here’s the thing. As we mentioned in the Pinterest algorithm section above, even search engine results are driven by algorithms, so the two topics are absolutely interrelated. To succeed with either one, you really need to understand both.
And remember how popularity was a key ranking signal for so many of the social media algorithms we outlined above? Social SEO is an important way of getting your content in front of more people, increasing your popularity and feeding good signals back to the algorithm. Like we said, it’s all connected.
In general, social SEO focuses on increasing discoverability through tactics like keywords and alt text. We’ve got a whole post to walk you through it.
2. Get to know your audience
Content relevance is a ranking signal for all the social algorithms. That’s because the entire point of algorithms in social media is to show people high-quality content they’re likely to be interested in.
Spoiler alert: People are not generally interested in content that can be deemed irrelevant.
The details of relevance can vary by platform, but it’s always about understanding your target audience and creating content that appeals specifically to them.
3. Know the best times to post
Many of the algorithms include recency and early engagement as key ranking signals. That means you need to know when your audience is most likely to be online and actively engaging with each social platform.
For general recommendations, check out our post on the best times to post on every social network. But remember that while these times are a good place to start, they won’t necessarily be most effective for your followers.
To get custom recommendations for the optimal time to post for maximum engagement based on your own followers’ behavior, check out the best time to post recommendations built into Hootsuite.
4. Encourage comments, saves, and shares
As we just said, engagement – especially early engagement – is a key ranking signal for all the social media algorithms. One easy way to get more engagement is simply to ask for it.
We’re not suggesting you plead with followers to like or share your posts. Instead, create content that naturally encourages followers to engage, both with your content and with each other.
One tried-and-true way to encourage engagement is to run a social media contest. But, of course, you don’t want to run a contest in every post.
Another great way to boost engagement is to ask a question or start a debate. And make sure that you get involved in the conversation yourself! Reply to all relevant comments and DMs (ideally from a centralized social inbox that allows you to distribute this work among team members) to keep both algorithms and followers happy.
When you create especially informative content, encourage followers to share with others who could benefit from the resources, or to save the post for their own future reference.
5. Experiment (a lot)
Working with the social media algorithms is part science, part art, and a tiny bit of magic. While we can give you tips to help send the right signals to the algorithms, there’s no universal formula for success.
That means you need to try new things, see what works, and refine your strategy over time. All good digital marketers know the manta “Always Be Testing.” It’s the only real way to learn what’s working right now, for your brand, in real time.
And don’t let the idea of running complicated experiments discourage you — testing doesn’t have to be complicated. Keep an eye on trends you notice in your own social media feeds to look for inspiration.
We’ve got a blog post that outlines how to run social media tests. For inspiration, check out the experiments playlist at Hootsuite Labs.
6. Post more video
Social platforms are leaning hard into video – especially short-form video. Posting more video content aligns your brand’s social strategy with the direction the platforms are headed.
In particular, Meta platforms provide lots of opportunities for users to discover short-form video content (i.e., Reels) from brands and content creators they don’t follow. Reels are an important way to reach new users and send relevance signals to the algorithms.
That said, you shouldn’t focus exclusively on video if still images are your comfort zone. Still images make up nearly half of content consumed on Instagram. If you want to use stills, try making them more interactive with carousels or music effects. Carousels still have huge reach on Instagram, and multiple photos mean multiple chances for the post to show up in the feed.
And of course, you can always use a Reels template to make a video out of stills, too.
7. Lean into trends – but not too hard
Trending topics keep people scrolling and engaged, so the social platforms want to serve up more of that content.
You don’t want to leap on every trend that comes along. But if something emerges with real potential to align with your brand messaging, it’s worth putting some of your best social minds on it. Use tools like Google Trends to see what’s trending online.
Source: Google Trends
A social listening program allows you to understand what’s happening in your industry specifically.
Also watch for ways to incorporate trending sounds and effects for short-form video like TikToks, Instagram Reels, and YouTube Shorts.
Make your social media marketing strategy work with algorithms and save time managing all your accounts using Hootsuite. From a single dashboard, you can schedule and publish content, engage your audience, and measure performance. Try it free today.