We have been told that algorithms are stopping our content. And now reaching your followers is harder than ever. The problem isn't the algorithm. 

 

While it is true. Reaching your existing followers is now hard due a change in social media algorithms. But reaching new ones is getting easy. I know that is kind of like saying you can’t meet your mom on mothers Day but you can meet the mother of your friend living in Germany. Organic Mother Reach!

 

Kind of strange isn’t it? But don’t take the stress off it. I’m here to explain to you how social media algorithms work.  And what you can do to learn social media algorithms and master organic marketing on social media. 

 

Before you move further, these insights I am writing are from my work with TikTok, Instagram & others. And the opportunities I had in the past to meet engineers behind these apps. 

 

And working at agencies & running my one to help businesses reach their ideal audiences. The goal here is to share why social media algorithms exist to help you!

 

History of Social Media Algorithms 

 

Social Media marketers can only adapt to these algorithm changes, we can’t control them. Every Time an online algorithm changes, The best strategy is to analyze how these social media changes impact user engagement. 

 

Because you need to always have a direct social media user flow to connect and build a better community. 

 

Meta, Twitter, Tumblr, and all these platforms changed their algorithms twice completely in the last decade. Here is why social media apps use algorithms now:

 

Curated Algorithm 

At the start of Social Media, We had curated algorithms. These had the job to curate content from the people & accounts we follow or want to receive content regularly. 

 

This format of Curated Algorithm had loopholes because social media apps felt they were at a loss. Because after a time point, users had nothing to consume and they would leave the platform. 

 

Mark & the social media brothers thought they were at a huge loss. And Product Managers started to brainstorm new ideas, giving birth to a new algorithm. 

 

Recommendation Algorithm 

 

Most social media platforms started to track user behavior. And other activities in terms of what they talk about in the comments, Stories & other interactive elements of social media. 

 

Originally this was around 2015-17 when Meta, Pinterest & other social media apps started to come into focus for Businesses. This algorithm initially kicked the era of influencers & online Brands. 

 

Businesses with great content marketing & Organic Influencers started to become a bigger part of social media. Because these recommendation algorithms started to recommend engaging content in masses. 

 

Since 2015, we have seen a huge development to better these algorithms. But as many know growing organic is now harder than ever. As these algorithms now have more quality & guideline levels than before. 

 

Below I have tried to break down some of the algorithms created by ML engineers to run platforms like Meta, Tiktok, Twitter, Pinterest, etc. 

 

 

The Moderation Algorithm (IG, Twitter & Tiktok) 

 

Most of these social media platforms use “write-path classifiers” to moderate your content and this algorithm analyzes your post the minute you upload it. 

 

Meaning this is the first system of social media you need to crack to make your content go further and reach more people. 

 

What is the job of Write path classifiers?

 

When you upload your post on Instagram, Pinterest, or Facebook. 

 

First, This algorithm analyzes your content to make sure it is not violating any content guidelines. 

 

Then, the algorithm compares your content to the previous content available on the platform to find if it is relevant to the platform. 

 

How does social media algorithm figure out that?

 

Write-path classifiers are content filters used to make sure quality & relevant content reaches the user. Nothing inappropriate. 

 

And these filters analyze your content pixels, visual identity, captions & text in your images to analyze if the content is eligible. To get pushed to your existing followers. 

 

This first algorithm does not use engagement factors to push the content further to more people. 

 

Only these factors are taken into account to push your content to Home or following feeds on social media. 

 

This social media system is not associated with helping you reach new followers through Explore or For you page. But your content needs to pass this first test to get engagement & reach. From your existing followers & then get pushed to more people. 

Fun Fact: A lot of your favorite platforms in 2015-17 only had this algorithm. That is why many creators & businesses grew faster in that era when compared to 2023. But now improving social media algorithms, we have a new content distribution system to prevent irrelevant and harmful content from going viral. Because unwanted content is always ready to beat social media algorithms. 

The Growth Algorithm (IG, FB & Twitter) 

 

This is the most important algorithm you need to understand in 2023. To reach new followers and grow your sales on social media. 

 

This algorithm exists on every single platform but the role of this model is less significant on Tiktok & Youtube Shorts. 

 

This algorithm is called the “Read-path model”. Unlike Write-path classifiers, this algorithm does not leave the engagement factor on the table. 

 

After analyzing the content based on the visual & Text identity. This social media algorithm moves to the engagement factor. Meaning when someone posts a piece of content and it is eligible for the first push to existing followers. 

 

The next step algorithm takes is factoring engagement metrics like comments & likes mainly to give the content a second push to more followers. 

 

And after that validation, your content reaches organic social media feeds. 

 

Here are a set of social media feeds that use read-path models to push your content further: 

 

• Instagram: Explore Feed, Hashtags Top Section & Reels Tab. 
• Twitter: Topics Section, For You Tab & Trending Hashtags aka “Trends for You”. 
• Reddit: Only to show the best posts from subreddits you have interacted with but did not join. 
• Facebook: To push your recommended aka “Suggested For You” content. 

I can confirm this ML engineering “read-path model” is being utilized by these platforms to the fullest. Regarding other apps, they use similar models but how people engage with the content on those apps is different. 

 

Like Pinterest, You can have an engaging piece of content or strategy based on what I mentioned here. But elements like your website & Industry play a huge role in helping you grow. 

 

That is why I recommend taking this data and insight into best consideration for IG, Twitter & FB. Try to understand who engages with your Content regularly. 

 

Have a look at their profile, find their unique interests, and understand the key personas who engage with your content. This might help you reach the audience & customers that root for your business. 

 

And the best advice is not to use your friends and family to get the engagement. There are other ways your connections can support you by resharing your content in stories or DMs with their friends. But relying on your friends to get more engagement will leave you with an unwanted audience. 

Generative AI Algorithm (Meta, Tiktok, LinkedIn & Twitter) 

This is the latest social media algorithm causing a change in content distribution and content quality measurement systems.

Meta was the first to embrace generative AI to boost in-app engagement and make users spend more time on relevant content. And we got confirmation at the starting 2023, with Mark revealing that AI is helping Facebook get back on track. And users are more active on the app than ever. This change in social media algorithms caused by Ai isn't limited to Meta as you will find by reading further. 

 

Woah! Wasn’t that shocking news for everyone? Yes, Facebook is not getting adopted by Gen-Z. But for now, CTO & Mark shared that AI is helping them keep users engaged. 

 

And as I tracked Meta’s latest updates in the newsletter, During the Q1 2023 Meta Earning Call. Meta announced they are doubling down on the usage of AI to recommend the content. Pointing out again, this directly impacts your current engagement & growth on Meta platforms: IG & Facebook. 

 

Generative AI is tracking the user interactions with your content to promote it to a newer and wider audience. 

 

The current State of Meta’s Use of AI from User Experience is different! 

 

Besides Meta claiming the in-app engagement is up and growing. The new AI algorithm is not serving people or creators well. Many have complained about their content not reaching the right audience. 

 

And often the creators are getting their content shown to the same list of followers. Neglecting a majority of their old followers. 

 

In terms of Non-follower Engagement, Many are experiencing that they have been getting more Explore/organic reach. They are not reaching their target audience. And the same statement comes from content consumers, irrelevant content on explore pages & Facebook is making them leave the app. 

Now, these are different user experiences, I have gathered from Reddit, Twitter, and Instagram. And validated with insights from Product Managers. 

 

July Update: Facebook (Meta) shared this week how AI algorithms work to share and refer your content to newer audiences. You can read more about it here and find the content recommendation factors mentioned in the Image below:

 

 

Why do I use Product managers for validations of AI algorithms?

 

The reason is simple, product managers today at Spotify, Netflix, Pinterest, and others know that AI is making mistakes. And they know consumers generally have a low tolerance for irrelevancy. But they have to look forward and embrace the change to build better products for the future. 

 

That is why I went to Product Managers. And My best insight is the usage of Generative AI to distribute your social media content is in development at Meta and other platforms too. 

 

The algorithm is making a few mistakes but with time. Generative AI will reward content with great quality, with greater distribution. 

 

Am I too empathetic about AI distribution for Social Media or Meta? What is My Goal? 

 

No, I am only giving you a message about the involvement of AI in social media to distribute your content. Previously, it was only content systems doing their job. Those systems got better and AI will too. 

 

Your Job is to know how these algorithms work & state of them. Because now you know the AI algorithm at Meta is learning and still in development. 

 

You can be aware of this and try to connect with your followers directly through Direct messages. Or encouraging them through stories to connect with you on a higher level like an email newsletter or private community. 

 

Get that burden off social media algorithms. And adapt to manual content distribution for 2023 on Meta platforms. 

 

Coming back to the Use of AI at TikTok 

 

Starting with Tiktok, They kind of introduced this AI algorithm to social media space by having an algorithm. That continuously adapts to user interaction 

 

I talked about the TikTok Algorithm in one of my older posts, explaining how tiktok suggests new content. You can find the post here, and read a detailed version. 

 

The important thing I want you to know about is the Growing and dying interest on Tiktok. The data shows TikTok users are on the platform, to develop new interests. Rather than connecting with friends or building deep connections. That does not mean users don’t build those connections. 

 

They do if the content or creator is engaging and appeals to the user. I’m only introducing you to the regular sentiment. 

 

That regular sentiment is what People at Tiktok know and have their algorithms designed for! They show you content frequently based on your recent interactions. And sometimes those interactions change, causing your whole tiktok feed to change.

 

And this phenomenon occurs more frequently than you would assume. That is also the reason why you search “Tiktok Trend” on Google. Often you will see a glimpse of negative TikTok trends impacting children, showing a different side of the app. 

 

This very targeted algorithm is a blessing and curse at the same time. That is why Tiktok needs a better moderation algorithm, to control their AI algorithm. 

 

Can you translate this into a marketing language? 

 

I’d be really happy to do that. In conclusion, The algorithm is very targeted and you can reach your target audience. But escape that audience very easily because the changes of Interests on TikTok are fast. 

 

That is why always have effective CTAs and advise your audience to always leave a follow. That will help you reach them again. 

 

Can we clap for the newbies: Twitter & LinkedIn?

 

The case of Twitter is simple, they even have their Algorithm open-source on Github. You can pay that a visit to understand how they use engagement factors to recommend your content to more people. 

 

You will find the read-path model is explained in more detail specifically for Twitter. And I don’t need anything to add here. Besides the user experience and insight from Twitter employees. 

 

The user experience shows that Twitter is recommending irrelevant content on the new “For You” page. A selective style of content is trending more on Twitter and alternative/different content often gets unrecognized. Showcasing, the algorithm is going through changes. 

 

My Advice for businesses is to identify what performed well on Twitter's “For You” Feed for your feed. And pick the best formats to recreate the content with your take or context. 

 

LinkedIn, AI is our savior 

 

The bigger picture is I’ve covered all LinkedIn updates since the beginning of the year 2023. And 80% of them were about AI or UX. 

 

Why is the usage of AI for LinkedIn the new vision? First massive layoffs at Big Tech and other industries have caused a huge spike in the number of daily active users on LinkedIn. 

 

And as I told you at the beginning, the Curation vs recommendation algorithm is the idea.

 

 LinkedIn still works on the curation algorithm. You can look at how the LinkedIn algorithm Works explained by their engineering team. Here is the link to the original post. Nothing special, after passing through classifiers. It is all curated content from your network based on the engagement factor. 

 

Well, As you know that curation does not lead people to stay on the app for a long time. They have also started recommending new content based on your previous interactions. Now you get Generative AI content picks in your feed, to keep you more engaged. 

 

Are there any issues with this change? 

 

Well, Organic reach on LinkedIn will increase. The only user experience backlash here is the current state (March-May 2023). Many users are seeing content from the same network of people more frequently like 5-7 posts from that same LinkedIn influencer. And content from other connections is buried. 

 

And the new content recommendations in the notification tabs are 5 out of 10 times irrelevant. 

 

Packing it up, Organic reach is up. But as I am writing today. My initial research suggests a lot in terms of AI developments is required at LinkedIn. 

Am I Stressing you out with all the Social Media AI? Or leaving you with no clue about what to do? 

 

There is nothing you can do. Other than hiring me. I’m kidding. But imagine you never knew about all this, you would be part of 1000s of others blaming algorithms. Or 100s of Gurus selling quick growth. 

 

When in actuality there is no hack, no loophole. All there is is your understanding of how these platforms work. Which I tried to tell here! 

 

And then come back to marketing basics: Doing the market research, building Customer personas & then implementing a strategy to achieve the ideal goal. 

Social Media is marketing served in an alternative form. Now you know the AI form of Social Media, it is a newborn. Taking the attention away from your content by crying like a baby. But you want peace, to find that create a strategy to bypass all the crying and reach the audience. 

 

To help you do that, I have written guides on Tiktok marketing , Instagram, and more writing. 

Common Algorithm Ranking  Factors 

 

There are common algorithm factors cited by Hootsuite, Sprout Social & many others. In their posts about social media algorithms, You can find them pointing at engagement factors like posting times, Hashtags, and relevancy. These factors can also make an impact on how an algorithm promotes your content. 

But Aren’t these factors the same on every single platform? 

 

On Instagram, Interactions/ Connection Engagements are important. And the important ones are “Direct engaging Interactions”. Like Messaging or Comments. 

 

We can say the same for Facebook, Twitter, and LinkedIn. Every connection-focused platform will focus on network interactions to push relevant content. 

 

On Tiktok, Content Engagement is important. Any form of engagement with a piece of content. Leads to a similar style of content. 

 

We can say the same for Pinterest, Youtube, or even Netflix. Previous content engagements impact your current feed and what suggestions you get in the future. 

 

The only change is people on these different platforms, prefer different styles of content. 

 

Even though they engaged with a LinkedIn text post about the Table Clothing business. They would not want the same type of text posted on Instagram. 

 

That is another factor coming into a play called “Content preferences”. 

Even though you understand these two factors: Connection & Content engagement. 

 

The most important factor to take into account is the “content preferences” of your audience. These are the three common factors of all social media algorithms you need to break the social algorithm jail or at least escape. Trying to reach your target audience. 

My Strategies Based on Social Media Algorithms 

 

I have built different strategies around these algorithm insights for my clients. And I never hold back my secrets. Because social media strategies require execution, 

And many fail with perfect execution. And I don’t wish that upon you. 

Take One Out 

In the first social media algorithm breakdown, we learned that Instagram and other platforms use Classifiers to validate the content and push it forward. 

 

From that learning, I came up with a social media strategy called “Take one out”. 

 

How this strategy works is very simple: 

 

First, You do social listening to learn what type of content is making a spark in your industry. Try to make a list of competitors and pieces of relevant content. 

 

Second, Try to find the common structure or frameworks being utilized in those selective posts. For Example, It could be an Instagram aesthetic, Twitter Thread Hook, or LinkedIn Topic. 

 

Pick that one element and add it to your original piece of content. Why you do that is simple, A similar concept is already validated and working on the platform. You take the working element of that content and add your value to the original post.

 

But why not? Add everything from that viral content with a small change. Mainly because that takes the spark from your content away. You will be like all the other creators & businesses, jumping on a trend. 

 

And this is not a trend. This is a strategy to integrate working elements of another piece of content. To increase the possibilities of your content discovery on social media. 

Network Bubble 

 

During the pandemic in the NBA bubble, Jimmy  Butler took the benefit and started a coffee business. And charged heavy prices for the same $4 coffee. Because the players did not have any changes. Only Big bills to hand over to Jimmy. But that was a one-time opportunity. 

 

A similar opportunity you get when you start your LinkedIn, Instagram, or Twitter account. The network plays a huge role in kick-starting your growth on social media. 

 

During the 'The growth' social media algorithm breakdown, we learned that your network can make or break your account. That is why from the start try to build relationships with industry businesses and creators. 

 

Get that targeted engagement from the start to reach the most valuable audience faster. I know this is hard to implement because you might have an old social media handle. 

 

It is a little late, but not too late. Try to refine your network and create a new bubble to revive your organic growth. 

CTAs, CTAs & CTAs

 

I know there are well-known theories on Tiktok and Instagram that calling out “link in my bio” is social media sin. The social media algorithm will never forgive you if you say that! 

 

No, it is not the social media algorithm. Most of the time it is Human Psychology. 

 

First, many types of research suggest that people treat each other differently on social media. And this also differs from platform to platform. The more anonymous people can be on social media, the more different they treat each other. 

 

Second, We all don’t like to take orders from anyone. Using a simple CTA like buy my stuff “link in the bio” sounds harsh to many followers. They don’t express that feeling but ignore the content. 

 

That is why always have Indirect Call-to-actions and be creative with how you represent what you sell or want them to do. 

 

Coming back to the bigger point! 

 

The strategy is to utilize CTAs to the fullest, trying to connect with your audience in DMs or outside social media apps is possible with them. 

 

You can use templatized CTAs, they are only one Google search away. But the best ones will be crafted by you with practice & understanding of your audience. 

 

This is all you need to implement this strategy and own your social media audience. 

 

Takeaway 

 

The initial goal behind writing this article was the same as usual. I feel like social media algorithm is a component of social media marketing often misunderstood. 

 

You only need an understanding of the technical parts of it and track how these algorithms change. And you will find new changes either here in an updated version of this post or in my weekly newsletter. 

 

And after understanding these parts. You need creative strategy and execution to do successful social media marketing. It is not easy or quick, but anyone with this knowledge will execute social media campaigns and grow better. 

This was Jaskaran, Your Favorite social media & Influencer Strategist. Hidden in the dark mountains of Reddit as u/lazymentor. 

Frequently Asked Questions

 

I will be updating the list of FAQs based on requests and queries about social media algorithms shared on my reddit & newsletter.

 

What is the problem with social media algorithms?

 

Your content doesn't rank on hashtags, explore page or fails to reach viral growth. The reason could be you have an unoptimized content strategy and no account authority. You build your account authority by being a relevant source for X topic. And by having a scattered content strategy that shifts on regular basis. You are now stuck in a constant phase of no organic reach. To fix this, I would advise start focusing on building a social media content loop to discuss key topics of your industry on regular basis to build the authority first and then experiment with virality content. This will fix your personal problem with social media algorithms.

 

Are Social Media Algorithms AI?

 

 

No, Social media algorithms in 2023 are designed content systems by programmers to filter content based on your tracked interests. But the role of AI is increasing as AI can do a better job of working with those social media distribution to share more relevant content. Meta and LinkedIn are the first ones to adopt AI for social media.