Last week, LinkedIn announced a change to its content algorithm that might finally show how AI can benefit readers, rather than just content creators. Assuming it works, the platform will more deeply personalise its content feed to show posts that are more directly relevant and useful for you.  

The shift is explained in a fair amount of technical detail here, but the core point is that it aims to use the power of large language models to better understand what people are posting about, and to link this more closely with what you prefer to read about. As the article explains, “this means discovering content from authors you don't follow, on topics you care about, expressed in different terminology than you might use - dramatically expanding your access to relevant professional insights”. 

In short, LinkedIn is moving beyond simple keyword matching. By using LLMs, the platform aims to get a semantic understanding of a post’s value. Based on your past activity (not just your likes, but also what you stop to read), it aims to deliver more accurate and helpful recommendations for you–and not just from your network, but from others with similar professional interests among LinkedIn’s 1bn+ users. 

If you’re a marketing professional working in thought leadership, it should start to recommend content that is directly or indirectly relevant to you: about research and insights, about content distribution, about impact and influence, and so on. Based on what you choose to read or interact with, the platform looks for semantic links with other relevant content.

A new hope
If this works as promised, it could be a genuine win for content consumers. This adds to other updates that LinkedIn has been making to try and target the wider plague afflicting social platforms today: the deluge of low-quality AI-generated content. It aims to limit “click bait” or “dwell bait” content (the success of this change remains unclear, but my own experience is that this is generally moving in the right direction; the sooner the “comment YES below” spam can go, the better). 

Of course, there is a risk that we end up in Facebook-land: an echo chamber of posts that only reinforce our worldview, rather than expanding it. But the approach LinkedIn is describing should (hopefully) avoid that. It amplifies posts that cover related topics, not necessarily just posts that include similar keywords and arguments. In addition, the algorithm is designed to adapt on the fly, while you’re browsing your feed. So if you stop to view a post on a new topic, it will suggest related posts within the same session (of course, this does little to discourage a different issue: doomscrolling).   

If this all works, it’ll help promote LinkedIn as a useful tool and platform for sharing professional insights, and showcase a more positive role for AI.  

“Hello, I’m over here” 

While these changes should be good news for content consumers, what does it mean for marketers seeking to get their content out to a wider audience? 

The most immediate benefit is that the reach of a high-quality post can now extend well beyond those in your immediate network, especially if people choose to read or engage with it. This is excellent news for experts who are willing to share genuine domain knowledge. 

There is plenty of speculation around what specific attributes can boost, or bury, your insights–and it’s difficult to verify these precisely, but the core points appear to include:

• Dwell time and/or engagement. This is pretty well established: if people stop to read your post and/or engage with it, it’ll get greater prominence in other people’s feeds. People saving or sharing your post are particularly helpful. 
• Quality classification. LinkedIn categorises your post as either high or low quality, and prioritises accordingly, as this guide outlines. It’s difficult to be certain, but originality and insight is believed to be amplified ahead of low-insight posts, repetition, excessive tagging, and an overdose of hashtags, all of which is great news for people actually seeking useful insights. 
• Niche relevance. The latest LinkedIn update confirms that the algorithm prioritises knowledge and advice from experts in related fields to you. As part of this, the platform takes into account your stated interests and experience from your profile details. 
• Link penalty. This has been widely A/B tested by experts like Richard van der Blom, but the short version is that LinkedIn prioritises posts that keep you within LinkedIn, so be careful when linking externally within your post. Keeping the external links in your comments or bio is typically safer.   
• Useful comments. When a relevant and insightful comment is added to a post, the reach is amplified; conversely, a simple “thanks” or “good post” comment is not prioritised. 

Do not comment YES if you liked this update, but thank you for reading.

‍