Changes in LinkedIn’s algorithm

Image: Generated with IA-Artilist.io under license

LinkedIn has announced changes to the algorithm that gives visibility to feeds in the timeline.

For some time now, it has been using AI to decide which users and content are shown and in what position. Until now, one of the main visibility variables was past “engagement”. Now the change focuses on:

• understanding post content in a more detailed way

• how that content fits with users’ interests

• content freshness — that is, focusing on relevant, current news rather than past content, no matter how much engagement it had

At first glance, this can be summarized as a promise to improve the “matching” between the user profile, what they read/interact with, and the content shown to them — and, on the other hand, a greater demand on publishers/content creators to continuously generate new content.

On the other hand, LinkedIn has also announced that it will focus on “engagement bait” that is:

• limiting the generation of repetitive clicks

• posts that do not provide value to the audience unless a link is clicked (in other words, content must be published within LinkedIn; similarly, reposting requires adding opinion)

This includes putting an end to typical posts asking “comment XXX to get…” or “comment yes if you agree”… or posts that use unrelated or meaningless videos simply to attract attention.

Another point that will lose value is reposting content from thought leaders without adding any value — meaning it’s not that it can’t be done, but if done, it is very important to add your own opinion.

In theory, these AI-driven changes should increase the reach of posts across a broader user base (rather than concentrating it among a few) and curb the “perverse” tactics that are starting to appear on the platform.

But that’s the theory — we’ll see if it actually happens or if things remain the same. What is clear, however, is that:

A) For LinkedIn, content creation is critical — they need users to create content continuously. This has always been a weak point of the platform; it is estimated that only 7% of users generate content regularly.

B) LinkedIn needs this content to be as unique as possible and not AI-generated “copycats”. It is estimated that today 54% of content on LinkedIn is generated using AI tools.

C) LinkedIn needs to take care of content quality/relevance, as it is a “self-reinforcing loop”: if content creators do not receive views/interactions, they lose interest in creating content. Therefore, it is key that this content is shown to the right people and is relevant and differentiated. Ultimately, it is a matter of “editorial quality management”.

Another interesting point is that, for now, LinkedIn continues with its strategy of “penalizing” company pages versus individual posts. The average reach of these pages remains at a very low 2%… and LinkedIn gives much more weight to posts from employees of these organizations, which can achieve between 5x–10x the reach of corporate pages.

There is a key point here — people trust people, and we follow people, not entities… but the money (which is ultimately what LinkedIn cares about) is with companies. So there is a dilemma on the table. LinkedIn needs companies — they are its clients — and now it “sells” corporate pages as a reputational asset, but in terms of visibility, people are key.

So, considering the upcoming algorithm changes and how LinkedIn treats corporate pages:

1- Start encouraging your employees/collaborators to give visibility to content. Not just with a simple “repost”, but by publishing directly or, in the case of sharing corporate content, adding opinion.

2- Corporate content, if published, should avoid being AI-generated — it must be as unique as possible. Sharing the most “human” and “unique” side of companies is key.

3- Carousels perform better than videos and better than single posts.

4- Do not overuse commercial content — try to maintain the 5-3-2 rule. Out of every 10 posts: 5 should be high-value unique content, 3 medium-value content, and 2 self-promo/low-value content.

LinkedIn in B2B environments remains an interesting option to explore, but:

A) The rules of the game so far have not been very oriented toward giving visibility to corporations

B) Standing out (or achieving “virality”) is more complex than on other networks, especially because gaining visibility in a feed format is very limited

C) Paid advertising solutions are very basic and extremely expensive when analyzed in terms of performance and return on investment (ROAS) — which is logical if we consider that inventory is very limited

In short, we will see whether everything LinkedIn is proposing actually materializes and has real impact… and what is clear is that the priority is high-quality, human-generated content.