Social Media - 24 de agosto de 2023

LinkedIn Expands Engagement Signals to Better Personalize User Feeds

Have you noticed a change in your LinkedIn engagement of late? Maybe you’re seeing more posts from the same people repeatedly, or you’re getting fewer notifications?

There’s a reason for that. As outlined by LinkedIn expert Richard van der Blom, LinkedIn’s been tweaking its algorithm and notifications of late, which has significantly altered some aspects of how it distributes posts, and shows people what they’re more likely to engage with.

According to van der Blom’s most recent analysis:

  • LinkedIn is showing you more content from the people and Pages you engage with most
  • LinkedIn is putting more focus on hashtag engagement to highlight relevant topics to users
  • LinkedIn’s reduced notifications for certain actions
  • Posts are now seeing more reach over time, as opposed to gaining the most traction on the first day (van der Blom says that posts are now getting far more traction on the second and third day after posting)

So why is this happening?

According to a new overview from LinkedIn, it’s recently updated its algorithm to factor in more engagement signals, including how users interact with hashtags, who they engage with in the app, and even what they interact with, in terms of individual posts.

As explained by LinkedIn:

Our Homepage Feed produces billion-record datasets over millions of sparse IDs on a daily basis. To improve the performance and personalization of the feed, we have added the representation of sparse IDs as features to the recommendation algorithms which power these products.”

Which is a technical way of saying that it’s added more signals into the mix, with “sparse IDs” in this context referring to hashtags, users, and posts, among other interactions including Likes and shares. Indeed, LinkedIn says that it’s increased the parameters for its feed recommendation architecture by 500x.

“Our focus is on transforming large corpus sparse ID features into embedding space, using embedding lookup tables with hundreds of millions of parameters trained on multi-billions of records. Embeddings represent high-dimensional categorical data in a lower-dimensional continuous space, capturing essential relationships and patterns within the data while reducing computational complexity. For example, members who share preferences or often interact with the same type of content or a similar group of other members tend to have similar embeddings, resulting in a smaller distance in the embedding space. This capability enables the system to identify and recommend content that is contextually relevant or aligns with member preferences.

That’s a lot of words, yes, and technical papers are not ideal for trying to get to the bottom of what they practically mean for you and I. But essentially, you’re likely seeing posts from smaller groups of people, and on more focused topics, because that’s what you’re likely to engage with most, and LinkedIn’s algorithm now has more measures to factor in, in order to predict likely engagement.

LinkedIn feed algorithm

Which should mean that your LinkedIn feed is more interesting, and more aligned to your actual interests. Which may or may not be ideal for discovery, because a lot of people engage with their colleagues, former and current, as opposed to their current areas of interest, but the weighting of hashtag engagement, for example, will be critical in this respect, ideally presenting a balance of people you know directly and the topics of most relevance to you right now.

But yes, you may see more of the same people than you used to in your feed as a result.

The solution? Engage with more hashtags, add your comments to relevant discussions, and take part in the app. The more you interact, the more signals you send to the algorithm about your interests, and it’s now more attuned than ever to your specific focus subjects.

And it’s clearly doing something right. Sharing of original content on LinkedIn increased by 41% year-over-year in 2024, while the platform continues to report “record levels” of engagement within parent company Microsoft’s quarterly performance updates.  

As more people look for an alternative to Twitter, and the changes being implemented by Elon Musk, it seems that LinkedIn has been a key beneficiary, while its continued algorithm updates are also driving more engagement, and bringing users back more often.

As such, focusing on key topics of interest could be the key to optimizing your LinkedIn experience, while from a posting perspective, it’s also worth underlining the value of community engagement, and building on this where you can, by replying to comments, using relevant hashtags, sharing topical updates, etc.

There’s no secret code, as such, to cracking the algorithm, but LinkedIn now knows more about who’s interested in your content, and it’s increasingly likely to show it to them in-stream.

Source link

Personas a las que les gusta estox