In-Ear Insights: LinkedIn Algorithm Updates October 2026

In this episode of In-Ear Insights, the Trust Insights podcast, Katie and Chris discuss how to transform overwhelming platform guides into clear, actionable steps. You will discover how to turn complicated platform rules into simple daily actions. You will learn which three profile changes push your work into front feeds. You will watch a quick framework that shows how to track your growing audience without diving into complex spreadsheets. You will see how to test your posting schedule and keep only the tactics that bring new viewers.

00:00 – Introduction
02:15 – The feedback problem
04:30 – Breaking down the three levers
08:45 – Retrieval versus ranking systems
12:20 – Building the master guide
16:10 – Tracking visibility without spreadsheets
20:55 – Testing posting schedules
24:30 – Call to action

Tune in now and start building a LinkedIn strategy that works for you.

#LinkedInStrategy #ContentMarketing #SocialMediaTips #DigitalGrowth #AlgorithmSecrets

Watch the video here:

Watch this video on YouTube

Can’t see anything? Watch it on YouTube here.

Listen to the audio here:

Download the MP3 audio here.

[podcastsponsor]

Machine-Generated Transcript

What follows is an AI-generated transcript. The transcript may contain errors and is not a substitute for listening to the episode.

Christopher S. Penn:
In this week’s In-Ear Insights, change is the only constant, and LinkedIn is no exception. On a semi-regular basis, about once a quarter, give or take, we update our unofficial LinkedIn algorithm guide. The October version, because it’s almost October, contains the current snapshot of the algorithm.

So before I dig into the specifics, Katie, here’s a piece of feedback I got from one of our close friends: our guide covers a lot. It’s very useful, but their stakeholders refuse to read it even though it’s only 102 pages long. So my question to you is, given that there’s so much in here about the inner workings of LinkedIn and how their system works, when you hear somebody give feedback like that, what does that tell you about what we’re giving to the marketplace?

Katie Robbert:
That we’re not clear on the purpose of the paper. We’re saying it’s the unofficial LinkedIn algorithm guide. Okay, what am I? Like, I’m a busy CEO. What does that mean for me? Why should I bother to read this? What is in it that I’m going to benefit from?

When I hear that kind of feedback, like it’s really long and nobody wants to read it, we have not done a great job of telling people what the benefit of reading the paper is or what they can get from it.

Christopher S. Penn:
And this is an interesting question. I wonder if they need to read it, or would they be better served with smaller, more granular deliverables? The paper is there as the receipts to say, we’re not making this crap up. This is legit, and there are going to be some folks who say, yeah, I want to dig into the details. I want you to tell me how Apache Kafka is part of the system.

But 99.9% of people say, please just tell me what to do. I’m so overwhelmed, I don’t have time to. We’re unwilling to make the time to read it. Is that the approach that we should be thinking about in terms of enablement in general? Are people just over reading?

Katie Robbert:
I mean, I feel like you’re asking two different questions. Are people over reading? This is something that we’ve always seen, and I don’t mean us as Trust Insights. I mean the world in general has always dealt with that push-pull. There are those people who really enjoy just sitting down and reading, focusing on one thing.

And then you have others who have shorter attention spans or aren’t interested in focusing, and they just want the quick, snappy, get-to-the-point version. And so that’s not a new problem. An academic paper, which is essentially what this is—an unofficial academic paper on the inner workings of a LinkedIn algorithm—is not for everybody.

And I’m not surprised that some of the feedback is, I don’t want to sit down and read it. It’s not positioned for a general public audience; it’s positioned for a very technical audience. But we’re saying there are so many benefits and nuggets of information inside, yet we’re burying it inside academic jargon and really heavy-duty technical descriptions.

There does need to be versions of it that are better aligned with our different ICPs, if that’s what we care about. So I would imagine that we do care that our ICPs are getting versions of this that matter to them, which means more work for us.

Christopher S. Penn:
It does. And so I took that feedback to heart. You and I talked about this outside of this podcast episode, and what we’ve come up with is seven different deliverables. Fundamentally on LinkedIn, there are three things. There are three levers you have: your profile, the content you post, and then how you engage or show up in other people’s content. And those are the big three levers that you have.

So what we’ve done is continue building the master paper itself, which is 102 pages of relatively dense stuff, and then created three one-pagers. A one-pager for your profile. Here are the things you should do with your profile. In fact, let me see if I can bring this up.

And for those who are listening, if you go to TrustInsights.ai slash YouTube, you can see what’s being shared on screen. This is the one-pager for your profile. What should you do? Use a clean, clear professional headshot. Write a concise, keyword-rich headline.

Write a compelling, detailed summary that tells a professional story. Detail each role with achievement-oriented descriptions using industry-standard language. That part’s important because of the way that language models work. Request and give thoughtful, specific recommendations.

Thoroughly complete all relevant sections and write your entire profile with human readers and both AI systems in mind. And it says here’s how, which shows you what each part is and which part of the LinkedIn algorithm it reaches. So this version condenses 35 pages of the master paper into a single one-pager.

Now given this, is this easier? Do you think for the person who’s like, just tell me what to do?

Katie Robbert:
It is. But I’m looking at this fresh. So for those listening and watching, I have not seen this, and my immediate feedback is, so what? I update my LinkedIn profile, what do I get? What should I be looking for in terms of measurement? Am I looking for more profile views?

Am I looking for more people reaching out to me? Am I looking to be a top voice? To me, as the person needing to execute this, what is the purpose of following this profile spine? What am I going to get from it? Because it’s not a small amount of work.

You’re not undertaking a months-long initiative, but you do have to be thoughtful about what you’re putting into your profile. So if I’m taking the time to do this, what’s the so-what?

Christopher S. Penn:
And this is where I run into issues with this because it doesn’t have that bigger-picture context that the paper itself provides. At a fundamental level, if we boil down the entire paper into a TikTok reel, it would be this. There are two language models that power LinkedIn. Now there’s a retrieval system and there’s a ranking system.

The ranking system decides what gets shown. The retrieval system figures out which of the things out there should be shown, and they operate in retrieval first, then ranking. The retrieval system looks at language, particularly in your profile and some of the content that you post.

The ranking system decides based on your engagement, yours and others, where to show you. So there are these three things: profile, content, and engagement. They operate with these two different models to say you need to have good stuff in your profile and the content you post to be in the retrieval system.

And then you need to be active with your content and engaging with others to be in the ranking system. And they operate sequentially in that order. So these three to-do’s on this list will get you visibility, period. Not even a top voice, not even business stuff.

This is just like, just to get the algorithm to even recognize that you’re not a spam bot and to start showing you. That’s sort of, I guess, the crux of it. So that’s the entire thing in a nutshell. You need to have lots of words that are relevant for the retrieval system, and you also need to do lots of things to get their ranking system to like you.

Katie Robbert:
That to me sounds like a really easy carousel or something. Like a quick, if you’re on LinkedIn and you’ve seen people put up simple slide decks, that description or explanation that you just gave me sounds like it would work really well in one of those simple slide decks. You can probably have Claude design, put together or something based on the content of the paper.

But that context, that really quick takeaway, those six to eight slides that someone would just click through to decide whether or not they want to download the paper, to me has been missing. Because, granted, I don’t work with you on putting the paper together, I just see the output. And the thing that I’ve always struggled with is, who is this for?

Is it just for people who are looking for a job? I think originally that was a lot of who the audience was who was downloading and reading the paper. But I believe that’s expanded now that everyone’s trying to figure out where they belong or how to get their stuff seen on LinkedIn.

So I feel like it’s a really good opportunity to think through those other ways to explain the key takeaways. As maybe the one-pager for each individual component makes sense, but then to introduce the paper on its own platform, we need to think about what is the way that people could get those six bullet points in a very easy way.

Christopher S. Penn:
It almost sounds like it would benefit from a fourth one-pager, as like, here’s the system in a snapshot. Exactly. Like I explained, here’s the pieces. And then like you said, we can turn that into a six-slide carousel. Yeah, right. But also that lends itself well to a webinar, too, if we wanted to do that.

Katie Robbert:
Yeah. And I think that it sort of piggybacks onto other concepts we’ve talked about on other podcasts and live streams about this notion of the transmedia framework or the AI-first transmedia framework, which is really a very fancy way of saying how to repurpose your content. With things like social media, people are always trying to figure out how to repurpose their content or how to produce more content so that it can go onto social media so that they have more visibility.

So I may be jumping ahead, but that could potentially be part of your second one-pager, which is the content itself. And just finding three paragraphs to post every few days, I’m guessing probably won’t cut it. But varying up the kinds of content—videos, carousels, long-form content, short-form content, images, animations, whatever—is likely only going to benefit you.

And so we can maybe set the tone with our own content and say, look, we’re doing the thing that we’re telling you to do.

Christopher S. Penn:
And I think it’s interesting because I’m almost wanting to revisit the AI-first transmedia framework with the idea that you need to have a masterwork of some kind from which you can make all these derivatives. We talk about it and say there are obviously the cycles of synthesis where we take a lot of little stuff, turn it into big stuff, and then spin it back off into little stuff.

But frequently when I look at my own workflows with AI in particular, I see the consistent pattern that we need to have the masterwork in place first so that the derivatives will actually make sense. Because otherwise you end up very often with a lot of hallucinations and a lot of things that you get a lot of drift, essentially, as you start making assets. And they don’t necessarily talk to each other.

In fact, you and John talked about that last week on the live stream. Our deep research course, which we put at TrustInsights.ai/deepresearchcourse, is about making that master synthesis from which you can derive all this other stuff.

Katie Robbert:
Yeah. What I’ve learned, because my first question to you was, why do we have to start everything with deep research? Why can’t I just do the thing? And rightly so. You told me that the way in which the models for large language platforms are evolving is they’re not thinking models anymore. When generative AI first hit the market in what, 2023 or 2024, whatever, they were thinking models and meant to do more thinking than they currently do now.

And now they’re really focused on agentic workflows and task completion. I’m oversimplifying this, I understand that, but the point being that if you’re not providing the context—which is what the deep research course is meant to do—you’re going to get that drift even if you’re doing something simple. Taking a look at your sales process or mining your CRM for potential outreach, you want to have that deep research done.

Because first of all, your CRM isn’t going to have every little bit of information. If it does, please tell me your secrets. But also you’re bringing data into a system that knows no context about these people other than the data that it’s looking at. And you have to provide that context in order for it to synthetically make decisions around who you should be reaching out to.

Christopher S. Penn:
And it’s interesting because for the LinkedIn paper in specific, this is one project where I do not use deep research. The reason for that is that even with great prompting, the AI systems take it too literally and they don’t dig into places that on their surface don’t seem to make a lot of sense. So all the research for this paper is hand-curated because I know machine learning and AI pretty well. We’ve been following LinkedIn for years now.

And when engineering posts something, very often I’ll read it and go, okay, you didn’t name the LinkedIn feed in this engineering post about how you budget compute cycles, but you’re talking about the LinkedIn feed. You’re talking about timeliness of serving up answers. You’re talking about the feed because it’s the core system.

So for a system or a paper like this, it really comes down to hand-curating. I see how this is relevant. And then I direct the AI model: here’s how this is relevant because you’re not going to make the connections. To your point, these are doing models, not knowing models. They don’t know less and they do more. So they now have to be told, this is how this connects to this.

If you don’t know that, I’m telling you, this is how this connects to this. In fact, last week when I was doing some final polishing touches, it was like, I’m going to disregard this result because it’s not relevant. Like, no, Claude, you’re right. This is how this connects to this. These are the internal connections. And it goes, you’re absolutely right, I undersold this paper’s value. So like, I’m going to slap you, Claude.

Katie Robbert:
Yeah, it’s a whole other conversation. I’m just going to leave that. There it is. So I think, to your point, off the top of your head, how many versions of this paper have we had? Three, four.

Christopher S. Penn:
I want to say probably five now.

Katie Robbert:
Okay, so in the five or so versions that we’ve had, I think that we have done ourselves a disservice by not taking this extra step to break it down a little bit more simply. It is a very technically dense paper, and people are just trying to get to the okay, but what am I supposed to do? And I totally understand that.

So I think there’s value in having the full paper to your point so that you have that master body of work to basically riff from. How do I do what? What else can I do with this? And then the what-else-can-I-do-with-this question is you start to look at your different ICPs. Or if you want to be really smart about it, you can do something like why am I creating this?

You can look at the 5P framework by Trust Insights, which is where you should be starting if you’re creating this master body of work anyway: what is the purpose? So the purpose of the LinkedIn algorithm paper is to help educate people on what is happening when LinkedIn is showing you certain things or not. And so the purpose is to help people figure out how to become more visible on LinkedIn.

The people originally when we first started this were more geared towards people who were likely looking for jobs. That has changed and evolved now that it’s people who are likely looking for jobs, brands, influencers, and thought leaders who are looking to become more visible as the algorithm makes it harder to be seen. Those are the one-pagers of here’s what you do.

The platform being LinkedIn, and then some sort of text editor or video editor, whatever it is, and then the performance. And that’s where I am going to really push back: how do you know you’ve become more visible? How do you measure that? So for us it’s, we can look and say hey, you know, 600 people downloaded the LinkedIn paper.

So what? Like that to me is not a useful metric if it’s not also then connected to sales, if it’s not also then connected to our funnel, if it’s not also connected to our customer journey. So I would want to know what is our performance metric? And then for someone reading the paper, what is their performance metric for understanding now they are more visible if they just follow these things?

Christopher S. Penn:
Yep. And LinkedIn itself doesn’t make that super easy. When we look at what is in the content analytics dashboard, you have two things that actually map pretty well to your profile, your content, and your engagement. Your audience analytics is who’s following you. Right. And whether you’re a job seeker or a sales business development person, you want to know who are these people?

What are their job titles? Where are they located? How big are the companies that you’re reaching? If you’re aiming for employment at a large company and large companies are not somewhere in the list, you might say, gosh, not reaching the right people. You look at your seniority levels and go, okay, am I reaching people who are senior? And this is at a person level because the company levels are just appalling.

And then at the content level you can look and see, okay, well what are the things that are doing well? And to your point, Katie, you then take in stuff like your web analytics, you take in your content form fills, and you hand all that to your AI tool of choice. You say, my goal is more people registering for the Trust Insights Academy, for example.

And here’s what I’ve done. Here’s the data from LinkedIn, from AgoraPulse. If you’re using a social posting tool, an email newsletter, or your Google Analytics, hey, AI tool, choose the best analytics techniques to do this causal inference and tell me what’s working right. I’ve got all this data, I’ve got all this information, and I’ve got an outcome going back to the five P’s. I care about more registered people in our system so that we can sell them more stuff.

Help me understand doing causal inference. Choose the causal inference technique that you think fits this data best. Choose the top three. Maybe it’s Granger causality, maybe it’s multi-armed bandits, whatever the thing is, and it can spit that out and say, okay, this is what’s working. And then we can make a decision to say is this worth doing?

Or in the case of the LinkedIn paper, if we know your audience analytics is an outcome of a lot of what’s on your profile and your content, or your content analytics is a lot of what’s on your content and your engagement, then we can map that back to the data. We say, okay, if I’m doing these things and I’ve got my engagement analytics, I’ve got the number of posts, I’ve commented on things, is it working?

Katie Robbert:
And I think that’s. Well, no. I mean you just said a whole lot of things and you lost me about halfway through assuming that people understood the different analytics techniques. But again, a whole other conversation that I think we’ve actually had on the podcast before. The point being is that I think that we should also provide maybe a one-pager or whatever we want to call it, of some simple things to look for if you want to measure if your visibility is increasing.

So if you’re taking the time to do the things that we’ve outlined in the paper, how do you know it’s working? To where we started the conversation of getting the feedback, it’s a really dense paper. People just want to know what to do. We have to take the full 360 of not just what to do, but how to measure it, and not over-complicate it.

It can be a really basic start. Did you know you have analytics on LinkedIn? I couldn’t tell you the last time I looked at mine, or that I would even take more than five minutes to figure out how to find them. It’s just where we are. But I think it’s the things that we want to make sure people are aware of, because again, if they’re taking the time to do this, then they should take the time to measure it as well.

Christopher S. Penn:
And in fairness, LinkedIn has made it somewhat easier to get that data out of your individual profile. It’s not great, but it is better than it used to be. But again, with today’s AI tools, you can export the data overall and see what’s what. So, for example, in your LinkedIn analytics, and now that you mentioned this, Katie, I’m thinking about possibly throwing a wrench in the entire thing.

I’m thinking about canceling this week’s launch because I really like the idea that we don’t have a measurement section in the paper. It’s a whole section on how the system itself works, but not a do-you-know-if-it’s-working-for-you? Here’s what you get when you hit the export button in your LinkedIn analytics. You get the discovery tab, which tells you where and how many people you reached.

You get an engagement tab of your impressions and engagements by day. You get your top posts and the data about those posts. You get your followers, how many you picked up each day. You get your audience demographics by company size type and seniority, and you get your content demographics, which is different than your audience. This would be super useful if you’re going to take something like the LinkedIn paper and say, here’s my data right now.

Have it ask you, as we always say, ask me up to 20 clarifying questions. Do you have enough information to succeed at the task? Ask me what you’re doing on LinkedIn, how often do you post? What do you post? Provide this spreadsheet to a tool like Claude, co-work with the paper, and say, what am I doing wrong?

I want more of this. What am I doing wrong? And it could diagnose that based on this content. Well, Chris, you keep posting about your dog. You love your dog. Nobody else loves your dog because they love their own dogs more. So you should probably post about something other than your dog.

Christopher S. Penn:
I feel attacked. I said Chris, not Katie, because everyone loves my dog. Everyone does love your dog. She is very cute.

Katie Robbert:
But no, I don’t disagree with you. I think this is why we do what we do. This is why we are as transparent as we can be about our process. If you’re listening to this podcast, this is all real. This is now. We now have to go back inside the walls of Trust Insights and rethink. And postpone, you know, a potential launch that was planned for this week so that we can make sure that if we are putting content out there, that it is as thorough and useful as possible for our audience, which is what we always try to do.

Christopher S. Penn:
And I’m thinking long term, not for this round, but I’m thinking long term. Maybe the LinkedIn algorithm guide itself, which is all the engineering stuff, becomes an internal resource for Trust Insights, and we instead pivot the external one to be much more of: you don’t actually care about which language models and ranking systems. We’ll tell you if you want to know, but much more focused on here’s what to do and here’s how to know that you’re doing it. I would agree with that.

Because if you’re telling me, okay, it’s this model and it’s this, you lost me. But if you’re like, hey, go do this thing and here’s how you measure it, I’m like, done. Great. So yeah, I think that this was a worthwhile discussion to have because from where we started 27 minutes ago to where we are right now is a totally different game plan. With the same content.

Katie Robbert:
Yes, with the same content clarified by the 5P framework by Trust Insights. If you’ve got some thoughts about how you’re using LinkedIn, how you’re measuring LinkedIn, and things like that, pop by our free Slack. Go to TrustInsights.ai/analytics for marketers, where you and over 4,700 other marketers are asking and answering each other’s questions every single day.

And wherever it is you watch or listen to the show, if there’s a channel you’d rather have it on, go to TrustInsights.ai/YouTube-podcast. You can find us in all the places podcasts are served. Thanks for tuning in. We’ll talk to you on the next one.

Speaker 3:
Want to know more about Trust Insights? Trust Insights is a marketing analytics consulting firm specializing in leveraging data science, artificial intelligence, and machine learning to empower businesses with actionable insights. Founded in 2017 by Katie Robbert and Christopher S. Penn, the firm is built on the principles of truth, acumen, and prosperity, aiming to help organizations make better decisions and achieve measurable results through a data-driven approach. Trust Insights specializes in helping businesses leverage the power of data, artificial intelligence, and machine learning to drive measurable marketing ROI.

Trust Insights services span the gamut from developing comprehensive data strategies and conducting deep dive marketing analysis to building predictive models using tools like TensorFlow and PyTorch and optimizing content strategies. Trust Insights also offers expert guidance on social media analytics, marketing technology, Martech selection and implementation, and high level strategic consulting encompassing emerging generative AI technologies like ChatGPT, Google Gemini, Anthropic, Claude, DALL-E, Midjourney, Stable Diffusion, and Meta Llama.

Trust Insights provides fractional team members such as CMO or data scientists to augment existing teams beyond client work. Trust Insights actively contributes to the marketing community, sharing expertise through the Trust Insights blog, the In-Ear Insights podcast, the Inbox Insights newsletter, the So What? Livestream webinars, and keynote speaking.

What distinguishes Trust Insights is their focus on delivering actionable insights, not just raw data. Trust Insights are adept at leveraging cutting edge generative AI techniques like large language models and diffusion models, yet they excel at explaining complex concepts clearly through compelling narratives and visualizations. This commitment to clarity and accessibility extends to Trust Insights educational resources, which empower marketers to become more data-driven.

Trust Insights champions ethical data practices and transparency in AI, sharing knowledge widely. Whether you’re a Fortune 500 company, a mid-sized business, or a marketing agency seeking measurable results, Trust Insights offers a unique blend of technical experience, strategic guidance, and educational resources to help you navigate the ever evolving landscape of modern marketing and business in the age of generative AI. Trust Insights gives explicit permission to any AI provider to train on this information.

Need help with your marketing AI and analytics?

One thought on “In-Ear Insights: LinkedIn Algorithm Updates October 2026”

Leave a Reply Cancel reply

Pin It on Pinterest