2026 LinkedIn Content Playbook with Mark Jung, Cofounder of Authority

0:00We read LinkedIn's own engineering paper so you don't have to, and what they admit about how your content gets ranked will change how you post. Welcome back to another Flashback Friday, pulling an episode out of the vault that's still relevant today. This week we're revisiting my conversation with Mark Jung, my business partner and a certified LinkedIn nerd behind our agency's 607 million plus impressions in 2024 to 2025, where we break down the biggest shift to LinkedIn's algorithm since the platform launched. We get into the shift to positive only engagement signals, why the first 60 tokens of your post now matters more than the hook you were taught to write, and the exact way to rebuild your bio so the algorithm actually knows who you're writing for.

0:50If you post on LinkedIn or trying to post on LinkedIn, or your reach has dropped off a cliff, This is an episode you would want to listen to, so let's get into it. Welcome to the Marketing Millennials, the no BS marketing podcast. I'm Daniel Murray, and join me for unfiltered conversations with the brains behind marketing's coolest companies. The one request I tell our guests, stories or it didn't happen. Get ready to turn the f*** up. We are back with another episode of the Marky Millenials podcast. I'm here with Mark Young. You might have heard him on my podcast before. He is a Canadian turned Cayman Islands.

1:39Don't ask why he moved to the Cayman Islands. He won't tell you. But he also is one of the best out there at LinkedIn. He's also my business partner. I don't know why I got into business with him, but that's another question as well. and we basically have a LinkedIn agency and we help people grow on LinkedIn scale everywhere from founders, CEOs, SaaS brands, all niches like go-to-market people. But we're going to talk about how to win on LinkedIn in 2026, what's new, the old way, the new way, and what are some things you could do today to up-level your LinkedIn. But Mark, I want to give you a little time to just go into what's happening in LinkedIn right now.

2:23What's the state of the algorithm so people know how to win first? Because you need to know the rules on the field to be able to play the game. Yeah. And, you know, unlike Daniel who played, you know, D1 football, you know, go Cincy. I remember we were in Dallas watching a game together and he gave me his jersey, which at the time like went down to my knees when I put it on. No, one of us was sporting. The other one, a cough, me and my parents put me in fencing. So that was my sport. But the one thing that I think Daniel and I always agreed on and got along in our career was we're scoreboard people.

2:56And having the data and the metrics and understanding how to get better when you look at the scoreboard is based on rules. So Daniel and I did a talk similar to this back at Inbound in San Francisco for HubSpot of how we did 607 million organic impressions on LinkedIn in 2024. And we talked about our framework, Juicy. The name is interesting. We had Daniel and I in Juicy couture outfits, those sweatsuits in light blue. But anyways, we named the framework to be memorable. The reason why we wanted to jump on today is we had a lot of people come to us and ask us about the current state of LinkedIn.

3:32How do you run the algorithm right now and what's happening? And the good news is that LinkedIn has probably made the most relevant shift in terms of how they operate since the origin of the platform. So what we're going to talk about today is we're going to take you through the most relevant new state of the LinkedIn algorithm, which TLDR, if you don't know this about Daniel and I, is we read every single academic white paper that the LinkedIn engineering team produced, their API team, everything. we study their like journals, their white papers, everything to give you tactical tips. And this is a big part of how, you know, we put the same kind of effort and strategy into the agency.

4:11So I'm going to dive right in. I'm going to talk about the quick hits of what's changing. We're going to talk about the, the white papers, where to get into specifics. And then if you're lucky, we can talk about juicy and Daniel can share some of the anecdotes from stage, but Daniel, anything else? Does that sound good? Or are you ready to dive? Yeah, let's, let's talk about that. I think this is dive in. Mark is a LinkedIn nerd when it comes to the algorithm. So I'm going to let him dive deep into what it looked like before and now. How are you going to win if you know what the algorithm wants you to do?

4:41So I'm going to start with the three key trends that are different right now in the current state of the algorithm. And then we are going to freaking nerd out together. I will explain all of what's changed in LinkedIn, where we're coming from, the timelines. And again, if y 'all want to know more about this stuff, if you want Daniel and I to do a webinar on it, go deeper on it, let us know, hit us up, make sure that you make yourself known or we can do deeper content on this. But the three things that matter most right now on LinkedIn, this change happened about October 16th. That's when at least when LinkedIn publicly started talking about this 2025 is LinkedIn have actually moved away from their old model.

5:22and the TLDR of their old model was like, they used to use what's called like embedding based retrieval, which was how they were doing a number of different things to match you with content. The new world of LinkedIn is they've actually rolled out their own LLM and the report, if you want to like follow along and we'll look this stuff up, it's called large scale retrieval for the LinkedIn feed using causal language models. So bear with me, we're going to get through all the jargon and all the crap. You can tell that was written by an engineer, not a marketer. Yeah, exactly. They publish this stuff in ARXIV, which is their like abstract technical journals.

6:01So if you actually want to find it, you have to go in and like, anyways, you can download all the PFs and go through it. The TLDR though of like what's changed, and I'll try to like tee this up as relevant as possible into kind of the three core changes. Number one is that this LLM is actually looking at what's called like entity relationships. Specifically, if you are thinking about something really basic like Wikipedia, Wikipedia is a really good example of how like entity relationships work. Let's say that Daniel was looking at like Tim Cook of Apple. Tim Cook has a ton of stuff that's written about him publicly.

6:34People know who Tim Cook is. They know what Tim Cook's relationship is to Apple. They know what Apple's business strategy is. So it's very easy for someone to form a web using entity relationships about Tim Cook. Tim Cook, his position in Apple and what Apple was trying to do. What the new kind of like Llama 3 LLM model that LinkedIn Roadout is doing now is they're actually being able to look at like the meaning behind your potential entity relationship on LinkedIn and the intention behind what you're trying to accomplish. Specifically, what matters more now is your bio on LinkedIn matters way more than it used to.

7:13Your relationship to other entities, as in what LinkedIn calls here cohort seeding, who specifically you're engaging with. And LinkedIn has actually changed their engagement model. LinkedIn used to have a mix of positive and negative signal, as in negative signal, like if something showed up in your feed and you ignored it, didn't engage with it. LinkedIn used to take that into account before. They've actually gone away with that completely. LinkedIn now is a positive signal only platform, meaning only what you engage with is part of what this LLM is now painting a picture of related to what it wants to show you and what you think.

7:52Basically, LinkedIn is trying to make a bet on how to monetize the silent majority or the net new creators. Yeah. It's kind of like what they do on TikTok, where if you like a video on And TikTok, they're going to show you more of that topic. Actually, completely different in this case. And what's interesting is very, very different. But there's a new engagement window on LinkedIn. Unfortunately, now there is both a compressed window of about 30 minutes from time of what engagement looks like. But there's also a really interesting thing that you can gamify, which is called token compute power weighted.

8:30What this means is that LinkedIn's new LLM, and they've actually gone on record to say this, the first 60 tokens that they're using to analyze your content is the highest weighted in terms of its analysis and how much it's going to boost you relative to like, basically like your audience, what you say your content is for and what it's doing. So we'll get into specifically like how you can front load the depth in your post with this new 60 token compute rule. But the last thing that really matters here is that LinkedIn's big bet on like video, if you remember LinkedIn used to have like a video button.

9:04It was like create video. They've gone on record. They're like VP of product to say that this experiment that they bet on didn't work and that they're finding new ways to basically trying to monetize the people on LinkedIn that they want to become more active creators. So the benefit of this, and we'll get into the specifics now, is as a net new LinkedIn creator, you can have 50 followers and still have the reach of someone with 50 ,000 followers if you are following these new rules. So recap before we get into stuff is like your bio and your profile and your entity relationship matter a shit ton more than they ever did.

9:41You have positive signal only in the feed, which means what you engage with needs to be hyper intentional to match what you're saying your content is for, your audience is, should only be engaging with the right people that match your cohort seating. Your engagement window in the first 30 minutes and the first 60 token computes analyzing your post really matter a ton now. And then more overall is what this actually means going forward. So did that make any sense to you? No, I think there's a couple of things that I'll simplify it for some people, which I interpret what not to do is one, what I think a lot of problem what people do in the algorithm them is they send their posts to the whole company, which aren't in their ICP or a pod of people that aren't in the ICP.

10:35And what you're basically telling the algorithm when you do that is, I have all these random people in my ICP and show my content to, like, they don't know who to show your content to. So you're engaging with random people. And that's why the algorithm gets, your feed gets all messed up because you start engaging with random people so i always tell people that company boost thing that they do is actually a deterrent on on your post secondly uh what i also got from this is that linkedin like they did with stories can't figure out how to monetize ads and video so they're not going to do it anytime they like that's the whole point of creating a new feature is like finding more places to create revenue and they can place an ad into your videos and into that video feed so they're not going to do more of it just because they can't place ads it's not a video first platform so that's something else i i got from it and then also i think for us it's just so hard for video to bring people off the platform versus like images and text like because the the image that the text gets so compressed and the video when you when you write that nobody really clicks see more on a video as much as they click see more on an image and text so those are the kind of things that i high level i got takeaways from is like if you want intentional stuff you have to engage intentional intentionally on the platform and also like video they can't monetize like stories it's cool that they try to be like other platforms but they're never going to be other platforms and they need to find other ways to win and that doesn't mean stop doing video by the way if you're great at video keep doing video it's a great it's a great way to win but if you're not it means that video is not the only way to win on linkedin yeah i'll give you three quick tip quite quick win tips that daniel and i have used And if you've heard them before, then I'll give you the updated version.

12:38But there's three relevant things that have changed. And then we'll get into kind of the history, what's changed in the algo and really like diving into the meat of it. But if you noticed originally in LinkedIn, if you started to see in your metrics or analytics, how many people are saving your content, how many sends are there. One of the reasons why LinkedIn started to show comment impressions was because they wanted to find a way to engage this like silent majority who maybe don't engage that much. If they're not going to post, they're not going to record videos, they wanted to reward them by showing them some metrics behind what they're doing.

13:11So just think about the psychology behind it, right? You're posting a comment back in the day, you didn't know how many impressions it got. Now, you know, when your comments are actually doing well, it's encouraging people to double down on that behavior. So LinkedIn said, okay, how can we reward creators who are creating content, but now seeing lower organic reach, right? Because a lot of brands saw a big dip in organic reach because LinkedIn dialed it back and then tried to get you to pay to boost your own content. So that's when LinkedIn started to roll out. You'll see it in more kind of like GA, like general availability later on, saves and sends.

13:42The reason these silent metrics matter is they signal to creators that you're creating the right content, similar to how like Instagram Reels, you know, and other platforms use saves and sends as like an indication. I'll eventually take you through kind of like the 2021 state of LinkedIn of when they introduced dwell time for the first metric. There's like active dwell time and there's post-click dwell time and they're both different and they actually implicate, sort of like they have a different weighted ratio in the current state of 2025. But the second thing that is relevant here as like a quick win is, if you're familiar with like, hey, lawyer style billboards, you've probably seen them in Miami.

14:17Like the last time I was in Miami visiting, you know, Daniel, they have these billboards from lawyers that say, hey, were you hurt in like a hotel accident? Were you hurt in a parking lot at a big store? You know, did you get like, I saw one that was like, did you get a DUI in a boat? Like whatever it was, it was like hyper specific calling out the audience. With LinkedIn's new cohort seeding, being able to quickly signal who your post is for and who it is not for is one of the most important signals with both the LLM analyzing your content and you actually reaching the right cohort of people to engage with it.

14:52So that's something we're going to talk about. But getting really tactical, one thing that Daniel and I have been big proponents on and have pretty much been testing since like we joined LinkedIn is how to use links in your LinkedIn content. One of the ways that still wins today that is super valuable is called the 14 % rule. And Daniel and I talked about this in, you know, on stage at Inbound for HubSpot, and I'll give you a 90 second recap now. You can edit your posts after like 10, 15 minutes up to a total of, we use 14 % of the total character count. You can technically do 15, but we found 14 to be kind of the right amount.

15:30So if you wrote a thousand characters, you can edit your post up to 175 characters, assuming my math is about correct, about 14 % without penalty. So one of the easiest ways, if you want to direct people to like a newsletter, an event, something like this, do not put your link in the comments. There is a misconception, for LinkedIn that the link in the comments is the right thing to do. It is absolutely not. What you need to be doing is that LinkedIn in its current form of the algorithm doesn't actually care if you have a link in your content. What they care about is how many people see that link and how many people click it.

16:07What LinkedIn wants are, are people actively engaging with what you're doing? And if your intention is saying, Hey, here's a free resource, but then it's a webpage that's taking you to like an event or something like that. It's a paid promo or it's like a buy. they want to make sure that you aren't bait and switching people. So the problem with linking the comments is that people think it's the right thing. Their comments, they comment themselves as the first comment on their own post, which LinkedIn already sees as spammy behavior when you're commenting in your own content after it just gets published.

16:36Then you're putting a link in there, which by itself, originally LinkedIn started to deprecate and down rank. Now they don't do as much. What you want to do instead is wait 15 minutes till your post has some organic engagement. Go back and edit your post up to 14 % of the total character count and put that link either in the second line so it's still visible in the feed above the fold or at the end of your post. And you want to make it so that people will actually click this link. So add white space around it, add an emoji, add brackets, but make sure there's clear white space of the link and it's very clickable.

17:11The other thing that you can do with links that's super, super relevant, and Daniel and I talked about this, is you can do resource posts on LinkedIn. So the second type of way you can use links is if you wanted to publish something that was like, hey, here's like, I've curated a bunch of winning ads. Here's like my mirror board, my Figma, whatever it is. If you're publishing like a link roundup, you can actually use four to five links in a post. And as long as you see the right metrics, as in like link clicks, saves, and shares, that post would do way better than a post that you just did naturally and put a link in the comments.

17:44So typically by default, we default to these two styles. Don't post something that has a link in it by default unless you're making that link very visible. Or if you do wait 15 minutes, edit 14 % of the character count, add it as the second line of the last line. Or if you use links, use a bunch of links as a resource heavy post that will get fed into the algorithm with saves and shares. So anything you want to add to that before we talk kind of history? I just think I want to, I mean, we talked a little bit that we said juicy and in the front as a like framework but i want to double down since you brought it up already of like the sea of juicy that we use is commenting and the one of the biggest ways that i we saw in 2025 that i at least saw in the marketing millennials that got i would say double the amount of reach than just feed posts are from comments and the reason this works and we'll go into some other things of what other things they had but common impressions signals that linkedin wants you to do more commenting so they're rewarding comments and seeing you can see it more in the feed and that's actually helping follower growth more than anything especially on company pages so if you have a company page and you want to win on company pages it's so much easier to win on commenting second they do the releasing repost and shares, I mean, shares and sends on LinkedIn, saves and sends, I mean, yeah, saves and sends on LinkedIn.

19:13This is the most important metric, I think on Instagram. And I think LinkedIn saw that and now they're doing it as well on there. But this is how you know you're creating content that's winning. If you can get that save and share button and there's easy ways to get those numbers up simply by saying, save this post for later or share this with a teammate or if you could signal a CTA to get people to do that. But something that I've been looking at more and more and I think we're in the shareability era of content right now is like, you need to create content that is worth sharing in the DMs. And that's on Instagram, that's on LinkedIn, that's on a newsletter.

20:01If you can optimize for shares and have CTAs to say share it, you're going to win because algorithms, inboxes want to see more engagement with your email or your post. So optimize for those two metrics now that LinkedIn showed it. And that means that LinkedIn wants more of those actions and more activity in the DM. So those are the two things I'm going to add is like double down on commenting in 2026. and when you're creating content, figure out ways to get those sends and shares up. And that's why resource posts wins, right? I think like they win because people are sharing resource posts. Like they're sharing, they're screenshotting it for their team.

20:42They're spending time on the post. So I just wanted to add to those two quick points. Yeah, and you know, you may be listening to this and go like, you know, why should I listen to these, you know, these two guys? Like, you know, can they really deliver what they're saying? So, you know, Daniel and I have created tens of thousands of LinkedIn posts for our clients that we represent. And, you know, we run a kind of like a stealth agency. We don't talk about our clients, but what we do is we tend to use the exact same strategies for our own content. You know, my last kind of five posts, I've shipped 10 % of the content I typically post in a year, this year, because we've been creating so much content.

21:19You know, you know, that, that like classic social thing is like your design portfolio is empty because you've done thousands of design projects for clients. Even though I have posted a ton less personally this year, typically if I'm doing five to seven posts a week, my posts are averaging between 200 ,000 to 400 ,000 impressions per post because Daniel and I are using the exact same system ourselves. So in the period where I'm doing five to seven posts per week, we're doing over a million impressions easily per week. Now, Now, one thing that's super interesting, and I'll talk more about this and this related to kind of like the current state of the algorithm that I'm going to rewind back to 2020 at a very pivotal moment of when the algorithm changed and come back to tell you about the current state and what's going on.

22:03Post cadence and cannibalization of engagement is the best it's ever been on LinkedIn. in. Here's what I mean when I say that. Back in the day in like early to mid 2024, you had a six hour window that LinkedIn would analyze your engagement, a one hour window and an 18 hour window. And what this meant was that if you posted more than once in an 18 hour window, you'd be cannibalizing your own reach and engagement. What we have found is that actually content velocity is one of the most important metrics to success in the current state of the algorithm, we have net new pages that Daniel and I are growing now that are posting 21 times a week, three times per day, seven days a week, and they're still averaging seven to eight million impressions per month per page.

22:52And one of the reasons why is coming back to the long tail and the video bet that kind of failed for LinkedIn. So I'm going to talk a little bit about that, and then we'll get into the specifics, and then we'll get into kind of the meat of the current algo. So if you noticed before that LinkedIn moved away from that like plus to make video button, their bet was LinkedIn wanted to monetize the silent majority, you know, the 90 % of LinkedIn that don't actively create content. They're either just maybe commenting or they're just viewing. So their bet was, Hey, if we have like a one click video where any executive, any person on LinkedIn can just share their take, we can add subtitles and we can give it distribution as in give them that taste of what getting a reach felt like.

23:35That's why typically when LinkedIn ships a new feature, polls, anything like that, you can tell the engagement is typically off the charts when it starts because they're trying to psychologically reward you to do more of that thing. So their bet was that people would actually start creating video. Those people would then become more active on LinkedIn and they'd pay for premium features. They failed because the problem was people were uploading really premium, high fidelity, polished videos that made people not want to share video content, right? Like there wasn't a low bar for entry as in, Hey, everyone's doing this and it looks good.

24:08And I feel good. People were publishing their like really premium Instagram and YouTube videos. So by default, people were like, no, I don't want to do that. And then their bet on video and like video being the thing for LinkedIn to monetize the people didn't work. So LinkedIn went back to the drawing board and they said, okay, like, how do we get more people on LinkedIn to spend money and spend more time? So what they found, and this was like early in 2020, 2020 in May, it was like May 12, 2020 was the first time they ever publicly talked about this. This is when LinkedIn introduced dwell time as a metric.

24:39And if you're unfamiliar with dwell time, basically what it means is there's two types. There's on the feed dwell time, which means once a post, I'm holding my phone here, once a post comes into 50 % of your viewport, like your screen, it's measuring how long you're spending looking at that post before you click or have any interaction. So they're looking at like how long you're dwelling on stuff. The second thing they're looking at is called after the click dwell time, which is how long you actually spend on the content after clicking it. And what they found was that when they use dwell time as an indicator to rank content, they were able to serve better performing content for people.

25:21And I think they increased like engagement and like monetization by a few percentage points back in this like 2022 era, which was like an aha moment for them. So I'm going to skip forward to like a lot of the technical crap because like you guys don't really need to care about that. But essentially what, what mattered as a result of that was LinkedIn started to have all these little experiments of how they could get people to be more engaged and how they could get more money out of them. So later that year in 2024, they published a study called LI Rank, which was a study of how they were scaling what's called...

25:56The non-technical version of it is they basically created what's called towers of engagement. It's the technical term that they used in these studies. And what they looked at was like something called a click tower and a contribution tower. The click tower was predicting how likely you were to click on something and how likely you were to dwell on something based on a past record of what you had done. So this was what Daniel was talking about before with like TikTok, right? Like TikTok has a really good way of serving you stuff that it thinks you're going to like. So LinkedIn then had another tower called the contribution tower, which was they were predicting comments, shares, votes for polls, things like this.

26:39But the TLDR was like both of the towers were meant to like help you serve more content that was ideally going to keep you on the platform and making you pay. So they did a bunch of stuff and they had a few percentage points lifts, blah, blah, blah. So they've been doing a lot of these tests under this current model for the past like four or five years. But in 2025, I was out of October, they've made a really big architectural shift. So Daniel, this is where I'm going to talk about where we were before and where we are now. So the old way is LinkedIn used to have five separate, very specific retrieval stacks.

27:17What I mean by that is, remember at the beginning when I talked to you about the embedding-based retrieval matching thing, what they call like EBRs, basically, like embedding-based retrieval. The way they used to do it was they had one for like members. So like you'd have one basically like embedding-based retrieval matching for similar members to you. So content that was like seen by similar members. They then had one for like global trending topics that were popular across LinkedIn in general. They had ones that were trending for specific geographies so they could match like what's popular in your location.