CloseX
Linkedin personal branding tips ask you to treat your Linkedin commenting as work. As the 2026 algorithm update tries to make that work pointless it's a good moment to ask what LinkedIn is actually for.
Somewhere in the last decade, the advice about building a professional reputation changed. It used to mean doing work worth talking about. Now, if you follow the mainstream Linkedin personal branding playbook, it apparently means showing up in the comments, ten times a day, within the first thirty minutes, fifteen words minimum, so the algorithm registers that you exist.
Sit with the arithmetic of that for a second: you create your own content, promote it, answer every notification... And then you spend hours scrolling feeds you don't care about, leaving obligatory comments, so that an algorithm will "promote" you in return. That is not a side effect of having a profile. That is a second, unpaid, full-time job, bolted onto the actual career you spent years building.
Which raises a question nobody in the advice industry wants to answer: if your days are spent feeding the feed, when do you do the real work? When do you think, lead your team, or solve the difficult problems for the exact clients you're supposedly on LinkedIn to win? A professional network was meant to support your work but somewhere along the way it started look like we are competing with it.
The mechanism is simple, and that's why it's so effective. Reach on LinkedIn is rationed by early engagement: a post is shown first to a small slice of your network, and only "graduates" outward to strangers and prospects if that slice reacts fast. So the entire game compressed into one instinct: get comments, quickly and everything downstream of that instinct got optimised, including the parts that should never have been optimised at all.
When a single number becomes the goal, people stop pursuing the thing the number was supposed to represent and start pursuing the number directly. This has a name: Goodhart's Law. When a measure becomes a target, it stops being a good measure. The principle was first observed in economics, but it describes the LinkedIn feed perfectly. Engagement was only ever a proxy, a rough stand-in for the thing that actually matters, which is whether your work reaches and persuades the right people. The moment the platform started rewarding that proxy directly, people stopped chasing the underlying goal and started chasing the number itself: not "did I say something worth reading?" but "did I hit ten comments in thirty minutes?" Optimise a proxy hard enough and it detaches from the reality it was supposed to represent. You can run up a perfect engagement score and have it mean nothing, which is exactly how you end up with a feed full of activity and empty of value.
Linkedin is created this way so it can, as any social network, create lots of engagement and active users, which is the metric needed for doing successful ads sale.So it always encouraged this version of Linkedin personal branding advice: the more active you are the algorithm more rewards you. So your first obligation is to create content and the second to promote that content by engaging with content of other people. With the rise of AI both of the tasks started to contribute to the communication noise:
By independent analysis from Originality.ai, more than half of long-form LinkedIn posts, roughly 54%, are likely AI-generated, a share that jumped after ChatGPT launched and has stayed above half since. LinkedIn itself has confirmed that content creation is up around 14% year over year, much of it low-quality. So the feed got louder, faster, and emptier. AI now also contributes to the commenting task by generating automatic commentary.
First "the commenting duty" meant compulsive manual commenting but afterwards it even "evolved" into engagement pods, skipping the pretence of interest entirely.
If it ever feels like everyone on LinkedIn is quietly more popular than you, here is the very very uncomfortable explanation: a large slice of that popularity is pre-arranged. They're called engagement pods: private groups whose members agree to like and comment on each other's posts, on schedule, specifically to manufacture the early engagement the algorithm rewards.
I'm invited into one of these almost every day.
The tell is easy to find once you know where to look. A profile running this play looks electric: hundreds of reactions, threads of comments, relentless activity. But trace it. Follow the first commenter to their own page and you'll find the original poster commenting right back. Follow the second commenter, the third, the people commenting on them and it's the same small cast in a loop, applauding in a circle.
So a meaningful portion of the platform is running on staged enthusiasm. And because of that people get hired, trusted, and put in front of serious clients on the strength of numbers that were assembled, not earned, bringing thin expertise and a borrowed network into rooms that needed neither.
And this is exactly how term personal branding lost all of its meaning and got a synnonim for "fake it until you make it" approach. People who have strong personal brands according to Linkedin personal branding tips are those who have a huge following and an active community - their achievements with respect to the work they do and services they offer are usually abstractly presented or skipped completely.
What the manufactured-crowd strategy misses, and it matters most in Europe, where the business culture is markedly more conservative than in the US is that the people you actually want, the buyers, the decision-makers, the institutions are running the opposite filter.
The 2025 Edelman-LinkedIn B2B Thought Leadership Impact Report, drawn from nearly 2,000 decision-makers, found that 79% are more likely to champion a vendor in an RFP when that vendor consistently publishes genuinely good thinking, and that 53% say strong thought leadership makes brand recognition matter less. The same study is blunt about the inverse: there are tangible risks to publishing low-quality content. And a long line of behavioural research on trust in professional markets points the same way: the moment people sense an interaction is built to extract value rather than deliver it, trust falls instead of rising.
So when a serious buyer lands on a profile and finds a trail of generic comments sprayed across fifteen unrelated feeds a day, they don't read "in-demand expert" but the opposite. Performance, at that level, is self-defeating. An algorithm can fake your reach. It cannot fake your revenue.
LinkedIn claims they never optimised purely for likes. Alongside behaviour, it tries to measure whether people found a post genuinely useful, and it weighs relevance and demonstrated expertise. That second track is what exposed the rot: engagement at the top of the feed kept climbing while satisfaction with the feed fell. The number had been gamed until it stopped measuring anything real, textbook Goodhart's law, where a measure that becomes a target stops being a good measure.
Linedin's VP of Product made it public that their goal is to make engagement pods entirely ineffective, and to keep widening the net that detects them. And detection no longer means catching a lazy "Great post!" It means reading the structure of engagement, are the comments coming from the same accounts every time, do they have unnaturally consistent timing, are there reactions from people with no real connection to you or your topic. Coordination is a structural fact, and structure is precisely what the system is now being re-built to see. At least they say so.
This is why the clever evasions don't rescue anyone. The tricks that beat the content filter, an AI trained to write in your voice, varied comments, fake back-and-forth threads only tighten the cluster, because a small group performing a conversation looks more like a pod, not less. And the tricks that beat the structural filter, rotating the pool, recruiting relevant people, spreading engagement over two days, dilute the one thing that made the pod work: reliable, mutual, early lift. Follow that arms race to its end and you've rebuilt a fake audience the hard way, with worse content and higher cost.
A year or two ago, the pod was a room with a door. The same twenty people piled in early, that noise was the password, and the post walked out into the wider network to strangers and potential prospects.
Now the door is being bricked up. The 2026 system reads "the same twenty people, again, fast" not as a launch signal but as the definition of a closed loop, so the post never graduates. The exact engagement that used to be the way out is now the evidence that keeps you in.
And the penalty is not obvious and with this more brutal: it isn't that your likes disappear but you're shown only to the people already in the room, and the people in the room are the pod. So the comments keep flowing, the dashboard still glows green, the numbers still read like a win. But every reaction is coming from inside the bubble, and the post reaches no one new. This is what a shadow ban actually is: no warning, no notification, just a cap. Reach that used to run into the thousands collapses to a few dozen and stays there.
The person running the pod is the last to know. A hard ban would tell them something's wrong, and they'd adapt. Capping the distribution while the applause keeps playing means they keep feeding the machine, watching the numbers and feeling successful while reaching nobody new. The vanity metric becomes the leash and the loyalty of the pod seals the room: the more faithfully those same people show up for each other, the tighter the cap. They are trapped in an echo chamber they can't see the walls of.
The algorithm didn't build that chamber but they them selves, the day they chose to engage out of obligation instead of interest. For a while the platform subsidised the exit, letting the staged signal buy real reach. The 2026 update simply stopped paying for the way out.
The conclusion I keep arriving at, and it's bigger than the bots.
LinkedIn cannot be your personal branding, and it shouldn't be your focus. It is not a stage you're obligated to perform on every morning. At its best it's a directory and a discovery surface, the place people come to find you, check that you're real, and see the work. The work itself happens elsewhere: in the projects, the results, the expertise, the relationships you build off-platform. LinkedIn is the shop window, not the workshop. Treating the window as the whole business is how you end up polishing glass while the shelves go empty.
And the problem isn't only the engagement bots and the pods. It's the entire commenting culture they grew out of the everyday, widely-repeated advice to comment compulsively on strangers' posts to please an algorithm. For a professional network, that ritual is senseless. It manufactures noise, mistakes motion for momentum and converts experts into content workers.
If you suspect this is the wrong setting for a serious business network, you're not alone, and the most useful thing you can do is say so out loud. Seek out and amplify the people already making that argument. Norms like this survive on silence: everyone privately rolls their eyes at the comment quota, and almost nobody admits it in public, so the quota stands. The fastest way to dismantle a senseless convention is to make dissent from it normal.
We were handed extraordinary new tools to communicate with. The rules of business didn't change to match them. We still play by old-school rules, just with new digital tools. The system didn't kill manipulation; it made manipulation more expensive than sincerity. At the point where faking it costs more than doing it, the person who actually has something to say wins on cost alone.
So stop optimising for a closed loop. Build the real work, and let LinkedIn do the one job it's good at: helping the right people discover it. An algorithm can fake your reach. It cannot fake your revenue.
Sources referenced:
Originality.ai analysis of AI-generated content on LinkedIn (2024–2025);
2025 Edelman-LinkedIn B2B Thought Leadership Impact Report;
public statements from LinkedIn's VP of Product and VP/Executive Editor on engagement-pod detection and feed quality;
independent algorithm-reach analyses (2025–2026).