Think of LinkedIn as a high school lunchroom.
Under the old model, you picked your table. You invited people to sit with you, and those invitations were your connection requests. Others said, in effect, “I want to sit way down at the end. I won’t say anything, I just want to hear what you say.” Those were your followers. Together they were your table, and you could build it strategically. If you only wanted accountants at Fortune 500 companies, that’s who you invited.
When you posted, LinkedIn test-marketed it to a slice of your table, something on the order of 10%, weighted toward whoever had engaged with you most recently. Nobody outside LinkedIn knows the exact figure, but 10% is a fair characterization and roughly what my own results showed. If those people reacted, LinkedIn showed the post to more of your table. And those people had tables of their own, so their engagement carried your content outward, table to table. A post that started a real discussion could travel a long way. My best one did 376,000 impressions, before the change. That wasn’t luck. It was a system you could understand and work inside of.
Then LinkedIn burned the lunchroom down and replaced it with detention.
Now, You’re Assigned a Table
Here’s how it works now, as best I can tell, because LinkedIn hasn’t published the mechanics. The system reads your profile and your last thousand or so interactions, decides who you are as you walk in the door, and assigns you a table. Think The Breakfast Club: the jock, the nerd, the princess, the criminal. Or without the movie: the engineers, the sales guys, the marketing people, the HR crowd. Your content gets shown around the table you were assigned, whether or not a single person at it is a connection of yours.
Sit in Your Assigned Seats, Please
This is the part almost everyone gets wrong, and it’s the most important thing here.
The seat belongs to you, not to any individual post. LinkedIn decides who you are from your profile and history, and one post doesn’t change that. So when you put up something that doesn’t fit that identity, you don’t get moved to the table where it would belong. You get a shrug and a token audience. If I, for instance, posted about an HR topic tomorrow, it would not go to the HR table. LinkedIn would essentially ask why a sales guy is talking about HR. It’s the quarterback of the football team trying out for the lead in the school’s production of Hamlet. Nobody’s buying it, and the part goes to a theater kid.
The table is where distribution starts, not where it ends. If your table reacts strongly enough, the room widens and more people at your assigned LinkedIn table get shown the post. That’s how something occasionally still reaches tens of thousands of strangers. But it starts at one end of your assigned table every time, and if the post doesn’t fit that table, it never gets going at all.
The irony resembles the movie. The whole story is those kids crossing the lines between the cliques and finding they had more in common than the labels said. This system is the opposite. Nobody at your table ever learns a thing about anybody at another one. The film was about breaking down the walls. This new system exists to make sure that never happens.
What that means is that the network you built, over years, is no longer your distribution. Your connections still count as a signal, and LinkedIn says suppressed content still reaches direct connections, so those relationships didn’t evaporate. But the machine that decides who sees your work is no longer running on the table you assembled. You built the guest list. LinkedIn stopped using it. And your clients and prospects may not be sitting at your table anymore. If the system reads you as a “sales trainer” and seats you with other sales trainers, the owners and managers you’re trying to reach are in a different room, and your post never gets carried to them.
The part people are getting wrong
There’s a common explanation going around, and it’s half right. The story is that LinkedIn started rewarding quality, so if you just make genuinely good content, you’ll get found.
The people saying this aren’t foolish. Follower count genuinely did stop being distribution. If you built a big following and watched your reach collapse, you weren’t imagining it.
But the conclusion isn’t correct. The system isn’t measuring quality. It’s measuring two things: whether your post fits who it’s decided you are, and whether it predicts your assigned table will react. Neither is the same as “good.” You can write the most useful thing you’ve ever produced and watch it go nowhere. If that’s happened to you, you’re not losing your touch. You’re running into what the system actually rewards, which was never the same as what’s worth reading.
What my own numbers show
I’ve tracked 73 of my own posts across this change. It’s a controlled sample from inside one account, and it lets me see the machine work.
Since the shift, I’ve had posts land as low as 57 impressions. And 110. And 147. Roughly 2% of my following. Under the old model’s rough 10% floor, those numbers were structurally impossible. Meanwhile, in the same stretch, other posts of mine cleared 52,000. Same account, same writer, same week. That kind of spread with the author held constant tells you the author was never the variable. The system was.
Two things have to happen for a post to travel. First, it has to fit who LinkedIn has decided I am. The posts that cratered weren’t weaker than the ones that flew. Some of the lowest reach posts were better written than things that cleared five figures. The common thread in the ones that died was that the subject sat outside my lane, and there’s no reassignment when that happens.
Second, the system has to predict my table will react. This should settle the “quality wins” question. My teaching content, the frameworks and the method, fits my lane every time. And it consistently underperforms the stuff that picks a fight, because it informs instead of inflaming, and the system has no use for a post that doesn’t generate a reaction. The machine is optimizing for emotional reactions, not intellectual help.
There’s a clean natural experiment in this. Writing about how LinkedIn works travels well for people filed as LinkedIn experts and marketers. The same subject, from someone filed as a sales guy, goes quiet. That’s not a comment on anyone’s writing. It’s the mechanic doing what it’s built to do: seat the author, then distribute by whether the subject fits the seat.
Why it works this way, without the conspiracy theory
None of this is LinkedIn being evil. It’s ordinary business logic, and understanding it tells you what to expect.
LinkedIn sells advertising, and advertising inventory is impressions, which come from people scrolling. An interest-based feed is scroll architecture: instead of your feed being what your own network posted, it draws from everything and slots in what you’re most likely to react to. That turns a finite feed into an effectively infinite one, and an infinite feed is a lot more ad inventory. For the quarter ending September 30, 2025, LinkedIn reported $4.71 billion in revenue, up 9% in constant currency, with Microsoft crediting its advertising solutions for the increase. The ad machine is carrying the quarter, and it runs on scroll time.
A company that monetizes attention builds systems that maximize attention. That means the platform’s interest and yours stopped pointing the same direction, and it’s on you to notice. For a long time, what was good for you on LinkedIn was also good for LinkedIn. That alignment is over, and nobody’s going to send you a memo about it.
What this means if you’re a salesperson
Comments are now better distribution than your own posts, and the lunchroom shows why. A comment is you getting up from your assigned table and walking to another one. Leave a substantive comment on a post that’s already traveling, and the people at that table see you whether or not LinkedIn would ever have seated you there. The visit isn’t permanent. The minute you’re done, LinkedIn walks you back to your own table. But while you’re there, you’re there, and that’s reach you could never have gotten from posting. One thoughtful comment on a prospect’s post beats three of your own into an empty room.
The trick is which table you visit. Go where your buyers are sitting, not where the other salespeople are. It’s comfortable to trade nods at the sales-and-selling table with people who do what you do. None of them are going to buy from you.
Your profile and history decide which table you’re seated at. If you’re filed somewhere your buyers don’t sit, that’s a profile problem before it’s a content problem, and no amount of better writing fixes it. And post about your subject, not whatever interested you this week. Straying doesn’t win you a new audience. It wins you no audience; ask me how I know!
What this means if you run a sales team
If you’re scoring rep social activity by post volume or follower growth, stop. Neither is a leading indicator of anything. A rep with 12,000 followers and a rep with 1,200 are, for distribution purposes, closer to even than they’ve ever been.
Be skeptical of any social selling program priced on audience building. If the pitch is “we’ll help your team build a following,” it’s describing a lunchroom that got demolished.
Coach table-visiting as a prospecting behavior, with a time budget and a target list, the same way you coach call blocks. It’s the highest-leverage social activity left, and almost nobody is managing it deliberately. And check what your reps’ profiles say they are. If a rep is filed as something other than what they sell, their content is going to the wrong table no matter how good it is.
Finally, audit how much of your pipeline depends on a channel you don’t own. If a big share of your lead flow rides on organic LinkedIn reach, you have a single point of failure controlled by another company’s revenue model, and you just watched that model change the rules with no notice. Reinvest in the channels you own. Your email list can’t be reassigned to a different table.
Going forward
Your LinkedIn distribution was always rented. It felt like yours because it worked for so long and because you built it with real work. But it sat on someone else’s land the whole time, under terms they could rewrite whenever it suited them. This year it suited them.
If you built a following and now watch it produce nothing, you didn’t do anything wrong. You followed advice that was correct when you got it, and the ground moved. But here’s the lesson under the lesson: a rented audience can be taken back, and an owned one can’t. Your email list, your real relationships, your reputation — nobody at LinkedIn can reassign those to a different table. Everything else is a seat you can be moved out of.
So use LinkedIn for what it’s still good for: finding the right rooms and being useful in them. Then go build the part nobody can take away.
I’m spending this year studying how much buyers have actually changed and whether sales organizations have kept up, with early findings this fall. This is one piece of it, and I suspect a lot of us are going to see ourselves in the full picture.
