Stop Calling LinkedIn a Professional Network

Now hold that next to what the people who make LinkedIn worth reading are seeing. Richard van der Blom studied 1.3 million posts from 50,000 creators for his 2026 algorithm report. His finding: reach for active creators is down about 60% over two years.

So the feed is getting watched more, and the experts are getting seen less. That’s not a glitch. That’s the strategy working.

The last refuge

There was a time when LinkedIn was the one place online you could go to be a grown-up. Facebook had your aunt’s politics. Twitter had the fights. Instagram had the beach. LinkedIn had the people who knew things about your industry and the jobs that came from knowing them. It was dull, and the dullness was the point. You went there to learn and to be found.

That wall is gone. Scroll your feed this morning and count what you see. A vacation selfie with a leadership lesson stapled to it. A newborn baby photo that somehow ends with a reason to buy a CRM. A founder who learned everything about sales from a toddler’s tantrum. Six people who aren’t AI experts explaining AI. A post that ends with “Agree?” And somewhere in the middle, one person who knows their field, sharing something real, with 11 likes.

I wrote about the AI part of this in September, when LinkedIn built the slop machine and then handed out smoke detectors. It shipped a tool to help you write posts with AI. Then it shipped a button to flag posts that “seem like AI slop,” and people clicked it more than a million times in a few weeks. That piece was about one feature. This one is about the whole product, and why it can’t undo what it built.

Engagement Debt

I call it Engagement Debt.

Every choice a platform makes to raise time in the feed borrows against the trust that made people show up in the first place. Open the feed to any kind of content, and engagement goes up. Push posts from strangers into everyone’s view, and engagement goes up. Give people an AI writing tool so they post more, and engagement goes up. Sell a tool that lets anyone filter a billion members and message them at scale, and engagement goes up.

Each move looks like a win on the next earnings call. Each one also takes a little away from why the place mattered. The debt piles up out of sight, because trust doesn’t show up on a dashboard until it’s gone.

Here’s the trap. You can’t pay Engagement Debt back without the number going down. To fix the feed, LinkedIn would have to show you fewer posts, less often, from fewer people. Time in the feed would drop. And time in the feed is the number the CEO of Microsoft reads out loud to investors. No product leader gets promoted for making that number smaller.
That’s what I mean when I say LinkedIn has become a beast of its own making. It isn’t that nobody there sees the problem. It’s that the fix looks like failure on the scorecard they’re judged by.

Is it the people, or the platform?

It’s tempting to blame the users. The people posting baby photos as business lessons. The people farming reactions with personal photos and “thoughts?” The people who discovered AI last Tuesday and now teach it every day.

But people do what the system pays them to do. That’s true on every platform, and it’s true here. If the algorithm rewards a personal photo with ten times the reach of an expert’s breakdown, you’ll get more personal photos. If it rewards AI hot takes because AI is the most searched topic of the year, you’ll get AI hot takes from accountants, chiropractors, and recruiters. Nobody needs to be told. They watch what gets seen, and they copy it.

So yes, the people did it. The platform taught them to.

I’ve written before that your anger is part of the media plan. The same is true for your attention on LinkedIn. Every scroll past a post fishing for reach is still a scroll, and every scroll counts toward that 15%.

The most expensive room in the building

Here’s the part that should bother every B2B marketer who signs the checks.
LinkedIn ads are the priciest attention most of us buy. Industry benchmarks put the median LinkedIn cost per thousand views near $39 for broad B2B campaigns, and close to $78 for account-based campaigns. Meta’s figure in the same roundup, measured for ecommerce brands, sits around $14. The audiences differ, but the gap is wide. You pay close to three times as much to show up in a feed that now looks a lot like Meta’s.

That premium used to buy you context. Your ad sat next to industry news and expert posts, in front of people in a work frame of mind. Now it sits next to a gender reveal with a sales lesson. The price didn’t come down when the context did.

In my own work with B2B teams, LinkedIn is the line item that’s hardest to defend in a budget review. It’s the most expensive media we buy and the toughest to tie to revenue. If you’ve ever caught yourself buying back your own customers and calling it performance, LinkedIn is the easiest place to do it.

Your data, your inbox, their product

The inbox tells the same story from another angle. Open your messages and count the ones from a real person who knows you. Most of what’s there is a sequence: a connection request, a “thanks for connecting,” a pitch, a follow-up, a “bumping this to the top,” and a breakup message. Sometimes from the same company, three people deep.

That flood didn’t come from nowhere. LinkedIn sells Sales Navigator, a tool that lets paying users slice the member base by title, seniority, company, and job changes, then reach out. Pair that with the automation tools LinkedIn’s rules ban but can’t seem to stop, and you get an inbox that works like a spam folder with your name on it.

The data side has a paper trail too. Here’s what’s on the record, and you can draw your own conclusion.

• The fine. In October 2024, Ireland’s Data Protection Commission fined LinkedIn €310 million for using members’ personal data for behavioral analysis and targeted ads without meeting the law’s rules on lawfulness, fairness, and transparency. LinkedIn is appealing.
• The AI training default. Since November 3, 2025, LinkedIn has used member profiles, posts, and activity to train its AI in the UK, EU, Canada, and other regions. You were opted in. You have to find the setting to opt out.
• The scrapers. In October 2025, LinkedIn sued a company it called an “industrial-scale fake account mill” that LinkedIn says scraped millions of member profiles and sold the data for up to $15,000 a month.

None of that proves bad intent. It does show what your profile is worth to everyone around it, LinkedIn included.

The wall for builders

There’s one more tell, and it shows where the priorities sit. If you want to build a tool on top of LinkedIn, you go through a review process that asks for business verification, a narrated screen recording of every feature, and test logins. If you get rejected, you can’t fix it and resubmit. You start over with a new app. And you wait. I’ve waited more than two weeks for an approval response, while most other platforms answer in minutes or days.

A company that wanted builders to make its network more useful would make that easy. A company that makes its money from attention inside its own walls has no reason to.

So what is LinkedIn now?

Here’s my opinion. LinkedIn is the most expensive attention market in B2B, wearing the suit of a professional network.

It still has the best list of professionals on earth. It still has the job graph. It’s still where your buyers spend time, which is the cruel part. I’m posting this article on LinkedIn, because that’s where the people who need to read it are. That’s the trap in one sentence. The platform is worse, and we still can’t leave, and LinkedIn knows it.

That’s also why there’s no real incentive to fix it. LinkedIn’s revenue grew 12% in the last quarter; Microsoft reported it on its own, “primarily driven by Marketing Solutions,” which is the ads business. Feed time is up. Members are up. And now the revenue doesn’t even get its own line. From where Microsoft sits, nothing is broken.

Run the refuge test

This takes five minutes, and the result is worth posting.

The feed. Open LinkedIn and look at the first 20 posts. Give each one a mark.

• Signal: someone who knows the topic taught you something you didn’t know.
• Noise: personal photos for reach, life events turned into lessons, AI takes from non-experts, engagement bait, ads.

The inbox. Open your last 20 messages. How many came from a real person who knows you or had a real reason to write?

Add up your signal. If you got fewer than five of 40, your LinkedIn isn’t a professional network. It’s a feed you pay for with attention. Post your number. I’ll bet most people land under five.

What to do about it

You can’t fix LinkedIn. You can stop letting it run you.

• Treat it as rented land. Post there, but send people somewhere you own: a newsletter, an email list, your site. If LinkedIn changed the algorithm tomorrow, would you still reach your audience?
• Post less, say more. Van der Blom’s data says the ideal cadence dropped to two to four posts a week. Fewer posts with something real in them beat daily filler.
• Audit the ad spend. Make LinkedIn prove its premium against what it costs elsewhere. If you can’t tie it to pipeline, cut it until you can.
• Turn off AI training. Go to Settings, then Data privacy, and switch off “Data for Generative AI Improvement.” It takes 30 seconds.
• Use Sales Navigator for research, not volume. If your outreach would embarrass you if the prospect saw the whole sequence, don’t send it.
• Follow people, not the feed. Ring the bell on the ten experts you trust, and let the algorithm show you whatever it wants for the rest.

LinkedIn used to be the last place online where expertise beat attention; it traded that for a bigger number. The good news is the experts are still there. You just have to go find them, because the feed won’t.

Jeremy Mays is the founder of Transmyt, a marketing firm for B2B SaaS, health and wellness, real estate, and hospitality brands, and a cofounder of Paciva, which builds AI agents. Transmyt offers fractional CMO and digital strategy work. Browse more Transmyt Insights.