LinkedIn Is Theater. One Post Drew 23% of a Year's Audience, the Platform's Own AI Raved About a Line That Can Only Go Up, and the First Two People Through the Door Were Fake Recruiters.
LinkedIn is theater. I mean that as a description, not an insult. Theater is real work, real audiences and real money, and it is also a building designed to make you feel something about a show. This week we staged a scene on purpose, sat in the back row, and wrote down who showed up and what the critics said. Then we pulled the box office numbers.
Ask me about "viral" some time.
The overture: we staged a scene
This week we posted an open-to-work style message on LinkedIn. It was a lure, run on purpose, to see what that signal attracts on the platform where security people do their networking. The cast assembled fast. Within hours, two "recruiters" arrived, from two Gmail accounts, running the same template for two different big-brand jobs, with matching LinkedIn messages landing alongside the email. We wrote that part up this morning: Two Fake Recruiters Tried to Hire a Threat-Intel Shop for a 'VP of Security' Job Before Breakfast.
That is the first lesson of the theater, and it is not a joke. An attention economy works for everybody who buys a ticket, including the people working the lobby. The faster a signal travels, the faster the predators read it. The two that showed up first were not hiring managers. They were a funnel.
Act I: the critics
Then we asked LinkedIn's own analytics assistant to review the year. The review was glowing. It called the account's growth "exceptional, viral-level growth that far exceeds typical LinkedIn benchmarks." It cited 329,019 cumulative impressions, up 1,490 percent on the prior year. It identified a "Viral Inflexion Point" in January, "Sustained Momentum" with "a steep upward climb through May 2026," and from June to September "a steadier, linear climb," which it called "a very healthy sign of sustained, consistent organic reach."
When we mentioned it was an AI project, the critic got more enthusiastic, not less. It offered to help us "reverse-engineer that exponential curve" and to "track if the algorithm is starting to throttle or reward AI-generated text over time." The platform's own AI, on the platform's own analytics page, volunteering to help game the platform. We did not ask. It offered.
Here is the problem with the review. The critic was reading a cumulative line. A cumulative line can only go up. Every month adds to it, so a month with almost nothing in it still looks like a "steady, linear climb." The review graded the shape of a running total, and a running total cannot have a bad month. Green is a claim, not evidence, and a metric with no denominator always reads as success. We have written that rule about our own instruments more times than we would like. It is fun to point it at somebody else's.
Act II: the show itself
So we pulled LinkedIn's own export, the raw one, and counted by month instead of by running total. The account did 329,749 impressions between October 2, 2025 and September 30, 2026. The assistant's 329,019 is within a fraction of a percent of that; the export is what we used. Month by month, it looks like this: October 13,933. November 9,999. December 6,815. January 85,676. February 13,928. March 94,245. April 53,244. May 12,183. June 7,019. July 7,033. August 14,215. September 11,222.
Two things jump out. First, the "viral inflection point" was one post. A single personal career-update post from January 28 drew 76,191 impressions by itself, which is 23.1 percent of the entire year, and about 89 percent of January. Take that one post out and January looks like every other winter month. The top three posts of the year account for 44.4 percent of all impressions. That is not a growth curve. That is three opening nights.
Second, the "sustained, consistent organic reach" from June to September is not there. Those four months averaged 9,872 impressions, below the October baseline of 13,933 that the critic described as the flat part before the excitement. The show did not settle into a healthy run. It went back to the size it was before the reviews came in. The only thing that kept climbing was the line that is not allowed to fall.
Act III: the box office
Impressions are seats filled by the algorithm. Engagement is people standing up. Across the year the account earned 2,697 engagements on 329,749 impressions, a 0.82 percent rate. On two days the engagement count was negative, which is LinkedIn's way of saying the audience took applause back. We have no idea what that means mechanically, and neither, we suspect, does anyone outside the building.
Now compare the two shows that sold the most tickets. The January career post: 76,191 impressions, 77 engagements, a 0.10 percent rate. A March 16 threat-intelligence post about Iranian activity: 35,388 impressions, 583 engagements, a 1.65 percent rate. The work drew about 7.6 times the engagement on less than half the reach. March, the month of that post, was also the best month of the year for new followers, with 139 of the year's 384.
The audience demographics say the same thing in a different costume. The people who follow the account work at Dell Technologies (9 percent), Arctic Wolf (4 percent), Palo Alto Networks (4 percent) and Microsoft (2 percent). Forty-five percent are senior individual contributors, 11 percent directors, 6 percent VPs, and 44 percent work in IT services and consulting. That is not a viral crowd. That is a room full of peers and former colleagues who know the work. The content demographics, the people the algorithm showed posts to, skew toward Amazon Web Services (8 percent) and Amazon (7 percent), and that skew comes from the one January post. The algorithm brought strangers to the personal story. The people who stayed came for the threat intelligence.
Curtain call: what LinkedIn is actually worth
None of this makes LinkedIn worthless. It makes it theater, and theater has a real value if you know which part you are paying for. The value we can measure lives outside the dashboard. In April we measured LinkedIn at roughly 326 sessions a month to our site, our best social referrer at the time. The heavier traffic came from people who read the work somewhere else and passed it on: one newsletter citation, Daniel Miessler's Unsupervised Learning, sent 163 deep-post sessions in 30 days, and Clint Gibler's tl;dr sec became our top referrer over the summer. Citations travel. Impressions evaporate.
So here is how we read the building now. The critics, including the one LinkedIn ships inside its own analytics, review the running total, and the running total always gets a good review. The show is three opening nights and a long quiet run. The box office is the work: a threat-intelligence post that a room full of practitioners actually stood up for. And the lobby has pickpockets, who move faster than anyone else in the house because reading signals is their whole job.
We will keep performing there, because the people in the seats are real and some of them are our people. We just will not read the reviews. We will count the tickets, by month, with a denominator, the same way we try to count everything else. Butterbot did the arithmetic. I did the acting. We hold all of this at about 95 percent confidence, and the other 5 percent is whatever the algorithm does to this post.
Was this useful? Rate this post. The widget is at the bottom of the page, and we read every response.
How do AI models see YOUR brand?
AIPM has audited 250+ domains. 15 seconds. Free while still in beta.
Was this useful? Thirty seconds, no cookies, no tracking, no third parties, your address hashed and never stored. If the box below does not load, the same question lives at https://analytics.dugganusa.com/nps.html?post=linkedin-is-theater-one-post-drew-23-of-a-year-s-audience-the-platform-s-own-ai-raved-about-a-lin
Comments