How audiences get built · Part 1 of 4

There are five ways to build a LinkedIn audience. Only one is on the menu.

We read the posts behind 49 creators on our own list, 164 big accounts and 1,102 mid-sized ones. Five ways of growing kept coming back. Only one can be chosen on purpose, and it is the one that fades.

September 2026 · 7 min read

A speaker with one hand raised, facing a packed hall of people raising theirs
43%

Big audiences that are not real

71 of 164 accounts above 50,000 followers, once we read their posts.

55%

Gone before anyone read a post

Five fixed rules removed 1,327 of 2,429 accounts at 10,000 to 50,000 followers.

22%

Audiences the posts built

532 accounts in that group grew from their own posts.

4–5%

Worth paying

Around a hundred passed every check.

01 · The filter

Half the list is gone before anyone reads a post

The follower count is the first number anyone looks at and the least useful. So we read the accounts one at a time and sorted them by how each audience was built.

Three groups. The roster is our own 49 creators, from Sammie Cooper at 10,814 followers to Paul Storm at 667,975. The index is our database of LinkedIn accounts; 164 have more than 50,000 followers, the biggest Alex Xu at 1,031,564. The band is 2,429 accounts at 10,000 to 50,000 followers, of which we read 1,102.

Before anyone judged the writing, five fixed rules removed 1,327 of the band, 55%. The biggest rule is the simplest. Engagement means likes, comments and reposts added together. If everything an account has ever had adds up to less than its follower count, it is out. That rule alone removed 795. The other four catch software and job boards rather than people.

What is left of 2,429 accounts at 10,000 to 50,000 followers

1,327

Removed by rule

The numbers alone ruled them out.

896

Real audience

The posts and the audience fit each other.

201

Doubtful

Something in the posts does not add up.

5

Not real

The following did not come from anything posted.

The two headline numbers measure different things. The 43% above 50K is a judgement made after reading the posts; the 55% in the band is a filter run before reading any. Do not compare them.

02 · The five paths

Five ways up, and only one you can choose

Two terms first. A typical post is the middle one when you rank an account's posts by engagement. Lifetime engagement is every like, comment and repost an account has ever had, added up.

The template grind

Slow to start · fades when the posting stops

The same post shape, 20 to 40 times a month, for years. Sandra Pellumbi ends 87% of her 631 posts with the same three-part request to her readers. Sandhya Mishra took 64.9% of her lifetime engagement in one year of 416 posts, then posted 25 times in 2025 and faded.

You can choose it300 to 1,000 posts

Career as content

Instant when the job changes · lasts as long as the title

The job writes the posts. On every career account on the roster, “I’m happy to share that I’m starting a new position” beats everything else the person has written. Three roster creators each took a decade this way. Sandy Carter’s typical post fell to a third the month she left Amazon.

Half a choiceNeeds a job people know

Personal story at volume

Fast once the format lands · slow to find

A photo and a story told as “I”, 500 to 2,000 characters, told again whenever it works. Muskan Agarwal has posted the same story about leaving Amazon four times, scoring 15,542, 5,426, 9,652 and 1,374. She wrote 304 posts before it landed.

Half a choiceThe format copies, the story does not

Newsjacking

Instant when it hits · lasts only on your own subject

A short opinion on that day’s news, with a company name and a real number in the first line. Neil Saunders has earned 97.6% of his lifetime engagement since September 2023 by writing about retail companies’ results. Pete Huang took 67.7% of his in the 2023 rush around ChatGPT and posted nothing in 2025.

The method, not the hit

Curation

Fast while LinkedIn favours it · quick to fade

Reposting things already spreading, with a short caption and a credit. Matthew Kerbel’s reposts brought in 75% of his lifetime engagement in 2020 to 2022 and roughly 4% a year since, at the same posting rate.

The method, not the hit
49%

of the roster is the grind

Only 20% of the band is.

36%

of the band is career as content

The largest path in the wider market, 40% at 10,000 to 20,000 followers.

0.1–0.3%

of followers answer a career post

The audience that lasts longest reacts least.

How each group was built

share of accounts on each path · roster 49 · 50K+ index 164 · 10K–50K band 1,102 read

Roster (49) 50K+ index (164) 10K–50K band (1,102 read)

Grind

49%
29%
20%

Career

20%
13%
36%

Mixed

0%
26%
14%

Narrative

14%
3%
8%

Newsjack

6%
2%
3%

Curation

6%
7%
3%

Not posts

4%
12%
12%

Automation

0%
7%
2%

In the wider market the job, not the writing, brought the audience.

The one path you can choose is the one that fades. Every other path needs something a brand cannot ask for.

03 · Not a path

Some followers never came from posts at all

The costly shape for a buyer is followers that never came from posts. There are two kinds.

People with lanyards talking over paper cups at a conference break

The first is a long career on LinkedIn: every accepted connection also becomes a follower, so the number grows while the posts do nothing. Recruiters are the usual case, along with people famous before LinkedIn. The second is automation, software that posts on its own. Twelve index accounts publish 20 to 200 times a day, and all but one have a typical post under 10 engagements.

That is 4% of the roster and 12% of both the index and the band, where the true share is higher: those 131 survived rules that had already removed 1,327.

The clearest test is one sum: all the engagement an account has ever had, divided by its followers. Above 50,000 followers the middle account scores 2.74.

0.05

Engagements per follower, ever · Michael Zhang

883,823 followers, built on a 2016–21 run of lifestyle photos.

2,905

Likes, comments and reposts, in total · Ronnie Lau

Twelve years of posting, on an account above 50,000 followers.

2.74

Exactly the middle mark · Brian Solis

34% of his ten years of engagement came from one 2020 run of feel-good videos, shrinking every year since.

Each post below is the best that account ever managed.

Roheela Khan

Roheela Khan

roheela-khan-1a66281 · 52,585 followers · 13 Apr 2025

Automated feed

https://lnkd.in/es9BW47e We're hiring a Senior Process Engineer - Cement

Want to lead performance on a large-scale kiln? Interested in carbon-cutting tech, AI tools, and real-time data systems?

Location: Buxton, Derbyshire…

A job advert card: Senior Process Engineer, Cement, UK, £45,000 to £70,000
172likes44comments16shares Open post
Michael Zhang

Michael Zhang

michaelzh99 · 883,823 followers · 24 Apr 2019

Fossil

持续坚持15年在单调且坚毅的撸铁中边流汗边快乐 [跳跳]

A mirror selfie in a gym, flexing one arm
415likes57comments0shares Open post
Brian Solis

Brian Solis

briansolis · 368,390 followers · 4 Nov 2020

Famous before LinkedIn

No matter what happens in life, the sun will still rise. Life is hard. But you really can make a dent in the universe.

What will you do with the gift of each day you're given?…

30,981likes2,071comments4,791shares Open post

Roheela Khan’s best post out of 76,975 job adverts drew 172 reactions; her typical post over the last twelve months got zero. Brian Solis is the only one here a brand might pay, and that engagement is six years old.

04 · What to do

Three simple rules

  1. 1

    Price an account on what its typical post got in the last twelve months.

    The follower count describes a past the account may no longer have.

  2. 2

    Ask which path built the audience before asking what it costs.

    A career audience lasts but reacts little; a grind audience reacts but does not last.

  3. 3

    Expect any shortlist of people who post to lean towards the grind.

    Half of ours does, against a fifth of the market.

Next in the series: the accounts that did this well, and what the turn upwards looked like from outside.

Sources · show
  1. linklabs.wiki, How LinkedIn audiences get built, September 2026. 49 roster creators, 164 accounts above 50,000 followers, and 2,429 at 10,000 to 50,000, of which 1,102 were read.
  2. Post totals as recorded in the index: 556,237 above 50,000 followers, 2,584,010 at 20,000 to 50,000, 1,750,326 at 10,000 to 20,000; 4.89M in all. None on record for the roster.
  3. Both the 164 and the 2,429 cover accounts still posting, with no stated cut-off date, so read that as a rough filter. Post texts, dates and counts from the same index.
  4. The band filter runs before any post is read, so every band share after it describes the filtered group and looks tidier than the whole market would.
  5. Photographs: Pexels, free licence. The three posts and the two post pictures are public LinkedIn posts, linked above.

Next in the series

The creators who did it well

The names worth approaching, with the posts that turned them.