A strong company page post on LinkedIn used to reach 10 to 15% of its followers. It now reaches 2 to 3%. It explains most of the frustration I hear from owners who feel like they're posting into a void.
LinkedIn stopped rewarding broadcast and started rewarding presence. And the rules for a person are now different enough from the rules for a company page that running one playbook on both accounts is a waste of time.
Overview: How the LinkedIn platform has changed
The platform is charging brands for reach
LinkedIn throttles organic reach for company pages on purpose. Sponsored posts are the advertising/business model on LinkedIn. It is how it makes money. So free distribution for a company does not make financial sense for the platform. The feed is engineered to favour individual humans, because humans are what keep people opening the app. LinkedIn's own published figure puts an individual's personal profile reach at roughly 8x company page reach.
Writing better articles won't fix that. It's the LinkedIn business model.
Readers now filter for humans
The second force is behavioural or reader preference. People have been saturated with brand-approved corporate messaging and those unrealistic rags-to-riches success posts. They now scroll past most of them. Across every thread we read, the same preference shows up: readers want a real practitioner explaining how they solved a specific problem, and better still, one that looks like the reader's own. Not a brand announcing an award.
One founder posting a 2-year growth breakdown found that honest "this didn't work" posts got their furthest reach of anything they published. Not the failure-to-success stories, or latest achievement posts. Just unresolved, mid-project takes. Multiple commenters said the same thing. And with AI posts flooding the feeds, everyone is sounding the same. The AI slop flood has made a real story stand out more, not less.
In fact LinkedIn have just introduced a “Seems like AI slop” button for readers to report posts.
Likes matter less.
A post with 10 likes and heavy dwell time will out-reach a post with 50 fast likes that people double-tapped while scrolling.
The metric that moves distribution is dwell time. The platform tracks how many milliseconds you pause, whether you tap "see more", and how long you stay after expanding. Posts with good dwell time routinely reach 5 to 10x their engagement count in actual views.
I started my career working with a leading internet discussion forum. If I remember my stats correctly, only about 5% of members would actually post and comment. Which is. It to say that the rest were not engaged, they were, just quiet.
Your lurkers outnumber your likers by an order of magnitude, and they're the ones who eventually reach out to you when they need you.
A study from Flypost.io looked at 130,000 posts by 2,500 B2B creators, July 2025 to July 2026, which compares every creator against their own posts rather than against other people.
The typical B2B post earns 34 reactions and three to four comments. A 10,000-follower account gets about 33 reactions. Comments are the scarce currency: most accounts under 25,000 followers get between one and three per post.
Optimise for the person who reads the whole thing and says nothing. They're often the buyer.
LinkedIn is now a source for AI answers
Interestingly, posting on LinkedIn is no longer just to build and engage your network and market. It is now a way to be found via AI (GEO, Generative Engine Optimisation, like SEO but for AI engines).
Semrush analysed 89,000 LinkedIn URLs cited by ChatGPT, Google AI Mode and Perplexity. LinkedIn came out as the second most cited domain across all three, appearing in 11% of AI responses. Only Reddit ranked higher. Of the LinkedIn content actually getting cited, 95% was original data, statistics or first-hand insight. Between 50 and 66% was long-form writing in the 500 to 2,000 word range. And 75% of the cited authors publish at least 5 times a month.
Reshares and engagement on your posts don't get cited. They're invisible.
This is important for a business and individuals. Before anyone hires you, they search your name or your offering. That search is increasingly happening inside an LLM like ChatGPT or Claude rather than Google search. Whatever you've put on LinkedIn is part of the raw material that the answer gets built from.
At this point, we are often asked how to write to be picked up by SEO and GEO, and I always say don’t! You can twist yourself in knots; there is no evidence that changing your writing style has any effect, and the harder you make the process, the less likely you are to write anything.
Instead, get your team talking about things they've actually done, in their own words, regularly. Let the passion, knowledge and expertise shine through. The citations follow naturally.

1: The playbook for individuals
Your LinkedIn profile is a search index, not a CV
A CV is a static list of duties. Your LinkedIn profile is a searchable database that recruiters and buyers query with specific strings.
So reverse engineer that.
Pull the exact terminology from the job specs or briefs you want, and put those phrases in your profile. Not your clever internal name for what you do. Use the words people type. If you sell a ‘garlic press’ and you list it as a ‘culinary bulb crusher’, nobody finds you because no one searches for that term.
The mechanics of a successful individual profile:
- Set a custom professional URL. The auto-generated string of numbers looks like a database error when it's pasted anywhere. I have failed on this point already. Instead of my name “Alan O’Rourke”, my professional URL is “https://www.linkedin.com/in/spoiltchild” (connect and say hi). It was set up 20 years ago, named after an old business and has been linked continuously from various sources over all that time. I am sure it has affected the view of the page, but I am not sure I am ready to let it go due to nostalgia reasons.
- Fix the photo. Recruiters see the blurry focus and bad lighting in photos and read it as bad work. AI headshot tools are accepted now, but the margin is thin, and over-polished faces trigger the uncanny valley and cost you trust. It should look like you on a normal workday.
- Fill the featured section. It's clickable media, and most people leave it blank. It is another section to fill with relevant information to help you get found and highlight why you are the ideal person for the job.
- Header: your value proposition in 4 to 8 words.
- Summary in first person, saying who you help and how.
Comments compound faster than posts
While you are trying to build your own profile, you can leverage other people's active and bigger audiences to grab attention. The current rule is 5 to 1. For every post you publish, leave 5 meaningful comments on other people's. A strong comment on a busy post puts you in front of other people’s active, engaged audiences.
Break down your list of ideal targets:
- List A: 10 to 15 accounts in your niche who reliably get 50 to 60 comments per post. Big enough to have an audience, small enough that you're visible in it. Turn the notification bell on for these.
- List B: your dream clients. Relationship building.
- List C: warm leads. People who replied to anything. Staying front of mind.
Timing matters because the algorithm judges a post's velocity in the first 60 to 90 minutes and expands or kills distribution accordingly. Being early puts your comment at the top while that surge is happening.
Two hard limits on speed, though.
A comment needs roughly 16 to 20 words minimum to register as human effort rather than bot noise. What the system appears to look for is friction: a counterpoint, a specific example, an addition. Applause (“Great post” or “I agree”) doesn't qualify.
And don't automate it. People can smell AI from 20 miles away, and the consensus was that a thoughtful comment left 4 hours late beats fast garbage every time. Several people said they now block accounts outright when they detect generated comments. There's no upside here.
People may suggest LinkedIn engagement pods to you, and in our research, engagement pods came up repeatedly, but the verdict was consistent.
A pod is a group of users who coordinate to artificially boost each other's content by liking, commenting and sharing posts and trying to fool the algorithm into distributing it more widely.
Our research shows a temporary spike in numbers from many people who will never buy from you and not much else over time. Though it looks good to your audience.
How often to post.
The general advice is 3 to 4 times a week.
It often depends on your audience and the market you are in.
One person we researched posted almost daily for 6 months. 270 posts, 3.6 million impressions, 7,000 followers, averaging 128 reactions. Total revenue attributable to it: about $1,000 in side gigs.
In comparison, a freelance financial writer said they'd posted fewer than 10 times in 2 years and had won close to $200,000 of work from the platform.
Volume grows an audience.
Specificity wins work.
Your market may have 10 customers who are likely to buy from you.
Whether you have those 10 connections on LinkedIn or 10,000.
You still only have 10 potential customers.
Decide which you're doing before you set a cadence.
Which posting formats work?
The general consensus is that documents and carousels perform the best. They earn the highest save rate, and swiping drives dwell time hard. Good for walking someone through something step by step. But again, test this with your own audience.
In the previously mentioned study of 130,000 posts by 2,500 B2B creators, Flypost compared every creator against their own posts and found carousels worth about 13%, and only for half the people who used them. An image is the thing that actually moves the needle. The same person's posts with an image attached earned 33% more reactions than their own text-only posts. It is the most reliable finding in that study and the least clever one. Stop publishing naked text. A screenshot, a photo of the thing you are describing or a plain chart all clear the bar.
Text with a sharp opening line still works. You have about 3 seconds before the scroll continues, so the first line has to break the pattern. Then pay it off immediately with something useful. The hook buys attention; the body is what earns the profile visit.
Short casual video outperforms polished video, which sounds wrong until you remember that studio-grade footage reads as an advert and triggers everyone's internal ad blocker. Phone video reads as a peer with something to say. However, one founder with 10k followers reported that video underperformed written content for them in their market. So like all of these suggestions, test it.
First-person posts whose first line opened with I, my, we or our earned about 20% more reactions than the same creator's other openers. What readers reward is a personal view.
Photos of you. Real photos in real settings, consistently high engagement. This will not be a big surprise to those from design and advertising who know that faces grab the most attention of any other type of image.
Announcements got 3 to 5x the reach of standard posts for that same account. Anything framed as news signals momentum and the algorithm amplifies it.
Which is the opposite of business page posts. I guess people want to celebrate an individual's success.
External links and plain reshares are widely said to get buried, on the reasoning that LinkedIn loses the ability to serve ads once you send someone away. Worth a caveat: HubSpot’s own head of brand says he has put a link in every post this year and has 43 million impressions to show for it, and HubSpot’s published founder playbook treats the question as open. Test it on your own account. The reshare half is less contested, since a reshare with no original commentary adds nothing to the feed.
Choose your topic and say one thing from many angles
The creators who build something durable stop chasing trending topics. They take one firmly held view about their field and come at it from a different direction each time. A teardown of a bad example on Tuesday. A mistake they talked a client out of on Thursday.
Consistency of perspective is what makes you memorable. As one commenter put it, consistency creates recognition but specificity creates selection.
And set the timeline honestly. Across our research, it takes about 3 to 6 months of consistency before inbound shows up, and around 6 months before traction becomes consistent. People who quit at week 5 are the norm and it is them who often report that LinkedIn does not work.

Part 2: The playbook for company pages
Your company page now has a different job.
Stop treating the page as a distribution engine. This year, organic reach from company pages is at an all-time low of 2 to 3%. No amount of posting will change that.
The job of your company page is now is verification and validation. When your sales team sends a message, or someone gets your email, or a referral hears your name, the first reflex is a background check. They won't buy from a post appearing in the feed. But they will absolutely click through to your page before replying to you.
To a potential customer, it's the health inspection grade in the restaurant window. It doesn't pull anyone off the street. But if it looks low, abandoned or reads like generic corporate filler, they turn around and leave.
Credibility beats chasing reach.
LinkedIn's own Senior Director of Product Marketing, Gina Kleiner, made the point publicly: stop optimising for engagement and start optimising for sales. Profile visits, DMs started, lead quality. Follower count tells you nothing about whether the right 40 people associate you with the problem you want to own.
What to cut, and what to post instead
The single question your page has to answer for a prospect scanning it is whether you understand their daily problems.
Awards don't answer that. Polished win announcements don't answer it. They prove you're successful without proving you can solve their thing. All three read as advertising and register as noise.
The mix that came up as a working ratio is 67% open-sourced knowledge and 33% practical tools. Lessons from a project that nearly derailed. A raw case study. Something that went wrong internally and the exact steps taken to fix it. Then checklists and templates people can actually use.
Cadence: 2 to 3 times a week, maximum
Company pages get penalised for overposting through content decay. If you post on Tuesday, the algorithm is still testing that post with small user cohorts on Wednesday. Post again Wednesday morning and you cannibalise your own Tuesday post. Distribution stops on the first to push the second, and often neither gets anywhere.
Long-form goes on your own site
If you want to prove deep expertise, write the 4,000-word guide. Put it on your blog, where it builds your own domain authority and gets indexed.
LinkedIn is a feed, not a library. Pull 2 to 5 paragraph extracts from that piece, or turn the core of it into a carousel. Give away a genuinely useful chunk of the thinking. People who want the whole thing know where you live.
Hub and spoke: Use your team for the content distribution
Your page is the hub. Founders and employees are the spokes, because a human profile out-reaches a company page by a wide margin and always will.
This is the mechanic that actually moves numbers for small companies. The page posts the content. Real people add their own perspective in the comments in the first 30 minutes. One marketer in these threads named that first-30-minutes employee commenting as their single biggest unlock.
The 5 to 1 rule applies here too. For every post the page publishes, your team leaves at least 5 substantive comments on prospects' posts across the platform.
And the page itself can comment. A company page leaving a 20-word insight on a big creator's post in the first hour puts your name in front of thousands of buyers without needing a single follower of your own. It's one of the most underused things available and it costs nothing.
Outbound on Linkedin
This advice comes from an operator managing 47 LinkedIn accounts (for founders and executives) who has sent over 100,000 connection requests:
- Connection requests with a note: 29% acceptance. Requests sent blank: 36%. Which goes against accepted wisdom but the numbers are compelling enough to test yourself.
- Target people who've been active in the last 72 hours. Static lists pulled from a database are full of inactive people who haven't logged in since 2019. Don’t waste time dropping flyers into empty houses.
- Withdraw pending invitations every 3 to 4 months. One operator hit an unexplained rate limit, found 2,000 invitations sitting pending, withdrew them, and had normal limits back within 24 hours. Whether 2,000 is the actual threshold is unknown.
- Related, and worth knowing: LinkedIn accounts appear to carry an internal trust score that governs how many connection requests you can send in a week. It is why two people doing the same outreach hit different ceilings, and why clearing old pending invitations helps.
- If you do add a message, make it personal. A comment they left, a job change, a post in your area. Then offer something specific rather than a meeting.
A handy tip is to target LinkedIn events. 60% reply rate: 15 quality responses from 25 accepted connections.
Search your industry keyword, open the Events tab, register for the busiest ones, and work the attendee list. Everyone there is active and already interested in the topic. Connect, wait 24 hours, comment properly on something they've posted, then message.
That sequence is important. By the time you message, you aren't cold. They recognise your name from the comments, they click your profile, they click through to your page, and the page does the job it's actually for.
There is a second use for events on the company side. Your own team can share an event across their networks in a couple of clicks, with a limit somewhere around 1,000 invitations each per week. For a small sales team, that is meaningful reach that costs nothing and does not touch the page limits.
Don't become the bottleneck
Read back through all of that and notice what the winning tactics have in common. Failure posts beat polished wins. Founder photos beat stock imagery. A comment with a real opinion beats 50 generated ones. Recorded speech beats ghostwritten prose.
Every mechanic that works today works because a specific human took the time to record it and get it written down. Which creates an obvious problem. All of this needs the words out of your head, several times a week, forever, and you have a business to run. That's why most people stall around week 5.
The best advice I picked up for this came a few years ago from a comment about how business books get made. You don't hand a busy business expert a blank page. You sit them down for an hour, record them talking about a deal they lost or a bottleneck they solved, and transcribe it. When people talk about work they actually know, they name real cases, use real numbers, and argue with conviction. All the texture that's impossible to fake, and all the detail an AI needs to cite you.
They had someone to ask them the questions.
That's the whole idea behind Louderlines. One question arrives by email. You answer it in 2 to 4 sentences, in your own words, from your phone, between other jobs. That answer goes out in a variety of formats from one single source. A newsletter, a formatted social post, a blog entry on your own site, and a permanent post on your public profile.
Louderlines doesn't write a word of it. It can't, and it shouldn't, because your customers can tell. Everything else on that checklist- the formatting, the images, the scheduling, the distribution to 4 places at once- is work a system should be doing for you. LinkedIn stays the one you paste in yourself, for the reasons in item 11, and the plan tells you when to do that.
You get to clear the bottleneck and start growing.
