The direct answer: stop copying templates, stop posting raw AI output, and stop opening with anything that could be on a motivational poster. LinkedIn's 360Brew algorithm now measures how long real people actually read your posts, and wally content gets three seconds before it disappears from the feed forever.
This guide gives you the 2026 playbook: the eight wally moves that kill reach, a Do/Don't breakdown, hook rewrites, a 10-step system for writing like a human, and the fresh data behind all of it.
TL;DR
Sound like a human by opening with a specific moment, writing in your actual voice, and giving the algorithm a reason to keep distributing you: a post people actually stay on, and at least one moment worth genuinely commenting on.
These are the patterns 360Brew has learned to associate with low dwell time. Each one is measurably costing you distribution.
| Wally Move | Why It Kills You | Severity |
|---|---|---|
| Opening with 'I am humbled to announce' | This phrase is a guaranteed scroll trigger. Readers have seen it thousands of times, scroll straight past, and a post nobody pauses on earns none of the dwell time LinkedIn's ranking system rewards. | High |
| Posting a wall of bullet points with no story | Bullet-only posts read as slide decks. No tension, no voice, no reason to stop scrolling. A meaningful comment carries roughly 2x the ranking weight of a like, and nobody comments on a slide deck. | High |
| Using 'it is not X, it is Y' frameworks | This template is so overused that readers swipe past it in under a second, and a ranking system trained on engagement signals learns to expect weak engagement from it. | High |
| Copying ChatGPT output without editing | By July 2026, roughly 81% of long-form LinkedIn posts were flagged as likely AI-written, up from about half in late 2024. The feed is drowning in the same vocabulary, and LinkedIn now lets members flag posts that 'seem like AI slop' to train its own detection models. | Critical |
| Starting posts with 'In today's world...' | A filler opener that signals low effort in the first three words. You have 200 characters before the 'see more' cutoff. Burning them on platitudes is a wally move. | High |
| No specific number, name, or story in the hook | Vague hooks earn vague engagement. LinkedIn's own engineering team has confirmed that dwell time, how long someone actually stays on your post, is a direct ranking input. A vague hook gives a reader nothing to pause on. Specificity forces a pause. | Medium |
| Posting a link in the first comment to dodge the penalty | This was a valid workaround until early 2026. LinkedIn closed the loophole. Links in first comments are now penalised alongside links in the post body. Both drop reach by roughly 60%. | Medium |
| Ending every post with 'Drop a comment below, what do you think?' | Explicit engagement bait is detected and suppressed. A large 2026 study of LinkedIn hooks also found question-led openers are the weakest performer of any hook type, so bolting one on at the end for the sake of it does not reliably help either. Curiosity has to come from the story, not a tacked-on prompt. | Medium |
Severity ratings are based on observed dwell time and engagement impact per LinkedIn's 360Brew-era data signals.
Five things that separate a post people stop for from one they slide past in under two seconds.
Open with a specific moment.
Name a number, a date, a person, or a place in the first line. Specificity earns the micro-pause that triggers dwell time.
Write the way you talk in a pub.
Short sentences. Contractions. The occasional fragment. If it sounds polished to the point of inhuman, cut it back.
Include one real anecdote per post.
Posts without a personal story read as interchangeable, and interchangeable content earns the weak engagement signals that feed ranking punishes. One concrete incident beats three abstract lessons every time.
Ask a question that you actually want answered.
A question only works when it is genuine. A large 2026 study of LinkedIn hooks found question-led openers actually score the lowest engagement of any hook type tested, so treat a question as a real conversation starter, not a formula. Readers can smell a fake one from a mile off.
Leave something unresolved.
The best hooks create a tiny itch the reader can only scratch by reading on. Tension drives dwell time. Dwell time drives distribution.
Do not polish out all the rough edges.
A typo or a clunky sentence signals a human wrote this. Ironically, small imperfections boost trust and dwell time.
Do not start with 'I am excited to share'.
The phrase is the LinkedIn equivalent of a form letter. It tells the reader nothing and trains them to skip your next post too.
Do not use three or more metaphors in one post.
One metaphor grounds your point. Three make you sound like you swallowed a TED Talk. Pick the strongest one and cut the rest.
Do not summarise your post in the caption.
If the first three lines tell the reader the whole story, they have no reason to click 'see more'. The 200-character hook needs to open a loop, not close one.
Do not post at cadences you cannot sustain.
Posting daily for two weeks then going silent for a month resets your audience faster than posting three times a week indefinitely. Consistency beats frequency.
These are the 360Brew-era data points that explain why sounding human is now a distribution strategy, not just good manners.
Originality.ai's July 2026 audit of 5,000 LinkedIn long-form posts found 81.2% scored as 'Likely AI', up from roughly half in late 2024. That is your competition for a reader's three seconds.
Originality.ai study →LinkedIn's own Q3 algorithm data, compiled by SayWhat, puts the median drop in impressions at 65% since 2023 as the feed shifted from social graph to interest graph.
SayWhat Q3 Algorithm Report →Forbes reporting on LinkedIn's 2026 distribution changes found links in the post body and links in the first comment are now penalised equally, cutting reach by roughly 60% either way.
Forbes / Jodie Cook →AuthoredUp's analysis of LinkedIn's ranking signals found a meaningful comment carries roughly twice the algorithmic weight of a like. Effort earns distribution.
AuthoredUp analysis →Lifast starts from your rough idea and your actual voice, not a template, so your posts pass the 'would a real person say this out loud' test and the algorithm keeps reading.
Try Lifast Free90 days of consistent posting. No ads.
Jasmin Alic breaks down what is actually working on LinkedIn right now, useful context for why generic, over-polished posting has stopped paying off.
"Aside from a few notable exceptions, we assume that members value their time, and will spend it appropriately on feed content that they're interested in."
Follow this in order before you hit publish. It takes five minutes and prevents 90% of wally moves.
Start with a specific moment, not a statement.
Name a date, a person, a number, or a place. 'Last Tuesday I watched a founder lose a deal because of one word on their LinkedIn' is a hook. 'Personal branding matters in 2026' is wallying.
Write your first draft in one sitting without editing.
Editing while you write produces over-polished, bloodless prose. Get the raw version down first, then refine. The voice lives in the first pass.
Read it back aloud.
If you stumble on a sentence, cut it or rewrite it. If it sounds like a press release, it is a press release. Your ears are the most honest editor you have.
Check the hook is under 200 characters and opens a loop.
Count the characters. Does the first 200 make a reader curious about what comes next, or does it summarise the whole post? If the latter, rewrite.
Add a specific number or named detail if you have not already.
Vague posts feel fake. '48,000 impressions on my third post' feels real. 'Lots of engagement on my early content' feels like a wally.
Cut every sentence that does not add information.
Filler sentences are wally tells. 'In today's competitive landscape' adds nothing. Delete it. Every sentence should carry weight.
Remove the last sentence if it is explicit engagement bait.
'Drop a comment below!' is the wally closer. Replace it with a genuine question that stems from the story you just told.
Check for AI vocabulary: 'leverage', 'delve', 'nuanced', 'robust'.
These are LLM fingerprints. Swap them for the words you actually use when talking to a colleague. 'Use' instead of 'leverage'. 'Dig into' instead of 'delve'. For a longer list of tells and fixes, see our guide on how to humanize AI LinkedIn posts.
Remove any link from the post body and the first comment.
Both placements reduce reach by roughly 60% per LinkedIn's 2026 algorithm. Post the resource separately or share it via DM to engaged commenters.
Publish and then go offline for at least two hours.
Early engagement signals matter most in the first two hours. Replying to every comment in that window tells 360Brew this post is generating real conversation, which compounds distribution.
Six of the most common wally hooks, rewritten to open a loop, earn dwell time, and sound like a person wrote them.
| Wally Hook (skip) | Human Hook (use this) |
|---|---|
| I am humbled to share that I just hit 10,000 followers. | Ten thousand strangers chose to follow me. Here is the one post that started it. |
| In today's fast-paced world, building a personal brand is more important than ever. | I ignored LinkedIn for three years. Here is what I lost, and how I got it back in 90 days. |
| It is not about what you know, it is about who you know. | The hire that changed my company came from a comment someone left on a post I almost deleted. |
| Here are 5 tips to grow your LinkedIn presence (save this for later). | My third post ever got 48,000 impressions. I had 200 followers at the time. This is what I did differently. |
| Excited to announce I will be speaking at [Conference Name] next month! | I said yes to speaking at a conference last year when I had nothing ready. What happened next was genuinely uncomfortable. |
| The key to success is consistency, hard work, and never giving up. | I quit my 'consistent' posting schedule for six weeks. My engagement went up. Here is the data. |
The irony of the AI-content problem is that you still need to post consistently, three to five times a week, to build any kind of compounding reach. Most people burn out trying to do that manually. The fix is not to dump raw AI output on your feed, it is to use a tool that starts from your idea and voice and builds outward from there. That is what Lifast does: you give it a rough thought, a story fragment, or a topic you want to cover, and it generates a draft that actually sounds like you rather than like every other post in the feed. You still edit it. You still own the voice. But you do not spend 45 minutes staring at a blank page every morning.
Readers identify these in under two seconds. Trained LinkedIn readers can spot a wally post before they reach the third line.
The suspiciously perfect paragraph
Every sentence is the same length. No contractions. No fragments. No voice. When you read it aloud and it sounds like a press release, it is pure AI and readers know it.
Emotion words with no emotion backing them
Phrases like 'I am incredibly grateful' or 'this journey has been transformative' appear in thousands of posts per day with zero specific detail to support them. Gratitude without a story is noise.
Three separate hashtag lines at the bottom
Hashtags do not help LinkedIn distribution in 2026. A block of ten hashtags at the end of a post signals that the author Googled 'how to get LinkedIn reach' circa 2019 and ran with it.
The 'value dump' carousel with no hook
A 10-slide carousel where slide 1 says 'The 7 things I learned from X' and every slide is a one-liner tells the reader nothing they could not skim in 10 seconds. No hook, no dwell, no reach.
Motivational quotes as posts
Posting a famous person's quote with 'This really resonated with me' as the full caption is a wally move of the highest order. You are not adding a perspective, you are reblogging someone else's thinking.
A quick diagnostic for every common wally symptom. Run your draft through each branch before publishing.
IF
If your hook is a platitude (e.g. 'success requires hard work')...
THEN
then rewrite it around a specific moment from your own experience. Name a number, a date, or a person.
IF
If your post has no personal anecdote...
THEN
then add one sentence describing a real situation you were in. Even a single line of 'I remember when...' anchors the whole piece.
IF
If you ran your post through an AI tool and posted it unchanged...
THEN
then add two sentences in your own informal voice and remove any word you would never actually say out loud.
IF
If your post opens with 'Excited to' or 'Humbled to'...
THEN
then delete the first sentence entirely. The post almost always works better without it.
IF
If your post contains a link and you want reach...
THEN
then remove the link from the post and do not add it to the first comment either. Post the link in a follow-up post or DM it to people who comment.
IF
If you have not asked a genuine question anywhere in the post...
THEN
then add one only if it's something you would genuinely enjoy answering yourself. Fake questions get zero responses, and a tacked-on closer will not fix a weak post either. Let the story create the curiosity.
This is the wally habit that gets defended the hardest, because it feels like a growth hack. The data says otherwise.
Every LinkedIn growth thread repeats the same advice: end your post with a question and you will farm more comments. It has been repeated so often it now reads as fact, so people bolt a question onto the last line of every post whether it fits or not.
AuthoredUp's analysis of 309,614 LinkedIn posts found question-led hooks score the lowest median engagement of any opener type tested, and closing questions barely move the needle either. See the full AuthoredUp study. Story-led openers consistently outperform question-led ones.
The lesson is not "never ask a question." It is that a question tacked onto the end of a post as a formula is not a substitute for a post that earned attention on its own. If your story is doing its job, the curiosity should already be there before you get to the last line.
The payoff for sounding human is not one viral post. It is a permanent shift in how the algorithm treats your account.
Dwell time signals establish your baseline
360Brew builds a relevance and quality profile for every creator. If your first 20 posts consistently earn 60+ second dwell times, the model establishes you as a high-quality signal in your topic area and distributes subsequent posts more aggressively from day one.
Comment depth increases reach multiplier
As your posts generate genuine comments, and you reply promptly, the comment thread itself becomes a dwell-time signal. Readers read the discussion. The post stays in feeds longer. Each reply you write compounds the distribution of the original post.
Interest-graph audience locks in
The interest-graph shift means LinkedIn now shows your posts to people interested in your topic, not just your connections. After 5 to 6 months of consistent human-voice posting, you accumulate topic authority that helps new posts reach cold audiences automatically.
Inbound replaces outbound
The compounding effect of dwell time, comments, and topic authority means that followers, connection requests, and inbound enquiries arrive without any promotion. The opposite of a wally posting strategy: you stop chasing and start being found.
Authenticity is not a feel-good idea. It is what the algorithm pays out for now. These four numbers explain the cost of getting it wrong.
Ordinal's Algorithm InSights Report found personal profiles generate 561% more reach than company pages sharing identical content. The feed was redesigned around people, not logos.
Ordinal report →Metricool's 2026 study of 673,658 LinkedIn posts found personal profiles average 2.60% engagement versus 1.60% for company pages, a roughly 63% relative gap on otherwise comparable content.
Metricool 2026 study →AuthoredUp's analysis of 309,614 LinkedIn posts found question-led hooks score the lowest median engagement of any opener type tested, well behind story-led openers. A bolted-on question is not a hack.
AuthoredUp hook study →360Brew is LinkedIn's real 150-billion-parameter feed-ranking model, described in the company's own research paper. It reads dwell time and comment quality, not just likes.
360Brew research paper →These are not style opinions. They are patterns the engagement-bait classifier and 360Brew's dwell model have both flagged based on observed reader behaviour.
| Banned Phrase | Why It's Suppressed | Say This Instead |
|---|---|---|
| Comment YES if you agree! | Engagement bait pattern from the 2022 era. LinkedIn's classifier now detects and suppresses explicit vote-solicitation regardless of context. | What's been your experience with this? Curious where I'm wrong. |
| Like this post to get the PDF in your DMs! | Triggers the engagement-bait classifier directly. No workaround exists as of 2026. | I've put the full breakdown at [link]. Save for later. |
| Drop a 1 in the comments | Classic engagement bait, suppressed by the same classifier that caught the 'YES' pattern. Low dwell, zero conversation. | If you've tried this, what did you learn? |
| In today's fast-paced world | An AI-tell phrase that earns near-zero dwell time. Readers identify the filler in under one second and scroll on. | A concrete number or named situation from your own experience. |
| Leveraging synergies to... | Corporate AI-slop pattern. 360Brew's dwell model has seen this construction tens of millions of times and correlates it with sub-3-second reads. | A plain English verb describing exactly what you did. |
| It's not [X], it's [Y] | 360Brew explicitly downranks this cliched template. The binary reframe format is now one of the most over-used structures in the feed. | A sharper, specific point with a number or a named example backing it. |
| Couldn't be prouder to share... | Status-bragging opener with low dwell. Readers recognise the pattern as a personal press release and keep scrolling. | A story sentence about HOW the thing happened and what you got wrong along the way. |
| "Thoughts?" as a solo closer | Forces the reader to invent the angle themselves. No framing, no curiosity trigger, no reason to type anything. | A specific question the reader has direct personal experience with and an opinion on. |
Suppression data based on LinkedIn's 2026 engagement-bait classifier and 360Brew dwell-time signal patterns.
In 2026 audiences are tired of polish. LinkedIn's feed measures dwell time, and behind-the-scenes content satisfies curiosity while building trust simultaneously. Polished corporate updates resolve in three seconds because the reader already knows the ending. Behind-the-scenes posts create tension that keeps readers in the post longer, which is exactly what 360Brew rewards.
LinkedIn's own research paper describes 360Brew as a 150-billion-parameter, decoder-only foundation model trained on LinkedIn's own data, designed to replace the thousands of separate ranking models that previously scored content. LinkedIn has not published an exact public rollout date, but industry analysts estimate it now powers a large and growing share of LinkedIn's feed surfaces. The difference matters because 360Brew does not just count likes. It measures signals that are much harder to fake: dwell time, comment depth, saves, and the quality of engagement relative to the size of your audience.
Before 360Brew, gaming LinkedIn meant posting at the right time, using the right hashtags, and hitting publish on a template format that had worked last quarter. After 360Brew, the game is simpler and harder simultaneously: write something a person actually wants to read. The model can identify content that earns a genuine pause versus content that people scroll past in under two seconds. And it distributes accordingly.
The practical upshot is that sounding like a corporate wally now has a direct, measurable cost. AI-generated slop, clichéd frameworks, and motivational-poster copy all produce short reads and fast scroll-pasts. Human posts with a specific hook, a real story, and a genuine question tend to hold attention far longer, and that difference in dwell time is exactly what the algorithm is built to notice and reward.
Originality.ai's ongoing tracking of LinkedIn long-form posts found that roughly half were likely AI-written in late 2024, and by July 2026 that figure had climbed to 81.2% across a 5,000-post sample. If you are posting the raw output of any LLM without meaningful editing, you are indistinguishable from that majority. And that majority is not performing well, which is a big part of why LinkedIn built a feature letting members flag posts that 'seem like AI slop'.
The problem is not that AI writing is bad. The problem is that it produces the same vocabulary, the same sentence cadence, and the same structural patterns across millions of posts. When a model like 360Brew has seen the same phrasing ten million times, it implicitly learns that this type of content earns weak engagement and distributes it accordingly. You are being penalised not for being wrong but for being predictable.
Human-AI hybrid content tells a different story: a real person contributes the hook, the personal story, and the specific detail, while AI helps tighten phrasing or add structure. The model cannot tell exactly how you wrote the post. But it can measure whether real people stopped to read it, and real people stop for real voices. For more on where the line sits, see our guide on whether LinkedIn penalises AI-generated content.
LinkedIn truncates posts at approximately 200 characters with a 'see more' prompt. Everything before that cutoff is your hook. Everything after it is your content. If you burn the first 200 characters on an announcement, a platitude, or a generic setup, you have already lost most of your potential readers. A specific, concrete opener is what earns the click through to 'see more' in the first place.
The 200-character rule has not changed, but the stakes have. In an interest-graph feed where median impressions have dropped roughly 65% since 2023 according to LinkedIn's own Q3 algorithm data, every reader you retain in those first two seconds matters more. Posts that generate a micro-pause at the 'see more' threshold earn longer dwell times than posts where readers never click through at all.
A good hook creates a small, specific itch. It names something unexpected, raises a question it does not immediately answer, or drops a number that needs explaining. A wally hook announces something, lists something, or states the obvious. You have 200 characters. Use them to open a door, not to hand the reader a brochure.
Sharp answers to the queries ChatGPT users ask about sounding human on LinkedIn in 2026.
Not through a blanket filter, but the effect is close to one. LinkedIn's ranking model does not officially label posts as AI or human. What it measures is engagement quality: dwell time, comments, saves. Pure AI-generated content consistently earns weak scores on these signals because it produces predictable, low-personality text that readers scroll past quickly. LinkedIn has also added a feature letting members flag posts that 'seem like AI slop' to help train its detection systems. So the penalty is mostly indirect, through weak engagement signals, but it is very real.
No personal specificity. Posts that use phrases like 'in my experience' or 'I have learned that' without naming a specific moment, person, number, or outcome read as generic regardless of whether a human or an AI wrote them. The fix is always the same: add one concrete detail. A date, a number, a name, a specific thing that happened. Specificity is the single fastest way to sound like a real person.
Because commenting requires more effort and intent than clicking a thumbs-up, and LinkedIn's ranking model treats that effort as a stronger relevance signal. Independent analysis of LinkedIn's ranking behaviour puts a meaningful comment at roughly twice the weight of a like. A post that generates genuine discussion is demonstrably more valuable to the platform's interest graph than a post that collects passive likes. This is why posts that share a perspective people want to push back on tend to get disproportionately large distribution, though forced or bolted-on questions do not reliably produce the same effect.
No, as of early 2026 this workaround no longer works. LinkedIn closed the loophole: links in first comments are now penalised similarly to links in the post body, reducing reach by roughly 60%. If you want to share a URL, post it in a follow-up post, DM it to people who engage, or mention the resource by name without hyperlinking.
There is no single ideal length, but the data points toward posts that take 60 to 90 seconds to read performing best on dwell time metrics. That translates to roughly 200 to 400 words of actual content, not counting the hook. What matters most is that the post earns the reader's attention all the way through rather than padding word count with filler. A tight 150-word post with a strong hook and a real story outperforms a bloated 600-word post with no narrative.
Consistency matters more than frequency. Posting three times a week for 52 weeks delivers better compounding returns than posting daily for six weeks and then burning out. The 360Brew interest graph rewards sustained presence in a topic area over time. Pick a cadence you can hold for a year, not one that impresses you for a fortnight.