Most advice about the LinkedIn algorithm is folklore passed between people who post twice a month. Post at 8:14am. Never use links. Reply to every comment within sixty minutes. Some of it is half true. Most of it is a superstition built on one post that did well once.
Here is what we can actually say about how the feed decides who sees your post, based on what LinkedIn has published about its own ranking systems, and on what anyone who posts weekly can observe for themselves.
The feed is a ranking problem, not a broadcast
When you publish, nothing is "sent" to your network. Your post enters a pool of candidates. For each person who opens LinkedIn, the system assembles a feed from thousands of possible items — posts from connections, posts from people they follow, posts other people engaged with, suggested content, ads.
That means your post is never competing for "your" audience. It is competing, one reader at a time, against everything else that reader could see in that slot. A post that gets 400 views has not been "suppressed". It has simply lost most of those individual auctions.
This reframing matters more than any posting tip. You are not trying to please an algorithm. You are trying to be the most relevant thing in a specific person's feed at a specific moment. Everything below follows from that.
Dwell time is the signal people underrate
LinkedIn's engineering team has written publicly about dwell time — how long a member spends with a post — as a ranking signal. The logic is straightforward and worth understanding, because it explains a lot of otherwise confusing behaviour.
Likes and comments are noisy. People like things reflexively. They comment for social reasons. And most people who find a post genuinely useful do nothing visible at all — they read it, think about it, and scroll on. If you rank only on clicks and reactions, you systematically over-reward bait and under-reward substance.
Dwell time catches the silent reader. Stopping to read something for twenty seconds is a real, hard-to-fake signal of interest. Expanding a post with "see more" is a smaller version of the same thing.
What this changes in practice:
- The first two lines are not a hook, they are a filter. They decide whether someone stops. Not whether they are entertained — whether they recognise their own problem in your words.
- Length is not the enemy. A long post that people actually read produces more dwell time than a one-liner. A long post nobody reads produces a scroll-past, which is a negative signal.
- Bait costs you twice. A curiosity gap that pays off with nothing gets the expand, then the instant exit. That pattern is legible.
I have watched posts with nine likes outperform posts with sixty in raw reach. The difference was almost always that the nine-like post was read carefully by people who cared, and the sixty-like post was skimmed by people being polite.
The golden hour is a symptom, not a cause
The standard claim: you have sixty minutes to gather engagement, and if you fail, the algorithm kills your post. Hence comment pods, hence "drop a comment and I'll send you the template", hence people refreshing their phone through dinner.
The mechanism is real but the story is backwards.
Early signals matter because ranking systems need evidence fast. When a post is new, there is no engagement history to learn from, so it gets shown to a small slice of people likely to care — recent interactors, close connections, people who open LinkedIn at that moment. How that slice responds informs how widely it gets tested next. This is ordinary exploration, not a countdown timer.
But the slice is the thing. If your post goes out to fifty people who never engage with you, strong early engagement is not going to happen no matter how clever the first line is. And if it goes out to fifty people who read everything you write, you barely need to do anything.
So the golden hour is mostly a measurement of something that was decided before you posted: who LinkedIn already believes cares about you. Engineering a pod-driven burst gives you a spike of engagement from people with no topical relationship to your content, which is a worse signal than it looks — reach without relevance does not compound.
The useful version of the advice: be present in the hour after you post, not to farm comments, but because replying properly is when your best conversations start. Also, do not post and vanish for a week. Consistency is what builds the warm slice that makes the first hour work at all.
Who you are to the reader outranks what you wrote
The strongest predictors of whether a specific person sees your post are relational, and you cannot fix them with formatting.
Connection strength. First-degree connections with a history of interacting with you are the easiest people to reach. Someone who liked two of your posts last month is very likely to see the next one.
Topical affinity. If a reader engages consistently with posts about, say, procurement, and you write about procurement, you are a better candidate for their feed than a stranger writing about mindset. This is why niche posting outperforms generic posting even when the generic post is better written. Narrow topics make you legible to the ranking system.
Second-degree spread. When someone comments, their network becomes eligible. This is why comments have outsized effects — not because "comments are worth more points", but because a comment opens a door into an adjacent graph. A like does far less of that work.
Recency of contact. People you have messaged, met or interacted with recently are more likely to be served your content. Prospecting and posting are not separate activities. The conversations you have in the inbox change who sees you in the feed, and the posts people read change how they respond in the inbox. If you are running both from one place — which is the whole reason a workspace like BIZLINKED exists — you stop treating them as two disconnected habits.
The quiet filters that flatten a post
Some things reliably reduce distribution. Not because of a penalty list, but because of what they do to reader behaviour.
Sending people off-platform. An external link in the post body means the best possible outcome is someone leaving. The feed is not motivated to reward that. The common workaround — link in the first comment — works less magically than people claim, but it does keep the post itself self-contained. The better fix is to make the post complete on its own and let the link be optional.
Documents and video that don't earn their format. A carousel can generate excellent dwell time because people swipe. A carousel of screenshots with no payoff gets one swipe and an exit. Format does not create attention. It can only hold attention that the content earned.
Tagging people who don't respond. A tag that goes unanswered is a visible non-signal. Tag two people who will genuinely reply, not twelve who will ignore you.
Engagement bait. "Comment YES below." LinkedIn has stated publicly that it demotes content designed to solicit engagement artificially. Beyond the policy question, it attracts the wrong readers.
Reposting your own post without comment. It rarely revives anything. Rewriting the idea from a different angle two weeks later usually works better.
One thing worth saying plainly because it comes up next to every algorithm conversation: tools that automate posting, commenting or engagement on your behalf sit against LinkedIn's user agreement, and accounts do get restricted. Whatever you gain in throughput you are borrowing against the account itself.
What to do Monday morning
Open your last ten posts and note two numbers for each: impressions, and the ratio of comments to likes. Ignore everything else.
Find the two with the highest comment-to-like ratio. Those posts made people think hard enough to write something. Read the first two lines of each out loud. That is your actual hook style — not the one from a swipe file.
Then pick one narrow topic you want to be known for and commit to it for six weeks. Not six topics. One. The affinity effect only works if you give it something to learn.
And for the next month, before you post, open your messages and reply to three conversations you have let go cold. Warm relationships are the input to feed distribution, not a separate department.
