Post for the First Hour, Not the Perfect Hour

The question everybody asks first

Best day and time. Everyone asks this one first, because it feels like a lever you can pull without changing a word of what you write.

The day-of-week answer is the one durable finding in this whole space, and it has held for years. For B2B, it is Tuesday through Thursday, and weekends are dead. Wednesday is the safest single bet. People are heads-down on Monday and checked out by Friday afternoon. That part is real, and you can take it to the bank.

The exact hour is where it falls apart, and this is the part nobody tells you. The studies do not agree with each other. They are not even close.

Read that again. The largest single dataset out there, Buffer's 4.8 million posts, ranks Tuesday among the worst days to post. Another major 2026 study calls Tuesday one of the best. They cannot both be your answer. And on the hour, one camp says mid-morning and the other says late afternoon, and both have real data behind them.

When the biggest studies contradict each other this cleanly, that is the finding. There is no magic minute. There is a general shape, which is midweek, during waking work hours, in your audience's own time zone. Past that, precision is a story we tell ourselves because it feels like control.

Buffer says it plainly in their own report: posting at four on a Wednesday will not rescue a post that is not relevant or valuable to begin with. That is the whole game in one sentence, and it comes from the people selling you the timing chart.

So here is what I told my client. Post Wednesday or Thursday, mid-morning where your buyers actually live, and then forget the clock. It is a tiebreaker between two good posts. It is not the difference between a good one and a dead one.

The first hour is a tryout

Which brings us to his second question, and it is a sharp one. Does the first ninety minutes still matter?

Yes. Just not the way the panic version tells it, the one that says engage in the first ten minutes or your post is dead. That version is wrong, and it has been wrong for a while. Here is what really happens.

When you hit post, LinkedIn does not show it to everyone who follows you. It shows it to a slice, and then it watches. It is running a tryout. The question it is asking is not "how many likes did this get." It is asking, of the people who saw it, how many did something, and how fast. Engagement rate, not raw count. That is why a small account can outrun a big one on a single post. A post shown to a hundred people that earns ten real reactions beats a post shown to a thousand that earns fifteen.

1. 01
   A slice of your network sees it
2. 02
   They stop, read, and comment
3. 03
   The feed widens the circle
4. 04
   It keeps widening as long as new people engage

The signal underneath all of it is the clearest one LinkedIn has actually confirmed it measures: dwell time. How long people stop and look. Their own engineers have been writing about it since 2020, and the reason is simple. A fast like is a weak, noisy signal. Someone quietly reading your post for a solid minute is a much better sign it was worth their time. Almost everything else you will read about the algorithm is reverse-engineered guesswork by smart outsiders. Dwell is the part LinkedIn told us itself.

So the first hour matters because that is when the tryout runs. Be there. Reply to the early comments, because that keeps the conversation moving while the machine is still watching.

But here is the 2026 update, and it takes the pressure off. A good post is no longer dead at the end of the day. One 2026 analysis of millions of posts put a typical LinkedIn post's half-life at around 23 hours, and the bigger shift is what happens after: the platform now actively resurfaces a post that keeps sparking conversation, and the firms that track reach put that tail at two or three weeks. Last summer LinkedIn leaned into that on purpose, testing relevance over recency by floating older posts it judged still useful, then easing off to keep some of the freshness people expect. The takeaway: work the first hour, then let it breathe. The tail is real now, and it rewards the posts people are still talking about, not the posts you happened to publish at the anointed minute.

So which is better, a like, a comment, or a share

His third question is where the real leverage lives, so let me slow down here. He asked it the way most people do. Is a comment better than a like? Is a share better than a comment? The honest answer has some texture to it, and it is worth the two minutes.

First, one caveat, because I will not hand you numbers I cannot stand behind. LinkedIn has never published its real weightings. The one signal its own engineers have written about, at length and by name, is dwell time. Everything else below is the best read of people who study hundreds of thousands of posts for a living. Directionally solid. Not gospel. I will tell you which is which.

Is a comment worth more than a like? Clearly, yes. AuthoredUp studied more than six hundred thousand posts and found that one comment is worth about two likes, and it climbs from there once the comment turns into a back-and-forth. The reason is human. A like costs a thumb. A comment costs someone stopping, reading, and typing. You may have read that a comment is worth fifteen likes. Nobody can source that number. The one people can source is closer to two.

But not every comment counts the same. "Great post!" is worth almost nothing, and the new feed actively pushes down that kind of filler. A comment that actually says something, fifteen words that add a thought, carries far more weight than a one-liner. And your own replies do double duty. Each reply is another comment on your post, and it pulls the first person back to keep talking. Which is exactly why the hour after you post is for working the comments, not admiring the post.

Now the share question, because this is the one people get backwards. Is a share better than a comment? It depends entirely on one thing. Did you add your own words?

The bare repost, the one-click "repost" with nothing added, barely travels. The reason is mechanical, not mystical. An empty repost never becomes a post of yours. It is a megaphone pointed at someone else's post, and it never gets its own tryout in the feed. Add a couple of real sentences of your own and it becomes a new post from you, which does get its own shot and can travel on its own merit. One study of reposts found that adding a real take earns about four times the engagement of an empty one, still short of an original post of your own. The exact number is fuzzy. The pattern is not. And the words have to be real. "Worth a read" gets you the downside without the upside.

And the two most undervalued signals of all: the save and the send. A save is someone telling the feed "this is worth coming back to," and it is worth several likes on its own. A send, where someone forwards your post into a private message to a colleague, is the closest thing to a personal referral that exists on the platform. Nobody has put a clean number on the send, so I will not invent one. But if you have ever had a post turn into a real conversation, odds are it traveled by send, not by like.

So here is the order that actually matters, best to worst:

One honest note on that table. Dwell is not really a rung on the ladder, it is the floor the whole ladder stands on. Every signal above it starts with someone stopping to read. It sits low only because it is the one thing on the list a reader does not choose to do.

Does the type of engagement change how the algorithm treats your post? Yes. That is the entire point. The feed does not add a like, a save, and a thoughtful comment into one number called "engagement." It reads them as different signals that mean different things, and it distributes accordingly.

What AI actually changed

One more layer, because my client's real question, the one under the other three, was whether any of this still holds in the age of AI. It does. AI did not rewrite the rules. It raised the bar and thinned the room.

Two facts set the scene. More than half of long-form LinkedIn posts now get flagged as likely AI-written. And reach is down hard across the board. Richard van der Blom's annual analysis, the most-cited one in this space, put views down by roughly half year over year. More noise, less oxygen. Everyone reached for the same easy button at the same time, and the feed got more crowded and less rewarding at once. I wrote about the same thing happening to website copy in "Strip the Logos". It is the same disease, different channel.

Now the myth I most want you to drop, because half the advice out there is built on it. LinkedIn is not running an AI detector and punishing you for using AI. It does not care whether a human or a model typed the words. It watches what people do with the post. Do they stop, read, comment, save. A generic AI post does not get penalized for being AI. It dies because nobody finishes it, which produces the one signal that actually sinks you: no dwell, no conversation. That is a behavioral outcome, not a detector. When LinkedIn describes what it is actually cutting, the list is all about behavior: engagement bait, recycled and click-driven posts, and low-substance filler. The word "AI" is not on it. And there is a quieter tell. LinkedIn once ran its own AI-seeded Collaborative Articles across the feed and pushed them hard, then stepped back from the program. Make of that what you will.

Here is the part that should change how you think about the quarter ahead. The feed now runs on an AI that reads your post for meaning, and it can put a genuinely useful post in front of people who do not even follow you, if it decides they would find it relevant. Sit with that. The machine can now carry a great post past your own network.

One consequence is worth naming. The old trick of trading likes and comments in a pod now works against you. When the AI reads who engaged to decide who else would care, a burst of comments from people with no connection to your world tends to muddy the signal instead of boost it. The right ten comments beat the random fifty.

Which flips the whole game. The scarce thing is no longer a clever posting time or a bigger following. It is the one thing the machine cannot manufacture for you: your firsthand specifics. Your actual numbers. The client situation only you have sat across the table from. The mistake only you made, and can name. That is what earns the dwell, the save, and the send. And it is the one thing every AI in the feed, including LinkedIn's own, is structurally unable to invent.

What I told him

So here is what the quarter ahead actually looks like, stripped of the myths. Six things, in order:

1. Pick two or three lanes you can speak to firsthand, and stay in them. The feed's new AI is trying to figure out what you are known for so it can carry your posts to the right strangers. Help it.
2. Post a few times a week, not daily and not once a month. Consistency beats the calendar, and a second post inside the same day tends to step on the first one's distribution.
3. Favor Wednesday or Thursday, mid-morning where your buyers live. Then let the clock go. It is the tiebreaker, not the strategy.
4. Write every post to be finished and to be saved. If it cannot earn a real comment or a save, no posting time on earth will rescue it.
5. Spend the first hour in the comments, not on the analytics tab. Reply to every real one. That is the tryout, and you get to influence it.
6. Use AI to sharpen your thinking, never to replace it. The second your post sounds like anyone could have written it, you have handed away the only edge you had left.

The clock question is the one everybody asks first because it promises a result without asking anything of you. The things that actually move a B2B post ask for more. Your judgment. Your specifics. An hour of your attention. And the nerve to sound like a person instead of the feed. That is more work. It is also exactly why it still works.

And a post that lands makes the reader look you up. The profile they find is the first handshake, and making it earn the call is what a Profile Power-Up does.

Post for the first hour, not the perfect hour.

All signal. No noise.