LinkedIn comment strategy: how to use the platform's scarcest signal

Comments are the scarcest signal on LinkedIn, and that is the entire case for having a comment strategy. In our analysis of 12,988 posts, the top-decile median post earned 235 reactions and 72 comments. The bottom-half median earned 80 reactions and 5 comments. Reactions differ by roughly 3x between the best and worst posts. Comments differ by more than 14x. Comments are also the only engagement type LinkedIn names as a ranking objective that a reader has to think before producing, which is why a post with 40 comments usually travels further than a post with 400 likes.

Why are comments the signal that matters most on LinkedIn?

Three reasons, and only the first one gets discussed.

They are named as a ranking objective. LinkedIn's March 2026 engineering post on the feed lists what its ranking model is trained to predict: “passive tasks (click, skip, long-dwell) and active tasks (like, comment, share).” Comment is on the list. LinkedIn has never published relative weights, and anyone quoting you a ratio invented it.

They redistribute the post.A comment surfaces your post to the commenter's network in a way a reaction does more weakly. That is second-order distribution you did not have to earn twice.

They compound. A comment invites a reply, a reply invites another comment, and each new comment gives LinkedIn a fresh reason to resurface the post to people who already scrolled past it. Nothing else in organic LinkedIn has that property. A reaction is terminal.

There is a fourth reason that has nothing to do with the algorithm. A person who writes three sentences under your post has told you what they care about, in their own words, with their job title attached. A thousand impressions cannot do that.

What do 12,988 posts say about comment-heavy posts?

We scored 12,988 English LinkedIn posts by engagement rate and compared the top decile with the bottom half. Three text patterns in the visible hook separate them, and all three point the same direction: comments come from posts that address a person rather than a topic.

The shape of a comment-earning opener is my story, your problem. The author supplies a specific experience, the reader supplies the stakes, and the question is the door. Note the size of the effects: the “you” gap is large, the question mark gap is real but modest. A question is not a magic switch, it is one of several ways to leave something open.

The example below is a real post from the dataset, and one of very few in the cohort with more comments than reactions.

Comments by format

Format changes how many comments a post earns, independently of what it says. These are medians across our whole cohort, with the sample size for each.

The document row rests on 60 posts, so treat it as directional rather than settled. The rest is a clean ordering: formats that carry an argument earn conversation, formats that carry a headline do not. A shared article gets a third of the comments a plain text post does. If comments are the goal, the biggest single change you can make is to stop posting links, which we cover in do external links kill your LinkedIn reach.

Comments by topic

Subject matter moves the number as much as format does. Medians across our cohort, with sample sizes.

Subjects where readers have a personal stake and a differing opinion produce four to five times the comments of subjects where they mostly agree or mostly do not have a view. Technology posts get read; personal branding posts get argued with.

Is the first hour after posting really the golden hour?

Early engagement matters. The sixty-minute figure is folklore, and it is worth separating the two because people plan their days around the second one.

No LinkedIn help page or engineering post names a window. What is documented is that engagement arrives fast: Metricool's study of 673,658 posts found roughly 40% of a post's interactions land on day one, and about 50% of lifetime impressions arrive within the first two days. Fast, but not a cliff. Half the impressions arrive after day two.

The mechanism that makes the first hour genuinely useful is not a scoring threshold, it is you. LinkedIn's 2019 engineering post on community-focused feed optimization describes adding “contribution” as a ranking objective and tuning for a creator-side effect, on the observation that “while viewers are less sensitive to the freshness, content creators do care about receiving prompt responses.” A comment answered in three minutes usually produces another comment. The same comment answered three days later produces nothing.

So the practical rule stands even though the folklore behind it does not: publish when you can be at your desk for the following hour. Not because a timer flips, but because that is the hour when replying multiplies.

How should you reply to comments on your own posts?

Every substantive comment gets a reply, and the reply continues the thought. That is the whole rule, and almost nobody follows it.

• Never reply with thanks alone. It closes the thread. It adds a comment count without adding a conversation, and the person who wrote three sentences learns not to bother next time.
• Reply with a question or a specific.“How did that play out for you?” or “We saw the opposite at 40 people, here is why.” Both invite a second comment from the same person.
• Reply to the disagreement first. Disagreement is the highest-value comment you can get. It brings the commenter back, brings their network, and is the only thing that reliably produces a thread rather than a list.
• Use the person's name. Cheap, obvious, and it visibly changes reply rates in any account we have run.
• Keep replying on day two and three.Half of a post's impressions arrive after the first two days, which means new readers are still arriving after most creators have stopped paying attention.

One thing to skip: replying to your own post with an extra comment to “boost” it. It does not read as conversation, and if the extra comment is the link you removed from the post body, say so plainly instead of dressing it up.

What to do when a post gets more comments than you can answer

This is a good problem and most people handle it badly, either by burning an afternoon on one-word replies or by abandoning the thread entirely.

Work in three passes. First pass, reply only to disagreements and questions, in the order they arrived. Second pass, reply to anyone who added a specific fact or their own story, using their name. Third pass, react to the rest and stop. Then come back the next morning and do the same again with whatever arrived overnight, because half of a post's impressions land after the first two days and the late comments come from a different set of people.

One thing worth doing that almost nobody does: when a comment is better than anything in your post, say so publicly and quote it. It costs nothing, it is true often enough to be honest, and it produces the highest reply rate of anything in this section.

What makes a comment worth replying to?

You will not reply to all of them once a post does well, so it helps to have a filter that is not simply first-come-first-served.

Does commenting on other people's posts grow your audience?

It is the most reliable growth channel available to a small account, for one structural reason: a comment on someone else's post is shown to their audience. You are not limited by your own distribution.

This matters most at the bottom of the follower curve. AuthoredUp's engagement rate analysis of 476,781 posts found small accounts convert attention better than large ones (a median 2.68% for profiles with 1,001 to 5,000 followers against 1.53% above 100,000). Small accounts do not have an engagement problem. They have an exposure problem, and commenting is the cheapest fix for exposure that exists.

What a comment worth writing looks like

• It stands alone. Someone who has not read the post should still get something from it. That is the bar, and it eliminates 95% of what gets posted.
• It contains something only you know. A number from your own work, a case where the opposite happened, a mistake you made doing the thing being described.
• It is two to four sentences. Long enough to say something, short enough to read in the feed.
• It does not pitch. The profile link is right there. Let the comment do the work.
• It sometimes disagrees. Politely and with a reason. Agreement is invisible; a well-argued disagreement gets read to the end.

Where to spend the comments

Pick 10 to 20 accounts whose audience overlaps the people you want to reach, not the biggest accounts you can find. A comment on a 200,000-follower generalist post competes with 400 other comments. A comment on a 4,000-follower post from someone your buyers actually follow gets read.

Timing helps here in a way it does not for your own posts: comment early on other people's posts, while the comment is still near the top and the author is still replying. That is where the golden hour idea does real work.

Why do company pages get so few comments?

Because nobody argues with a logo. This is the most reliable finding in the personal profile versus company page literature, and it is specifically about comments rather than engagement in general.

Metricool's study found personal profiles generate 238% more comments per post than company pages. The same study found company pages get shared up to 17 times more often across all formats, which is a real strength and a completely different one. Pages get forwarded; people get replied to.

The mechanism is not algorithmic, it is social. A comment is a public statement with your name and job title attached. Replying to a named person is a conversation; replying to a brand account is a customer service ticket, and most people would rather not be seen doing it in the feed.

The practical consequence for anyone running both: put the argument on the founder's profile and the announcement on the Page. If a Page post needs comments, the only reliable source is employees who genuinely have something to add, commenting under their own names. Note that organising that is only legitimate if nobody has agreed in advance to engage regardless of content, which is exactly the line LinkedIn's policies draw.

Do polls count as a comment strategy?

No, and the data on polls is the clearest example of a format that buys reach without buying conversation.

AuthoredUp's analysis of more than 3 million posts found polls at 1.78x baseline reach, the highest of any format, and 0.37x baseline engagement, the lowest. Their explanation is the right one: votes are not discussion. Socialinsider's benchmarks put polls at a 4.20% engagement rate on business pages, below documents, multi-image, video, and images.

A vote costs one tap and reveals nothing. It does not surface the post to the voter's network, it does not produce a reply, and it does not tell you who the person is. If your objective is reach on a slow week, a poll is a legitimate tool. If your objective is conversation, a poll actively substitutes for it: people who would have written a sentence tap an option instead.

The one poll pattern that does earn comments is a poll with a deliberately incomplete set of options and a line asking for the missing one. That converts the tap into a sentence, which is the whole point.

Are LinkedIn engagement pods worth the risk?

No, and this is one of the few places in LinkedIn advice where the rule is written down in plain language rather than inferred.

The Professional Community Policies say: “Don't do things to artificially increase engagement with your content. Respond authentically to others' content and don't agree with others ahead of time to like or re-share each other's content.” That is a direct description of a pod. LinkedIn states that depending on severity it may limit the visibility of content, apply labels to it, or remove it, and that repeated or egregious violations can restrict the account.

The risk is asymmetric. The upside is a temporary bump in numbers that do not convert. The downside is account-level and lands on the asset you spent years building. We go through the mechanics and the alternatives in the guide to LinkedIn engagement pods.

There is also a quieter problem. Pod comments are generic by construction, because the commenter has not read the post. Generic comments do not produce replies, do not produce threads, and do not bring anyone to your profile. You end up with the metric and not the thing the metric was a proxy for.

The legitimate version of the same idea is a small group of people in your field who genuinely read each other's posts and comment when they have something to say. That is not a pod, that is a professional network, and the difference is whether anyone agreed in advance.

How do you write posts that earn comments in the first place?

Four things move the number, in descending order of effect size in the data we have seen.

1. Take a position someone could disagree with. Posts that state a view get argued with; posts that summarise get scrolled. This is also the one lever with no downside if you are right.
2. Ask one real question, at the end. Metricool found posts containing a question earned 77% more comments, and posts with a clear call to comment earned 80% more. One question, not three, and one you actually want answered.
3. Leave something out. A post that answers everything has nothing for the reader to add. A post that describes a problem and one attempt at solving it invites the other attempts.
4. Address a person, not an audience.Nearly half of top-decile posts say “you” in the visible hook against a third of the bottom half. Details in how to start a LinkedIn post.

A caution on question hooks specifically, because this is where the field oversells. AuthoredUp's analysis of 417,550 posts classified by opening style found question hooks at a median 2.38% engagement rate, below contrarian hooks at 3.07% and flat statements at 2.58%. Questions are good at producing comments and not automatically good at producing engagement overall. Put the question at the end and let something sharper open the post. More on that trade-off in the LinkedIn hooks guide.

What does a weekly comment routine look like?

Concrete, and about 25 minutes a day.

• Morning, 10 minutes. Five comments on posts from your target list. Two to four sentences each, at least one containing a number or a counterexample.
• Publish, then stay for an hour. On the days you post. Reply to everything substantive as it arrives.
• Afternoon, 10 minutes.Clear the replies on yesterday's post. Almost nobody does this and it is where second conversations start.
• Friday, 5 minutes. Count comments received for the week and note which posts produced them. That is the only comment metric worth tracking.

The rule that makes this sustainable: never comment to be seen commenting. If you have nothing to add to a post, react and move on. Five real comments beat fifty empty ones, and the empty ones are indistinguishable from the automated behaviour LinkedIn's spam policy exists to catch.

How do you measure whether your comment strategy is working?

Three numbers, none of which is total engagement.

1. Comments per post, as a median over the last 15 posts. Medians, because one post that took off will make an average meaningless.
2. Comments per 1,000 members reached. This separates a comment strategy working from a post that simply got more distribution. Use members reached rather than impressions, for the reasons in LinkedIn impressions vs reach.
3. Profile viewers from this post. LinkedIn reports this per post, and it is the closest free measure of a comment turning into interest. If your comments are rising and profile views are not, the comments are coming from people who were never going to look you up.

For context on what those numbers should look like: the median post in our cohort of creators with 1,000+ followers earned 12 comments. The top decile earned 72. If you are posting into an audience of a few thousand and clearing 10 comments consistently, you are doing better than the middle of a cohort made of established creators.

Two caveats we will keep repeating. Our findings are correlational, the cohort skews toward creators who already had an audience, and engagement counts are lifetime totals at scrape time. They describe what strong posts have in common, not a guaranteed mechanism. The distribution logic behind why any of this moves reach is in how the LinkedIn algorithm works, and the broader pattern set is in our study of 34,000 LinkedIn posts.