Do hashtags still work on LinkedIn? We checked 34,000 posts.

Short answer: hashtags are no longer a growth lever on LinkedIn, and using a lot of them actively signals a weak post. In our analysis of 34,000 LinkedIn posts, four or more hashtags showed up in 41.9% of the worst-performing posts against 25.9% of the best. Meanwhile LinkedIn itself has removed most of the machinery that once made tags worth adding, including hashtag following and prominent hashtag search. Use zero to three, and spend the saved effort on the first line.

What do 12,988 posts say about LinkedIn hashtags?

We took a public dataset of 34,012 influencer posts, filtered to 12,988 English posts from 65 creators, scored each one by (reactions + 4 x comments) divided by follower count, and split the ranking into a top decile of 1,298 posts and a bottom half of 6,494.

Read those two rows together, because they tell different stories and most summaries only quote one of them.

The first row says most strong posts did carry hashtags. That is unsurprising: the dataset comes from a period when tags still fed a discovery surface, and skipping them entirely was leaving something on the table. Anyone who tells you hashtags were always useless is arguing with the record.

The second row says the moment you cross into stuffing, the correlation flips hard. Four or more tags is a bottom-half marker. Nearly half of the weakest posts are tagged like a garage-sale flyer, against a quarter of the strongest.

Across the whole cohort, 57.9% of posts carry at least one hashtag and 35% carry four or more, so this is not a fringe behaviour. It is the single most common thing people do to a post after writing it.

Why does hashtag stuffing travel with weak posts?

Our read, which the data cannot prove: hashtag walls are what people do instead of writing a strong opening line. They are a reach ritual, cheap to perform and satisfying to complete. Ten tags feel like work. Rewriting the first sentence for the fourth time does not.

There is a selection effect underneath that. Posts assembled quickly tend to get the full treatment: a stock image, a link, a block of tags, a generic opener. Posts somebody actually cared about tend to get an opening line and nothing else. So the tag block is a symptom of the effort level rather than a cause of the outcome. Our data cannot separate the two, and anyone who claims to has not looked closely at their own dataset.

The practical conclusion survives either way. If tags are a symptom, removing them will not help but the effort you redirect will. If tags are a cause, removing them helps directly. There is no version of the analysis where adding six more tags is the right move. The patterns that actually separate the groups are in the full engagement study, and almost all of them live in the first 210 characters.

Why you should not trust per-hashtag engagement statistics

Search for LinkedIn hashtags and you will find lists ranking tags by average engagement, or naming the tags that get you the most reach this quarter. We could publish one from our dataset in about ten minutes. We deliberately do not, and the reason matters more than the list would.

In our cohort, 218 hashtags appear on at least 25 posts. Only 60 of those are used by 8 or more distinct accounts. The rest are dominated by a handful of prolific creators, and in many cases by exactly one. When that is true, an average engagement rate for the tag is really an average engagement rate for that person.

Look at the first four rows. Over a thousand posts carry #digitalhealth, which sounds like a healthy sample until you notice they come from four accounts. #themedicalfuturist appears on 726 posts written by one person. If we computed an engagement average for it and published it in a ranked table, we would be publishing a fact about a single creator's audience and calling it a fact about a hashtag.

This is not a quirk of our dataset. It is what hashtag data looks like everywhere, because tag usage is a habit, and habits belong to people. Any tool that surfaces per-tag engagement is aggregating over whoever happened to be in its sample. That is why two hashtag reports published in the same month will rank the same tags in a different order and neither will explain why.

The tags in the last two rows are the kind you could say something about, because they are used across twenty-plus accounts. Even there we would only be describing correlation between a topic and its audience, not the effect of typing a tag. So the number we would be publishing is uninteresting even when it is defensible. We report which tags are broadly used and leave it there.

How many hashtags should you use on LinkedIn?

Zero to three. Here is how that compares to the advice you will find elsewhere, and why we land lower than the consensus.

The gap between three to five and zero to three is smaller than it looks, because the real disagreement is about the floor rather than the ceiling. Sprout treats three as a minimum. We treat zero as a perfectly good answer, since nothing in our data suggests a post is worse off without tags, and the features that made tags useful are gone.

If you do use them, three rules keep you out of trouble. Put them at the very end, after your closing line. Choose tags a human would plausibly search, not compound inventions nobody types. And never let a tag be the first thing above the fold, because that spends the only real estate that decides whether the post gets read. More on that in how to start a LinkedIn post.

What has LinkedIn actually changed about hashtags?

The platform has moved against hashtags on its own, which is the part of this story that makes the dataset finding less relevant rather than more. According to Sprout Social's current rundown, LinkedIn “removed hashtag following, eliminated hashtags from profile displays and stopped surfacing them in the dropdown search bar”, and recent updates have repositioned hashtags “as SEO signals rather than standalone discovery tools”.

Work through what each of those removals took away. Hashtag following was the mechanism by which a stranger who cared about a topic could end up seeing your post without following you. That is gone. Hashtags on profiles were how a visitor could tell what you were about at a glance. Gone. The search dropdown was how anyone found a tag feed in the first place. Gone.

What remains is a keyword hint. Your tags tell LinkedIn's topical systems roughly what the post is about, which they can also work out from the text. That is a real function, and it is a small one. If you want the fuller picture of what the ranking system does and does not weigh, see how the LinkedIn algorithm works.

Did LinkedIn hashtags ever work?

Yes, briefly, and it is worth being clear about that rather than pretending the advice was always nonsense. For a few years LinkedIn ran a genuine hashtag product. You could follow a tag the way you follow a person, tags appeared on profiles as a statement of what you cared about, and a tag feed was a browsable place. During that window, adding a relevant tag put your post in front of people who had opted in to the subject, which is a real distribution mechanism and worth two seconds of typing.

Our dataset comes from exactly that period, which is what makes the finding above more damning rather than less. Even when tags were doing their most useful work, a wall of them still marked a weak post. The behaviour was already a proxy for effort back when the tags themselves had a job.

Then the product was dismantled piece by piece, and the advice did not update. That is the normal life cycle of a social-media tactic: it works, it gets written up, the platform changes, and the write-ups keep circulating because nobody goes back to retire them. If you read a hashtag guide that does not mention hashtag following being removed, it was written from other guides rather than from the platform.

Do hashtags help you get found in LinkedIn search?

A little, and less than your own words do. With the browsable tag feed gone, what is left is a topical hint: a tag tells LinkedIn's keyword systems something about the subject of the post. Those same systems can read the post. If the word appears in your first sentence, the tag is telling them something they already know.

We cannot measure LinkedIn search from a post-level dataset, so treat that as reasoning rather than evidence. What we can say is where the effort pays off with more certainty. Search on LinkedIn is dominated by profiles rather than posts: people look for a person or a skill and land on a headline. A specific, human-readable headline does more discovery work than any tag you will ever append, and it works for every post you publish afterwards.

The other durable answer is topical consistency. An account that publishes about the same few subjects becomes legible to both the ranking system and the reader. You can see what that looks like at scale on the post examples hub, where every topic has its own set of high-performing posts and none of them are winning because of their tags.

Does hashtag use change by content format?

Nobody segments this, and it turns out to be the clearest illustration of what tagging behaviour actually signals. Here is how tagging varies across the four common formats in the cohort.

Text-only posts are the least tagged format by a wide margin. They are also the format with the highest share of first-person openers. Both facts point at the same thing: a post that is nothing but words has to earn its attention with the words, so the person writing it is already thinking about the sentence rather than the metadata.

Image posts are the most stuffed after video, and they are the format that skews hardest toward the bottom half overall. That combination is the reach ritual in its natural habitat: a stock graphic, one line of caption, and a block of tags. Video is heavily tagged too, but it is the strongest format in the study, which is a useful reminder that the tags are not the deciding variable in either direction.

Do the top creators tag differently from everyone else?

Yes, and in a direction that surprises people. Splitting the cohort by author role and comparing each group's top decile against all of its posts shows the same shape repeatedly: the strongest posts use hashtags at least as often, and stuff them far less.

Look at the coaches and consultants row. Their top-decile posts are the most tagged group in the entire table at 88.2%, and simultaneously the least stuffed at 4.1%. Almost none of their best posts carry four or more tags, while almost half of their posts overall do. That is a clean picture of what the whole finding means: one or two deliberate tags is a different behaviour from a tag block, and only the second one marks a weak post.

The usual caveat applies harder here than anywhere else on this page. These role groups contain between 11 and 21 accounts each, so the exact percentages are soft. The reason to trust the shape is that it repeats in every group we can cut, not that any single figure is precise.

Do branded and campaign hashtags work?

The concentration data above answers this by accident, and the answer is more interesting than a yes or no. The single-account tags in our cohort are branded tags in the wild: #themedicalfuturist on 726 posts from one creator, and a long tail of similar personal tags used consistently by exactly one person.

That is a filing system, and a good one. If a reader already cares about you, a consistent tag gives them an archive to browse and gives you a way to reference your own back catalogue. What it does not do is bring anyone new, because the surface that used to let strangers browse tag feeds has been dismantled. A branded hashtag in 2026 is closer to a folder label than a channel.

Company campaign tags face the same problem with an extra one on top: they only work if lots of people use them, and the mechanism that would have shown those posts to strangers is the one LinkedIn removed. If you run a campaign tag, judge it by whether your own community adopts it, not by reach.

Should hashtags go in the post or the first comment?

In the post, at the end, if you use them at all. The first-comment habit came from platforms where a wall of tags in the caption looked unprofessional, so people hid it. On LinkedIn the move solves a problem you can solve better by using fewer tags.

There is a second reason to keep them in the body. Tags now function mainly as a topical keyword hint, and the post body is where LinkedIn most reliably reads your text. Pushing them into a comment is a small bet that the comment is indexed the same way, with no upside if it is. Meanwhile your first comment is genuinely valuable real estate, and there are better uses for it, including the one covered in the external links analysis.

What should you do instead of hashtags?

• Put the keyword in the first sentence.LinkedIn's systems read your actual text. A post that opens by naming its subject in plain words is better targeting than any tag block, and it also helps the human reader decide the post is for them.
• Be consistent about your topics. Accounts that post about the same few subjects teach both the ranking system and the reader what they are for. That consistency is what hashtag following used to approximate, and it is more durable because no product decision can remove it.
• Fix your profile instead. Headline and about-section keywords do the discovery work that tags used to, and they apply to every post you will ever publish. See profile optimization.
• Spend the effort on the hook. The visible first lines decide whether the post earns engagement at all. Our guide to LinkedIn hooks covers what separates the top 10% there, with real examples.
• Show up in the comments.Replying under other people's posts puts you in front of an audience that is already reading. It is the closest thing to the discovery channel hashtags used to be.

How do you test hashtags on your own account?

You do not need our dataset to settle this for yourself, and your own account is the only sample that describes your audience. The protocol that gives a usable answer:

1. Fix everything else first. Same posting cadence, same rough post length, same formats. A test that also changes the format is not a test.
2. Alternate, do not batch. Tags on, tags off, tags on, tags off. Batching confounds the test with whatever else changed that month.
3. Run at least ten posts per condition. Five is not enough. LinkedIn engagement is heavily skewed, so a single viral post will swamp a small sample.
4. Compare medians, not totals. The mean will follow your best post around. The median tells you what a typical post did.
5. Judge on comments. Comments were by far the widest gap between strong and weak posts in our data, and they are less noisy than impressions, which move for reasons that have nothing to do with you.

If you find nothing, that is the expected result and it is worth knowing. It means you can stop thinking about hashtags entirely and put the attention somewhere it pays. If reach has fallen and you are wondering whether tags are the cause, the reach troubleshooting guide covers the likelier explanations.

What this analysis cannot tell you

These are correlations. Heavy tagging travels with weak posts. We cannot prove the tags caused the weakness, and we think the more likely story is that both come from the same low-effort process. The cohort skews toward established creators, all above 1,000 followers, across 65 accounts. Engagement counts are lifetime totals at the time of collection, and post ages vary. The dataset also predates LinkedIn's removal of hashtag following, which means the 78.0% figure describes a platform that no longer exists in that form. That cuts against hashtags, not for them: tags were more useful then than they are now, and they still did not separate the top decile from the bottom half in the direction people assume.