LinkedIn engagement rate by industry: what is actually published

There is one LinkedIn engagement rate by industry table we could actually trace to a publisher, and it runs from 2.8% for government to 4.7% for marketing agencies, published by Hootsuite and measured on each industry's best-performing content format. It states no sample size, no date range and no denominator. The three biggest LinkedIn benchmark studies, covering 1.3 million posts, 673,658 posts and 58,000 posts respectively, publish no industry split at all. We cannot publish one either, because 65 creators cannot be cut into industries honestly. What we can add is the thing our data does hold: engagement by topic of post across 12,988 posts, which is a different variable and we will keep saying so.

What is the average LinkedIn engagement rate by industry?

We chased every source we could find for this query and ended up with a single table. It comes from Hootsuite's social media benchmarks page, last updated 14 April 2026. Reproduced as published, including the format qualifier, which matters a great deal and which almost everyone who copies this table drops.

The spread from top to bottom is 1.9 percentage points, or roughly 1.7 times. That is a modest spread by the standards of this field, and it is smaller than the spread the same publishers report between formats on a single platform. Hold onto that, because it is the argument against organising your content strategy around your industry benchmark.

What does the Hootsuite table leave out?

Four things, all of which change how you should use it. We are listing them because the page does not.

• No sample size. The page does not say how many posts or how many accounts the figures come from.
• No date range. The page carries a 2026 update date. The measurement window is not stated.
• No denominator. Engagement rate per follower and engagement rate per impression differ by roughly an order of magnitude on the same posts. The table does not say which it used.
• These are best-format figures, not averages.Every row is the industry's best-performing content type. Retail's 4.3% is a video number, not a retail number. Comparing your all-format average against it is comparing your mean against somebody else's maximum.

The table also carries a posting frequency in brackets for several rows, and those brackets are informative in an unintended way. Retail's figure is attached to 15 posts per week, marketing agencies' to 19, utilities and energy's to 10, while most other rows sit at two. An industry posting nineteen times a week and an industry posting twice a week are not two samples of the same behaviour, and any difference between their engagement rates is partly a difference in cadence. Our own read on that trade-off is in how often to post on LinkedIn.

Which large studies publish a LinkedIn industry split, and which do not?

We fetched each of these and checked. The absences are as informative as the presences, so we are publishing both.

Four of the six largest published LinkedIn datasets have no industry dimension in their public output. That is not laziness on their part. Industry is genuinely hard to assign at scale: a page belongs to a company, a company belongs to several categories, and the categories differ between data providers. Follower tier and post format are unambiguous and machine-readable. Industry is neither.

Why does the most-cited source contain no LinkedIn data?

Rival IQ's Social Media Industry Benchmark Report is the single most quoted document in social benchmarking. It samples 150 companies from each of fourteen industries, covers more than four million posts and nine billion interactions, and defines engagement rate as all interactions divided by follower count. It is a serious piece of work.

It covers Facebook, Instagram, TikTok and X. LinkedIn is not in it.

Rival IQ does publish a separate LinkedIn report, and that is the one worth citing: 58,000 posts from January to September 2023, from handles with at least 2,000 followers, reaching across fourteen industries. It reports a median engagement rate by follower of 0.41%, an engagement rate by impression of 4.73%, about 8.63 impressions per 100 followers per post, and a median cadence of 3.3 posts per week. It reports those figures in aggregate. It contains no engagement rate per industry.

So when you read “according to Rival IQ, LinkedIn engagement in financial services is...”, the safest assumption is that somebody merged two documents. The 0.41% and 4.73% figures on the same posts are also the clearest demonstration of the denominator problem that runs through every benchmark on this subject, which we unpacked in our breakdown of average LinkedIn engagement rates.

What can we publish instead? Engagement by topic across 12,988 posts

Our dataset is 12,988 English posts from 65 creators with at least 1,000 followers, scored as (reactions + 4 x comments) divided by followers, reported per 1,000 followers. It carries post text, media type, reactions, comments and follower counts. It does not carry an employer field, so it cannot tell you anything about industry.

It can tell you about subject matter, because subject matter is in the text. Every row below is the topic of the post. The author count is printed next to each figure so you can see how many different people stand behind it.

The spread here is much wider than the published industry spread: 3.72 against 0.20 is more than eighteen times, where Hootsuite's industry table spans 1.7 times. If subject matter moves engagement eighteen-fold and industry moves it under two-fold, industry is not the variable to organise around.

Look at what sits at the top. Personal branding, job search, networking, career growth, personal story. The subject of each is a person, usually the author. The bottom of the table is subject matter about the world: technology, productivity, finance. Note that finance carries 2,795 posts from 51 authors, so its 0.20 is not a sampling accident. It is the largest topic in the cohort and the weakest performer.

Is topic of post the same as industry of company?

No, and the difference is the most common error in this category of article. Industry describes the poster. Topic describes the post. A software firm publishing an employee's promotion story is technology by industry and career growth by topic. A bank publishing a market outlook is financial services by industry and finance by topic.

Our data can only see the second. That is a limitation, and it is also why the table above is more actionable than the industry table. You cannot change your industry before Thursday. You can change what your Thursday post is about.

There is one more honest caveat. Creators who write about personal branding are frequently creators who study the platform, so some of the 3.72 is selection rather than subject. What the table reliably tells you is that if you post about technology or finance and your numbers look weak against a generic benchmark, part of the gap was assigned before you wrote a word.

What is the closest thing we have to an industry split?

Author role. The dataset lets us bucket creators by what they do rather than by what sector their employer sits in, which is one step nearer to industry than topic is and still not the same thing. Author counts are small in every bucket, so read this as five groups of people rather than five markets.

The founders row is the one to study. Founders and CEOs earn more reactions and more comments per post than executives do, 132 and 12 against 101 and 8, and post a lower rate, 0.19 against 0.46. Nothing about their content is weaker. Their follower counts are larger, and rate is a ratio. Any industry benchmark that does not control for account size is reproducing this artefact and calling it an industry effect.

Does the topic gap survive when you compare like with like?

This is the cut that answers the obvious objection, which is that the topic table might just be sorting good creators from bad ones. So we split each topic against itself: that topic's own top decile against that topic's own bottom half. Same subject, same broad population, different outcomes.

Median engagement rate per 1,000 followers, within-topic. Every one of these comparisons holds subject matter constant, and every one shows a gap of forty times or more between a topic's best posts and its weakest. The gap between topics, large as it is, is smaller than the gap inside a single topic.

The practical reading: a finance post can reach 8.93 per 1,000 followers, which is above the 90th percentile for the whole cohort. Finance is not a ceiling. It is a harder starting position, and the difference between the top and the bottom of it is decided by the same things that decide every other topic.

What separates the best posts from the worst inside one industry-like topic?

Take finance, the weakest topic in our table and the one whose practitioners most often say the platform is not built for them. Comparing its own top decile (279 posts, 24 authors) against its own bottom half (1,398 posts, 25 authors):

• First-person openers: 12.9% in the top decile against 2.1% in the bottom half. A six-fold difference, the widest of any topic we looked at.
• Second person anywhere in the hook: 46.2% against 14.9%.
• Four or more hashtags: 33.7% against 74.0%. Three quarters of the weakest finance posts are hashtag-stuffed.
• Native video: 21.9% of top-decile finance posts against 3.1% of bottom-half ones.

Technology looks the same in the hashtag column: 49.4% of its top decile carries four or more hashtags against 79.6% of its bottom half, across 162 and 814 posts respectively. If you work in a technical field and your posts underperform, the hashtag habit is the first thing we would look at, and what the data says about LinkedIn hashtags is where we went through it in full.

Which format works best in which industry?

The one genuinely useful column in Hootsuite's table is not the engagement rate. It is the format column. Video is the best-performing format in eight of the thirteen industries listed, photos in three, and a tie between the two in the rest. No industry's best format is text, and none is a link post.

That lines up with the format hierarchy in the studies that do publish their methodology. Socialinsider, across 1.3 million posts from 16,645 company pages, reports native documents at 7.00% engagement, multi-image at 6.45%, video at 6.00%, image at 5.30%, text at 4.50%, polls at 4.20% and link posts at 3.25%. Buffer, from a different panel, reports LinkedIn carousels at a 21.77% median engagement rate against 7.35% for video, 6.52% for images, 3.81% for link posts and 3.18% for text.

The two disagree about the size of the carousel advantage and agree completely about the order. Documents and carousels at the top, link posts at the bottom, in every industry that has been measured. Our own format table, scored per follower rather than per impression, produces the same ranking from a third direction: document posts at 2.42 per 1,000 followers on a thin sample of 60 posts from 13 authors, native video at 0.60 across 1,613 posts, text at 0.58, and image and shared links tied at the bottom on 0.33.

The practical consequence is that a format change is available to every industry, while an industry change is available to none. If you are in one of the lower rows of the industry table, moving from link posts to document carousels or native video is a bigger lever than the entire spread between the best and worst industries.

Should you benchmark a company page or a personal profile?

Whichever one you actually post from, and the answer changes the number you should compare against by a lot.

Almost every industry table in circulation is built from company pages, because page data is what social analytics vendors can collect at scale. Socialinsider's 16,645 accounts are all business pages. Rival IQ's handles are brands. If you post from a personal profile and compare yourself to those figures, you are comparing two populations that behave differently.

Metricool measured both in the same window and found personal profiles at a 2.60% engagement rate against 1.74% for company pages, roughly 63% higher, on almost identical impressions (817.67 against 812.64 per post). It also found that pages collect more shares while profiles collect substantially more comments. Given that comments are the scarcer and more discriminating signal in our data, that difference matters more than the headline rate. We broke down the impressions side of this in our impressions benchmark by follower tier.

Why do published industry benchmarks disagree with each other?

Five reasons, in rough order of how much damage each does.

1. The denominator. Per follower and per impression differ by an order of magnitude. Rival IQ reports 0.41% and 4.73% for the same posts.
2. Pages against profiles. Most large studies are built on company pages, because page data is available at scale through official APIs. Metricool, which measured both, found personal profiles running 63% higher engagement than pages.
3. Best format against all formats.A table of each industry's strongest format looks nothing like a table of each industry's average.
4. Mean against median. Engagement is heavily skewed, so means read high and almost nobody hits them.
5. Industry taxonomy. Providers assign categories differently. A fintech company can be technology in one dataset and financial services in another.

Before you compare yourself to any published industry figure, look for the sample size, the date range and the denominator. If a page does not print all three, you are reading a ranking, not a measurement.

How do you build an industry benchmark that is actually about you?

1. Pick eight to twelve accounts in your field whose follower count is within roughly a factor of two of yours. Peers, not the three giants everyone names.
2. Record their last ten posts each. Reactions, comments, follower count, format, topic. All of it is public.
3. Score every post the same way: (reactions + 4 x comments) divided by followers, times 1,000. Our engagement rate calculator will do the arithmetic.
4. Compute the median and the 90th percentile of that pool. You now have a benchmark built from accounts that share your industry, your rough size and your calendar.
5. Tag each post by topic, not by company. Then check whether the top posts in your peer set cluster on a subject rather than a format. In our data they usually do.
6. Refresh it quarterly. Platform-wide engagement is moving, and a peer benchmark drifts with it automatically while a published figure does not.

This takes about an hour and gives you something no published table can: a comparison set that shares your constraints. It also pairs with the account-level method in our benchmarks guide and with the planning approach in our LinkedIn content strategy guide.

What are the limits of everything on this page?

• We have no industry data. Not a small sample of it. None. The dataset has no employer field.
• Our topic labels are derived from post text, so a post can sit in more than one topic and the labels are approximate.
• Correlational, not causal. Nothing here shows that switching topic causes a rate change.
• Our cohort skews established: 65 creators, all above 1,000 followers.
• Engagement counts are lifetime totals at scrape time and post ages vary.
• Text findings describe the visible hook only, because the source truncates at the see-more fold.
• The Hootsuite figures are unverifiable in the sense that matters. We can verify that Hootsuite published them. We cannot verify what they were measured on.

LinkedIn engagement rate by industry, restated

The published range is 2.8% to 4.7%, from a single table with no methodology attached and measured on each industry's best format. The largest LinkedIn studies do not publish an industry split, and the report most often cited for one covers four platforms that do not include LinkedIn. That is the honest state of the evidence.

The variable that does move engagement in data we can inspect is what the post is about, not who wrote it: 3.72 per 1,000 followers for personal branding against 0.20 for finance, across 12,988 posts. And the gap inside a single topic is wider still. Whatever industry you are in, the distance between your best posts and your worst is larger than the distance between your industry and somebody else's, which is the whole argument of our study of what the top decile does.