LinkedIn impressions vs reach: what every metric means and which ones matter

LinkedIn impressions count screens; LinkedIn reach counts people. An impression is recorded every time your post is shown, so the same person scrolling past it three times generates three impressions. Reach, which LinkedIn labels “members reached,” is the “number of distinct members and Pages that saw your post.” The gap between the two is large and consistent: AuthoredUp measured members reached at a median of about 46.5% of impressions across 42,493 posts, meaning the typical post reaches roughly half as many people as its impression count suggests. Everything else on your analytics screen is one of three things: a count of a specific action, a ratio built from two of these numbers, or a metric that tells you almost nothing.

What is the difference between LinkedIn impressions and reach?

LinkedIn's post analytics help page defines impressions as the “number of times your post was shown on LinkedIn” and members reached as the “number of distinct members and Pages that saw your post.” On the Pages side, LinkedIn adds two qualifications worth writing on a sticky note: the impression count “is an estimate and may not be precise,” and members reached “is an estimate and does not include repeat displays.”

Both numbers are estimates. Neither is a headcount. The ratio between them is the useful part, and it tells you something the raw numbers do not: whether your post found new people or circled the same ones repeatedly. A post at 3,000 impressions and 700 members reached went round and round a small audience. A post at 3,000 impressions and 2,200 members reached travelled.

There is also a naming trap. AuthoredUp notes that LinkedIn “relabeled the post metric from ‘impressions’ to ‘views’ in its feed.” That matters because “views” already meant something stricter for video, where LinkedIn counts a view only after two or more continuous seconds of watching. Two metrics with different definitions now share a label, and third-party tools do not all handle it the same way. Before comparing your numbers to anyone else's, find out which definition their tool used.

What does every LinkedIn post metric actually mean?

This is the table the rest of the article hangs off. Every definition below is LinkedIn's own wording from its post analytics help page, with a plain-English note on what it is good for.

Two footnotes from LinkedIn's analytics documentation that change how you read all of the above. First: “Your own views, social engagements, link engagements, saves, and sends are counted towards the analytics of your content.” You are in your own numbers. On a post with 200 impressions that is not a rounding error. Second, retention differs by metric: discovery and social engagement counts stay for 1,000 days, members reached for 400 days, demographic breakdowns for 180. Long-range comparisons quietly lose columns.

Why do impressions and members reached diverge so much?

Three mechanisms, and each one tells you something different about the post.

• Repeat exposure. LinkedIn will show a post to the same member more than once, especially when it is getting engagement. High impressions with low reach means the same audience saw it repeatedly.
• Comment resurfacing. Every new comment gives the post another reason to reappear for people who already saw it. Comment-heavy posts naturally show a lower reach-to-impression ratio.
• Distribution shape. A post pushed to a wide, shallow audience racks up reach without repeats. A post held inside a tight network racks up impressions.

The median ratio of about 46.5% is a useful anchor. Far above it and your post is finding new people once each. Far below it and you are being shown repeatedly to a small group, which is not necessarily bad (it is what happens when a niche post lands hard) but is worth knowing before you read the impression number as audience growth.

How do you calculate LinkedIn engagement rate?

There is no standard. That is not a complaint, it is the single most important fact about engagement rate, because two credible studies can publish opposite rankings using identical data.

We weight one comment as four reactions. That number is a judgement call, not something LinkedIn published, and we say so plainly. The reasoning: a comment costs the reader seconds rather than a tap, it puts the post into the commenter's network, and it creates replies that create more comments. In our data the choice was validated after the fact, because comments turned out to separate strong posts from weak ones far more sharply than reactions did.

The denominator matters more than the numerator. Divide by impressions and you are asking “did the people who saw this respond?” Divide by followers and you are asking “is this account worth its audience?” Those are different businesses. If you are trying to justify time spent on LinkedIn, the follower-based version is closer to the question you are actually asking.

Why do two studies rank company pages and personal profiles differently?

Because of exactly that denominator, and it is the cleanest worked example of why formula choice is not a detail.

Metricool's study of 673,658 posts across 63,108 accounts found personal profiles beat company pages on engagement by 63%, and generate 238% more comments per post. AuthoredUp's analysis of 476,781 posts reports the reverse ordering: company pages at a 2.60% median engagement rate against 2.38% for personal profiles.

Both are correct. AuthoredUp divides by impressions, which is LinkedIn's own method for Pages. Company pages get fewer impressions per follower, so a per-impression rate flatters them; a per-follower rate does not. If you ever find two LinkedIn studies contradicting each other, check the denominator before you check anything else. It explains most disagreements in this field.

What is a good LinkedIn engagement rate in 2026?

Here are the benchmarks worth keeping, grouped by which formula they use so you compare yourself against the right one.

Notice that the Socialinsider average is roughly double AuthoredUp's median for the same kind of account. Different population (business pages only), different numerator (their interactions include clicks), and an average rather than a median on a skewed distribution. All three differences push the number up. None of them makes either study wrong.

The one relationship that repeats across every dataset: engagement rate falls as reach rises. AuthoredUp measured 2.86% at a median 496 impressions for typical posts and 0.89% at a median 34,121 impressions for viral ones. Their arithmetic is worth sitting with: a viral post at 34,000 impressions and 0.89% produces 303 interactions, while a typical post at 500 impressions and 2.86% produces 14. Twenty times the interactions for seventy times the reach.

Which of these metrics does the algorithm actually use?

Fewer than you would guess, and the overlap between “metrics LinkedIn shows you” and “metrics LinkedIn ranks on” is smaller than the analytics screen implies.

LinkedIn's March 2026 engineering post on the feed names the outcomes its ranking model is trained to predict: “passive tasks (click, skip, long-dwell) and active tasks (like, comment, share).” Six things. Three of them are invisible in your analytics.

The two rows that matter are dwell and skip. LinkedIn's earlier engineering work on dwell time explains why they exist: clicks and reactions are “sparse,” binary, and noisy, while dwell is “always measurable” and real-valued. LinkedIn built a model to predict the probability that a member's dwell on a post falls below a skip threshold, and reduces that post's ranking score accordingly.

The consequence for measurement is uncomfortable: the two strongest ranking signals are not in your dashboard. The best proxies you have are video watch time, and the reach-to-impression ratio, neither of which is the same thing. This is a good reason to stop over-fitting to the numbers you can see.

How do Page metrics differ from personal profile metrics?

Pages get a calculated engagement rate. Personal profiles do not, which is why every tool that reports one for a profile had to invent a formula.

On its Page content analytics page, LinkedIn defines engagement rate as “the ratio of interactions per impressions on your post,” where interactions include clicks, reactions, comments, and shares. Clicks in the numerator is the detail that trips people up: a Page post with a link can post a healthy engagement rate on the strength of clicks alone, while a profile post measured by any of the common third-party formulas cannot.

Pages also report clicks as a standalone metric, defined as clicks “on your content, company name, or logo by a signed in member,” explicitly excluding reposts, reactions, and comments. Profiles get link visits but not the company-name and logo component.

The practical upshot when comparing a Page against a founder's profile: you are almost certainly comparing two different formulas. Rebuild both from raw counts (reactions, comments, reposts, and one shared denominator) before drawing a conclusion. Otherwise the Page will look better than it is on link-heavy posts and worse than it is on everything else. The underlying performance difference, and why personal posts tend to win on a follower basis, is covered in the LinkedIn algorithm guide.

Which LinkedIn metrics are vanity and which predict business outcomes?

Ranked from least to most predictive of anything commercial. This ranking is our judgement applied to LinkedIn's own metric definitions, not a measured causal study, and we would rather say that than dress it up.

The two metrics in the outcome-adjacent row deserve more attention than they get. LinkedIn reports, per post, how many distinct people viewed your profile from it and how many followed you because of it. That is a per-post conversion rate from attention to interest, sitting in the app, free, and almost never mentioned in guides about LinkedIn metrics. If you only ever look at two numbers, look at those.

Why do impressions go up while results go down?

Because impressions rise fastest exactly when a post leaves the audience that can buy from you.

The engagement rate decline with reach is the visible half of this. The invisible half is composition. A post that goes wide on LinkedIn typically goes wide because it touched a general professional nerve, and general professional nerves are not segmented by whether the reader has your problem. Ten thousand impressions from people outside your market produce a good screenshot and nothing else.

This is also why our cohort's comment gap matters more than its reaction gap. 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, comments by more than 14x. Comments are where people self-identify, and a person who writes three sentences under your post has told you something a thousand impressions cannot. That is the whole argument of our LinkedIn comment strategy guide.

The mechanical explanation for why comments also buy you distribution, rather than just signalling interest, is in how the LinkedIn algorithm works.

What should you actually track every week?

Five numbers, on one row per week. Not per post, because per-post variance will drown you.

1. Median members reached per post. Median, not total, and not average. One outlier ruins an average.
2. Total comments received. The scarcest signal and the one most under your control.
3. Median out-of-network percentage. Whether the algorithm is still expanding you past your own audience.
4. Profile viewers from posts. Attention converting into interest.
5. Inbound conversations started. Count them by hand. LinkedIn will not do it for you and it is the only number your business cares about.

Impressions do not appear on that list. Neither do reactions. Both are fine as inputs to a rate and useless as goals, and treating them as goals is how people end up writing posts that get applause from people who will never hire them. If your numbers have fallen and you want to work out why, the diagnostic is in why did my LinkedIn reach drop.

Segment the five numbers by format once a quarter

The weekly row tells you the trend. A quarterly cut by format tells you the cause. Split those same five metrics by post type (text, image, video, document, poll, link, reshare) and by whether the post carried an outbound URL.

That one cut usually settles arguments that studies cannot settle for you. Independent datasets disagree about video and agree about links, so your own numbers are the only thing that resolves it for your account. In our cohort, article and link posts made up 32.7% of the bottom half against 22.1% of the top decile, and the wider evidence is in do external links kill your LinkedIn reach. Do the same cut for hashtag counts while you are in the spreadsheet; the pattern there is covered in do hashtags still work on LinkedIn.

How do you benchmark yourself honestly?

Against your own last quarter first, then against a percentile band that matches your account size, then against nobody.

Comparing against a creator with 200,000 followers is comparing across a gradient. If you want the percentile bands laid out by account size, they are in our LinkedIn engagement benchmarks. AuthoredUp found median engagement rate declines steadily with audience size, from 2.68% in the 1,001 to 5,000 follower band down to 1.53% above 100,000 followers. Large accounts reach further and convert worse. Neither number tells you anything about the other.

Within our own cohort, the useful frame is percentile rather than absolute. A median post sits at 0.40 per 1,000 followers. Reaching 1.27 puts you in the top quarter. Reaching 5.95 puts you in the top tenth. If you have 3,000 followers, that top-quarter threshold is roughly four engagement-weighted points per post, which is a handful of comments and a few dozen reactions. Written that way, the top quartile stops sounding mythical.

Two caveats we will not bury. Our cohort skews toward established creators with at least 1,000 followers across 65 accounts, and engagement counts are lifetime totals captured at scrape time, so post ages vary. Treat the percentiles as a map, not a scoreboard. The patterns behind them are in our study of 34,000 LinkedIn posts, and the writing mechanics that move them are in the hooks guide.