What are impressions on LinkedIn?

LinkedIn defines an impression, on its post analytics help page, as the "number of times your post was shown on LinkedIn". The combined analytics page says the same thing in slightly different words: "the number of times your post was displayed on screen." Two things follow from the definition. An impression counts a display, so the same person scrolling past your post twice is two impressions. And LinkedIn says on both pages that the figure is an estimate and may not be precise, so a difference of a few percent between two readings means nothing.

An impression is the top of the funnel and the loosest number on the page. It says the post was put in front of someone. It does not say they read it, stopped on it, or noticed whose it was. That is why every serious read of LinkedIn numbers starts with impressions and immediately divides by something.

Members reached is LinkedIn's name for unique viewers. The post analytics page defines it as the "number of distinct members and Pages that saw your post", and the combined analytics page adds the operative sentence: "repeat views aren't counted." So if one person saw your post five times, that is five impressions and one member reached.

The gap between the two numbers is informative on its own. Impressions well above members reached means the same people kept seeing the post, which usually means it circulated inside your own network. The two numbers close to each other means each display went to a new person, which is what a post that travelled looks like. LinkedIn also splits impressions into in network and out of network, the share from people who follow or are connected to you and the share from people who do not, which answers the same question more directly.

Views is a word LinkedIn uses for several different things, which is where most of the confusion starts.

• Video views. On a video post, LinkedIn counts a view as a play "watched for more than 2 seconds", per its video analytics help page. An impression is the post appearing on screen; a view is the video actually running for a moment. A video post always has more impressions than views.
• Profile views. People who opened your profile. On a post's analytics LinkedIn reports the distinct members who viewed your profile from that post; on your dashboard it reports profile viewers over a period. Neither has anything to do with how many times the post was displayed.
• "Post views" in advice you read. When a tool or an article talks about post views, check which number it means. LinkedIn's own analytics pages use impressions for displays of a post and views only for video plays, so advice about views is usually advice about impressions wearing an older word.

The rule of thumb: impressions measure the feed's decision to show you, views measure a reader's decision to do something, and the two should never be compared as if they were one number.

Under any post you published, LinkedIn shows a count in the lower left and a View analytics link in the lower right; either one opens the post's analytics, per the help page on viewing post analytics. Only you can see them. The same page says discovery and social engagement counts stay available for 1,000 days and video performance for 180, so an old post's reach can still be checked long after it stopped earning any.

For the whole account, the combined post analytics view rolls every post up over a period you choose, from the last 7 days to the last 365. These numbers exist on LinkedIn's pages and nowhere else: LinkedIn's public API does not return a member's impressions, which is why every tool that shows them either asks for your login or reads them in your own browser.

There is no good number, and anyone who gives you one without asking how many followers you have is guessing. Impressions scale with the audience LinkedIn can show a post to, and the size of that audience is mostly your follower count plus your network's. So the honest measure is a ratio.

Impressions against followers

Divide a post's impressions by your follower count. A post that reaches a fraction of your followers reached the people who already knew you. A post that reaches more people than follow you was shown beyond your network, and the out-of-network share on the analytics page will say so. Track the ratio rather than the count and a good week looks the same whether you have 800 followers or 80,000.

Impressions against your own history

The only benchmark you can actually see is your own. Take your last ten or twenty posts, work out the mean, and judge each new post against it. Two times your own mean is a post that did something. Half your mean is a post that did not, and if the last five are all under the mean the average has moved and the material has changed. Nobody else's numbers are on the page, and LinkedIn keeps everyone's analytics private, so a benchmark built from strangers' screenshots is built from a selection of their best days.

The two-day rule

Do not read a post's impressions on the day it went out. The count keeps climbing for days, sometimes longer for a post that gets picked up late, and a number read at hour six is a number read at a random point on a curve. Wait two days for the per-post read and a month for the account read.

Nobody outside LinkedIn knows the ranking in detail, and anyone claiming a formula is selling one. What can be said honestly, from the definitions above and from watching a founder's own numbers over months:

• The first line and the fold. A post that earns the click on see more keeps people on it, and time on a post is the plainest signal a feed has that the post is worth showing to more people. The hook is covered in How to write LinkedIn posts.
• Early comments from people who are not you. Members reached climbs with reposts and comments, because each one puts the post in front of a new network. Replies you write to those comments keep the thread alive.
• Consistency. An account that posts every week has an audience that expects it; an account that posts in bursts has to rebuild that expectation each time. Turning up does more for the monthly number than any single post.
• The hour, a little. Posting when your own readers are awake helps at the margin. It is a second-order variable, and the guide to it is honest about the size of the effect: The best time to post on LinkedIn.

Because LinkedIn's API does not return a member's impressions, Sonar reads them where they are: on your own analytics pages, in your own browser, while you are signed in. That is the whole job of the Sonar for LinkedIn Chrome extension. It runs only on your dashboard and your content and audience analytics pages, and it reports a number only when LinkedIn itself labelled it, the label names the metric exactly, and the period the number covers is known. It reads the value that sits beside the caption "Post impressions" or "Impressions", the one beside "Members reached", and the same way for social engagements, profile viewers, search appearances and total followers.

A reading with an unknown period is dropped rather than guessed, and every row it sends carries the literal LinkedIn caption the number was read under, so you can check the work. The numbers then appear in two places: the Reach tile on Home and the Impressions row in your Insights ledger, the row captioned with its window ("last 7 days") and its provenance ("your LinkedIn analytics, read by the Sonar extension"). Beside them, Sonar measures each of your posts' public engagement against your own average, which is the ratio this page keeps recommending. A switch in Settings pauses the extension; uninstalling it ends it. What it can and cannot do, in full, is on the LinkedIn page.