Why is my LinkedIn post not getting impressions: rejected, or ignored?

A post with no impressions has failed in one of two places

Two completely different failures produce the same disappointing total, and they need opposite repairs. Either the post was shown to a limited set of people and did not earn a wider round, or it earned the wider round normally and then stopped expanding a few hours later. The first is a problem with the opening, the format, or the account. The second is a problem with everything after the opening, which means the hook you are about to rewrite is the one part that already worked.

The standard article on this question opens with five causes: your hook was weak, you posted at the wrong time, you used a link, you posted too rarely, your content was too promotional. All five happen. None of them is attached to a number you can look at, so the reader picks whichever one sounds most like them and rewrites in the dark. The table below is the same information organised so that your own analytics choose for you.

Read the row that matches what you observed, and do only what that row says.

Notice that two of the four rows tell you to stop. That is the point of separating the failures. A rewriting habit that treats every weak post as a hook problem will eventually break the posts that were working, because you keep repairing the part that was never broken.

LinkedIn does not store the curve, so you have to record it

The data that separates the two failures does not exist anywhere you can reach, so it has to be captured while the post is live. LinkedIn reports a running total of impressions and unique members reached. It does not report impressions over time, so a post that gained everything in the first hour and a post that gained the same amount slowly across two days look identical the following week.

For a post that has already died, the honest answer is that the first hour figure is gone and cannot be reconstructed. What you can still read is the total against your median and whether any engagement arrived at all. A post with a total far below your median and no comments in the thread is more likely to be the first failure. A post with a total below your median but a normal looking early comment thread is more likely to be the second. That is an inference rather than a reading, which is exactly why the recording habit is worth starting today.

Your threshold is your own median, not a number from a blog

No universal first hour threshold can be published, because the figure scales with how many followers you have, how many of them are active at the hour you posted, which timezone they sit in, and how much of your audience is made up of people who follow you rather than people who once engaged with you. Any article that gives you a specific number of impressions to expect in hour one has invented it, because the author cannot see your account.

Here is the arithmetic, with every input labelled as an assumption you should replace with your own recordings. Assume ten posts recorded at the marks above, a median first hour of 180 impressions and a median twenty four hour total of 1,150. Divide the second by the first and your normal expansion multiple is roughly 6.4, meaning a typical post ends the day about six times bigger than it was after an hour.

Now run two posts through it. Post A recorded 45 impressions in the first hour against your median of 180, which is a quarter of your normal, and it finished the day at 210. Post B recorded 170 in the first hour, which is normal for you, and finished at 340, giving an expansion multiple of 2.0 against your normal of 6.4. Post A was refused at the first test. Post B passed the test and then stopped. Same disappointing total, two different diagnoses, two different weeks of work.

Two working thresholds, both derived from the spread rather than from any published study. Treat a first hour below half your median as a refusal. Treat an expansion multiple below half your median multiple as a stall. Half is not a magic number, it is simply wide enough that ordinary variation between posts will not trip it, and you should tighten or loosen it once you can see how much your own posts naturally vary.

The Ra-Aha Two Clock Test

Two clocks, four outcomes, and each outcome sends you somewhere different. Clock one is the first hour, which tells you whether the post was accepted for distribution. Clock two is the expansion multiple at twenty four hours, which tells you whether it was carried once it had been. Read both against your own median before you read a single piece of advice about hooks.

The fourth outcome is the one people resist, because it reframes the complaint. Somebody arriving at this question has already decided the problem is reach, and roughly a quarter of the time the numbers say the reach was ordinary and the post simply did not do anything to the people who read it.

Fixes for a post that was refused at the first test

A refusal is decided on the part of the post visible in the feed, so every repair belongs above the fold. The preview is short, shorter on a phone than on a desktop, and a blank line ends it early regardless of how many characters you have used. That last behaviour catches more people than any character limit, because the writer thinks they have three lines of preview and a stylistic line break has quietly reduced it to one.

One check sits outside the post itself. If every recent post is being refused rather than one, the problem is at the account layer and no amount of rewriting will move it. Compare your profile views over the last 90 days against your post impressions over the same period, because profile views holding steady while post reach collapses points at the posts, and both falling together points at the account. That split is worked through properly in matching a reach drop to its signature.

Fixes for a post that passed the test and then stalled

A stall means the people who read the post gave the feed nothing to act on, so the repairs live in the middle and the ending. The first round of readers arrived, opened it, and left without saving it, resharing it with a thought of their own, or writing a reply that another person would want to answer. Each of those is a different signal and each one is produced by a different part of the writing.

• The payoff arrives too late, so readers who opened the post left before the part that would have been worth keeping
• The middle is a run of general observations, and general observations cannot be argued with, which means no comment thread forms
• There is nothing worth saving, because the post describes a situation rather than giving a method, a threshold, or a number
• The ending asks a question the author does not care about, and readers can tell, so the replies are polite and one line long
• The post agrees with everything the reader already believes, which produces approval rather than discussion

How you answer comments belongs in this section too. A thread of one word thanks and a thread of substantive exchanges are different signals, and the second is largely produced by the author. Replying with a sentence that adds a detail invites another reply. Replying with appreciation ends the thread politely, which is a courteous way to close down the exact behaviour that would have carried the post further.

When impressions are the wrong thing to be looking at

If a post reached a reasonable number of people and produced no replies and no profile views, raising the impressions will not help, because you are multiplying a rate of zero. That is the wrong question, and it is worth saying plainly because the whole category of advice around this query assumes reach is the constraint.

Here is the arithmetic on assumed figures you should replace with your own. Assume a post recorded 1,200 impressions, 800 unique members reached, six profile views and no replies. The rate from members reached to profile view is 0.75%. Double the reach and you get twelve profile views instead of six, which is six extra strangers glancing at a page. Raise the rate to 3% at the original reach and you get twenty four, without a single additional impression.

The second route is cheaper and it is almost always available, because the rate is governed by things you control completely. Whether the post named who it was for. Whether it gave a reason to look you up rather than simply nodding. Whether your profile answers the question the post raised, which is the connection worked through in what a profile has to do after the read.

There is a version of this that gets misread as success. A post with high impressions, high members reached and nothing happening afterwards is usually a post that pleased a wide audience of people who will never buy anything from you. The impression count is the most satisfying number on the platform and the least connected to anything you care about.

The measurement traps that make you fix the wrong post

Four traps account for most wrong conclusions in this exercise, and three of them are about sample size. One weak post is noise and always will be. Posts vary by a large multiple for reasons that have nothing to do with quality, including who happened to comment in the first ten minutes and whether that person has a wide network. Judge a change across six posts, never across one.

The second trap is the moving baseline. Your median is only a fair comparison if the surrounding conditions held, and the composition of the feed does shift over time. LinkedIn has reported video watch time rising sharply year on year, which means the mix of what your audience is being served has changed underneath your text posts. A median built from posts eighteen months ago is a historical document rather than a baseline.

The third trap is changing two things at once. If you rewrite your openings and switch to documents in the same fortnight, the numbers will move and you will not know which change did it. The fourth is recording at inconsistent times, which quietly corrupts the median: an hour one reading taken ninety minutes after posting is not comparable to one taken at sixty, and the difference is large precisely because early distribution moves fast.