A post with 12,000 impressions and two comments did worse than a post with 900 impressions and fifteen comments. It feels wrong, and yet it is the first reflex to acquire if you want to read your LinkedIn stats properly.
The problem is that LinkedIn puts the most flattering and least actionable number front and centre.
Why impressions mislead
An impression means your post passed through a feed. Not that it was read, not even that it was looked at for half a second.
That number depends mostly on things you do not control: how many people were online that morning, one commenter with a large following, a topic in the news lifting the whole sector. Two posts of identical quality can differ tenfold.
Impressions are not useless — they tell you whether your distribution is growing over months. But post by post, they teach you almost nothing.
Engagement rate: the only honest comparison
Engagement rate relates interactions to the number of people reached:
(reactions + comments + shares) ÷ impressions × 100
It is the metric that lets you compare a post seen by 500 people with one seen by 20,000. On LinkedIn, 2 to 5% is solid; above 5%, your post genuinely landed.
A rate collapsing while impressions climb is the classic signal of a post distributed to the wrong audience.
The hierarchy of signals
Not all interactions are worth the same, either to the algorithm or to you.
Reactions cost one click. They signal approval, rarely deep interest.
Comments cost time and a little exposure — commenting means showing yourself publicly. A comment longer than a few words is worth far more than a like, and it is the signal the algorithm rewards most.
Shares and reposts put the sharer's own reputation on the line. That is the strongest signal, and the rarest.
Saves are visible only on your own posts, and are the best indicator of real usefulness: someone judged your content valuable enough to come back to.
Read those four in that order. A post with many likes and no comments was liked without being interesting.
Profile views: the metric everyone forgets
The point of a post is almost never the post itself: it is that someone goes and looks at who you are.
Watch your profile views in the 48 hours after publishing. A post with few reactions but many profile visits worked perfectly — it did the qualifying job you wanted.
This is especially true of expertise posts, which draw few comments but a lot of curiosity.
What you cannot measure, and why that is fine
A large share of your impact is invisible. Silent readers, the person who mentions you in a meeting, the one who reaches out six months later saying "I've been following you for a while": none of that shows up in a dashboard.
All the more reason not to steer day by day. Look at your numbers in batches of ten posts, and look for trends, not accidents.
Reading your numbers without losing your Sundays
A monthly routine is enough:
- Rank the month's posts by engagement rate, not impressions.
- Take the top three and look for what they share — format, topic, angle, length.
- Take the bottom three and ask the same question.
- Write down a single conclusion, and apply it next month.
One conclusion a month, seriously applied, beats ten observations noted and forgotten.
In Ego, published posts carry their counters — impressions, reactions, comments, reposts — directly under each publication, which removes the constant back and forth between your writing tool and LinkedIn.
The takeaway
Impressions measure luck, engagement measures quality. Compare rates rather than volumes, give comments and shares the weight they deserve, watch your profile views, and never draw a conclusion from a single post.
Generate your LinkedIn posts with Ego
Try for free
