I posted something last week and the impression count came back lower than I expected. Same topic I always write about, same effort, worse number on the screen. My first thought was that something broke. It hadn't. The numbers themselves just don't mean what they used to.
Here's the actual shift. Impressions across LinkedIn dropped somewhere around 63 to 66 percent since 2023. Meanwhile engagement per post went up, somewhere between 12 and 39 percent depending who's measuring it. Put those two together and it's not really a decline story. It's a trade. Fewer people are seeing your stuff, but the ones who do are actually reading it.
Two numbers you can basically stop worrying about
Impressions and follower count used to be the whole scoreboard. Check the app, watch the number, feel good or bad about your day. That's not really how it works anymore.
A guy I know has maybe 4,000 followers and outperforms accounts with 40,000. Not because he's tricking the algorithm, just because his 4,000 actually read what he writes. Follower count climbing feels like winning. On its own, it barely tells you anything.
The number that's quietly running everything
Dwell time. How long someone's eyes stay on your post before they scroll past. It starts the second your post lands on their screen, and it's become the biggest single factor in how far anything travels now.
The gap here is bigger than you'd guess. Posts held for 61 seconds or more land around 15.6 percent engagement. Posts glanced at for under 3 seconds sit around 1.2 percent. Do the math, that's roughly 13 times the difference, between two posts that might have gotten the exact same number of likes.
This is also, honestly, why carousels do so well right now. Nothing magic about them. They just physically take longer to swipe through, so dwell time goes up, and the algorithm reads that as "people actually cared."
About that comments-are-15x-a-like thing
You've seen this stat. Everyone has. Comment equals 15 likes, so go get comments. Mostly true, but the exact number is shakier than people treat it. One analysis that went through several million posts puts the real multiplier closer to 2 or 3 times, once you strip out the "Great post!" comments that add nothing.
Doesn't really change what you should do though. Comments still beat likes by a lot. Ten people leaving a real thought will move a post further than two hundred quick taps ever will. Whether the true number is 2x or 15x, chasing likes is still chasing the weakest signal on the whole platform.
Saves. Nobody checks this and they should
When someone bookmarks your post instead of just scrolling past it, that's about as strong a signal as it gets. And I'd guess most people never even look at this number.
Stuff that gets saved tends to be useful in a specific way. A checklist. A framework. Something you'd actually want to find again in three weeks. A funny story gets a like. A genuinely useful post gets saved, and saved is worth more.
Where to actually find these numbers
Here's the catch nobody mentions. LinkedIn doesn't hand you a "dwell time" number anywhere in the native app. It's a backend signal the algorithm uses, not something on your dashboard. So you're reading it indirectly, through the metrics that do show up.
The native analytics on a post give you impressions, reactions, comments, reposts, and for video, average watch time. Watch time is your closest direct stand-in for dwell time, use it if you post video. For everything else, treat comment depth and repost rate as your dwell-time proxy, since people don't leave a real comment on something they scrolled past in two seconds. Saves show up if you tap into a post's individual stats, they're easy to miss because they're not on the main summary view.
The honest workaround: stop trying to find one perfect number and start reading the combination. A post with modest impressions but a string of real comments and a decent repost count is quietly doing better than a post with triple the impressions and nothing but likes, even though the dashboard makes the second one look more impressive at a glance.
What to actually check once a month
Impressions, raw exposure, fine to glance at, not worth chasing.
Dwell time, the real signal, higher is basically always better.
Comments, weigh the real ones, ignore the one-word ones.
Saves, probably your biggest blind spot if you've never looked.
Follower count, nice number, tells you almost nothing on its own.
Once a month, go find your five best posts and look for what they have in common, format, topic, whatever. Usually two or three things jump out. Do more of those. Stop trying to figure out why the flop posts flopped, sometimes there's no lesson there.
Where ScoutHook comes in
ScoutHook's analytics were built around this exact list, not the old scoreboard. Instead of leading with impressions and follower growth, it shows dwell time, comment quality, and saves per post. So what you're looking at actually matches what the platform cares about right now, not what it cared about two years ago.
Track what actually moves the needle. ScoutHook surfaces dwell time, comment quality, and saves, so you're not optimizing for numbers that stopped mattering. Try ScoutHook
Frequently asked questions
What analytics actually matter in 2026? Dwell time, comments, saves. Impressions and follower count have drifted pretty far from what they used to signal.
How much more is a comment worth than a like? People say 15 times. A closer analysis puts it nearer 2 to 3 times once you account for comment quality. Either way, comments still win by a wide margin.
Why did my impressions drop with the same content? Platform-wide, impressions fell hard since 2023 while engagement per post climbed. Looks more like tighter targeting than a problem on your end.
Is follower count still worth tracking? Worth a glance, not worth chasing. A smaller, engaged crowd beats a big, quiet one almost every time.