The LinkedIn Metrics That Actually Matter in 2026 (And the Ones I've Stopped Tracking)

Author's Note:

​
This week's piece came out of a client conversation I have not been able to stop thinking about — one that gets at the question I'm hearing more than any other right now: why am I getting wild swings in impressions that have nothing to do with the engagement on my posts? It's also a follow-up to last week's deep dive on the 2026 algorithm. Where that one was about what to post, this one is about what to actually measure.

In the spirit of compounding work, I walked through all of this LIVE on May 19th, 2026 in a workshop called "The Winning Content Mix, Live: Build Your 2026 LinkedIn System in 60 Minutes."​

During the workshop, I showed exactly how I plug Stanley into my content operating system — including how I track these very metrics — so viewers could build alongside me, in real time. You can grab a free Stanley trial here.

---

A few weeks ago, a client of mine — who is, by every objective measure, blowing up on LinkedIn right now — sent me a message that made me laugh out loud, mostly because I have been thinking the exact same thing:

You can see my response. Pretty much an emphatic ¯\(ツ)/¯

Again, I reemphasize…

One of my recent posts: 159 reactions. 16,000 impressions.

My birthday post a few weeks earlier: Now 243 reactions. ~4,775 impressions.

On the surface, those numbers make no sense. More reactions should mean more reach. That's been the unspoken physics of LinkedIn for years.

Except that physics is pretty much gone. Darn! Or yay? Let’s find out…

Last week, I wrote about Richard van der Blom's 2026 Algorithm Insights Report and the seismic shift from a social graph to an interest graph on LinkedIn. (If you missed it, check it out here— and I'd start there before this one if you want the full context.)

But here's the specific piece of his data that explains the wild swings my client was seeing:

Median follower count is up roughly 31% year over year. Median impressions are down.

Those two numbers used to move together. They don't anymore. They are structurally decoupled — and the reason why is the algorithm itself.

According to the same report, in 2026:

• Only ~33% of your feed is comprised of people you follow.
• ~31% is first-degree connections (used to be the dominant share — not anymore).
• ~25% is second- and third-degree connections.
• The rest is suggested posts (which are a massive tell about the interest graph) and company pages (a sliver, because LinkedIn wants companies paying for that reach).

So when you follow someone, it is no longer a guarantee — or even a strong likelihood — that you will see their content. The algorithm is matching content to interest patterns, not relationships. So if the topic someone posts about are not in the cluster of topics you've been engaging with recently, you will not see them, even if you hit the follow button two years ago.

Which is exactly why one of my posts can get 159 reactions and reach 16,000 people, and another can get 238 reactions and reach 4,000. The first one hit an interest cluster the algorithm was actively serving. The second one was a birthday post — emotionally resonant for the people who did see it, but not a content match for anyone outside my immediate network.

Same person posting. Same account. Wildly different reach. Same algorithm doing exactly what it was designed to do!

So if reactions and impressions are no longer reliably coupled, what should we actually be looking at?

Here's what I'm tracking in 2026.

1) Saves 💾

This is, hands down, the metric I've reorganized my content strategy around.

A save tells you that someone found a piece of content valuable enough to bookmark and return to. In an interest-graph world, that is one of the strongest signals you can possibly get. The 2026 LinkedIn Algorithm Report is also clear on this: saves now carry significantly more algorithmic weight than likes, and meaningfully more than a comment. Industry analyses peg saves at roughly 5x the weight of a like and 2x the weight of a comment.

But the algorithmic boost is honestly the second reason I care about saves. The bigger reason is this:

A save tells me I'm reaching someone further down the funnel.

A like is a polite passive action. A save is "I want to come back to this." That is a different animal entirely. You can't see who saved your post — LinkedIn doesn't surface the names — but you don't need to. You just need to know that anytime a piece of content hits a save count higher than its average, you've reached people who are doing more than scrolling. They're studying.

I now design content with the question: Would someone bookmark this? If the answer is no, I rewrite it.

2) Premium button link clicks 🖱️

This one is small, and most people overlook it. But it's one of the most useful business-outcome indicators LinkedIn gives you natively — if you have a Premium profile.

When you're a Premium user, that blue-highlighted link that follows your name around the platform becomes a CTA button that LinkedIn tracks per-post. You can see, post by post, how many people read your content and then click the link.

This is the closest LinkedIn comes to giving you direct attribution: this specific post moved someone to take this specific action. I want to see that not only is someone interested enough to read what I'm saying, but they're also interested enough to click the link I'm pointing them to.

Even more importantly, LinkedIn just announced they're going to allow people to book and pay you directly from this link; meaning you can actually see, in real time, the dollars that posts generate. Check out this post from Jason Feifer breaking it down:

If you're on Premium and you're not looking at this number, you're leaving signal on the table.

3. Sends ✉️

A send is a private share — someone took your post and DM'd it to another person.

I treat this as the highest-trust tell in my entire LinkedIn dashboard, and it gets stronger every quarter.

Here's why: a send means someone found your content interesting enough that they wanted to privately have a conversation about it with someone else in their life. That is miles further down the funnel than a like or a comment. A send is almost always either (a) "this person reminded me of you" or (b) "I want to talk about this." Both of those are warm, qualified, conversation-starting signals — the kind that turn into discovery calls and clients if you're paying attention.

4. Reposts 🔄

Reposts are the public-facing version of a send, and they hold real algorithmic weight in 2026. When someone reposts your content without commentary, they are essentially co-signing it to their entire network. That is a public endorsement — and the algorithm reads it that way. Richard’s report says this:

I track instant reposts especially closely on text posts. They are a leading indicator of whether a piece of content has the kind of velocity that will compound over the next 48 hours.

5. Article view rate 📰

I've been saying this for years now, and the 2026 data has finally caught up to me: LinkedIn newsletters and articles are the single most underused asset on the platform.

As you can see:

• Article reach is up 13.1% year over year.
• Engagement multiplier on articles is up 51.7% — the single biggest positive shift of the year, by format.

But there's a second reason I'm tracking article view rate that wasn't on my radar even a year ago: articles get heavily indexed by AI answer platforms and search. LinkedIn itself has now published guidance on how to write articles that surface in AI chats and searches. When someone asks ChatGPT or Perplexity a question in your space, your LinkedIn article can become the source.

When I look at my own data, the correlation is uncomfortably clean: the months where my article views are highest are the months where my sales are highest. That is not a coincidence. Article views mean someone read 1,200+ words of my thinking. That is a person who is doing real research, and real research is what precedes real buying decisions.

​
6. Comments (but not the way we used to count them) 🗣️

Comments used to be the engagement currency of LinkedIn. Remember the whole "warm up the feed by commenting in the first hour" playbook? That advice is kinda dead. Comments are still important, but for an entirely different reason.

In a 2026 LinkedIn where DMs are absolutely bombed with cold spam and pitch decks, the comment section has become the most important place to actually build relationships on the platform.

A well-crafted comment — one with a story, a contrarian take, or a specific insight — does two things at once:

1. It builds rapport with the original poster and their audience. Better than a DM. By a mile.
2. It can out-perform your own posts in terms of impressions.

I have a student in my Brand Inner Circle with around 1,000 followers who is a serial commenter. One of her recent comments — not a post, a comment — pulled 21,000 impressions. With a thousand followers.

This is the most efficient way to get the right eyeballs on you that exists on LinkedIn right now, and almost nobody is doing it strategically.

Want to laugh? Here’s my highest impression comment of the last few weeks about a copperhead snake biting my client’s dog (the dog lived, just in case you were wondering).

1. Likes/Reactions 👍

Look — engagement is engagement, and we all judge ourselves on it. I won't pretend I don't still glance at the number.

But there was a time when I would check likes in the first hour to predict whether a post was going to take off. That correlation is mostly broken now. I have posts that get almost nothing in the first three hours and then catch fire two days later as the algorithm finds the right interest cluster. I have posts that get strong early likes and then plateau. The first-hour like count no longer tells me what it used to.

Likes are still nice. They are no longer exactly diagnostic.

​
2. Reach and impressions 👀

I know — this one is going to sound strange coming from someone whose entire client conversation that opened this article was about impressions. Let me explain.

It's always nice to see a post go viral. I'm not going to pretend otherwise. But in an interest-graph feed, the right 5,000 impressions are worth more than the wrong 50,000.

A 50K-impression post that reaches a bunch of people outside my target audience produces almost nothing for the business. A 5K-impression post that reaches founders, speakers, authors, and decision-makers who are actively trying to figure out their LinkedIn strategy produces calls, applications, and sales.

I would rather be precisely useful to the right people than mildly interesting to a lot of the wrong ones. And the interest graph is finally making that math possible.

3. Follower count 👥

This one is the hardest to let go of, because for over a decade, follower count was the social media currency. It is congruent with importance, with influence, with cache — at least, that was the story.

Three reasons I've stopped optimizing for it:

One: As I covered above, following someone doesn't mean you'll see their content anymore. The follow button used to be a guarantee of distribution. It isn't.

Two: In an interest-based graph, a huge share of the people who see my content when it performs well are not my followers in the first place. When a post hits an interest cluster, it reaches strangers who never hit the follow button — and a meaningful chunk of those strangers turn into clients.

Three — and this one might be the most counterintuitive:

Follower count can actively work against you when it comes to securing sponsorships.

I was in an RFP process for a sponsorship a few weeks ago, and the brand made it very clear: once you go over 50,000 followers, they were no longer interested.

That is not a one-off. That is a big direction the influencer market is moving in 2026. The data backs it up:

• 73% of brands now prefer micro and mid-tier creators over celebrities and macro-influencers. (Source)
• 40% of all influencer marketing budgets are now allocated specifically to the micro-tier (creators with 10K–100K followers). (Source)
• 76% of C-suite leaders say their influencer budgets are going up in 2026. (Source)
• Micro-influencer campaigns commonly generate 5x–8x ROI, versus 3x–5x for macro campaigns. (Source)
• And engagement-wise: micro-influencers average 3–6% engagement versus 1–3% for macro creators.

The reason is simple, and it's the same reason the interest graph exists: as follower count goes up, intimacy and trust go down. Marketing teams have figured this out, and they are rightly redirecting their budgets toward creators with smaller, more engaged, more trusting audiences.

​
If I had to summarize what's changed in 2026, it would be this:

We are moving from broadcast metrics to relationship metrics. (Hallelujah)

Likes, follower count, raw impressions — those are broadcast metrics. They tell you how widely something traveled.

Saves, sends, comments, article views, and premium link clicks — those are relationship metrics. They tell you who is leaning in. And in an interest-graph world, who is leaning in is the only number that turns into a measurable business outcome.

The good news is that the same content engine that drives the relationship metrics drives the broadcast ones, too. You don't have to choose. But you do have to know which numbers you're actually being graded on — and design your content for the right ones.

If this resonated, check out the May 19th workshop replay where I walked through exactly how I track these metrics inside my system and how I use Stanley to do it at scale. We went through the dashboard, the weekly review cadence, and the prompts I use to score my own content against this list.

✅ Grab this workbook to use during the workshop.

​
​