If your LinkedIn reach has dropped while your post quality has gone up, you are not imagining it. The LinkedIn algorithm in 2026 has shifted its weighting from reactions to attention, which means a post with 20 likes and a 45-second average read time now outperforms one with 200 likes and a two-second scroll-past.
That single change rewrites the playbook for B2B founders, salespeople and brand pages. Shorter is no longer automatically better, hooks matter more than headlines, and the "save" button has quietly become the most valuable action on the platform.
Key takeaways
- The LinkedIn algorithm in 2026 weighs dwell time, saves and shares far above likes and basic reactions.
- Posts that hold attention for 30 seconds or more get pushed to second and third-degree networks much faster than short, reaction-bait posts.
- Carousels (PDF documents), long-form text posts with structured reveals, and native video under two minutes are the three formats winning distribution right now.
- Comments still matter, but only when they are substantive. One-word replies are increasingly discounted.
- Profile and company page dwell time also feed the ranking signal, so bios and banners now do real distribution work.
What the LinkedIn algorithm in 2026 actually measures
LinkedIn has been moving in this direction for years, but 2026 is the point where the signals are stacked clearly. The feed ranker now leans on a small number of behaviours that prove a post was genuinely useful.
The main ones are:
- Dwell time - how long a viewer's screen rests on your post before scrolling.
- Saves - a private, high-intent signal that the content has future value.
- Shares to messages and feed - amplification with a personal endorsement.
- Substantive comments - replies long enough to suggest real engagement, not pod activity.
- Repeat profile visits - people coming back to see what else you post.
Reactions still count, but they are now closer to a tiebreaker than a core driver. A "like" takes half a second and tells the algorithm very little about whether the content was worth the space it took in the feed.
The practical shift: stop writing for the thumb-tap, start writing for the thirty-second pause.
Why dwell time became the dominant signal
Dwell time is hard to fake. You cannot pod your way to it, you cannot buy it from a bot farm, and it is strongly correlated with actual value delivered. For a platform that has spent years fighting engagement-bait and AI slop, it is the cleanest proxy available.
It also matches LinkedIn's business model. Advertisers want attention, not vanity metrics, so the organic feed is now tuned to reward the same thing paid campaigns are measured on.
Why saves punch above their weight
A save is the strongest implicit endorsement a professional network user can give. It says: I want this later. The algorithm reads that as evergreen value and often re-surfaces the post to similar profiles for days after publication.
If you can build a habit of writing "saveable" posts - checklists, frameworks, templates, breakdowns - you compound reach in a way reaction-driven content never does.
Format implications: what to post more of
The format of your post now matters more than the topic. Two posts with identical insights will perform very differently depending on how they are packaged, because packaging determines dwell time.
Here is how the main LinkedIn formats stack up in 2026.
Carousels (PDF documents)
Carousels remain the single strongest format for dwell time. A reader who swipes through ten slides has given you ten to forty seconds of attention, and the algorithm sees every swipe as an engagement signal.
What works now:
- 8 to 12 slides, not 20.
- A hook slide that promises a clear payoff ("5 cold DMs that actually get replies").
- One idea per slide, large readable type, minimal decoration.
- A final slide with a soft call to action: save, comment, or follow for more.
Avoid turning carousels into brochures. The moment a slide looks like an ad, dwell time collapses.
Long-form text posts with structured reveals
Text posts are not dead - they are harder. The winning structure in 2026 is a strong first line, a "see more" break at the right moment, and a body that uses white space, short paragraphs and mini-sections to keep the eye moving down.
A useful template:
- One-line hook that creates an open loop.
- Context in two lines: who you are and why this matters.
- The core idea, broken into 3 to 5 numbered or bulleted beats.
- A specific example or mini-story.
- A closing line that invites a reply, not just a reaction.
If a reader has to scroll through your post, you are winning dwell time. If they can see the whole thing without the "see more" click, you are probably losing it.
Native video under two minutes
Short native video is finally working on LinkedIn the way it does on other platforms, but the bar is higher. Talking-head clips with captions, a clear hook in the first three seconds, and a single takeaway outperform polished brand videos almost every time.
Upload natively, add captions, and keep the aspect ratio vertical or square. Linked-out YouTube videos still get throttled.
Single images and link posts
Both formats have slipped. A single image rarely holds attention long enough to trigger the dwell time signal, and link posts pull users off platform, which LinkedIn penalises. If you must share a link, drop it in the first comment and explain the value in the post itself.
How to engineer dwell time without clickbait
The temptation with a dwell-time algorithm is to lean into cheap tricks: fake controversy, bait-and-switch hooks, artificial suspense. These work for a week and then get flagged, either by users or by LinkedIn's own quality filters.
Honest ways to earn attention:
- Promise something specific in the first line. "Here is the exact email sequence that booked us 14 meetings last month" beats "Sales is changing."
- Front-load the useful part. Give the insight early, then expand. Readers who get value fast stay longer, not shorter.
- Use formatting as pacing. A line break is a breath. A bullet list is a slowdown. A short sentence is a punch.
- Write in scenes, not summaries. "On Tuesday, a prospect ghosted us after three good calls" holds attention better than "We lost a deal recently."
- End with a question people actually want to answer. Not "thoughts?" but something tied to their own experience.
A simple dwell-time audit you can run this week
Open your last ten posts in LinkedIn analytics. For each one, look at impressions, reactions and the newer "average time viewed" metric where available.
Sort by time viewed, not by likes. The top three are your real template. Copy their structure - hook length, paragraph rhythm, format - for your next few posts and compare.
Most people discover that their highest-reaction posts and their highest-dwell posts are not the same. The dwell winners are usually the ones that quietly keep bringing in profile visits and inbound messages days later.
What this means for comments and engagement pods
Comments still matter, but the algorithm has gotten better at reading them. A reply of "Great post!" now contributes very little. A reply that adds a specific example, a counterpoint or a question contributes a lot.
This has two consequences:
- Engagement pods are losing power fast. Pod comments tend to be short, generic and clustered in the first few minutes. That pattern is now a negative signal, not a positive one.
- Replying to your own commenters is more valuable than ever. A real back-and-forth doubles the dwell time of anyone reading the comment thread, and the algorithm sees that.
Block out 20 minutes after each post to reply thoughtfully. Treat the comment section as the second half of the post, not an afterthought.
Profile and company page: the overlooked dwell signal
When someone clicks your name and lands on your profile, the clock starts again. If they bounce in two seconds, that is a weak signal. If they scroll your About section, open your Featured posts and visit your company page, that is a very strong one.
For personal profiles:
- Rewrite your headline as a value statement, not a job title. "Helping B2B SaaS founders book 10+ sales calls a month" holds attention better than "CEO at Company."
- Use the About section as a short story, not a resume. Three or four short paragraphs with concrete outcomes.
- Pin three Featured posts that represent your best work. These are often what a new visitor reads before deciding to follow.
For company pages:
- Treat the tagline and banner as a landing page. A clear promise and a next step.
- Post from the page at least twice a week so visitors see activity, not a ghost town.
- Use employee advocacy: a post reshared by five team members with genuine commentary outperforms a single page post almost every time.
A quick weekly rhythm that works
- Monday: carousel with a framework or checklist your audience can save.
- Wednesday: long-form text post telling a specific story with a lesson.
- Friday: short native video or a lighter opinion post to keep the profile active.
- Daily: 15 minutes of substantive comments on other people's posts in your niche.
This rhythm is small enough to sustain and broad enough to hit different dwell-time patterns across the week.
Measuring what matters in 2026
If you are still reporting on likes and follower count, your dashboard is out of date. The metrics worth tracking now are:
- Average time viewed per post (where LinkedIn exposes it).
- Saves per post - a proxy for evergreen value.
- Profile views from post - shows whether content is building authority, not just noise.
- Qualified inbound messages per week - the real business outcome.
- Share of comments that are longer than ten words - a quality indicator.
Set a simple weekly review. If dwell time and saves are climbing while likes stay flat, you are winning the 2026 algorithm. If likes are up but dwell is flat, you are probably writing bait and will plateau soon.
Reach without attention is a vanity metric. Attention without reach is a format problem. Track both, fix whichever is weaker first.
Getting help if you want to move faster
Rewiring a LinkedIn strategy around dwell time takes a few months of consistent testing. If you would rather have specialists run the posting rhythm, carousel design, video editing and real engagement for you, the Henify LinkedIn growth plan covers profile and company page work with real people, not bots. It is a practical option for founders and B2B teams who want the output without staffing it in-house.
Either way, the direction is clear: build for attention, measure saves, and let the LinkedIn algorithm in 2026 do the distribution work for you.
FAQ
Does the LinkedIn algorithm in 2026 still show posts to my followers first?
Yes, your first-degree network still sees posts earliest, but whether the post breaks out to second and third-degree connections now depends almost entirely on dwell time, saves and substantive comments in the first 60 to 90 minutes. Followers are the launch pad; attention is the fuel.
How long should a LinkedIn text post be in 2026?
Long enough to earn a "see more" click and a scroll, usually 150 to 400 words. Very short posts rarely generate enough dwell time to trigger wider distribution, and very long posts lose readers unless the structure keeps them moving with white space, bullets and short paragraphs.
Are hashtags still useful on LinkedIn?
They have minor value for topical classification but almost none for reach. Two or three relevant hashtags are fine; stuffing ten at the bottom of a post looks dated and does not help the algorithm.
Do engagement pods still work?
Less and less. The clustered, short, generic comments pods produce are exactly the pattern LinkedIn's quality filters now discount. A handful of long, specific comments from real readers in your niche is worth more than fifty pod likes.
What is the single fastest way to improve my LinkedIn reach this month?
Switch one of your weekly posts to a carousel built around a framework your audience will want to save, and spend 15 minutes replying thoughtfully to every comment it gets. That combination tends to move dwell time, saves and profile visits at the same time, which is exactly what the 2026 algorithm rewards.