How the LinkedIn algorithm
LinkedIn doesn't publish its ranking code, but a decade of engineering blog posts, patent filings, and consistent creator experiments have pinned down how it actually works. Here's the practical, no-fluff version.
TL;DR
LinkedIn scores every post in four phases: quality check, test distribution to a small seed audience, engagement evaluation, then viral distribution if the signals are strong. The signals that matter most — in order — are meaningful comments, dwell time, content relevance, creator authority, and connection strength. Everything else is noise.
The 4 distribution phases
Your post moves through four automated checkpoints before it reaches a large audience. Understanding each one tells you where you're actually losing reach.
Quality check (spam filter)
Within seconds of publishing, LinkedIn runs the post through classifiers that flag spam, low-quality, or policy-violating content. Too many hashtags, external links without context, generic engagement bait, or AI-obvious patterns get downranked here — often silently.
Outcome: Low-quality posts are capped before distribution even starts.
Seed audience test
The post is shown to a small slice of your network (typically 50 to a few hundred people) — picked by connection strength, past engagement with you, and topical relevance. The first 60 minutes of this test window are critical.
Outcome: If engagement rate stays above a threshold, the post advances.
Engagement evaluation
LinkedIn weighs what happened during the seed test: comment-to-view ratio, dwell time (how long people stop scrolling), share volume, and the profile authority of everyone who engaged. A meaningful comment from an active creator is worth exponentially more than a passive like from a dormant account.
Outcome: Strong signals trigger expansion to the broader feed.
Viral distribution
If the post clears the engagement threshold, LinkedIn pushes it to second-degree connections and beyond — users interested in the topic, not just your network. This is where reach multiplies 10x to 100x. Viral posts can keep earning impressions for 48 to 72 hours.
Outcome: Reach compounds until engagement decays.
The 5 ranking signals that actually move the needle
LinkedIn considers dozens of factors. In practice, these five explain 90% of why one post takes off and another doesn't.
Meaningful comments
Not emojis, not one-word replies. Substantive comments — especially ones that trigger a back-and-forth — signal that your content is worth pausing for. This is the single heaviest positive signal.
Dwell time
How long people stop scrolling on your post. Long-form posts that actually keep attention beat short posts that get a skim-and-scroll. The 'See more' click is a measurable proxy.
Content relevance
The algorithm matches your topic to each viewer's inferred interests. Posts that stay in a consistent niche build topical authority and get shown to tighter, more relevant audiences over time.
Creator authority
LinkedIn tracks your historical engagement rate, post frequency, and network quality. Consistent creators with solid engagement earn a baseline boost on every post.
Connection strength
Weighted by past interactions — people who've liked or commented on your posts before are heavily prioritized in the seed audience. Reconnecting with dormant contacts costs you early-window engagement.
What changed in 2026
LinkedIn has been re-tuning distribution aggressively. Four changes this year have the biggest impact on what you should actually post.
External links penalty softened
Posts with external links no longer get auto-capped. LinkedIn started rewarding links when the post itself gets substantial dwell time — effectively requiring the link to be earned by the content.
Video gets priority in professional topics
Short-form vertical video posts (under 90 seconds) are seeing 2 to 3x the reach on B2B/career topics compared to text-only equivalents. Captions and strong hooks in the first 3 seconds matter more than production quality.
AI-generated content detection
Posts with generic AI patterns (excessive lists, boilerplate openers, hollow thought leadership) are increasingly downranked. The fix is specificity: real names, real numbers, real lived experience.
Comment quality weighting
Comments over 12 words count significantly more than short ones. Reciprocal commenting loops (pods) are still detectable and devalued — organic depth wins.
7 tactics that still work in 2026
These aren't hacks. They're the patterns that correlate with reach across thousands of posts analyzed in 2025 and 2026.
Win the first 60 minutes
Post when your core audience is actually online. The engagement your post earns in the first hour determines whether it escapes the seed-audience test. Check your analytics for your personal peak window.
Hook in the first 2 lines
Those are the only lines visible before 'See more'. Specific, unexpected, or slightly contrarian openers trigger the click — which LinkedIn measures as dwell time.
Trigger comments, not just likes
End posts with a question that requires an opinion, not a yes/no. Reply to every comment in the first 2 hours to keep the thread alive — thread depth is a strong signal.
Ship 3 to 5 times per week, consistently
Creator authority compounds. Posting Monday-Wednesday-Friday for 3 months beats 20 posts in week 1 and silence after. LinkedIn rewards the creators who show up.
Stay in a clear niche for 6 months
Jumping between unrelated topics resets your topical authority. Pick a narrow lane and dominate it — the algorithm will eventually push your posts to people who care.
Use native formats, not reposts
Original text, native video, native document, and native polls all outperform link shares. Documents (PDF carousels) specifically still have very high dwell-time averages.
Write like a human, not like LinkedIn
Formulaic 'leadership' posts, motivational fluff, and obvious ChatGPT structure are downranked. Specific stories, unusual opinions, and one-of-a-kind experiences are rewarded.
Common myths, fact-checked
These claims spread every year. None of them hold up against actual data.
More hashtags = more reach
Three to five relevant hashtags are fine. Beyond that, the spam classifier kicks in and caps distribution. Hashtags are a tiny relevance signal, not a reach multiplier.
External links kill your reach
Not anymore. Since Q1 2026, LinkedIn measures dwell time before deciding to cap. A well-written post with one contextual link outperforms a link-free post with no substance.
There's a best time to post
There's a best time for your specific audience. Generic '8 AM Tuesday' advice is worthless — check when your connections are actually online and engaged, which varies wildly by industry.
Engagement pods work
They used to. LinkedIn's classifier now detects repetitive cross-engagement between the same accounts and devalues the reach boost. Real engagement scales; artificial doesn't.
Posts with images always win
Text-only posts often outperform posts with stock images. What matters is dwell time — if the image doesn't add signal, it's just slowing down the read and reducing your hook effectiveness.
Skip the guesswork. Write posts the algorithm actually rewards.
Posty is trained on thousands of posts that went viral in 2025 and 2026. It generates hooks, structure, and CTAs calibrated for the signals above — in seconds.
Try Posty freeLinkedIn algorithm FAQ
How often should I post on LinkedIn for the algorithm?
3 to 5 times per week is the sweet spot for most creators. Daily posting can work if you sustain the quality — but one thoughtful post beats five rushed ones every time. Consistency over months matters more than raw frequency.
Does the LinkedIn algorithm punish external links?
Not since early 2026. External links used to cap reach automatically, but LinkedIn now measures dwell time on the post itself before deciding. If your post earns attention, the link ships with it. If the post is thin, the link is what gets blamed but the real issue is the content.
What's the best time to post on LinkedIn?
There's no universal answer. Check your own analytics for when your audience engages most — it varies wildly by industry, country, and role. Tuesday to Thursday mornings work for most B2B audiences, but your peak might be Sunday night if you serve founders.
Why do some posts get 0 views?
Three common causes: the spam classifier flagged the post (hashtag stuffing, engagement bait, or AI-obvious patterns), the seed audience didn't engage in the first hour, or your historical engagement rate is so low the algorithm isn't testing you with enough people. Fix the post pattern, then the timing, then the consistency.
Do comments or likes matter more?
Comments dominate. A single substantive comment is worth more than 10 likes. A reciprocal thread where the original commenter replies again is worth much more than a one-shot comment. The algorithm heavily weights dialogue over passive reactions.
Does LinkedIn penalize AI-generated content?
LinkedIn doesn't ban AI posts, but it downranks generic AI patterns: over-listed content, boilerplate openers, hollow 'leadership' advice, and obvious ChatGPT structure. Human specifics — names, numbers, unexpected opinions — are what gets rewarded regardless of whether AI helped draft it.
What counts as 'dwell time' on LinkedIn?
How long someone pauses on your post before scrolling past. Clicking 'See more' is the strongest proxy. Long posts that keep attention beat short posts that get a skim. This is measured silently — you can't see it directly, but it drives whether your post moves to the next distribution phase.
How long does a LinkedIn post keep earning reach?
Most posts peak within 24 hours and decay fast. Viral posts can keep accruing impressions for 48 to 72 hours. A very small subset — highly shareable evergreen content — keep earning engagement for weeks, but these are rare.
Understand the algorithm. Outsource the writing.
Posty drafts posts calibrated for the exact signals on this page — hooks for dwell time, structure for comments, tone for your niche. Generate your first post free.
Start writing better posts