A 2026 drop in LinkedIn impressions usually has one of two causes, and each needs a different response. Either LinkedIn changed distribution for everyone, or something about your own posts or account changed. On the first, LinkedIn rebuilt how the feed ranks posts on March 12, 2026, and in May it described systems that keep generic AI-written posts from travelling far beyond the poster’s direct network. Your post analytics can usually tell you which cause you’re looking at.
LinkedIn has not published a figure for how much reach fell across the platform. The “reach down 50%” numbers passed around in posts and blog articles don’t trace back to anything LinkedIn said.
This article is about diagnosis. For tactics once you know the cause, see the five tactics for increasing LinkedIn post impressions.
What did LinkedIn change in the feed in 2026?
On March 12, 2026, LinkedIn said it had begun ranking the feed with Generative Recommenders augmented with large language models, and that it was working to reduce “repetitive, low-substance posts and engagement bait.” In May it described systems that keep generic AI content from spreading far beyond a poster’s first-degree contacts. Both changes apply to every member.
The March announcement came as a LinkedIn Pressroom post and a longer engineering write-up. LinkedIn gave two examples of the bait it wants to show less of: “comment to agree” prompts and videos that don’t match the post text. The same Pressroom post said LinkedIn takes action against “comment automation, engagement pods, and unauthorized third-party tools.”
According to the engineering post, the ranking model processes “more than a thousand of your historical interactions,” drawing on what a member has “read, liked, commented on, returned back to, or simply scrolled past.” Ranking is built around each viewer’s own interests, so the same post can be picked for one reader and skipped for the next.
Writing for Social Media Today the same day, Andrew Hutchinson said that as every member’s feed changes, “that could also mean less overall reach for each post.” That line is Hutchinson’s own reading. LinkedIn’s announcement doesn’t say it.
The May change came through Laura Lorenzetti, LinkedIn’s VP and Executive Editor, in an interview with Entrepreneur published May 19, 2026. She said “Content creation on the platform is up 14% year over year” and “The feed is more competitive.” Entrepreneur reported that LinkedIn won’t delete posts it believes are AI slop but will depress their reach, “often not allowing it to be seen beyond a poster’s first-degree contacts.”
On July 30, 2026, chief product officer Hari Srinivasan told TechCrunch about new classifiers meant to cut AI slop and other low-quality content out of suggested recommendations from outside a member’s network.
Has LinkedIn said how much reach dropped?
No. LinkedIn has not published a platform-wide percentage for reach or impression declines in 2026. Figures like “views down 50%” or “company pages lost 60% of reach” circulate widely, and none of them has a LinkedIn source. The only view-related percentage LinkedIn has shared covers one narrow group: people who copy and paste AI-written posts.
In August, Srinivasan said that group was seeing around 40 percent fewer views than before LinkedIn launched its “Seems like AI slop” report button, as The Register reported. It’s a company-reported number about one behaviour, and it says nothing about a typical post.
How do you check whether your impressions really dropped?
Open analytics for individual posts and compare like with like. LinkedIn counts impressions and members reached differently and includes your own views in the totals. It also keeps some metrics longer than others. Once the numbers are comparable, the in-network and out-of-network split under impressions shows where the change happened.
- Separate impressions from members reached. LinkedIn’s post analytics help page defines impressions as the “Number of times your post was shown on LinkedIn.” Members reached is the “Number of distinct members and Pages that saw your post,” an estimate that “does not include repeat views.” If impressions fell while members reached held steady, the loss is in repeat views.
- Discount your own activity. The same page says “Your own views, social engagements, link engagements, saves, and sends are counted towards the analytics of your content.” Say you used to reopen every post a dozen times on launch day to answer comments, and now a colleague handles replies. Your impressions dip for reasons unrelated to the feed. LinkedIn also notes the numbers “are estimates and may not be precise.”
- Check how far back you can compare. For most post types, discovery and social engagement counts stay available for 1,000 days, members reached for 400 days, and demographic breakdowns for 180 days. Video analytics run to 365 days and article and newsletter analytics to two years. A year-on-year members reached comparison only works for posts under 400 days old, so write down the numbers you want to keep.
- Line up posts of the same format across the change dates. Put text posts from before March 12 next to text posts from after it, then do the same around mid-May.
- Read the in-network and out-of-network split. The help page defines In network as the “Percentage of impressions from members who follow or are connected to you” and Out of network as the percentage from members who don’t. Sam Corrao Clanon, LinkedIn’s director of creator products, said in June 2026 that the breakdown sits “in the discovery section under impressions,” according to Social Media Today. It was ramping up globally at the time, so not every account had it straight away. Our guide to reading in-network vs out-of-network reach covers how to interpret the two percentages across a batch of posts.
- Look for a private notice about AI use. TechCrunch reported that LinkedIn would begin privately flagging in users’ dashboards when their content comes across as inauthentic because of heavy AI use. A notice like that is about as direct an answer as LinkedIn gives.
Is the drop feed-wide, or specific to your posts?
Use the network split. If your out-of-network share fell while in-network impressions held, LinkedIn is recommending your posts to strangers less often, which fits its stated limits on generic and low-context content. If in-network impressions fell as well, look at the posts themselves first, then at the feed-wide ranking change.
LinkedIn’s Suggested posts help page says a post shown outside the poster’s network should give context for the topic and shouldn’t be “overly promotional in nature,” and every part of it should relate to the rest. LinkedIn’s own example of content that works for your network but not for strangers is a job-change update.
Suppose a consultant used to get about a third of their impressions from outside their network, and that share is now in single digits while in-network impressions are flat. That pattern points away from an account problem. Their posts are being picked less often for people who don’t know them, and the fix is writing posts that make sense to a stranger.
If both numbers fell around March 12 or mid-May and your habits stayed the same, the feed-wide changes are the likeliest explanation.
Which habits cut reach, according to LinkedIn?
LinkedIn has named engagement bait such as “comment to agree” prompts, videos that don’t match the text, recycled low-substance posts, generic AI-written content, overly promotional posts in suggested feeds, and automated comments. Links and editing get blamed often too, though none of the LinkedIn pages cited here describe a reach penalty for either.
Engagement bait and recycled posts
Both appear by name in the March Pressroom post. If your recent posts end with a reaction prompt your older posts didn’t have, test the same kind of post without it.
Heavy AI writing
Lorenzetti told Entrepreneur that LinkedIn doesn’t want to punish people for using AI tools, only for low-value content that lacks original insight, expertise, or perspective. The Next Web reported that LinkedIn singled out the “it’s not X, it’s Y” format as one example of generic writing. If you started leaning on AI drafts around the time your numbers fell, compare those posts with ones you wrote yourself. The HypeLab guide to using AI without flattening your LinkedIn engagement goes further.
Automated comments
LinkedIn’s comments help page says: “if we detect excessive comment creation or use of an automation tool, we may limit the visibility of those comments.” In February 2026, Gyanda Sachdeva, LinkedIn’s VP of Product Management, described action against comments “posted to LinkedIn through a third party auto-script or a browser plugin without any human oversight or review.” Social Media Today reported that detected comments are removed from the “Most Relevant” section and may be kept from reaching beyond the commenter’s network, and that members who keep posting them may have their accounts restricted. If a tool has been commenting for you to keep you visible, those comments may now reach far fewer people.
Links and edits
A “links in posts cut reach by 60%” figure gets repeated a lot, with no LinkedIn source behind it, and the LinkedIn help pages cited here say nothing about penalties for links or edits. Your analytics show “Visits to links in this post,” so compare a few link posts with no-link posts on similar topics before changing anything.
What does each symptom point to?
Start from the symptom. Each row begins with something you can see in your own post analytics, pairs it with the explanation that fits best, and gives you a check to confirm or rule it out before you change anything. Several rows can apply at once, so work through all of them instead of stopping at the first match.
| Symptom in your analytics | Likely cause | How to check | What to change |
|---|---|---|---|
| Impressions down, members reached steady | Fewer repeat views, including your own | Compare both metrics on the same posts | Nothing, if members reached is the number you report |
| Out-of-network share fell, in-network steady | Posts picked less often for suggested feeds | Discovery split before and after March 12 and mid-May 2026 | Give strangers context and cut promotional framing |
| Drop concentrated on AI-drafted posts | Generic content held within first-degree contacts (May 2026) | Compare with posts in your own words; check for a dashboard notice | Write from your own work; keep AI to proofreading |
| Posts ending in “comment YES” style prompts underperform | Engagement bait filtering (March 2026) | Compare posts with and without the prompt | Ask a real question tied to the post, or drop the prompt |
| Your comments on others’ posts get little response | Visibility limits on excessive or automated comments | Check whether any tool comments for you | Stop automated commenting; comment less often, with substance |
| Both network segments fell at once, habits unchanged | Feed-wide ranking change | Same-format posts across March 12 and mid-May | Judge posts over several weeks by members reached and reply quality |
| Link posts trail no-link posts | Unconfirmed; LinkedIn has published no link penalty | Similar topics with and without links | Test before believing an unsourced percentage |
| Restriction notice on your account | Account restriction, a separate problem from reach | LinkedIn notifications | See why LinkedIn restricts accounts |
Is LinkedIn shadowbanning you?
None of the LinkedIn help pages or announcements cited in this article use the word “shadowban.” What LinkedIn does describe is narrower and comes in separate pieces: visibility limits on automated or excessive comments, generic AI posts held close to a poster’s first-degree network, engagement bait filtered out of the feed, and account restrictions for members who break the rules.
Each of those shows up somewhere different in your data, so one label won’t help you find the cause. Held-back recommendations read as a falling out-of-network share, while comment limits hit your comments on other people’s posts and leave your own posts alone. Restriction is a separate matter altogether. LinkedIn’s prohibited software and extensions page says members who use prohibited tools “risk having their accounts restricted or shut down.”
Where does early conversation fit once the diagnosis is done?
Once you’ve ruled out the causes above, what’s left is the post itself and the people who reply to it. Comments from people who work in your field can start a proper discussion under a post that deserves one. They can’t rescue a weak post or undo a feed-wide change.
A lot has been written about the first hour after posting, and the HypeLab post on LinkedIn comment velocity covers it. Do the diagnosis first, fix whatever it turned up, and worry about who joins the conversation after that.
PostPilot fits at that last step. It’s HypeLab’s engagement tool for LinkedIn posts. You submit a post you have already published, and PostPilot matches it to network members whose profiles and interests fit the topic. It drafts the comments, and nothing goes live until you have read and approved each one. The engagement comes from real LinkedIn profiles. Interactions arrive spread out over time rather than in a single burst, and the tool runs in the cloud, so there is no Chrome extension to install. The how it works page walks through the steps. PostPilot won’t restore lost reach, and it isn’t a way around the changes LinkedIn has made. If your diagnosis pointed to bait or an automation tool, deal with that first.
If your analytics say the post was worth reading and it only needed people in your field to join the conversation, Get Started for Free with PostPilot and check the current plans before you scale it across more accounts.
FAQ
How long does a LinkedIn reach drop last?
LinkedIn hasn’t published a recovery timeline, and feed-wide changes don’t expire. The March Pressroom post said “these changes are already rolling out” with testing continuing, and Lorenzetti said the May AI content systems may take several months to fully affect users. Judge a drop across several weeks of comparable posts.
Why are my impressions so much higher than clicks to my website?
The two count different actions. An impression is each time the post was shown. Link clicks appear separately as “Visits to links in this post,” including repeat clicks. LinkedIn also shows no data for links it detects as spam, malicious, or in violation of its professional community policies, or for links pointing to content on LinkedIn itself.
Can one “Seems like AI slop” report cut my reach?
According to The Register, Srinivasan said no single piece of feedback is used to adjust how widely a post travels, inside or outside a user’s network. He said distribution is determined by a range of signals.
Does LinkedIn penalize posts written with AI?
LinkedIn says it goes after low-value content. Lorenzetti told Entrepreneur the company doesn’t want to punish users for using AI tools. TechCrunch reported that LinkedIn pulled its own “enhance your post” AI writing feature and replaced it with one that proofreads your words instead of changing your voice.
Sources
- LinkedIn Pressroom, “How LinkedIn Is Improving the Feed to Show More Relevant, Authentic Professional Content” (March 12, 2026): https://news.linkedin.com/2026/ImprovingTheFeed
- LinkedIn Engineering, “Engineering the next generation of LinkedIn’s Feed” (March 12, 2026): https://www.linkedin.com/blog/engineering/feed/engineering-the-next-generation-of-linkedins-feed
- LinkedIn Help, “Post analytics for your content”: https://www.linkedin.com/help/linkedin/answer/a516971
- LinkedIn Help, “Suggested posts in your feed”: https://www.linkedin.com/help/linkedin/answer/a1499047
- LinkedIn Help, “Comment on posts and reply to a comment”: https://www.linkedin.com/help/linkedin/answer/a524166
- LinkedIn Help, “Prohibited software and extensions”: https://www.linkedin.com/help/linkedin/answer/a1341387/prohibited-software-and-extensions?lang=en
- Social Media Today, Andrew Hutchinson, “LinkedIn Updates Its Feed Algorithm” (March 12, 2026): https://www.socialmediatoday.com/news/linkedin-updates-its-feed-algorithm/814638/
- Social Media Today, “LinkedIn Outlines Measures To Combat Engagement Pods” (February 16, 2026): https://www.socialmediatoday.com/news/linkedin-outlines-more-measures-to-combat-engagement-pods/812290/
- Social Media Today, Andrew Hutchinson, “LinkedIn Adds More Post Performance Insights” (June 7, 2026): https://www.socialmediatoday.com/news/linkedin-adds-more-post-performance-insights/822193/
- Entrepreneur, Jason Feifer, “LinkedIn Is Fighting Back Against AI Slop and AI Comments” (May 19, 2026): https://www.entrepreneur.com/business-news/linkedin-is-fighting-back-against-ai-slop-and-ai-comments
- The Next Web, “LinkedIn is finally cracking down on AI slop” (May 20, 2026): https://thenextweb.com/news/linkedin-ai-slop-crackdown-generic-content
- TechCrunch, Sarah Perez, “LinkedIn adds a button to report AI-generated ‘slop'” (July 30, 2026): https://techcrunch.com/2026/07/30/linkedin-adds-a-button-to-report-ai-generated-slop/
- The Register, “Users mash LinkedIn’s AI slop button 1M+ times in 3 weeks” (August 24, 2026): https://www.theregister.com/ai-and-ml/2026/08/24/users-mash-linkedins-ai-slop-button-1m-times-in-3-weeks/5291802
All URLs re-opened and wording confirmed on September 13, 2026.
