Why are my LinkedIn posts getting no views? (And what's actually worth fixing)

Organic reach on LinkedIn is down significantly for most people. Before you spend an afternoon tweaking your posting schedule or stuffing a carousel with bullet points, it’s worth understanding why, and what the platform actually wants from you now.

First, the honest context

Richard van der Blom, the most-cited independent analyst on LinkedIn’s algorithm, has tracked a 34–50% drop in organic reach year-on-year across 2024 and 2025. That’s not a glitch and it’s not something you can engineer your way around. The feed is more crowded, the algorithm is more selective, and the bar for what counts as “relevant” has moved.

Anyone claiming they’ve cracked it is probably selling a scheduling tool.

That said, there’s a real difference between posts that perform and posts that don’t, and the gap is mostly explained by a few things the algorithm has started to prioritise more explicitly.

What the algorithm is actually measuring

Dwell time, not just reactions

LinkedIn now tracks how long someone actually reads your post, not just whether they tapped a reaction. A post that collects 50 quick likes and no reading time will underperform one that earns 20 comments and holds people long enough to finish it. This matters for how you write. A post that buries its point or pads out the opening to create “suspense” trains readers to scroll past you faster. A post that gets to the point and then earns its length tends to do better.

Comments over everything else

One widely-cited breakdown puts comments at roughly 15 times the algorithmic weight of a like. A genuine back-and-forth in the comments section is worth far more to the algorithm than a stack of thumbs-up. The catch: “Great post!” comments are increasingly detected as low-value. The algorithm wants real conversation, not performance of conversation. If your posts generate that, you’re in good shape. If they don’t, adding “What do you think?” to the end won’t fix it.

The first hour

LinkedIn tests every new post with a small slice of your network, somewhere between 2 and 5%. How that group responds determines whether the post gets broader distribution. Strong early engagement sends it further. Weak early engagement and it quietly stops travelling. This is why timing still matters, though not in the obsessive “8:47am on Tuesday” way some tools suggest. Posting when your audience is likely online is sensible. Posting a good piece at 2am and wondering why nobody saw it is a different problem.

Topic authority

LinkedIn has shifted from rewarding activity to rewarding expertise. If you post consistently about one topic, the platform starts to categorise you as an authority on that topic and distributes your content to people interested in it, even beyond your direct connections. Scattered content across multiple unrelated themes confuses the system. Staying in your lane professionally pays off over time, which is convenient if you happen to have something genuinely useful to say in a specific area.

Saves and sends

A post someone saves to read later, or sends privately to a colleague, tells the algorithm something a like doesn’t. It says the content had enough value to keep. If your posts generate saves, that’s a sign you’re writing things worth writing.

Format: what performs and what doesn’t

Carousels (uploaded as PDF documents) consistently top the performance data, with engagement multipliers around 1.4–1.6x compared to plain text. They keep people on the platform longer, they get saved more, and saves are a stronger signal than likes. Polls are close behind, though overusing them looks lazy. Images outperform plain text. Plain text, if the writing is genuinely good, still works on personal profiles but struggles on company pages.

Video is genuinely contested in the research right now. Some 2025 data shows video underperforming; some shows it doing well. The consensus leans toward short, native video — uploaded directly to LinkedIn rather than linked from YouTube — being worth testing, but it’s not the obvious winner it once was.

What gets punished

External links

Posts with external links see a meaningful drop in reach. LinkedIn wants people to stay on LinkedIn. One link in a post reads as promotional. The workaround most people have used — putting the link in the first comment — appears to have also been picked up and penalised as of early 2026. If you need to share something, either write enough genuinely useful context around it to justify the reach penalty, or accept it and move on.

Engagement bait

“Comment YES if you agree” and “Follow for more” posts get flagged. The algorithm has got better at spotting the pattern. Even the softer version — ending every post with “What do you think?” when there’s no genuine reason for the question — is increasingly transparent. If the question is real and the post earns it, fine. If it’s a mechanical add-on, it reads that way.

Hashtags

LinkedIn removed clickable hashtags from desktop in late 2024. Using more than two or three signals that you’re following 2022 advice. Use one if it genuinely helps with search. Otherwise, drop them.

Generic AI content

LinkedIn doesn’t technically penalise AI-assisted writing, but research suggests AI-generated posts without a genuine human perspective average around 45% fewer interactions than posts with a real point of view in them. The algorithm doesn’t know you used a tool. Readers do. They scroll past content that sounds like a press release, and posts that get ignored early die quickly.

The practical upshot

The algorithm has moved in a direction that rewards exactly what you’d write anyway if you were being direct and specific: a clear point of view on something real, in your own voice, that gives someone something to think about or push back on. It penalises the LinkedIn performance art that was always embarrassing: the inspirational quote graphics, the “I’m humbled to announce” posts, the vague “thoughts on leadership” filler that says nothing and asks for nothing.

The one practical adjustment worth making is to think about format occasionally. A well-structured carousel that takes a specific problem and walks through it clearly will outperform the same content as a text post. That’s not about changing what you’re saying. It’s about choosing the medium that gives it the best chance of being read.

Beyond that, the fundamentals are simple enough: post on the same topic consistently, write posts that are worth reading rather than just eye-catching, respond to comments properly, and don’t chase reach with gimmicks the platform was designed to detect.

The posts that do well in 2026 are mostly the posts that would have been worth reading in 2016. The algorithm has just got better at telling the difference.

W4E helps technical businesses communicate more clearly. If your LinkedIn presence isn’t working as hard as it should, that’s usually a copy problem, not a platform problem.