The diagnostic · four branches, one afternoon

Your LinkedIn reach dropped. Find which of the four problems you actually have.

Ezekiah Livingston LinkedIn strategist and ghostwriter. I diagnose dead reach for a living, and most of what's written about it is astrology. This is the checklist I actually run. Find me on LinkedIn.

Published August 28, 2026

TL;DR

Reach drops have exactly four causes: a platform shift, a leaking profile, a content problem, or the wrong audience. Panic-posting treats none of them.

Your profile is part of your content now. LinkedIn's ranking model reads profile text as context when scoring posts, so a mismatch suppresses everything you publish. Confirmed

Each branch below has a test you can run free, in minutes, in order.

The most common finding: the posts are fine, but the profile disagrees with them, or the audience was built years ago for a different business.

Reach drops feel personal, and the feed ships no error message. So founders do the one thing guaranteed not to help: post more of the same, faster. Diagnosis first. There are only four places the problem can live, and you can test all four in an afternoon without paying anyone anything.

Branch 1: Is it the platform? (everyone's favorite suspect)

Sometimes it genuinely is. LinkedIn replaced its stack of task-specific ranking models with 360Brew, a roughly 150-billion-parameter foundation model that reads text and scores relevance. Confirmed Reach reshuffled hard around that transition, and many accounts that coasted on volume or engagement-bait got repriced downward. Observed

The test: compare 90-day windows, not week to week, because weekly volatility is noise. Then check two or three peers in your niche. If everyone's recent posts show the same slide, some share of your drop is the platform, and the fix is relevance, not desperation. The full plain-English breakdown is in the 360Brew guide.

You can't fix the platform. You can only be more relevant than the account the model replaced you with.

Branch 2: Is your profile leaking? (the one nobody checks)

This is the branch that surprises people, because it doesn't feel like a reach problem. But the model reads your headline, About, and history as context for every post you publish. Confirmed If your profile says 'CEO at Company' and your posts are about, say, data infrastructure, the model has to guess whether you're credible on the topic. Guessing costs you distribution.

The test takes two minutes: does your headline name the topic you post about most? Does your About mention it in the first three lines? If either answer is no, your profile is arguing with your content, and the model hears the argument. Then run the free 8-Second Test to see what a stranger, and by extension a model that reads like one, concludes from the profile itself.

Branch 3: Is it the content? (be honest, or let a tool be honest for you)

Three content patterns kill reach quietly. Generic cadence: if your posts read like everyone's AI drafts, readers skip them, the model notices the skipping, and distribution follows. Weak hooks: posts that die in the first two lines never earn the dwell time that ranking rewards, and hooks are fixable. And engagement bait: 'comment YES below' patterns get repriced brutally now. Observed

The test: paste your last five posts into the free Content Audit. It grades hooks, structure, readability, endings, and substance without knowing or caring whose posts they are. It also refuses to predict engagement, which is exactly why its craft verdict is trustworthy. If everything comes back strong and reach still fell, your problem lives in another branch.

One more honest check: would a stranger save your post? Saves and substantive comments signal depth. Likes from the same twelve supporters signal nothing. Observed

Branch 4: Is it the audience? (right posts, wrong room)

A network built three years ago for a different job is a distribution problem wearing a loyalty costume. Your posts surface first to your connections; their behavior teaches the model who else should see them. If your feed is full of peers, recruiters, and mass-connect leftovers instead of buyers, every impression teaches the model the wrong lesson.

The test: open your last three posts and look at who actually reacted. Count how many could buy from you, hire you, or refer you. If it's under a third, your reach number was never measuring what you thought it was, and growing it in this room wouldn't help. Audience repair is slower than copy fixes: connect deliberately with the people your content is for, and comment substantively where they already are.

The fix order (this part matters)

Profile alignment first. One afternoon, and it compounds: every future post inherits the fix. Headline names your topic, About backs it, Featured shows proof.
Hooks next. Apply to your next five posts. Cheapest possible test of whether craft was the leak.
Topic consistency for four to six weeks. The model updates on evidence, and evidence takes posts. One niche, held, beats five niches sampled. Observed
Audience, ongoing. Deliberate connections and substantive comments where your buyers already read.

Volume is deliberately last. More of what's being ignored is just more ignoring.

The uncomfortable version: sometimes reach didn't drop, it corrected. If the old number came from engagement bait, broad-topic posting, or a feed full of peers, the model wasn't wrong to reprice it. The way back is narrower and better, not louder.

Two free tests · ten minutes total

Run the diagnosis instead of guessing

The 8-Second Test tells you what your profile signals. The Content Audit tells you what your posts earn. Between them, three of the four branches are covered before you spend a dollar.

Test my profile Audit my posts

Questions people keep asking me

Why did my LinkedIn reach drop in 2026?

One of four causes: a platform shift (LinkedIn's 360Brew model repriced relevance), a profile that contradicts your content, content that reads generic or dies in the first two lines, or an audience built for a different era of your business. Each has a free test, and the fix order matters: profile first, hooks second, consistency third, audience ongoing.

Did LinkedIn penalize or shadowban my account?

Almost certainly not. True penalties follow policy violations and come with notifications. What feels like a shadowban is usually relevance repricing: the ranking model found posts it predicts your network cares about more. The distinction matters because a penalty needs an appeal, while repricing needs better-aimed content.

Does posting more often fix low reach?

No. Volume multiplies whatever signal already exists, so posting more of what's being ignored produces more ignoring. Fix the relevance problem first: profile alignment, hooks, and topic consistency. Then volume helps.

How long does it take for LinkedIn reach to recover?

Expect four to six weeks of consistent, aligned posting before the model has enough evidence to re-rank you. Profile fixes apply to every post from the moment you make them, which is why they come first, but distribution follows a track record, not a single good post.

Do hashtags still matter for reach?

Far less than they did. The current model reads the meaning of your text directly, so topic clarity in plain words does the work hashtags used to pretend to do. A tidy row of hashtags neither saves a vague post nor meaningfully boosts a clear one.

Keep going: how the 360Brew algorithm actually works, the hook-writing guide, the 11 profile mistakes that quietly kill trust, or the analytics guide to track whether the fixes worked.

Model facts labeled Confirmed come from LinkedIn's 360Brew paper (arXiv 2501.16450, January 2025). Reach observations are aggregated community reports, labeled Observed because LinkedIn has not published them.