What earns reach on LinkedIn for B2B accounts in 2026

LinkedIn reach is compressing for everyone in 2026, but not evenly. Platform-wide, median impressions per post dropped 47% in the twelve months to May 2025, and average reach now sits at 8-12% of a creator's follower count, down from 15-20% a year earlier. The accounts that hold or grow their reach share a pattern: consistent topic focus, format discipline, and deliberate first-hour engagement engineering.

What earns organic reach for B2B brands on LinkedIn in 2026

Reach on LinkedIn in 2026 comes down to four variables: format, account type, posting cadence, and topic consistency. Miss one and the others cannot cover for it. A founder publishing sharp carousels three times a week from a company page instead of a personal profile still underperforms, because account type caps the ceiling before format gets a chance to matter. These four move together, and the accounts that grow are the ones treating all four as a single system rather than chasing one at a time.

Format is the clearest opportunity on the platform right now. PDF carousels, which LinkedIn classifies as native documents, carry a 1.39x reach multiplier and a 7.00% engagement rate, the highest of any format. Yet only 4.88% of creators post them regularly. That gap between performance and adoption is unusual. Most format advantages get competed away quickly once they become known, but the production friction of building a carousel keeps supply low while demand for that content stays high. We treat that as the single best format bet a B2B account can make in 2026.

Account type is where most B2B brands quietly lose. Company pages receive roughly 1-2% of the average user's feed allocation. Personal profiles occupy around 65%. Per 1,000 followers, personal profiles generate 2.75x more impressions and 5x more engagement than company pages. This is not a tuning problem you fix with better posting times or a content calendar. The company page reach ceiling is structural, which means the organic engine for a B2B brand is its people: the founder, the executive team, or a deliberate employee advocacy program.

Cadence is the variable teams get exactly backwards. The reflex is to post more. Daily posting causes a 26% drop in average reach per post compared to a 2-4 posts per week rhythm. Richard van der Blom, who tracks this data across large post samples, revised his own recommendation downward from 5-6 posts per week to 2-4 for 2026, citing quality and topic focus over raw volume. Posting less, on a tighter set of themes, is one of the few reliable ways to raise per-post reach without producing more content.

Topic consistency ties the other three together, and it interacts with the first hour of a post's life. LinkedIn seeds each new post to only 2-5% of a creator's network in the first 60 minutes, and just 5% of posts that underperform in that window ever recover to broader distribution. The algorithm decides fast, and it decides based on whether the post matches what it already expects from that account. Format, account type, cadence, and topic focus are the levers you control before that 60-minute test begins. Everything after publish is largely out of your hands.

The rest of this guide takes each of these in turn, starting with the algorithm change that reset the rules in March 2026 and why it makes topic discipline more valuable than it has ever been.

360Brew rewrote LinkedIn's algorithm in March 2026

On March 12, 2026, LinkedIn replaced five separate content-ranking pipelines with a single unified AI system called 360Brew. It is a 150-billion-parameter decoder-only foundation model trained on LinkedIn's proprietary professional data, and it now handles more than 30 tasks across the platform, including feed ranking, job recommendations, ad targeting, and search. One model, trained on how professionals behave on LinkedIn, deciding what shows up in your feed. That consolidation is the most important structural change to how the platform distributes content in years.

The mechanics matter for anyone producing content. 360Brew processes over 1,000 of a user's historical interactions as a temporal sequence before it ranks any new post. A separate retrieval layer called LiNR first narrows millions of candidate posts down to roughly 2,000 per user, and 360Brew then re-ranks those by semantic relevance. The consequence is direct: the feed now has long memory. It reads a creator's posting history as an ordered sequence to predict what they cover and who cares. Posting off-topic content even once disrupts that sequence and muddies the classification the model has built.

The dwell-time system changed at the same time, and it is the less-discussed half of the update. LinkedIn abandoned fixed-threshold dwell scoring, the old approach where crossing a static skip threshold such as 30 seconds counted as a positive signal. In its place is an auto-normalized binary classifier that predicts whether a user spends more time on a post than they spend on a comparable percentage of similar posts. That comparison is recalculated daily, split by content type, creator type, and distribution method.

The practical implication of the dwell change is that absolute attention no longer means what people assume. A post gets rewarded for holding attention relative to its peers, not for holding attention in absolute seconds. A text post that outholds other text posts from similarly sized accounts can outscore a long video that underholds relative to other videos, even when the video's raw dwell time is higher. This is why format comparisons only make sense within a format, and why chasing a single vanity metric like average watch time misreads how the model works.

For the full technical picture, LinkedIn's engineering team documented how the unified model replaced the prior retrieval-pipeline architecture, including how 360Brew reads interaction history as a sequence. The short version for practitioners: the algorithm now behaves less like a set of rules you can game and more like a model that has formed an opinion about what your account is for. The work in 2026 is giving it a clear, consistent opinion to hold.

Do company pages or personal profiles get more reach on LinkedIn in 2026?

Personal profiles get substantially more reach, and the margin is not close. Company pages receive approximately 1-2% of the average user's feed allocation. Personal profiles occupy roughly 65%. Per 1,000 followers, personal profiles generate 2.75x more impressions and 5x more engagement than company pages. If you are deciding where to invest B2B content effort, that ratio settles the question before any content quality discussion begins.

This gap predates 2026, but 360Brew sharpened it. The model's interest-based graph is built around person-to-person signals, and it weights the relationship between a reader and a specific human creator far more heavily than the relationship between a reader and a brand entity. Publish the same carousel from a founder's personal profile and from the company page, and the founder's version reaches a multiple of the audience. Same content, same day, same followers where they overlap. The difference is entirely the account type the model sees.

None of this makes the company page useless. It makes it a different tool. The page works as a credibility anchor and a content repository: the place a buyer lands after a founder's post catches their attention, the source of brand-consistent announcements, the archive of case studies. It is where trust gets confirmed, not where reach gets created. Treating the page as your primary distribution channel is the most common structural mistake we see in B2B LinkedIn programs, and no posting schedule fixes it.

This site regularly receives the search query 'should I use my personal or company LinkedIn profile to post to,' and in 2026 the answer is clean. Post substantive content, the observations and analysis and carousels that carry your point of view, on personal profiles. Use the company page to reshare that content and to publish the announcements that genuinely need to come from the brand voice. Run the reach engine through people and the archive through the page.

The pattern holds across account sizes. Executive and CEO LinkedIn programs consistently outperform company page organic programs on a per-follower basis, because they sit on the right side of the interest graph. If your brand has invested heavily in growing company page followers and reach has stayed flat, that is not a signal to post more from the page. It is the structural ceiling doing exactly what it does. Move the content to the people connected to the business, and the same effort produces a different result.

PDF carousels, not short video, lead B2B reach on LinkedIn in 2026

Native documents, the PDF carousels that swipe through a feed, have the highest engagement rate of any format on LinkedIn: 7.00%, up 14% year-over-year, per SocialInsider's analysis of 1.3 million posts from 16,645 business pages. They carry a 1.39x reach multiplier and a 1.30x engagement multiplier relative to the platform median, which works out to 39% more reach and 30% more engagement than the typical post. In a year where most formats are losing ground, documents gained it.

The save rate is the signal worth understanding, because it explains why the algorithm favors the format. Documents account for 12.92% of all saved posts while only 4.88% of creators post them regularly. That is a 2.6x save rate relative to their share of content. Saves are a strong dwell-time proxy: a reader who saves a carousel spent real time in it and expects to return, and 360Brew weights that behavior heavily. Low creator competition combined with a high save signal is the clearest format edge available on the platform right now.

Video is the format most B2B teams misread. Despite video uploads growing 53% year-over-year, the format carries a 0.86x reach multiplier and a 0.93x engagement multiplier relative to the platform median. It underperforms documents, images, and text. The story is a flooded market: everyone is uploading short clips, so the supply is enormous while the dwell-time signal those clips generate is thin. Video reach dropped 36% year-over-year even as upload volume climbed. Producing more of a saturated, low-dwell format is not a reach strategy.

There is a real nuance inside the video numbers, and it points the opposite way from the short-clip trend. Videos over 3 minutes achieve a 1.21x reach multiplier and a 1.17x engagement multiplier, outperforming most formats. Videos under 30 seconds get only 0.96x reach and 0.91x engagement. If your team is going to invest in video at all, the data says depth wins: a 3-minute-plus piece with genuine substance outperforms a batch of quick clips by a measurable margin. Short video is the volume trap; long video is a legitimate format when the subject earns the length.

Ranked by engagement rate, the 2026 format order runs: documents at 7.00%, video at 6.00%, images at 5.30%, text at 4.50%, polls at 4.20%, and link posts at 3.25%. Polls deserve a note. Before LinkedIn's March 2026 Authenticity Update, polls functioned as a reach-inflation tactic, carrying a 1.78x reach multiplier. After the update, poll engagement collapsed to a 0.07% engagement rate. The format that used to buy cheap reach now buys almost none. Any content plan still built around polls is optimizing for a mechanic the platform deliberately removed.

What most B2B LinkedIn content strategy advice gets wrong about posting frequency

The most repeated advice on LinkedIn is 'post 5 days a week.' The 2026 data says the opposite. Daily posting causes a 26% drop in average reach per post compared to a 2-4 posts per week cadence. Richard van der Blom revised his own frequency recommendation downward from 5-6 posts per week to 2-4 for 2026, citing quality and topic focus over volume. Posting less is not a compromise here. It is the higher-reach strategy, and the accounts still grinding out daily content are paying a reach tax for the effort.

There is a within-day throttle that most content guides ignore, and we have watched it hold across a wide range of accounts. The second post published in the same 24-hour window consistently underperforms the first by 30-50% in initial distribution, regardless of content quality. It is not documented publicly, but the pattern is stable enough to plan around. Accounts that keep a minimum 20-24 hour gap between posts show meaningfully better per-post reach. If you have two strong posts, the answer is to space them across two days, not to fire both and split their audiences.

Going dark carries a cost that surprises people when they resume. Accounts that stay quiet for more than 10-14 days do not pick up where they left off. The first 2-3 posts after a gap behave like cold-start content, seeding to a reduced initial test audience and recovering slowly over 2-4 weeks. Jumping straight back to full cadence makes it worse. Ramping back gradually, one post, then two, then normal rhythm, produces better first-hour distribution than returning immediately to full volume. The algorithm effectively re-learns the account, and rushing that process suppresses the early posts.

Reach is also compressing platform-wide, independent of anything you do. Median impressions per post fell from 1,211 in June 2024 to 636 in May 2025, a 47% decline in twelve months. Average post reach now sits at 8-12% of a creator's followers, down from 15-20% a year earlier, and active creators have seen roughly a 60% reach decline over two years. The right benchmark in 2026 is not raw impressions, which are falling for nearly everyone. It is relevance-weighted engagement from the specific accounts that matter to your pipeline.

The practical cadence for most B2B accounts is narrow: 2-4 posts per week, a minimum of 20-24 hours between posts, and consistent topic focus across everything published in a given week. Van der Blom's data shows posting around three times a week with 24-hour spacing can yield up to a 120% visibility increase compared with sporadic or overly frequent posting. Volume above that threshold does not add reach. It divides it.

The 90-day topic authority window

LinkedIn's algorithm needs roughly 90 days of consistent posting on 2-4 defined topic themes before it reliably classifies a creator's content domain and expands distribution beyond their immediate network. Richard van der Blom describes this shift to an interest-based graph as the biggest change in 10 years, and calls the process 'topic fingerprinting.' The account is not being throttled during those first 90 days. It is being read. The model is assembling a picture of what you cover, and until that picture is stable, distribution stays close to home.

The mechanism traces straight back to 360Brew. The model uses a creator's posting history as a temporal sequence to predict which topics they own and which user segments care about those topics. Consistency is what makes that prediction confident. An account that alternates between supply chain logistics and personal development gets classified into neither interest graph, because the sequence contradicts itself. The model cannot form a clean fingerprint from mixed signals, so it hedges, and hedging means narrow distribution. Posting off-topic content even once introduces noise the model has to work around.

Some patterns are not just unrewarded but actively penalized. LinkedIn explicitly detects and deprioritizes engagement-bait prompts, videos paired with irrelevant text, recycled generic-insight content repackaged without a new angle, content generated by comment automation tools, and generic template-based posts. That last category catches more well-intentioned creators than the others. A team producing legitimate content on a fixed template can trip a filter built to catch low-effort output, because the model reads structural sameness as a spam signal regardless of the quality underneath.

Template-based content creates a second, subtler problem beyond the direct filter, and this is where we have consistent first-hand data. Accounts leaning on numbered lists with identical structures week over week see engagement decay begin around weeks 6-8, even when the underlying information is genuinely good. The audience learns the shape and stops slowing down for it, which starves the dwell-time signal. Voice-matched content, where AI-assisted drafts are calibrated to reproduce a specific person's sentence-length patterns, vocabulary, and hook style, retains readers and avoids that decay far better than template output. In B2B tech especially, voice consistency matters as much as topic consistency.

The practical instruction is disciplined and a little uncomfortable. Pick 2-4 content themes that map to the questions your actual B2B buyers ask, and post only within those themes for at least 90 days before you evaluate whether to expand or pivot. Premature topic diversification resets the classification clock, and you pay the full 90-day cost again. The temptation to chase a trending topic outside your themes is real, and it is usually a mistake. The account that says fewer things, more consistently, is the one 360Brew learns to distribute.

Prime the first-hour engagement window before you publish

LinkedIn seeds each new post to only 2-5% of a creator's network in the first 60 minutes, and the engagement that small test audience produces decides everything that follows. Only 5% of posts that underperform in the first hour recover to reach a broader audience. That makes the first hour a binary gate, not a slow build. A post that lands in the initial test gets expanded; a post that does not, in almost every case, does not. Whatever you do to influence reach has to happen before or during those 60 minutes, because afterward the outcome is largely set.

The most reliable lever we have measured is early comment velocity from real people. Accounts that gather 3-5 substantive comments from genuine connections within the first 15 minutes of publishing show substantially higher breakout rates than accounts waiting for organic discovery to find the post. The word substantive is doing real work here. The comments must run 15 words or more. Short affirmations like 'great post' do not generate the same ranking signal, and a flood of them can trip the engagement-bait classifier, which turns a priming attempt into a penalty.

The comment signals 360Brew weights positively are specific: comments of 15 words or more, replies between commenters rather than only between a commenter and the original poster, and responses from the original poster keeping the thread alive. Threads beat counts. Twenty one-word reactions carry less weight than four real exchanges where people talk to each other. This is why answering every early comment yourself is not just good manners, it is a distribution input. The conversation you sustain in the first hour is part of what the model scores.

Priming works when it is real relationship activation and fails when it is orchestration. Three to five connections commenting because they have a genuine opinion on the topic sends exactly the signal the model rewards. A coordinated pod of accounts leaving recycled phrases on a schedule sends a different one. LinkedIn's detection of comment automation is active, not passive, and content generated by comment automation is on the explicit deprioritization list. Accounts running pods can see reduced organic reach beyond the automated actions themselves, which means the tactic can suppress the very posts it was meant to boost.

A practical method that stays on the right side of that line: before publishing, message 3-5 connections who have real interest in the topic. Share the core point of the post and ask what they think. If the post is worth their time, they arrive already primed to engage early and with something genuine to say, because they have been thinking about it. This is relationship activation, not engagement bait, and the difference is not cosmetic. One builds the early-conversation signal the algorithm rewards; the other is a pattern the algorithm is built to catch.

Structure your B2B LinkedIn content strategy in 2026 around four reach levers

Lever one is format selection. Default to PDF carousels for high-stakes content, because they carry the reach and save advantage nothing else matches in 2026. Use text posts for personal observations and quick takes, where the format's low friction fits the content. Reserve video for material that genuinely benefits from demonstration, and if you commit to video, publish pieces over 3 minutes rather than short clips, since the long form is the only video segment that outperforms the platform median. Short video is where effort goes to underperform.

Lever two is distribution channel, and it is the highest-leverage decision most teams make once and forget. Route substantive content through personal profiles. Use the company page for reshares and brand-consistent announcements. If you have access to an executive's profile and their buy-in, that is where the B2B content pipeline should run, because per 1,000 followers a personal profile generates 2.75x the impressions of a page. The company page reach ceiling is structural. No amount of page-side effort moves a wall that the interest graph built on purpose.

Lever three is link placement, and the penalty here is steep. Posts containing external links receive approximately 60% less reach than equivalent posts without them, a penalty two independent datasets confirm: dataslayer's roughly 60% figure and van der Blom's -18.8% median reach reduction across 1.3 million posts. LinkedIn deprioritizes content that sends users off-platform. The standard mitigation is placing the link in the first comment instead of the post body. It is worth being honest about the caveat: the first-comment workaround is widely used, but its 2026 effectiveness has not been independently verified across a broad range of account types and posting contexts. Treat it as a reasonable default, not a proven fix.

Lever four is automation hygiene, and it is where reach quietly leaks. Tools that post through LinkedIn's native scheduling interface produce different distribution outcomes than tools posting through third-party API endpoints. Native-path posts appear to receive the same initial seeding treatment as manually published posts. API-path posts from non-approved integrations show a measurable lag in initial distribution that compounds as posting frequency rises. Beyond publishing, rate-limiting outbound automation, meaning connection requests, profile views, and comment triggers, to well below observed platform tolerances is necessary. Push those actions too hard and the throttling can spill into your organic post reach, not just the automated actions themselves.

Pulling the levers together points to a specific workflow, and it is one we have watched hold up in practice. The founder or executive writes seed ideas from customer conversations and sales calls, because that is where the non-commodity point of view lives. A content operator drafts in the executive's documented voice. The executive reviews and approves. The post ships through native LinkedIn scheduling or an approved integration. The load-bearing part is voice consistency in the draft: reproducing the executive's real sentence patterns and vocabulary is what prevents the engagement decay that template content produces starting around weeks 6-8. A hybrid workflow scales output, but only voice discipline keeps that output from reading like everyone else's.

Frequently asked questions

What type of content gets the most reach on LinkedIn for B2B accounts in 2026?

PDF carousels (native documents) get the most reach: a 1.39x reach multiplier and 7.00% engagement rate, the highest of any format. Only 4.88% of creators post them regularly, so competition for that attention is low. Text posts perform well for personal observations, while video underperforms the platform median unless it runs over 3 minutes. Short video (under 30 seconds) is the weakest performing format despite high upload volume.

How does the LinkedIn algorithm decide what content to show in 2026?

LinkedIn's feed now runs on 360Brew, a 150-billion-parameter AI model deployed March 12, 2026. It processes over 1,000 of a user's past interactions as a temporal sequence to predict what content they will find relevant, then scores new posts using a dwell-time classifier that compares relative attention against similar content, recalculated daily. Topic consistency and early engagement velocity are the two strongest creator-controlled signals.

How often should a B2B brand post on LinkedIn to maximize reach without getting penalized?

Post 2-4 times per week, with at least 20-24 hours between each post. Daily posting causes a 26% drop in reach per post compared to the 2-4 per week cadence. A second post published within the same 24-hour window consistently underperforms the first by 30-50% in initial distribution. Quality and topic focus drive more reach than volume in 2026.

Do company pages or personal profiles get more reach on LinkedIn in 2026?

Personal profiles get far more reach. Company pages receive roughly 1-2% of the average user's feed; personal profiles occupy around 65%. Per 1,000 followers, personal profiles generate 2.75x more impressions and 5x more engagement than company pages. For B2B brands, the company page works as a credibility anchor and content repository, while organic reach runs through personal profiles of founders, executives, or employee advocates.

Why is my LinkedIn reach dropping in 2026 and what can I do about it?

Platform-wide reach is compressing: median impressions per post fell 47% between June 2024 and May 2025, and average reach now sits at 8-12% of a creator's follower count, down from 15-20%. If your drop is sharper than that, the most common causes are posting off-topic content that disrupts the interest-graph classifier, posting too frequently, or returning to full posting volume after a gap instead of ramping back gradually.

Are PDF carousels or native video better for B2B reach on LinkedIn?

PDF carousels outperform video on every distribution metric in 2026. Documents carry a 1.39x reach multiplier; video carries a 0.86x reach multiplier. The exception is video over 3 minutes, which achieves a 1.21x reach multiplier and outperforms most formats. Short video (under 30 seconds) is the weakest performing format, with a 0.96x reach multiplier, despite video uploads growing 53% year-over-year.

Does posting external links hurt your LinkedIn reach in 2026?

Yes. Posts containing external links receive approximately 60% less reach than equivalent posts without them. LinkedIn's algorithm deprioritizes content that directs users off-platform. The common mitigation is placing the link in the first comment rather than the post body. That approach reduces the penalty, though its 2026 effectiveness has not been independently verified across a broad range of account types and posting contexts.

How long does it take for LinkedIn's algorithm to recognize your content niche?

Approximately 90 days of consistent posting on 2-4 defined topic themes. Before that window closes, LinkedIn's 360Brew model is still classifying your content domain, and distribution stays closer to your immediate network. Posting off-topic content during this window resets the classification. Richard van der Blom calls this process 'topic fingerprinting' and describes the shift to an interest-based graph as the biggest algorithm change in 10 years.

What counts as engagement bait on LinkedIn and how does it get penalized?

LinkedIn explicitly detects and deprioritizes: posts using engagement-bait prompts ('comment X if you agree'), videos paired with irrelevant text, recycled generic-insight content repackaged without a new angle, content generated by comment automation tools, and generic template-based posts. These are actively penalized, not merely unrewarded. Accounts using comment automation may see reduced organic post reach beyond the automated actions themselves.

Sources and further reading

• Engineering the Next Generation of LinkedIn's Feed
• How LinkedIn Ranks Feed Content
• Leveraging Dwell Time to Improve Member Experiences on the LinkedIn Feed