LinkedIn Pulse lost 89% of its estimated Google traffic while traffic to LinkedIn posts grew 250%. For B2B teams, that changes how LinkedIn should be treated: not only as a short-term distribution or lead-generation channel, but as a place where expert content can be indexed, discovered and cited.
The numbers come from Foundation Marketing's analysis of Ahrefs data for two LinkedIn URL structures: /pulse/, which contains long-form articles, and /posts/, which contains standard feed posts.
The divergence is striking. Estimated monthly organic traffic to Pulse peaked at approximately 33 million visits in March 2024. By March 2026, it had fallen to around 3.6 million—an 89% decline. The number of Pulse pages ranking in Google fell from approximately 5 million to 500,000 over the same period.
Source: Foundation Inc research
Posts moved in the opposite direction. Since October 2025, estimated monthly organic traffic to /posts/ URLs increased from roughly 3 million to 11 million—about 250% growth and more than three times Pulse's current estimated traffic.
These are third-party estimates, not LinkedIn's internal traffic figures. They also do not prove that publishing a post causes an AI citation, that Google has deindexed millions of Pulse articles, or that posts should replace company blogs. What they do show is a material shift in Google search visibility toward a format many B2B teams still treat as disposable.
Most B2B teams still budget for LinkedIn as two things: paid demand generation and a company page that proves the brand is active. Both have a place, but neither fully captures the value of expert content that can remain publicly available after it disappears from the feed.
A better way to think about LinkedIn is as a high-authority third-party publishing surface where your people can make what they know discoverable. Subject-matter experts can explain category problems, implementation trade-offs, customer patterns and the reasoning behind a point of view without every claim living on a corporate website that asks the reader to take the company's word for it.
This does not make owned content less important. Company blogs still build search equity, support deeper arguments and give the business control over presentation, conversion and measurement. The more useful distinction is between the roles the two surfaces play:
For an AEO strategy, the opportunity is to make the two reinforce each other.
Separate studies confirm that LinkedIn is already a significant source in AI-generated answers, although its exact position depends on the platform, query set and methodology.
In early 2026, Semrush analyzed 325,000 prompts and 89,000 LinkedIn URLs cited by ChatGPT Search, Google AI Mode and Perplexity. LinkedIn ranked second among cited domains in that dataset and appeared in an average of 11% of responses.
Profound reached a related but distinct finding. Its analysis of millions of real ChatGPT queries found LinkedIn moving from approximately the 11th-most-cited domain to fifth between November 2025 and February 2026. Using a separate synthetic prompt set across six AI platforms, Profound found LinkedIn was the most-cited domain for professional queries.
Those results should not be treated as a universal league table. A Peec AI analysis of 30 million citation sources, for example, ranked LinkedIn third overall across five AI platforms. The differences are a useful reminder that citation rankings are highly sensitive to the prompts, industries, products and dates being measured.
There is another important nuance: LinkedIn articles still account for a larger share of current LinkedIn citations than posts. In Semrush's study, articles represented 50–66% of cited LinkedIn content, while feed posts accounted for 15–28%, depending on the AI platform. Posts are becoming a more meaningful search and citation surface, but the evidence does not support declaring articles dead or posts universally superior.
The sensible conclusion is narrower: B2B teams should stop treating posts as temporary promotional wrappers around the “real” content. A useful post can be a retrievable document in its own right.
Publishing on LinkedIn does not automatically improve a brand's visibility in AI answers. Citation studies observe which sources appear in outputs; they do not isolate all the factors that caused a brand, source or claim to be selected.
There is also a difference between what an AI system uses and what it displays as a citation. An answer engine can retrieve or learn from more sources than it ultimately shows to the user, so citation share is an incomplete proxy for influence.
That limitation does not make publishing irrelevant. Buyers and answer engines both need public evidence of what a team knows. When the best material is trapped in sales calls, Slack threads, post-mortems, QBRs and internal decks, it has little chance of informing an external answer. Making that expertise discoverable is useful even if no single post-to-citation causal chain can be proven.
The goal, then, is not to chase citations as an isolated metric. It is to increase the amount of credible, specific and publicly retrievable evidence associated with your experts and your company.
Most company-page content contains little that a buyer—or an answer engine—can extract and reuse.
Consider a generic post about reducing churn in B2B SaaS:
Churn remains one of the toughest growth challenges for SaaS companies, especially once customer acquisition costs start rising. We help customer success teams identify early signs of risk, improve onboarding and build stronger customer relationships over time. If retention is a focus this quarter, we'd be happy to share what we're seeing.
It is not badly written. It is simply difficult to distinguish from thousands of similar posts. There is no specific failure mode, mechanism, observation or practical takeaway.
Now consider this explicitly hypothetical example of how a subject-matter expert might make the same topic more useful:
A common retention blind spot appears between the end of onboarding and the start of business-as-usual customer success. The onboarding team marks the account complete, but the success manager waits for the next scheduled review. If product usage drops during that gap, nobody owns the signal. One practical fix is to trigger a check-in from a defined usage decline rather than a calendar date. That gives the team a chance to intervene when customer behaviour changes, not several weeks later.
The difference is not polish or keyword density. The second version identifies a moment in the customer journey, an ownership problem, a behavioural signal, an operational mechanism and a takeaway another team could apply.
That is what “entity-rich” or citable content should mean in practice: fresh, original and specific information drawn from what the team actually knows. It does not require keyword stuffing or a forced list of product names.
The material usually exists inside the business before marketing turns it into content. It can be found in:
The challenge is turning that knowledge into a regular publishing habit without asking already-busy experts to become full-time creators.
That is where workflow matters. Marketing needs a way to identify questions worth answering, route them to the people with first-hand knowledge, shape their answers without erasing their voice and publish consistently across the right surfaces.
Start by measuring the formats separately rather than assuming every LinkedIn URL behaves the same way.
Pulse still has a role. Existing articles can continue to rank and appear in AI answers, and long-form LinkedIn content currently represents a substantial share of LinkedIn citations. But the search case for treating posts as durable assets has become much stronger.
Rocksalt is built to make expert-led social content function as part of an AEO strategy rather than as feed filler.
The workflow has three parts:
The Foundation data reinforces the need for that workflow. As LinkedIn posts gain search visibility, the constraint on B2B AI visibility is not simply publishing volume. It is getting specific, defensible expertise out of busy people consistently—and giving that expertise a format and distribution strategy that make it discoverable.
If you want to see which questions your buyers are already asking on LinkedIn and Reddit—and turn your team's answers into content for your website and social channels—book a demo of Rocksalt.