LinkedIn ranks second in AI citations for B2B searches

Meltwater has published research showing LinkedIn was the second most-cited source in AI-generated answers, based on an analysis of 9.5 million citations across 16 B2B categories.

Using Meltwater's GenAI Lens, the study examined which sources and content formats were cited by AI tools including ChatGPT, Claude and Microsoft Copilot. YouTube ranked first, while LinkedIn placed ahead of many company-owned and review-based sources in business-focused queries.

The findings suggest a shift in how B2B brands are discovered online, as buyers increasingly use large language models during research and purchasing decisions. Meltwater cited 6sense research showing that 94% of B2B buyers use LLMs during the buying process.

Source mix

Platforms built around third-party and user-generated material accounted for 47.5% of AI citations, according to the research. That compared with 15% from peer review sites and 18.7% from company websites, suggesting AI models often favour content outside a brand's own channels.

Within LinkedIn, individual users were the main source of visibility. Around 75% of LinkedIn citations came from member profiles, compared with 25% from company pages.

This pattern suggests personal expertise and direct commentary were cited more often than official corporate publishing. At the same time, the report said companies still needed a balance between individual and brand content.

Content patterns

The most-cited LinkedIn material included posts and articles with clear formatting, such as bullet points, numbered lists, strong headings, named entities and quantitative data. The report found that AI systems tended to reward relevance and structure rather than raw audience size.

More than half of LinkedIn citations, or 51%, came from users with fewer than 10,000 followers. That suggests visibility in AI answers does not depend solely on social reach, but on how clearly information is presented and how closely it matches the question being asked.

LinkedIn performed particularly strongly in professional and technical sectors, ranking among the top five citation sources for B2B searches across Technology and SaaS, Consulting and Professional Services, Financial Services and FinTech, Marketing and Advertising, and HR and Talent.

The research argues that these categories are more likely to generate professional, technical or decision-oriented questions, making LinkedIn content more likely to be selected by AI systems looking for practical and attributable information.

Chris Hackney, Chief Product Officer at Meltwater, said the results showed that discoverability in AI systems was becoming a commercial issue for brands as well as a communications concern.

"For the last twenty years, the job of a brand was to be discoverable. In an AI-first world, the job is to be the answer. LLMs are now the first stop for decisions that used to take hours of research - and if your brand isn't being cited, you're not in the consideration set," said Chris Hackney, Chief Product Officer at Meltwater.

He added that the analysis showed AI models were favouring credibility over volume.

"What this data makes clear is that AI models aren't looking for the loudest voice in the room - they're looking for the most credible voice. Organisations that are discoverable in reputable earned media, credible with structured, factual content, and consistent with the channels they engage on, are the ones that will show up when decisions are being made," Hackney said.

LinkedIn also presented the findings as evidence that product discovery is moving into AI-driven interfaces. Its marketing leadership said brands absent from AI answers risk being overlooked at the point where buyers form shortlists.

"Product and brand discovery doesn't happen in stages anymore - it starts with a question and ends with an AI assistant's answer," said Davang Shah, Vice President of Marketing at LinkedIn.

"If your brand isn't showing up in LLMs, you're not just missing awareness, you're missing the moment of decision. That's why being discoverable, credible and consistently cited isn't a nice-to-have, it's what defines your buyability and determines whether your products or solutions get considered and ultimately get bought," Shah said.

Meltwater said it analyses 1.3 billion pieces of content each day and has 27,000 customers worldwide. It positioned the research as a measure of how AI systems choose sources, with a particular focus on the types of content that gain visibility in business-related searches.