AI Search
LinkedIn forms cross-functional AI Search Taskforce after B2B traffic plummeted 60%, releasing framework that replaces "rank, click, visit, convert" model with new paradigm.
LinkedIn today revealed how the platform fundamentally restructured its marketing operations after B2B non-brand traffic collapsed up to 60% across awareness-driven topics, announcing a comprehensive AI search optimization framework that abandons traditional SEO measurement in favor of visibility-based metrics. The disclosure represents the most detailed account from a major platform about adapting to AI-powered search while traffic declined despite stable rankings.
“AI is rewriting the rules of discovery – and with it, how brands show up,” according to Inna Meklin, Director of Digital Marketing at LinkedIn, and Cassie Dell, Group Manager of Organic Growth, who co-authored the announcement on January 28. The platform documented how “our established playbook wasn’t broken but it needed updating” after observing that click-through rates softened while rankings remained stable, creating a measurement crisis where traditional success metrics no longer correlated with business outcomes.
The traffic decline LinkedIn experienced aligns with broader industry patterns affecting B2B channels. A 2024 SparkToro/Similarweb study estimated approximately 60% of searches in the United States and European Union end without clicks, with more recent reporting in 2025 showing that number rising. LinkedIn’s internal data revealed that “across the industry, non-brand, awareness-driven traffic declined by up to 60% across a subset of B2B topics” as Google AI Overviews and ChatGPT delivered answers instantly.
The fundamental transformation LinkedIn documented centers on where buyers conduct research. “Today, discovery no longer manifests on websites alone,” according to the announcement. “Increasingly, it takes place in AI-generated environments – inside AI answers, summaries, and assistants – often before or without a click occurring.” This pattern breaks the traditional chain that defined digital marketing success: rank, click, visit, convert.
LinkedIn’s B2B Organic Growth team began hands-on research into Google’s SGE – the beta version of AI Overviews – in early 2024. By early 2025, these changes began materially impacting traffic with immediate effects. The platform observed that “traffic, which was once a marker of SEO success, was disappearing as Google AI Overviews and ChatGPT delivered answers instantly.”
The shift created what LinkedIn characterizes as a “dark funnel” measurement problem. While analytics for LinkedIn’s B2B marketing websites revealed triple-digit growth in LLM-driven traffic with trackable conversions, the platform “simply couldn’t quantify how visibility within LLM responses impacts the bottom line.” This measurement gap meant LinkedIn’s down-funnel performance held early thanks to strong brand health, but the platform recognized this wouldn’t sustain growth without increased visibility inside generative systems.
LinkedIn’s response involved forming an AI Search Taskforce uniting SEO, PR, Editorial, Web Marketing, Product Marketing, Product, Partner Marketing, Social, Paid Media, and Brand teams. “Together, we aligned on a unified strategy to accelerate content activation across owned, earned, and social surfaces,” according to the announcement. This cross-functional structure represented a departure from traditional SEO operations where optimization typically remains siloed within search teams.
The taskforce methodology focused on three critical areas. Deep landscape immersion built networks of external and internal experts to understand ecosystem changes through conversations with partners, agencies, industry peers, and insights from SEO and GEO/AEO thought leaders posting on LinkedIn. Cross-functional activation identified priority topics using AI visibility software, SEO data, demand signals, and LLM insights while correcting misinformation surfacing in AI responses.
New measurement exploration replaced traditional KPIs that couldn’t capture impact inside AI answers. “We began building new frameworks centered on visibility, mentions, and citations – not just traffic,” according to LinkedIn. This work included using AI visibility software to understand brand appearances in LLMs across branded and non-branded contexts, defining new KPIs focused on visibility rate and citation share, updating internal analytics to create new traffic sources for LLM-driven traffic, and monitoring LLM bot traffic through CMS logs.
The taskforce executed specific tactical initiatives including identifying priority topics, correcting misinformation surfacing in AI responses, expanding depth and clarity on high-performing content, launching new owned content optimized for generative visibility, and publishing articles on LinkedIn to validate social content strength. “By uniting content owners and enablers, we created consistent, reinforcing signals across the entire ecosystem,” according to the platform.
LinkedIn’s early testing showed “meaningful lift in visibility and citations across the topics we focused on, with owned content delivering the fastest and most scalable gains so far.” The platform identified specific content-level patterns that influenced AI visibility. Headings and information hierarchy proved critical, as “the more structured and logical your content is, the easier it is for LLMs to understand and surface.”
Content accessibility and semantic markup improved AI readability, with clear HTML structure and accessible text helping large language models correctly interpret section purposes. The platform documented an early mover advantage where “building credibility quickly creates algorithmic stability with LLMs that make it harder for competitors to overtake position.”
Publishing authoritative, fresh content improved visibility according to LinkedIn’s testing. “LLMs favor content that signals credibility and relevance, authored by real experts, clearly time-stamped, and written in a conversational, insight-driven style on platforms like LinkedIn,” the announcement states. This finding emphasizes content attributes beyond traditional optimization factors like keyword density or backlink profiles.
Strategic learnings extended beyond individual content optimization. Cross-functional content consistency emerged as critical, with “LLM visibility requires alignment across every team producing content.” Shared signals across all surfaces help amplify authority in ways that isolated optimization efforts cannot achieve. Attribution challenges intensified as “discovery shifts into AI-driven experiences, tracking impact becomes more complex.” LinkedIn emphasized that integrating internal analytics with third-party visibility tools proves essential for measuring influence.
LinkedIn articulated a fundamental shift in how platforms should conceptualize marketing success. “We are moving away from ‘search, click, website’ thinking toward a new model: Be seen, be mentioned, be considered, be chosen,” according to the announcement. This framework acknowledges that traditional metrics tracking website visits undervalue visibility wins occurring inside AI-generated responses where brands receive mentions without corresponding click-throughs.
The measurement transformation extends to KPI definitions. LinkedIn recommends establishing metrics including LLM referral traffic, citation and mentions volume, and AI Overview win rate. “Start defining KPIs now – like LLM referral traffic, citation and mentions volume, or AI Overview win rate – so you can track influence as generative discovery evolves,” the platform advised.
This measurement philosophy addresses challenges documented across the AI search optimization industry. Traditional dashboards tracking only visits risk undervaluing visibility wins or hiding looming losses as discovery increasingly occurs inside AI environments before users reach websites. The pattern creates situations where brands lose market share despite maintaining strong traditional SEO performance.
LinkedIn’s visibility focus reflects broader B2B marketing transformations documented in recent research. The platform’s December 2025 analysis argued that B2B brands must shift investment from “rented prominence” through paid advertising toward “owned prominence” built through brand memory and distinctive assets, as AI-driven discovery increasingly mediates brand recommendations without displaying paid advertisements.
LinkedIn released detailed technical guidance on February 10 titled “How to Optimize Your Owned Content for AI Search,” providing specific implementation strategies for the frameworks developed through taskforce testing. According to Brooke Weller, AEO/GEO Consultant at LinkedIn, the guide outlines “concrete strategies we leveraged to boost visibility” developed while treating “our content strategy like a lab to understand how Large Language Models build trust.”
The 16-page guide introduces what LinkedIn characterizes as an “answer-first” approach replacing “keyword-first” strategies. “Traditional SEO isn’t enough. You must now ensure AI systems Trust your content AND Surface your content,” according to the guide announcement. This methodology requires content creators to structure information so each section stands alone when extracted by AI systems, which break content into passages rather than evaluating entire pages.
The guide provides 13 specific action areas spanning metadata optimization, content structure, visual elements, and technical implementation. Answer-block formatting represents the core structural recommendation, with content beginning major sections with 30-80 word summary blocks directly addressing main topics. Schema markup implementation, particularly FAQPage or QAPage schema combined with Article schema, emerged as “currently the highest-impact markup for increasing LLM discoverability in our internal testing.”
Technical requirements emphasize clarity over creativity, recommending sentences under 20 words and reading levels between Grade 9-11 for maximum AI comprehension. Heading hierarchy requires descriptive labels progressing from broad to specific in predictable order. Natural language optimization extends to metadata and URLs, with titles reflecting how users ask questions and meta descriptions functioning as short direct answers between 140-160 characters.
Visual content receives detailed treatment addressing images, videos, and PDFs. Images require relevant captions, descriptive filenames, and specific alt text. Video optimization requires embedding YouTube content with transcripts and structured data, optimizing titles to reflect content accurately, and minimizing page load impact. PDF optimization emphasizes machine-readable text, descriptive metadata, and providing HTML versions for core content.
The announcement revealed LinkedIn’s strategic positioning for AI visibility beyond owned web properties. “With LinkedIn already the #2 most-cited source, we’re seeing strong signals and significant upside at scale,” according to the platform. This citation rate creates opportunities for brands using LinkedIn as a content distribution channel to increase presence in AI-generated responses.
LinkedIn indicated forthcoming guidance on leveraging the platform specifically for AI visibility. “Next, we’ll go deeper on our own story, and the unique advantage LinkedIn brings to AI discovery,” the announcement states. “We’ll share how we’re using LinkedIn as a social channel, alongside broader social activation and content partnerships, to strengthen AI visibility and improve GEO performance.”
The platform emphasized its approach differs from theoretical discussions by focusing on data-driven results. “This post is part of our series on AI search and LLMs. While others theorize, we’re sharing data-driven results,” according to the announcement. This positioning contrasts with concerns raised by industry observers about AI search optimization tactics that cross from legitimate optimization into manipulation.
The strategic principles guiding LinkedIn’s ongoing work include staying agile by avoiding fixed strategies to enable pivoting as platforms evolve, testing and refining through expanding pilots across new surfaces, and keeping pulse on industry developments by tracking new platform launches and partner insights. “AI-led discovery is moving fast – and so are we,” according to the platform.
LinkedIn’s disclosure arrives amid significant competitive activity in AI search optimization. Multiple platforms launched specialized tools throughout 2025 including Amplitude’s AI Visibility tool in November, Zeta Global’s Generative Engine Optimization solution in September, and Similarweb’s dual tracking platform in July.
However, LinkedIn’s approach emphasizes organizational transformation over tool adoption. The cross-functional taskforce structure and new measurement frameworks represent systematic changes to how marketing teams operate rather than incremental adjustments to existing processes. This organizational focus distinguishes LinkedIn’s methodology from vendors promoting specialized software solutions.
The platform’s emphasis on owned content optimization also contrasts with manipulation tactics documented by investigations into generative engine optimization. LinkedIn’s framework centers on content quality, structure, and credibility signals rather than techniques specifically designed to game AI systems through planted brand authority statements or coordinated cross-website promotion schemes.
Research supporting LinkedIn’s visibility-first approach includes studies showing AI search visitors converting at rates 4.4 times higher than traditional organic traffic despite representing smaller traffic volumes. This conversion advantage stems from AI platforms pre-qualifying visitors through detailed information before generating click-throughs, potentially reducing bounce rates and improving match quality between intent and landing page content.
The transformation LinkedIn documented reflects fundamental shifts in how B2B buyers conduct research. Analysis of complete customer journeys shows typical B2B purchasing spanning 211 days from first touch to closed deal, involving 6.8 buyer stakeholders across 3.7 channels, and requiring 76 touches before completion. AI search optimization addresses these extended decision processes by ensuring visibility during early research phases when buyers form consideration sets.
The measurement transformation LinkedIn described addresses challenges extending beyond traffic metrics to fundamental questions about marketing attribution. “While analytics for our LinkedIn B2B marketing websites revealed triple-digit growth in LLM-driven traffic – and we can track how those visitors convert – the real challenge lies in the ‘dark’ funnel,” according to the announcement.
This dark funnel represents interactions occurring inside AI environments where brands gain visibility through mentions and citations without direct website visits. Traditional attribution models fail to capture this influence because no trackable event occurs when users read brand information synthesized into AI-generated responses. The pattern creates situations where marketing investments drive business outcomes through mechanisms invisible to standard analytics platforms.
LinkedIn’s response involves integrating multiple data sources including AI visibility software understanding brand appearances in LLMs, established SEO tools tracking traditional metrics, and internal analytics monitoring LLM-driven traffic sources. The platform also works to monitor LLM bot traffic and behavior through CMS logs, providing technical visibility into how AI systems access and process content.
The new KPI framework LinkedIn developed focuses on visibility rate, citation share, mentions across AI Overviews, and LLM citations rather than traditional metrics like organic sessions or bounce rates. “Traditional KPIs couldn’t capture impact inside AI answers, so we began building new frameworks centered on visibility, mentions, and citations – not just traffic,” the platform stated.
LinkedIn’s announcement emphasized ongoing adaptation rather than presenting finished solutions. “Our focus moving forward will be guided by the following principles: Staying agile. We are avoiding fixed strategies so we can pivot as platforms evolve,” according to the platform. This approach acknowledges that AI search represents rapidly moving territory where best practices continue developing.
The platform committed to expanding pilots across new surfaces to understand what drives AI visibility, including how content on LinkedIn itself shows up within AI-led discovery. This testing approach mirrors the methodology LinkedIn applied to owned web properties, treating the social platform as another optimization surface requiring systematic experimentation.
LinkedIn also indicated plans to track new platform launches, data trends, and partner insights to anticipate shifts before they happen. This environmental scanning reflects recognition that AI search represents an emerging ecosystem where new platforms and capabilities continue launching. Recent developments including Google’s agent-driven search transformation demonstrate the rapid pace of change requiring continuous monitoring.
The mental model shift LinkedIn articulated extends beyond tactical optimization to strategic positioning. “We are moving away from ‘search, click, website’ thinking toward a new model: Be seen, be mentioned, be considered, be chosen,” the platform stated. This framework acknowledges that traditional conversion funnels no longer accurately represent how buyers progress from awareness to purchase when discovery increasingly occurs inside AI environments.
Who: LinkedIn’s marketing organization led by Inna Meklin, Director of Digital Marketing, and Cassie Dell, Group Manager of Organic Growth, forming cross-functional AI Search Taskforce uniting SEO, PR, Editorial, Web Marketing, Product Marketing, Product, Partner Marketing, Social, Paid Media, and Brand teams to address fundamental shifts in discovery patterns.
What: A comprehensive strategic transformation abandoning traditional SEO playbooks after B2B non-brand traffic declined up to 60% across awareness-driven topics, implementing new measurement frameworks centered on visibility, mentions, and citations rather than traffic, and releasing detailed technical guidance titled “How to Optimize Your Owned Content for AI Search” providing 13 specific action areas for brands adapting to AI-powered discovery environments.
When: The transformation began in early 2024 when LinkedIn’s B2B Organic Growth team started researching Google’s SGE beta, accelerated through early 2025 as traffic impacts materialized, and culminated with detailed methodology disclosure on January 28, 2026, followed by comprehensive optimization guide release on February 10, 2026, representing approximately two years of testing, learning, and framework development.
Where: The changes affect LinkedIn’s owned web properties including product pages, company information, and blog content, with forthcoming guidance addressing how content on LinkedIn’s social platform shows up in AI-led discovery, leveraging the platform’s position as the second-most-cited source in large language model responses to strengthen generative engine optimization performance.
Why: Traditional success metrics including rankings and click-through rates decoupled as discovery shifted into AI-generated environments where buyers encounter brands inside AI answers, summaries, and assistants before or without clicks occurring, creating a “dark funnel” measurement problem where visibility within LLM responses impacts business outcomes through mechanisms invisible to standard analytics, forcing fundamental reconceptualization of how marketing success manifests when websites no longer serve as primary engines of discovery.
Your go-to source for digital marketing news. Get the latest updates from Google, Meta, Amazon, and The Trade Desk. Stay informed on ad tech innovations, programmatic trends, and policy changes.
