Why LLM-only pages aren’t the answer to AI search

Llm-Only Pages: a Costly Seo Misstep

Introduction to LLM Content Strategies

Companies pursue new tactics to boost AI-driven search visibility, implementing LLM-only pages. These include llms.txt files, markdown (.md) versions, and dedicated /ai/ directories. The premise: provide machine-readable content by eliminating distractions like ads and navigation. But how effective are these strategies in driving citations?

Expert Opinions on LLM Pages

Industry leaders express skepticism. Google’s John Mueller questions the relevance of LLM-only content, comparing it to outdated keyword meta tags. He states that LLMs are trained on typical web pages, rendering user-invisible formats largely irrelevant. Despite ongoing trials, no clear benefits emerge from these specialized formats.

Examining Citation Effectiveness

Testing by Malte Landwehr across 18,000 citations reveals dismal results for LLM-specific pages. llms.txt files garnered a mere 0.03% citation rate. Markdown pages? Zero. The only exception lies with sites like GitHub, where .md files function as standard URLs. In contrast, JSON metadata files achieved a 5% citation rate, primarily when they contained unique information.

Key Findings on Citation Rates

  • llms.txt: 0.03% of citations
  • Markdown pages: 0% of citations
  • /ai pages: 0.5% to 16% of citations, depending on content richness
  • JSON metadata: 5% of citations when unique

Limitations of LLMs in SEO

LLMs carry inherent limitations: hallucinations, context constraints, and retrieval fragility. Attempting to improve AI accessibility with optimized formats does not solve these fundamental issues. Clean formats fail to address semantic drift or the contextual understanding necessary for meaningful AI interactions.

Implications for Future Strategies

Instead of focusing on LLM-only pages, businesses should prioritize high-quality content on standard web pages and utilize structured data like schema.org. The future lies in hybrid approaches that combine traditional content strategies with AI capabilities. Relying solely on format gimmicks detracts from the real goal: producing valuable content that resonates with both users and AI systems.

Looking Ahead

Over the next 6–12 months, expect a shift away from LLM-only formats as SEO professionals recognize their limitations. The focus will return to creating robust, user-centric content that addresses the nuances of search algorithms, ensuring better visibility and engagement.

Post List #3

Google Shopping API cutoff looms, putting ad delivery at risk

Google Shopping API Migration: a Looming Deadline for Advertisers

Marc LaClear Jan 20, 2026 3 min read

Migration Deadline Approaches Google’s transition from the legacy Content API for Shopping to the new Merchant API is not just another update; it’s a hard cutoff that threatens ad delivery for unprepared advertisers. The fully deprecated Content API will cease…

Google’s Mueller: Free Subdomain Hosting Makes SEO Harder via @sejournal, @MattGSouthern

Google’s Mueller Warns: Free Subdomain Hosting Complicates Seo Efforts

Marc LaClear Jan 19, 2026 3 min read

Understanding the Warning John Mueller, a Google Search Advocate, recently highlighted the pitfalls of using free subdomain hosting during a Reddit discussion. His insights reveal that even well-optimized sites can struggle for visibility due to the spammy environments these services…

Google On Phantom Noindex Errors In Search Console via @sejournal, @martinibuster

Decoding Google’s Phantom Noindex Errors in Search Console

Marc LaClear Jan 18, 2026 3 min read

Overview of Phantom Noindex Errors Google’s John Mueller recently discussed phantom noindex errors that appear in Search Console. These errors indicate a discrepancy where a page marked for indexing is simultaneously flagged with a noindex directive that site owners can’t…

Daily Search Forum Recap: January 15, 2026

Google’s Latest AI Moves and Their Impact on Seo

Marc LaClear Jan 15, 2026 2 min read

New Features from Google: Personal Intelligence and Gemini Integration Google has begun rolling out Personal Intelligence within its Gemini app, which aggregates user data from Gmail, Photos, YouTube, and Search history. This feature aims to enhance user experience by delivering…

Microsoft adds new customer acquisition goals and deeper visibility to PMax

Microsoft’s Pmax Update: New Goals and Visibility for Targeted Customer…

Marc LaClear Jan 15, 2026 3 min read

New Customer Acquisition Goals Microsoft has expanded its Performance Max (PMax) capabilities by adding specific customer acquisition goals. This change allows advertisers to optimize campaigns not just for conversions, but explicitly for acquiring new customers. By leveraging AI, PMax can…