AI Search Visibility: Does Structured Data Fall Short?

New tests reveal that structured data does not improve AI search visibility for web pages. Despite claims within the SEO community, recent experiments show that schema markup provides no advantage in how AI models process content, challenging previous expectations of enhanced rankings.

Tests Reveal Limits of Structured Data in AI Search

AI Search Visibility

On September 15, 2025, SEO expert Mark Williams-Cook shared a test on LinkedIn, debunking the idea that structured data enhances AI search visibility. He explained that large language models (LLMs) like ChatGPT and Gemini tokenize content, breaking schema like “@type”: “Organization” into separate tokens. This “destroys” the schema’s structure, making it no different from regular text.

Also Read: Top Voice Search Statistics: Speaking to the Future!

As a result, LLMs ignore schema in their training data. Julio C. Guevara’s experiment further confirmed this. He created two product pages—one with visible text and schema, another with only schema. LLMs only pulled details like price or SKU from the text-heavy page, ignoring the schema-only page. These findings highlight that AI search visibility relies on visible text, not structured data. While this could change, current tests show no SEO benefit from schema in AI searches. Discussions on LinkedIn continue to explore this topic.

How to Test Structured Data for AI Search Visibility

  • Set Up Test Pages: Create two pages for a fictional product—one with text and schema, one with only schema.
  • Query LLMs: Use prompts to ask Gemini or ChatGPT for product details like price or colors.
  • Compare Outputs: Check if LLMs extract data from both pages or only the text-based one.
  • Examine Tokenization: Analyze how LLMs split schema into tokens to understand their limitations.
  • Stay Informed: Follow SEO forums on LinkedIn for updates on AI search visibility trends.

These tests prove structured data doesn’t yet boost AI search visibility. SEO professionals should prioritize visible text to improve AI search performance. Join the LinkedIn conversation to stay updated on future changes in AI search dynamics.

More News To Read: OpenAI Grove Opens Doors for Aspiring AI Innovators

Perplexity Growth Strategy 2025: Vital Tactics Shared

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top