Ahrefs Study: Schema Markup Fails to Boost AI Citations

Image credit: Search Engine Journal

A new Ahrefs report indicates that adding schema markup does not directly increase AI citations for web pages already receiving visibility from artificial intelligence systems.

The study challenged the belief that structured data directly enhances a page’s likelihood of being cited by AI Overviews, AI Mode, or ChatGPT, particularly for content already established in AI results.

Ahrefs tracked 1,885 web pages that implemented JSON-LD schema markup and compared their AI citation performance against a control group of similar pages. The research found no significant increase in citations across Google AI Overviews, AI Mode, or ChatGPT after the schema was added.

Google AI Overviews showed a 4.6 percent decline in citations for the treated pages relative to the control group, a statistically notable observation, the report stated.

While pages cited by AI are approximately three times more likely to include JSON-LD, Ahrefs attributed this correlation to overall site quality and authority rather than a direct impact of schema markup itself. The company suggested that higher quality sites are more likely to implement best practices, including schema, and also more likely to be cited by AI.

The study focused on pages already receiving more than 100 AI Overview citations, a limitation acknowledged by Ahrefs. This focus means the research may not reflect the potential impact of schema on pages not yet visible to AI systems.

The report also cited an experiment by searchVIU, which found that five different AI systems did not utilize schema markup when fetching web pages in real time. Instead, these systems reportedly extracted information exclusively from the visible HTML content of the pages.

Ahrefs concluded that for web pages already established and cited by AI, the direct addition of schema markup does not serve as a mechanism to further boost their citation frequency.

Source: Search Engine Journal