Schema.org Shifts to AI Data Integrity Amid Google Changes

Image credit: Search Engine Journal

Schema.org’s role in technical search engine optimization (SEO) is evolving from primarily generating Google‘s rich results to a foundational layer for data integrity, particularly for large language models (LLMs), as Google has deprecated support for nine ItemTypes from its rich result search gallery in the past two years.

This shift has prompted industry debate over schema’s future utility and is redefining its importance for web content comprehension by artificial intelligence.

Google’s removal of FAQ rich results, for instance, was a display feature deprecation, not an elimination of the underlying FAQ schema, which continues to serve as a comprehension layer for identifying entities and relationships within web content, according to industry experts.

Schema’s significance now extends beyond Google’s traditional index pipeline to include LLM pretraining and runtime retrieval, contributing to data integrity and combating ambiguity that can lead to AI hallucinations, said Gianluca Fiorelli, an SEO consultant.

Data integrity in SEO encompasses five distinct layers: Entities, Relationships, Format, Actions, and Perception, each playing a role in how various agents understand and interact with web content, according to Andrea Volpini, CEO of WordLift.

Chiara Carrozza, a data scientist, emphasized that these layers are important for establishing trust and accuracy in the information consumed by AI systems.

New protocols and standards are emerging to address the aggregation, guidance, and consumption of web data for agents, but a unified consensus or agreed standard, similar to Schema.org’s origins, remains absent, Suganthan Mohanadasan of Keyword Insights reported.

Organizations like Wikidata, GS1, ISNI, and ORCID are contributing to these efforts by providing structured data and identifiers, further highlighting the need for data integrity.

Emina Demiri-Watson, a product manager at Yoast SEO, noted that ensuring data is properly structured and validated is becoming paramount for SEO professionals.

The evolving situation requires technical SEO practitioners to focus more on the underlying semantic structure of content rather than solely on the visual presentation in search results, according to Search Engine Journal.

Source: Search Engine Journal