We Tracked 1,885 Pages Adding Schema. AI Citations Barely Moved.

05/11/26
Source: Data & Studies – SEO Blog by Ahrefs by Louise Linehan. Read the original article

TL;DR Summary of Does Adding JSON-LD Schema Markup Increase AI Citations?

Our extensive study reveals that adding JSON-LD schema markup to already AI-cited web pages does not significantly increase citations by AI systems like Google AI Mode or ChatGPT. While schema is more common on AI-cited pages, its presence likely correlates with other strong SEO factors rather than causing citation boosts. A small but statistically significant decline in AI Overview citations was observed, though the cause remains unclear. Overall, schema alone is not a reliable lever for improving AI visibility on pages already recognized by AI.

Optimixed’s Overview: Evaluating the True Impact of JSON-LD Schema on AI Citation Growth

Comprehensive Analysis of Schema Markup Influence on AI Citing Behavior

To determine if adding JSON-LD schema markup directly influences AI citation frequency, a large-scale study tracked nearly 2,000 web pages that introduced schema between August 2025 and March 2026, comparing their citation trends to matched control pages without schema additions. The AI citation sources analyzed included Google AI Overviews, Google AI Mode, and ChatGPT.

Key Findings

  • No Significant Citation Gain: Pages adding JSON-LD showed no meaningful uplift in citations from Google AI Mode (+2.4%) or ChatGPT (+2.2%), both statistically indistinguishable from zero.
  • Small AI Overview Decline: A 4.6% statistically significant but small decline occurred in AI Overview citations post-schema addition, though reasons remain speculative and may relate to platform-wide content shifts or recrawl timing.
  • Correlation vs. Causation: Schema markup is prevalent on well-optimized sites that naturally attract AI citations due to strong content, authority, and technical SEO — making schema a correlated rather than causative factor in AI citation frequency.
  • AI Systems and Schema Usage: Experimental evidence shows major AI systems do not currently parse or use JSON-LD schema data during real-time content retrieval, relying instead on visible HTML content.

Methodology Insights

The study employed a matched difference-in-differences (DiD) statistical approach, controlling for platform-wide citation trends and pre-existing citation levels to isolate the effect of schema:

  • Each treated page (adding JSON-LD) was matched with three control pages from different domains with similar citation histories but no schema addition.
  • Citations were tracked 30 days before and after the schema introduction to detect changes beyond general AI platform trends.
  • Multiple validation tests confirmed the consistency of findings across different statistical models and time windows.

Practical Recommendations for SEOs and Marketers

  • Don’t Rely Solely on Schema for AI Visibility: Adding JSON-LD to pages already cited by AI is unlikely to boost citations significantly. Focus on comprehensive SEO strategies including content quality, authority building, and technical optimization.
  • Consider Running Controlled Tests: Use tools like Brand Radar to track AI citations on test pages with schema additions and matched controls to measure impact specific to your site.
  • Schema’s Other Benefits Remain Valuable: Despite limited AI citation impact, schema supports rich search results, voice assistant integration, knowledge graphs, and downstream entity recognition.
  • Further Research Needed: Future studies could explore effects of specific schema types, longer post-addition observation periods, and differences between schema formats or injection methods.

Conclusion

While schema markup correlates with AI-cited pages, it does not independently drive increased AI citations on pages already visible to AI systems. Marketers should view JSON-LD as one element within a broader SEO ecosystem rather than a standalone solution for AI content visibility enhancement.