New benchmark compares ChatGPT, Perplexity, and Google AI Overviews citation behavior — structured data dominates shopping queries

A comprehensive multi-engine citation benchmark published on July 29, 2026, compared how ChatGPT Search, Perplexity, and Google AI Overviews cite ecommerce product pages across six product categories. The study, based on 1,200 product pages from Shopify and WooCommerce stores selling in US, EU, and APAC markets, provides the clearest cross-engine comparison of citation behavior to date.

The research team selected 1,200 product pages with varying levels of structured data completeness — from pages with no structured data to pages with full Product, Offer, ShippingDetails, and aggregateRating schema. Each page was submitted as a candidate source for 15 standardized shopping queries per category, and citation outcomes were tracked over a 14-day period in July 2026.

Structured data completeness is the top predictor across all engines. Pages with Product, Offer, and ShippingDetails schema were cited at 3.2× the rate of pages with only Product schema. Pages with no structured data were effectively invisible across all three engines — fewer than 2% received any citations.

ShippingDetails schema is the biggest differentiator. Among pages with otherwise similar product attributes, those with ShippingDetails schema showing at least three destination markets were cited 2.1× more frequently across all engines. Perplexity showed the strongest preference for shipping information — pages with shipping schema were 2.8× more likely to be cited than those without.

Google AI Overviews favors aggregateRating more than other engines. Pages with aggregateRating schema and 50+ reviews were cited by Google AI Overviews at 2.4× the rate of pages without structured review data. ChatGPT Search showed a preference for pages with brand entity consistency (1.8× lift), while Perplexity favored shipping and delivery information.

FAQPage schema provides consistent lift across engines. Pages with 5+ FAQPage entries had roughly 70% higher citation frequency than pages without structured FAQ content. The lift was consistent across all three engines, suggesting FAQPage is the single most portable structured data investment for multi-engine visibility.

Entity consistency matters most for ChatGPT Search. Pages where product names matched across title, H1, and Product schema name had a 1.9× citation advantage in ChatGPT Search queries. The same factor produced only a 1.3× lift in Perplexity and a 1.4× lift in Google AI Overviews.

Cross-border stores face a structural gap. Only 18% of cross-border product pages had ShippingDetails schema covering 3+ destination markets — yet those that did received citations at 2.5× the rate of the overall sample.

The benchmark confirms what Visora's audit data has been showing since early 2026: structured data completeness is not a nice-to-have for AI search visibility — it is the primary eligibility gate. A merchant with Product, Offer, ShippingDetails, aggregateRating, and FAQPage schema on their hero pages has a citation rate approximately 4× higher than a merchant with only Product schema, regardless of domain authority.

Visora's citation monitoring platform tracks citations across ChatGPT Search, Perplexity, Google AI Overviews, and Bing Copilot. The benchmark data aligns with patterns observed across Visora's merchant cohort: stores that complete their schema stack see citation lift within 2-4 weeks, with ShippingDetails and FAQPage delivering the fastest returns. Merchants can benchmark their own pages against the study's findings by running a free audit at geovisora.com/audit and comparing their structured data completeness scores against the study's 75-point threshold for frequent citation inclusion.