Review schema, structured data, and how to earn star ratings in Google search results.
Start with the pillar guide: Review Schema & Structured Data
ChatGPT, Perplexity, and Google AI Overviews all crawl the web differently than a human reader does. What is actually documented about how they retrieve pages, and what our crawl of 985 SaaS homepages shows about who is even offering them structured facts to retrieve.
Hand-coded JSON-LD, general-purpose schema generators, and Google's own tools all produce review markup that looks right and still fails. Here is what to check before you trust any generator's output.
Review snippets vanish for six recurring reasons, and only two of them are markup bugs. How to date the loss in Search Console, work out which cause hit you, and what recovery actually takes, backed by our crawl of 985 SaaS homepages.
The Rich Results Test tells you whether Google can read your structured data, not whether Google will show it. How to read its output, the four failures it never reports, and the debugging order that clears most schema problems, backed by our crawl of 985 SaaS homepages.
Google stopped showing FAQ rich results on May 7, 2026, breadcrumbs still render, and star ratings depend on who controls the reviews. Our July 2026 crawl of 985 SaaS homepages shows which markup SaaS sites actually ship, and how often they get it wrong.
A property-by-property reference for SoftwareApplication JSON-LD: the fields Google actually reads, the strict rules on offers and aggregateRating, the self-serving ceiling on your own domain, and a full valid example.
A hands-on walkthrough for adding AggregateRating JSON-LD: choosing the parent type, required and recommended properties, validation, the errors that break rich results, and the self-serving trap to avoid.
Star ratings in Google come from review schema, but Google rejects self-serving markup, so most sites do it wrong. What review snippets require, the self-serving rule, and the setups that actually earn stars.
Three syntaxes can express the same schema.org data, and JSON-LD won. Why Google recommends it, what microdata and RDFa still look like in the wild, and the few cases where the older formats linger.
