How to Audit Schema Markup: JSON-LD, Entity Types, and Common Errors

Schema markup helps machines understand what a page represents. It can describe a product, organization, article, software application, person, FAQ, website, breadcrumb trail, or another entity. A schema markup audit checks whether that structured data is present, valid, complete, and connected to the visible page.

A page can have schema and still be weak. The markup may parse correctly but describe the wrong entity, miss recommended fields, duplicate conflicting nodes, or use generic types that do not clarify the page.

JSON-LD, Microdata, and RDFa

Most modern sites use JSON-LD because it keeps structured data separate from visible HTML. Microdata and RDFa embed structured data into markup. A good audit checks all three because legacy templates, plugins, and CMS themes can output more than one format.

Start with parse errors

If JSON-LD cannot be parsed, nothing else matters until the syntax is fixed. Common causes include trailing commas, invalid quotes, missing braces, unescaped characters, or dynamically injected markup that renders differently than expected.

• Find every structured data block on the page.
• Check whether each block parses cleanly.
• Inspect raw JSON-LD when a parser reports an error.
• Confirm that rendered HTML includes the schema you expect.

Check entity types

The @type field tells parsers what kind of thing a node represents. A page should use types that match its purpose. A blog post is not the same thing as a product page, a support page, or a software listing.

• Use Article or BlogPosting for editorial content.
• Use Product or SoftwareApplication for product-focused pages.
• Use Organization and WebSite to clarify brand context.
• Use WebPage or CollectionPage to describe the page itself.
• Use FAQPage only when the visible page contains matching FAQ content.

Look for missing recommended fields

Required fields help markup validate. Recommended fields help the entity become more useful. For example, an Article can be stronger with author, datePublished, dateModified, image, headline, and publisher. A WebSite can be clearer with publisher and a meaningful potentialAction.

Review entity relationships

Good schema is a graph, not a pile of disconnected objects. Stable @id values let you connect the WebPage to the WebSite, the Article to the author, the Organization to the logo, and the product page to the product entity.

• Use stable @id URLs with fragments like #website, #organization, or #article.
• Connect author, publisher, and organization nodes instead of repeating disconnected text.
• Make sure structured data matches visible page content.
• Avoid multiple conflicting Organization or Product nodes.

Schema audit checklist

1. Find JSON-LD, Microdata, and RDFa.
2. Fix parse errors first.
3. Confirm each @type matches the page.
4. Add missing recommended fields where useful.
5. Connect entities with stable @id values.
6. Make sure schema content matches visible content.
7. Review Article, Organization, WebSite, WebPage, Product, FAQPage, and SoftwareApplication types where relevant.
8. Retest after template or CMS changes.