Structured data validator

What structured data is

Structured data is a layer of information you add to the page for machines, not for people. The text says “Ergonomic chair, 899 lei, in stock”, and the markup says, in unambiguous language, that this is a product, that the price is 899, the currency RON and the availability “in stock”. The shared vocabulary is called schema.org and is maintained by Google, Microsoft, Yahoo and Yandex.

There are three formats: JSON-LD, a separate code block, recommended by Google and the easiest to maintain; microdata, attributes placed directly on HTML elements; and RDFa, similar to microdata but with different attributes. The validator above reads all three.

What the structured data validator checks

Syntax

The first cause of markup that “does not work” is code that cannot be read at all: an extra comma, typographic quotes inserted by a text editor, a newline in a title, content escaped as HTML entities by a plugin. Google does not notify you in such cases; it ignores the whole block, silently. The validator shows you the line and column and tells you exactly what does not fit.

The schema.org vocabulary

Next it checks whether the types and properties really exist and whether they are used where they belong. A property written “datePublised”, a type “Organisation” instead of “Organization”, a date written “19.08.2026” instead of “2026-08-19”, a price written “1.500,00 lei” instead of “1500.00”: all are mistakes that go unnoticed when reading, but make the data unusable.

Google's requirements

Markup can be perfectly valid for schema.org and, at the same time, useless in search, because Google has its own requirements for each type of rich result. A product without a price and without a rating does not get a product snippet, however correctly it is written. The validator tells you, for each feature, whether you are eligible and what is missing.

Google's policies

Here lie the traps that no syntax validator shows. Reviews a site assigns to itself are not allowed for businesses and organizations. Stars appear only on certain types, not on any page. Marked-up content must be visible to the visitor, not just present in the code.

What no longer has any effect

In recent years Google has retired several rich results: step-by-step instructions in 2023, the sitelinks search box in 2024, and in June 2025 another seven types, among them fact-check and vehicle listings. FAQs remain, but are shown only for government and medical sites. The validator tells you when your markup describes a feature that no longer exists, so you do not wait in vain for a result that is not coming.

What it cannot check

Markup added from JavaScript, through Tag Manager or by a React application, does not appear in the code delivered by the server. The validator flags the situation, but cannot run the page's scripts. If this is your case, open the page in a browser, copy the rendered code and use the pasted-code check.

The second thing no outside tool can do: tell you whether Google will actually show the result. Eligibility is a necessary condition, not a sufficient one. Google chooses what it shows depending on the query, the device and the overall quality of the page.

How to use the result

Start with the syntax errors, because they make the whole block invisible, so nothing else matters until they are fixed. Continue with the missing required properties, then with the warnings. Leave the recommendations for last, when you have time: there is no harm if they stay untouched.

After publishing the corrected markup, request reindexing of the page from Search Console and follow the enhancements report. The data shows up there within a few days, and rich results can take longer.

If you want someone to go through your whole site, not just one page, you can write to us through the contact page. See also the E-E-A-T audit, which checks the site's trust signals.