Schema checker: find and fix structured data errors

What is schema markup, and why check it?

Schema markup is code that describes the things on a page, such as an organisation, product or article, using the Schema.org vocabulary. Google says it uses structured data to understand the content of a page and to show some rich results, and it recommends JSON-LD as the format.

Broken markup is ignored, and wrong markup is worse: it states facts that disagree with your page. That matters for search engines and for AI answer engines that read your pages, because a clear, consistent description of who you are and what you sell is easier to quote accurately.

Be careful with the promise, though. Google says structured data isn't required for generative AI search, and that there's no special markup you need to add. It also says it's a good idea to keep using structured data as part of your overall SEO. Correct schema is good housekeeping, not a shortcut into AI answers.

What does the schema checker test?

The checker fetches the page as your server sends it, finds each structured data block and runs four checks, reported separately.

Can the markup be read?

Missing commas, unclosed brackets and other syntax errors stop a parser reading a JSON-LD block. The checker shows the error with its line and column, and you can open "Show the markup we found" to see the exact code. Where it can repair the block (missing or trailing commas, curly quotes, comments, line breaks inside text), the other checks run on the repaired version and the corrected block uses it. Where it can't, it points you to the line to fix by hand. Other blocks on the page are still tested.

Are the types and properties real?

Each type and property is checked against the Schema.org vocabulary. Names are case-sensitive, so a misspelt type or property is an error, not a style choice. Where there's one clear intended name, the checker suggests it. A real property used on a type it doesn't belong to is a warning.

Does it meet Google's requirements?

The checker tests required and recommended properties for 15 Google features: Article, Breadcrumb, Event, FAQ, How-to, Job posting, Local business, Organization, Product snippet, Profile page, Recipe, Review snippet, Site name, Software app and Video. For example, it raises an error for a Product with none of offers, review or aggregateRating, and a warning for a Product with no brand. Missing optional properties, such as geo, openingHoursSpecification or priceRange on a local business, are shown as information notes.

Does the markup match the page?

Google's guidelines say not to mark up content that isn't visible to readers. On a URL check, the checker looks for FAQ questions and answers on the page, compares an article's headline with the page, and looks for marked-up prices in the page text. It also flags dates that aren't valid, relative URLs, and the same organisation described twice under different names.

Is there anything Google no longer uses?

The checker notes markup that no longer brings a Google result. For example, Google stopped showing FAQ rich results from 7 May 2026 and how-to rich results in 2023. Both are still valid Schema.org types.

Should I check a URL or paste my code?

Check a URL when the page is live, because themes, plugins and caching can change what is actually served. It's also the only way to run the check against the page: pasted code has no page to compare with, so that result shows "Needs a URL".

Paste code to test markup before you publish it. You can paste JSON-LD on its own, or HTML that contains it, up to 1 MB. The limit is 30 checks an hour.

What do the results mean?

Results are grouped by test. Each finding names the block, the property and the rule behind it, so you can see why it was raised.

Worked example

We ran the checker on a product page on a test site we set up, shown here as example.com. The page shows a price of £49, but its JSON-LD says 39. The JSON is valid and the vocabulary passes, so the problem only shows up in the check against the page:

Warning: The price 39 isn't shown on the page
Google expects the marked-up price to match the price visitors see.

The fix is to change that one value and leave the rest of the block alone:

"offers": { "@type": "Offer", "price": "49", "priceCurrency": "GBP" }

The corrected block doesn't change this for you, because only you know which price is right.

Which schema problems should I fix first?

1. Blocks that can't be read. Nothing else in a broken block counts.
2. Facts that disagree with the page, such as price, address or opening hours.
3. Duplicate entities. A theme and an SEO plugin can each output their own Organization block with different names or logos. Choose one source and switch the other off.
4. Missing required properties for a Google feature you want.
5. Optional properties, once you have the facts to support them.

What can't the schema checker tell you?

• It does not run JavaScript. If a tag manager or script adds your markup after the page loads, this checker won't see it. Use the Schema Markup Validator, which can extract structured data injected by JavaScript.
• It checks one page at a time. Use a site crawler to find the same issue across templates.
• It is not a Google validation service. To test a Google feature, use the Rich Results Test.
• Its check against the page looks for text, so it can miss a mismatch that's worded differently, and it can't tell you which of two conflicting facts is right.
• Valid markup does not guarantee a rich result or an AI citation.

How does this compare with other schema checkers?

Most sites use more than one tool. This summary uses each provider's own description, reviewed on 5 November 2026.