The schema toolkit reads the structured data on a page, checks it’s valid, checks it matches what a visitor actually sees, and generates correct JSON-LD where it’s missing. Structured data is one of the few parts of SEO with a published specification, so most of the problems are fixable with certainty rather than judgement.
Built in-house. It is my own tooling rather than a reseller’s dashboard, which is why it checks the things I act on rather than everything a generic tool can measure.
What it does
Three jobs. Detection, which finds what markup exists across a site and where. Validation, which checks it against the specification and against Google’s requirements for the features it supports. Generation, which produces the JSON-LD for the types a page should carry.
The fourth job matters most and gets least attention: checking the markup agrees with the page. Structured data describing a price, a rating or an address that doesn’t appear on the page is a problem whether or not it validates.
What it checks
| Check | What it looks at |
|---|---|
| Presence | Which types exist, on which templates, and which pages have none |
| Validity | Required and recommended properties, correct types, valid JSON-LD syntax |
| Agreement | Whether the marked-up values match what’s visible to a visitor |
| Consistency | Whether the business name, address and details match across the site and beyond it |
| Connections | Whether Organization, WebSite, Article, Product and Breadcrumb are linked rather than isolated blocks |
| Retired types | Markup for features search engines no longer show |
Reading it properly
A detail that catches tools out: structured data embedded by a content system often contains raw line breaks inside strings, which makes the block invalid JSON even though it looks fine. The toolkit parses every block properly rather than pattern matching it, because a block that fails to parse is one search engines can’t read either, and that’s a real finding rather than a tooling artefact.
On retired features
Some markup no longer earns anything. FAQ rich results stopped showing in Google’s results in May 2026, and the documentation was later removed. Several other structured data features were retired in June 2025, with more phased out in November 2025.
That doesn’t always mean removing the markup. FAQPage schema is still worth keeping on pages with a visible FAQ, because AI systems read it, but it should be kept for that reason rather than in the belief it will produce a rich result. Knowing which is which is most of the value here.
Example output
TEMPLATE: Product pages PRESENT: Product, Offer, AggregateRating VALID: Yes AGREEMENT: Fails. Offer price is the ex-VAT figure. The page shows the inc-VAT price. MISSING: BreadcrumbList on all product pages. ACTION: Correct the price source, add BreadcrumbList to the template.
What I review
Markup is checked against the rendered page, not just against the validator. A page can pass validation while describing something the visitor never sees, which is the version of this problem that causes trouble.
Where schema is generated, it’s generated to match the page rather than to maximise the number of properties. Adding a rating that isn’t on the page is not an optimisation.
Limitations
- Structured data doesn’t improve rankings on its own. It helps search engines understand a page and makes certain features possible.
- Features change. Search engines retire rich results, so a type that earned something last year may not now.
- Templates vary. A check on one product page is a check on one template, so the sample has to cover each one.
Services it powers
The schema toolkit supports SEO, AI citation readiness and every web development build. See the rest of the tools on the tech stack page.
FAQ
Does schema markup improve rankings?
Not directly. Structured data helps search engines understand what a page is about and makes certain result features possible, which can affect how a listing appears and how often it’s clicked. It doesn’t work as a ranking boost on its own.
Is FAQ schema still worth adding?
For rich results in Google, no. FAQ rich results stopped showing in May 2026 and the documentation was later removed. It’s still worth keeping on pages with a visible FAQ, because AI systems read structured data, but expect no rich result from it.
What’s the most common schema problem you find?
Markup that validates but disagrees with the page, usually a price, a rating or an address that doesn’t match what a visitor sees. Validators don’t catch it because the syntax is correct, so it survives for years until someone compares the two.
See the whole toolkit
The tech stack page lists every tool, and the free assessment includes a structured data check.