Schema markup generator: build JSON-LD you can trust

What is schema markup?

Schema markup is code that describes the things on a page, such as your organisation, a product or an event, in a shared vocabulary called Schema.org. JSON-LD is a way to write that code in a single block. Google recommends JSON-LD as the easiest solution to implement and maintain at scale.

Why does schema markup matter for Google and AI answers?

Google says it uses structured data to understand the content of a page and to show some rich results. For your organisation, markup such as name, logo, contact details and official profiles states the facts in one clear place.

That clarity also helps any system that reads your pages, including AI answer engines. But don't over-promise: Google says structured data isn't required for generative AI search. Good markup supports clear pages. It doesn't replace them.

What can the generator build?

Organisation, local business, article or blog post, product, FAQ page, breadcrumbs, event and person. Each type shows the properties that matter most for it, and marks the ones you need as Required.

Organisation

Name and website URL are required. Add your logo, description, email, telephone and address. Put each official profile on its own line, such as your LinkedIn or Companies House page. These become sameAs links, which tell readers and machines that those profiles are yours.

FAQ and article

For an FAQ, use "Add a question" for each question and answer that visitors can see on the page. For an article, add the headline, author and dates shown on the page.

Fill from page

Enter a URL and the generator pulls the facts it can find in that page's HTML: the title, main heading, meta description, social sharing image, canonical URL, article dates and author, and any organisation details in structured data the page already has. It doesn't read the rest of the page's text. Each fact it fills is marked "From page". Check every one, because the tool can only find what your server sends.

Copy, download or check

Copy the code with its script tag ready to paste into your page, copy the JSON alone, or download it. "Check this markup" runs our schema checker on the code and shows the results under it, along with any required field you left empty.

How do I read the generated code?

The preview shows one JSON-LD block. Each line is one fact.

Worked example

We filled in the form for a firm called Example Ltd, with its name, website and two profiles. The generator produced:

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Example Ltd",
  "url": "https://www.example.com/",
  "sameAs": [
    "https://www.linkedin.com/company/example",
    "https://find-and-update.company-information.service.gov.uk/company/00000000"
  ]
}
</script>

"Check this markup" then finds no errors and one warning: the recommended logo is missing. The firm adds its logo URL, pastes the block into the head of its home page and checks the live URL.

What should I get right first?

1. Accuracy. Every fact must match what visitors can see. Google's guidelines say not to mark up content that isn't visible to readers.
2. One source. If a plugin or theme already outputs Organization markup, replace it rather than adding a second block.
3. Required fields. The form marks them, but it doesn't stop you copying incomplete markup, so run "Check this markup" first.
4. Official profiles. Only list profiles you control.
5. Extras. Add optional properties when you have the facts.

What can't the generator do?

• It can't know facts that aren't on your page. "Fill from page" reads the HTML your server sends and does not run JavaScript.
• It doesn't publish anything. You paste the code into your site yourself.
• It doesn't guarantee a rich result. Google decides what to show.
• It can't earn you an FAQ rich result. Google stopped showing them from 7 May 2026. Use FAQ markup only for a visible FAQ.

How does this compare with other schema generators?

This summary uses each provider's own description, reviewed on 5 November 2026.