Structured Data for AI: A Practical Guide

Structured data is one of those topics that sounds more complicated than it is. At its core, it's just a way to tell machines "this is who we are, this is what this page is about."

We've seen sites add basic Schema.org markup in an afternoon and see immediate improvements in how AI systems represent them. It's not magic, but it removes guesswork.

The quick version

If you're short on time:

• Use JSON-LD (it's what everyone recommends, and for good reason)
• Start with Organization, WebSite, and BreadcrumbList
• Add Article or BlogPosting to your blog posts
• Validate before you ship
• Scan again to see if it helped: Scan your site.

What it actually does (and doesn't)

What structured data can actually do for you:

• Makes things explicit — "this is our company name, this is our logo, this person wrote this"
• Helps with page type — search engines and AI know this is a blog post, not a product page
• Can unlock rich results — those fancy cards in search results? This is how you become eligible
• Reduces AI guesswork — less chance of being misrepresented

What it won't do:

• Magically boost your rankings (Google has said this repeatedly)
• Fix thin content — if your page has nothing to say, schema won't help
• Replace actual internal linking and site structure

New to all this? Start here: What is AI-readiness?.

Why AI systems care about this

Think about it from the AI's perspective. It's processing millions of pages, trying to figure out:

• Who are you? (Organization, Person)
• What's this page about? (Article, Product, FAQ)
• Who said what and when? (author, dates)
• How does this connect to other things? (relationships)

Without structured data, AI has to guess. With it, you're giving explicit answers. We've found that sites with good schema tend to be represented more accurately in AI responses—fewer "hallucinated" company descriptions, correct product names, right dates.

JSON-LD: the format you should use

There are technically three ways to add structured data (JSON-LD, Microdata, RDFa), but everyone recommends JSON-LD these days. Why? It lives in a <script> tag, separate from your HTML. Cleaner, easier to maintain, harder to break.

What it looks like

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Example Company",
  "url": "https://example.com"
}
</script>

The schemas that actually matter

You don't need to implement everything Schema.org offers. Here's what we recommend starting with—and what we use on this site.

1) Organization

This is your "who we are" markup. Put it on your homepage, or better yet, include it globally.

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Example Company",
  "url": "https://example.com",
  "logo": "https://example.com/logo.png"
}

2) WebSite

Tells machines "this is a website, here's its name." You can add SearchAction if you have site search, but don't fake it—only add it if there's a real, crawlable search URL.

{
  "@context": "https://schema.org",
  "@type": "WebSite",
  "name": "Example Site",
  "url": "https://example.com"
}

3) BreadcrumbList

This one's underrated. It tells AI exactly where a page sits in your site hierarchy. We add this to every page.

{
  "@context": "https://schema.org",
  "@type": "BreadcrumbList",
  "itemListElement": [
    {
      "@type": "ListItem",
      "position": 1,
      "name": "Blog",
      "item": "https://example.com/blog"
    },
    {
      "@type": "ListItem",
      "position": 2,
      "name": "Structured Data",
      "item": "https://example.com/blog/structured-data-for-ai"
    }
  ]
}

4) Article / BlogPosting

For blog posts and articles. The key fields are headline, author, and dates. This is what you're reading right now—and yes, we have BlogPosting schema on this page.

{
  "@context": "https://schema.org",
  "@type": "BlogPosting",
  "headline": "Structured Data for AI: A Practical Guide",
  "description": "A practical guide to Schema.org JSON-LD.",
  "author": {
    "@type": "Organization",
    "name": "Example Team"
  },
  "datePublished": "2025-01-21",
  "dateModified": "2025-01-21"
}

5) FAQPage (use carefully)

FAQ schema can get you those expandable FAQ rich results in Google. But here's the catch: the questions and answers must be visible on the page. If you mark up hidden content, that's a spam signal.

{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "What is AI-readiness?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "AI-readiness is how well AI systems can access and understand your website."
      }
    }
  ]
}

Tips from our own implementation

Don't overthink it

We started with just Organization and WebSite. Added the rest over time. You don't need perfect schema on day one.

Only mark up what's real

This is important: only mark up information that is:

• Actually true
• Currently accurate (not last year's pricing)
• Visible on the page (not hidden in a comment)

Google explicitly warns against marking up invisible content.

Remember to update it

The most common mistake we see? Outdated schema. Someone added it once, then changed the page content but forgot to update the JSON-LD. Now the schema says one thing and the page says another.

Mistakes we see constantly

Invisible content markup If users can't see it, don't mark it up. Google treats this as spam.

Stale data That product price from 2023? That event date from last year? Update or remove it.

Broken nesting Schema has relationships—author belongs to article, article belongs to website. Mess up the nesting and it all falls apart.

Copy-paste without customizing We've seen sites with "example.com" still in their schema. Check your URLs.

How to check your work

Before you ship, validate:

1. Google Rich Results Test — shows what Google sees and whether you're eligible for rich results: https://search.google.com/test/rich-results
2. Schema Markup Validator — catches syntax errors: https://validator.schema.org/
3. Your AI-readiness score — see if it improved: Scan your site

We run all three whenever we add new schema.

Keep reading

• How LLMs Choose Which Websites to Recommend — training data vs. retrieval, per-platform differences
• What is AI-readiness? — the bigger picture
• Understanding AI Crawlers — who's actually visiting your site
• How to Improve Your AI-Readiness Score — the full checklist
• What Is AI Visibility? — why it matters for your brand

Not a silver bullet, but one of the easiest wins for AI Visibility.