Schema Markup for AI: A Step-by-Step Guide

What you'll learn:

• What Google and the one controlled study say about schema and AI
• Which schema types still do something in Google Search, and which are retired
• Step-by-step implementation with JSON-LD code examples
• How to test and validate your schema
• Common mistakes to avoid

Why Schema Matters for AI

Schema markup adds a machine-readable layer on top of your human-readable content. The honest question is whether any AI engine reads that layer. Google says it is not required: "Structured data isn't required for generative AI search, and there's no special schema.org markup you need to add." Ahrefs' matched-control test (1,885 pages, three controls each, published 2026-05-11) found no uplift on any platform and a statistically significant 4.6% drop in AI Overview citations. No engine documents reading a fact that exists only in JSON-LD.

What schema still does: power the Google rich results that remain (Product, Organization), and disambiguate your brand name. What matters for AI is parity — every fact in the markup must also be in the visible text. A fact that is only in the schema is a fact no AI crawler will read, and a fact that disagrees with the page is a published contradiction.

FAQ Schema — Retired

Google deprecated FAQ rich results on 2026-05-07 and removed the FAQPage documentation on 2026-06-15. The markup no longer produces anything in Google Search. Leaving existing markup in place does no harm, per Google; adding it gains nothing. The example below is kept for reference.

When to use it

Any page that answers questions — product pages, pricing pages, help pages, category pages, blog posts with FAQ sections.

JSON-LD Example

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "What does your product cost?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Plans start at $49/month for Starter, $99/month for Growth, and $149/month for Pro. All plans include a 7-day free trial."
      }
    },
    {
      "@type": "Question",
      "name": "Which AI platforms do you monitor?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "We monitor ChatGPT, Perplexity, Google AI (Gemini), and Claude. Each platform is audited separately with platform-specific scoring."
      }
    }
  ]
}
</script>

Organization Schema

Organization schema's honest job is disambiguation in Google Search: if your brand name is also a common word, sameAs links tell Google which entity you are. It is not a documented input to any AI engine's citation decision — treat brand entity strength (consistent NAP, Knowledge Graph presence, third-party mentions) as the thing worth building, and Organization schema as a Google Search nicety that should simply agree with what's on the page.

When to use it

Your homepage and About page. Only add it once per page — duplicates cause confusion.

JSON-LD Example

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Your Company Name",
  "url": "https://yourcompany.com",
  "logo": "https://yourcompany.com/logo.png",
  "description": "A one-sentence description of what your company does.",
  "foundingDate": "2024",
  "sameAs": [
    "https://twitter.com/yourcompany",
    "https://linkedin.com/company/yourcompany"
  ],
  "contactPoint": {
    "@type": "ContactPoint",
    "email": "hello@yourcompany.com",
    "contactType": "customer service"
  }
}
</script>

Product Schema

Product schema still powers product snippets and merchant listings in Google Search. For AI, the useful part is parity: the price in the Offer must be the same price a crawler reads in your raw HTML. Wrong pricing is the #1 type of AI inaccuracy, and the plausible cause is your own site stating it two ways — not missing markup.

When to use it

Pricing pages, product pages, and any page that describes your offering with specific attributes (price, features, ratings).

JSON-LD Example

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Growth Plan",
  "description": "AI visibility monitoring for growing brands. Includes weekly audits across all platforms.",
  "offers": {
    "@type": "Offer",
    "price": "99",
    "priceCurrency": "USD",
    "priceValidUntil": "2026-12-31",
    "availability": "https://schema.org/InStock"
  }
}
</script>

HowTo Schema

Step-by-step formatting genuinely helps with "how do I..." queries — but that's a property of the visible content structure (numbered steps, one action per step), not of the HowTo markup wrapped around it. No AI engine documents reading HowTo schema, and Google's own AI-features guidance doesn't list it as a requirement. Write the steps clearly in the page; add the schema only if you also want Google's classic how-to rich results.

When to use it

Any content that walks through a process with distinct steps — setup guides, tutorials, implementation instructions.

JSON-LD Example

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "HowTo",
  "name": "How to Run Your First AI Visibility Audit",
  "step": [
    {
      "@type": "HowToStep",
      "name": "Enter your domain",
      "text": "Add your website URL and brand name to start the audit."
    },
    {
      "@type": "HowToStep",
      "name": "Select target queries",
      "text": "Choose 5-25 queries that your customers typically ask."
    },
    {
      "@type": "HowToStep",
      "name": "Review results",
      "text": "View your Visibility Score and per-platform breakdown."
    }
  ]
}
</script>

Testing Your Schema

After adding schema to your pages, validate it:

1. Google Rich Results Test — Paste your URL to confirm Google can parse your schema.
2. Schema.org Validator — Tests for structural errors in your JSON-LD.
3. Manual check — View the page source, search for "application/ld+json", and confirm the data is accurate and current.

Also read our blog post on common schema mistakes that undermine AI visibility.

Common Mistakes

• Outdated data in schema: Your schema says $79/month but your pricing page says $99/month. This makes hallucinations worse, not better.
• Duplicate schema on the same page: Two Organization schema blocks on one page confuse parsers.
• Schema without matching visible content: Google penalizes schema that doesn't reflect what users see on the page.
• Missing required properties: Each schema type has required fields. Partial schema is often ignored.

Key Takeaways

• Schema markup is not an AI ranking lever. Google says none is needed; the one controlled test found a 4.6% drop.
• FAQ schema is retired — Google deprecated FAQ rich results on 2026-05-07.
• Every fact in your schema must also appear in your visible text. Parity is the check that matters.
• Organization schema disambiguates your brand name in Google Search; every field must match your About page.
• Always validate schema and keep the data current.

See if your schema agrees with your page. Run a free audit with 99Visibility to find facts that are in your markup but not in your visible text.