Why schema matters more for AI than for Google

Schema markup is structured data — JSON-LD blocks you embed in your HTML that tell crawlers what a page is, not just what's on it. For traditional SEO, schema is a nice-to-have. For GEO, schema is the difference between getting cited and getting skipped.

The reason: AI engines don't read your page top-to-bottom. They retrieve passages, evaluate extractability, and decide whether to cite. Schema makes extractability trivial. A page with valid FAQPage schema is shipped to the model as a clean list of question-and-answer pairs. A page without it is shipped as a wall of text the model has to parse heuristically.

Google AI Overviews is the most schema-sensitive of the four major engines. ChatGPT and Perplexity rely less on schema (they emphasize entity signals and fact density), but still benefit from it. There is no engine where schema hurts you.

The four schemas that actually move the needle

Schema.org has hundreds of types. Four of them produce 90% of GEO impact:

1. Organization (entity foundation)

Organization schema tells AI engines who your brand is as an entity. Ship it on your homepage and every page with the sameAs property pointing to every external entity reference (Wikipedia, LinkedIn, Crunchbase, your X/Twitter, GitHub if relevant).

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Your Brand",
  "url": "https://yourbrand.com",
  "logo": "https://yourbrand.com/logo.png",
  "description": "What you do in one sentence.",
  "foundingDate": "2024",
  "sameAs": [
    "https://en.wikipedia.org/wiki/Your_Brand",
    "https://www.linkedin.com/company/yourbrand",
    "https://www.crunchbase.com/organization/yourbrand",
    "https://x.com/yourbrand"
  ]
}

The sameAs array is the most important field. It's how AI engines verify "the YourBrand on your site is the same YourBrand on Wikipedia." Without it, the AI treats the entity as ambiguous and is less likely to cite.

2. FAQPage (question-and-answer content)

FAQPage schema turns any FAQ block on your page into AI-citable Q&A pairs. This is the single highest-leverage schema for GEO because it maps directly to how users query AI ("How do I...", "What is...", "Why does...").

{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "What is your product's pricing?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "Three tiers: Starter $29/mo, Pro $99/mo, Enterprise custom..."
      }
    }
  ]
}

Pair this with question-formatted H2 headings on your page (we cover that pattern in Lesson 1.4). The schema confirms what the page structure already implies. Both signals together = strongest extraction.

3. Article (long-form content)

Article schema goes on every blog post, case study, and long-form page. It tells AI engines who wrote it, when it was published, when it was last updated, and what it's about. The author and dateModified fields matter most:

{
  "@context": "https://schema.org",
  "@type": "Article",
  "headline": "Article title",
  "datePublished": "2026-01-15",
  "dateModified": "2026-05-15",
  "author": {
    "@type": "Person",
    "name": "Author Name",
    "url": "https://yourbrand.com/team/author-name",
    "sameAs": [
      "https://www.linkedin.com/in/authorname",
      "https://x.com/authorname"
    ]
  },
  "publisher": {
    "@type": "Organization",
    "name": "Your Brand",
    "logo": { "@type": "ImageObject", "url": "https://yourbrand.com/logo.png" }
  }
}

65% of AI bots prefer pages updated within the past year. dateModified is how you signal freshness. Update it whenever you genuinely refresh the content — but don't fake it. Some engines penalize stale-content-with-fresh-dates patterns.

4. HowTo (procedural content)

If your page teaches a procedure ("How to set up X", "How to choose Y"), HowTo schema is mandatory. AI engines aggressively pull from HowTo content for "how do I..." queries.

{
  "@context": "https://schema.org",
  "@type": "HowTo",
  "name": "How to do X",
  "step": [
    { "@type": "HowToStep", "name": "Step 1", "text": "Do this first..." },
    { "@type": "HowToStep", "name": "Step 2", "text": "Then this..." }
  ]
}

What not to do

Three common schema mistakes that hurt more than help:

1. Schema that doesn't match the page content. If your FAQPage schema lists Q&A pairs that don't appear visually on the page, that's spam. Google flags it. AI engines distrust the source.
2. Multiple competing schemas on one page. Pick the most specific type. A blog post is Article, not BlogPosting + Article + WebPage.
3. sameAs pointing to dead links. If your Wikipedia URL 404s, the entity verification fails. Audit quarterly.

Validation: how to know your schema is right

Three tools, in order:

• Google Rich Results Test (search.google.com/test/rich-results). Catches the schema Google can read. Required minimum.
• Schema.org Validator (validator.schema.org). Catches type errors Google's tool doesn't flag.
• Run a Reffed audit. We score your schema coverage across all pages and show which engines are picking it up. This is how you verify the schema is actually working for AI citation, not just SEO.

Implementation: shipping schema this week

Practical sequence for getting these four schemas live:

1. Day 1. Audit current state. Run Google Rich Results Test on your homepage, top 5 commercial pages, and your most-trafficked blog post. Note what's there.
2. Day 2. Ship Organization schema with full sameAs array. This is sitewide — put it in your footer template or layout file so it appears on every page.
3. Day 3. Add FAQPage schema to every page with a visible FAQ section. If your top 5 pages don't have FAQs, this is also your cue to add them.
4. Day 4. Add Article schema to your top 10 blog posts. Include dateModified set to the last real edit date.
5. Day 5. Add HowTo schema to any procedural content. If you don't have any, skip — don't fake it.
6. Day 6-7. Re-validate. Re-audit. Compare baseline to new state.

What comes next

Lesson 1.2 (Entity SEO: sameAs, about, mentions) goes deeper on the entity-defining schema properties — specifically how to use about and mentions to tell AI engines what your pages cover at a topical level, and how knowsAbout on the Organization schema becomes the "topic authority" signal AI engines look for.

← Back to course dashboard