Structured Data for AEO: A Practical JSON-LD Implementation Guide

Why Structured Data Matters for AEO

Structured data (JSON-LD) provides machine-readable annotations that tell AI search systems exactly what your content contains. Without structured data, AI systems must infer the structure of your page from HTML headings and paragraph text — a process with significant error rates. With structured data, the AI knows explicitly that "this page contains a definition of X" or "this page answers question Y."

The impact is measurable: pages with FAQPage schema are 4.2x more likely to be cited in AI Overviews than pages without it (Semrush, March 2026). Pages with Article schema are 1.6x more likely. Structured data is the single highest-impact technical change you can make for AEO.

Use JSON-LD format exclusively. Google, Perplexity, and Bing all prefer JSON-LD over Microdata or RDFa. Place the schema in a <script type="application/ld+json"> tag in the page's HTML body.

FAQPage Schema

The most powerful schema type for AEO. Wraps a page with explicit Question/Answer pairs, telling AI systems exactly which questions the page answers and what the answers are:

{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "What is AEO?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "AEO (Answer Engine Optimisation) is the practice of structuring web content so AI-powered search systems can extract, cite, and surface it as a direct answer."
      }
    },
    {
      "@type": "Question",
      "name": "How is AEO different from SEO?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "SEO optimises for blue-link rankings. AEO optimises for being cited as a source by AI-generated answers. Approximately 70% of best practices overlap."
      }
    }
  ]
}

Best practices for FAQPage schema:

• Include 3-8 questions per page. Fewer than 3 is thin; more than 8 dilutes focus.
• Match questions to "People Also Ask" results for your target keywords.
• Keep answers between 50-150 words. Direct, factual answers perform best.
• The visible FAQ content on the page must match the schema content exactly.
• Place the FAQ section near the bottom of the page, before references.

Article / BlogPosting Schema

Provides citation metadata that AI systems use to evaluate source credibility and freshness:

{
  "@context": "https://schema.org",
  "@type": "BlogPosting",
  "headline": "What Is AEO? Answer Engine Optimisation Explained",
  "datePublished": "2026-04-25",
  "dateModified": "2026-04-25",
  "author": {
    "@type": "Person",
    "name": "Sam Wong",
    "url": "https://samwong.com/about/"
  },
  "description": "AEO is the practice of structuring web content for AI-powered search systems.",
  "url": "https://samwong.com/blog/what-is-aeo/"
}

Always include datePublished, dateModified, author with a URL to the author page, and the canonical URL. The dateModified field signals freshness — update it when you revise articles. The author URL should link to a page with Person schema that establishes the author's identity and credentials.

HowTo Schema

For procedural content with sequential steps. AI engines extract HowTo steps with high accuracy and present them as numbered instructions:

{
  "@context": "https://schema.org",
  "@type": "HowTo",
  "name": "How to Optimise Content for AI Answer Engines",
  "step": [
    {
      "@type": "HowToStep",
      "position": 1,
      "name": "Add FAQPage Schema",
      "text": "Implement FAQPage JSON-LD on every informational page with 3-5 questions."
    },
    {
      "@type": "HowToStep",
      "position": 2,
      "name": "Rewrite Opening Paragraphs",
      "text": "Ensure every page starts with a one-sentence definition of the topic."
    },
    {
      "@type": "HowToStep",
      "position": 3,
      "name": "Add Comparison Tables",
      "text": "Replace prose comparisons with HTML tables for higher extraction fidelity."
    }
  ]
}

Use HowTo for any "how to" or "guide" content with 3 or more sequential steps. Each step should have a name and a text description. Keep each step to 1-2 sentences.

DefinedTerm Schema

Marks up term definitions for "what is X" queries. This schema type is underused but effective for glossary and definition pages:

{
  "@context": "https://schema.org",
  "@type": "DefinedTerm",
  "name": "Answer Engine Optimisation",
  "description": "The practice of creating and structuring web content so AI-powered search systems can extract, cite, and present it as a direct answer.",
  "inDefinedTermSet": {
    "@type": "DefinedTermSet",
    "name": "AEO Glossary"
  }
}

Implementation Checklist

1. Every blog post: BlogPosting + FAQPage (if FAQ section exists)
2. Every service/product page: Article + FAQPage with 5 questions about the service
3. Every "how to" page: HowTo + FAQPage
4. Homepage: Organization or WebSite schema with site name and URL
5. Author page: Person schema with name, job title, and social profiles
6. All pages: BreadcrumbList for navigation trail

Common Mistakes

1. Using Microdata or RDFa instead of JSON-LD: Google, Perplexity, and Bing all prefer JSON-LD. It is easier to implement and maintain.
2. Placing schema in a separate file: Schema must be in the HTML inside a <script> tag. External JSON files are not crawled.
3. Duplicating schema across multiple script blocks: Use one <script> block per page with all schema types combined using @graph.
4. Using FAQPage without visible FAQ content: Schema must match visible content. If you mark up FAQ questions, they must be visible on the page.
5. Not validating: Always run pages through Google's Rich Results Test before deploying. Invalid schema provides no benefit.