the playbook The AI SEO playbook: prepare your site for AI search in 4 steps

Question 1: Is the answer to the main query in the first 60 words of the article?

AI scrapers read the first 60 words under the H1/H2 and use them as the citation hook. If the first paragraph is a marketing intro, AI skips you.

The fix: if not, move the definition/answer up top as a bold answer block (40–60 words).

Question 2: Does the page contain concrete facts, numbers, quotes?

Fact-dense content typically has a higher chance of being cited in AI answers than general marketing copy. Count the “facts per 100 words”:

• ≥ 3 concrete facts (number / percentage / date / study) per 100 words = ✅ AI-friendly
• < 1 fact per 100 words = ❌ too general

The fix: add data, source citations, concrete numbers. Eliminate filler.

Question 3: Is the H2/H3 structure logical?

Each subheading should answer a user’s sub-query. A clear hierarchy usually helps AI systems and human readers identify what the page answers faster.

The fix: rewrite the H2/H3 as questions or definitions (“What is …”, “How to …”, “When to use …”).

Question 4: Does it have an FAQ section with schema.org/FAQPage markup?

FAQPage schema is a direct channel to Google AI Overviews — Google pulls Q&A from it into the AI Overviews panel.

The fix: if not, add it. Take the questions from People Also Ask (PAA) boxes and Google Search Console (queries with rising impressions but falling CTR).

Article schema — for every article

{
  "@context": "https://schema.org",
  "@type": "Article",
  "headline": "SEO vs. GEO vs. AEO",
  "description": "A practical guide …",
  "author": {
    "@type": "Person",
    "name": "Article author"
  },
  "datePublished": "2026-05-24",
  "dateModified": "2026-05-24",
  "image": "https://example.com/og/article.png"
}

FAQPage schema — for Q&A sections

{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [{
    "@type": "Question",
    "name": "What's the difference between SEO and GEO?",
    "acceptedAnswer": {
      "@type": "Answer",
      "text": "SEO optimizes a page for rankings in Google. GEO optimizes content for citations in AI answers."
    }
  }]
}

HowTo schema — for guides

{
  "@context": "https://schema.org",
  "@type": "HowTo",
  "name": "How to add FAQPage schema",
  "step": [
    {
      "@type": "HowToStep",
      "name": "Pick 5–7 real questions from PAA",
      "text": "Open Google Search Console → Performance → Queries…"
    }
  ]
}

Validation

Before deploying, validate with:

• Google Rich Results Test — https://search.google.com/test/rich-results
• Schema Markup Validator — https://validator.schema.org

Layer 1: The answer block (40–60 words)

Right after the H1 or H2 — this is where AI engines pull the citation. The format:

**X is [a one-sentence definition]. [A second sentence with a concrete
fact or context]. [A third sentence with an implication for the reader].**

An example (from this page):

Start in 4 steps: (1) audit your existing content for AI citability, (2) add schema.org markup (FAQPage, HowTo, Article), (3) structure answers into an answer block + long context, (4) measure via SERP feature tracking and AI citation monitoring.

Layer 2: The detailed context

Below the answer block — for human readers who want depth. This is where you put:

• A detailed explanation of each step
• Examples, case studies
• Tables, visuals, code blocks
• Internal links to related content

This site has a prominent bold definition in every section (“X is…”). It isn’t an aesthetic touch — it’s a direct signal to AI scrapers showing where the definition of a term lives.

A manual AI Overviews check

If you don’t have a budget for tools, do a manual check once a month:

1. Take your top 10 keywords (from GSC)
2. For each, open https://google.com/search?q=<keyword>
3. Note: did the AI Overviews panel appear? Are you in the citations?
4. Do you get a click from the AI Overviews panel?

A manual ChatGPT/Perplexity audit

The same process for AI engines:

1. ChatGPT (search mode): “<keyword>” → are you cited?
2. Perplexity: “<keyword>” → are you in the citations on the right?
3. Claude (with search): “<keyword>” → does it cite you?

Weeks 1–2: Audit and strategy

• Audit your top 20 pages against the 4 questions from Step 1
• Classify pages (short/long, informational/transactional)
• Choose a strategy (embrace or bypass AI Overviews) — see the Decision matrix

Weeks 3–6: Implement schema + answer blocks

• FAQPage schema on your top 20 pages
• Article schema globally (via the BaseLayout/template)
• HowTo schema on guide pages
• An answer block (40–60 words) on your top 20 pages
• Validate with the Google Rich Results Test

Weeks 7–10: Measurement and iteration

• Set up tracking (GSC + a manual AI Overviews presence sheet)
• A weekly manual AI Overviews check for your top 10 keywords
• A monthly AI citation audit (ChatGPT/Perplexity)

Weeks 11–13: Optimize based on data

• Analysis: which pages are AI Overviews “stealing” traffic from? → consider nosnippet
• Brand-mention building (Reddit, guest posts, podcasts)
• Iterate answer blocks based on CTR data