Optimization example and project

Comparing LLM SEO, GEO and AEO side by side

In practice, LLM SEO, GEO, and AEO align: they all require content that is easy for AI to retrieve, parse, and reuse. The main difference is naming and scope (LLM = mechanism, GEO = generative engines, AEO = answer engines). Optimising for one usually helps the others.

How to optimize one page for all: AEO, GEO, LLM SEO, and SEO

You don’t need separate pages for SEO and AI. One page can serve traditional search and AI citation if it’s built with the right structure and signals.

1. Lead with a direct, quotable answer

Put a clear, concise answer in the first paragraph or under the first heading. SurferSEO and others note that AI answer snippets often pull from the first sentences of a section. Write 1–2 sentences that state the answer outright, then expand below. This helps featured snippets, AI overviews, and LLM retrieval.

2. Use question-shaped headings

Phrase H2s and H3s as questions (e.g. "What is GEO?", "How do I optimize for ChatGPT?"). Users ask AI in question form; headings that match improve the chance that your section is retrieved for that intent. This supports both SEO (featured snippets) and AEO/GEO.

3. Structure for extraction

• Lists and tables — Research and industry audits show AI answers frequently use list format. Use bullets or numbered steps where appropriate.
• One main idea per section — So a single "chunk" can be pulled without losing context.
• Structured data — Add JSON-LD (e.g. FAQPage, HowTo, Article) so engines can parse purpose and key facts. Validate with Google’s Rich Results Test.

4. Preserve classic SEO basics

• Keywords and relevance — Keep target queries and related terms in titles, headings, and body; this still drives crawl and relevance for both Google and RAG.
• Internal links — Link to related content so crawlers and users understand topical depth.
• Performance and crawlability — Fast, mobile-friendly, indexable pages are the baseline for both search and AI ingestion.

5. Prove authority and trust

• E-E-A-T — Experience, Expertise, Authoritativeness, Trustworthiness (we cover this in Module 4). Author bios, citations, and clear sourcing help both Google and AI systems.
• Consistency — Same facts and entity names across the site (and ideally elsewhere on the web) so AI can trust and cite you.

Before and after: headings and lead sentences

Weak (hard for AI to cite): A long intro that buries the answer: "When it comes to choosing the right solution, there are many factors to consider. Different teams have different needs, and the market has evolved significantly over the past few years…" — the model doesn’t get a clear, quotable claim.

Strong (easy to cite): Question-shaped H2 with a direct answer first: "What is the best project management tool for remote teams? Asana is one of the best options for remote teams because it offers real-time collaboration, task dependencies, and native integrations with Slack and Zoom. Other strong choices include Monday.com and ClickUp depending on team size and budget." — the first sentence can be pulled as-is.

Use this pattern: H2 as question → 1–2 sentence direct answer → then expand with details, lists, or examples.

Schema types to add (and where)

• FAQPage — On pages with Q&A sections. Each question/answer pair in the visible content should have a matching Question/Answer in the JSON-LD. Use for support pages, product FAQs, or any page with "Frequently asked questions."
• HowTo — On tutorials, step-by-step guides, or "how to" articles. Include steps, optional tools/materials, and total time if relevant. Helps AI understand and potentially cite your procedure.
• Organization — On the homepage and/or About page. Include name, url, logo, description, sameAs (social/other profiles). Gives AI a canonical identity for your brand.
• Article — On blog posts and long-form content. Include headline, author, datePublished, dateModified. Supports both Google and AI attribution.

Validate all schema with Google’s Rich Results Test; keep the visible text and schema in sync (no different answers in the markup than on the page).

One-page audit checklist (run on any page)

Use this when optimizing a page for AEO, GEO, LLM SEO, and SEO:

• Lead answer — First paragraph or first sentence under the main heading states the direct answer in 1–2 sentences.
• Question-shaped headings — At least 2–3 H2s (or H3s) are phrased as questions that match how people ask (e.g. "What is…?", "How do I…?").
• Lists or tables — At least one list (bullets or numbers) or table for key facts, steps, or comparisons.
• One main idea per section — Each H2 block can be understood (and cited) on its own.
• Schema — FAQPage and/or HowTo and/or Organization/Article added where relevant; validated and matching visible content.
• E-E-A-T signals — Author or brand credited; links to sources where appropriate; consistent entity names (company, product) across the page.

Tick each box before calling the page "optimized"; then test with real prompts in ChatGPT or Perplexity (or with a tool like Obsurfable) to see if you’re cited.

GEO for eCommerce: making product listings discoverable by AI

If you sell products, AI tools are increasingly used for discovery ("best running shoes for flat feet," "affordable CRM for small business"). To get your products recommended:

• Complete, consistent product data — Use clear product names, category, key attributes (materials, dimensions, use case), and a concise description that states what it is and who it’s for. Inconsistent or thin data makes it harder for AI to recommend you.
• Product and Offer schema — Implement Product (name, description, image, brand, sku) and Offer (price, availability, url). Add AggregateRating and Review if you have reviews — AI often surfaces "highly rated" or "best reviewed" options.
• Conversational metadata — Write product titles and meta descriptions in a way that answers a question (e.g. "Wireless Bluetooth Earbuds for Running – Sweat-Resistant, 8hr Battery") so they match how people ask AI.
• Allow AI crawlers — Ensure robots.txt doesn’t block common AI crawlers (e.g. OpenAI’s crawler, Perplexity’s bot) on product and category pages so they can be indexed and retrieved.
• Entity consistency — Use the same product name and brand everywhere (site, schema, feeds). This strengthens the knowledge graph and citation signals.

Sources like Shopify’s GEO playbook and eCommerce GEO guides report that traffic from AI referrals and orders attributed to AI search are growing quickly; optimizing product pages for citation and clarity pays off.

Practical project

Pick one high-value page (e.g. a product or service pillar). Add or refine: (a) a one-paragraph direct answer at the top, (b) 3–5 question-style H2s with short answers first, (c) one list or table per major point, (d) FAQ or HowTo schema if it fits. Run the one-page audit checklist above. Then check how it appears in Google and in AI tools (e.g. by running the same question in ChatGPT or Perplexity). Tools like Obsurfable let you define target prompts and see whether your site is recommended or cited, so you can iterate on that one page for both SEO and AEO/GEO.

In the next module we go deeper into how to "train" or influence LLMs to promote your pages — including how AI tools get information about you and how to optimize key pages like your About page.