Google's official AI Optimisation Guide (June 2025) addresses this directly. Their position: GEO (Generative Engine Optimisation) is still fundamentally SEO. Their generative AI features - including AI Overviews and AI Mode - are rooted in their core Search ranking and quality systems. Optimising for AI search is optimising for the Search experience overall.

What this means practically is that the content quality, technical structure and authority signals that have always driven Google rankings are the same signals that drive AI citation eligibility. The difference is understanding how AI systems use that content - via RAG (Retrieval-Augmented Generation) to ground responses in indexed web pages - and building content that is uniquely valuable enough to be cited rather than paraphrased or ignored.

No - and Google explicitly confirmed this in their guide. They state: "You don't need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search (including its generative AI capabilities), as Google Search itself doesn't use them." LLMS.txt files are explicitly called out as something Google Search ignores entirely.

Similarly, Google confirmed you do not need special AI-specific schema markup, content "chunking" into small pieces, or rewriting content in a particular way for AI systems to understand it. These are widespread myths in the GEO space. The only things that work are the things that have always worked for high-quality search: unique expertise, non-commodity content, clear structure, crawlability and genuine authority.

Google's guide makes a clear distinction between commodity and non-commodity content. Commodity content is generic, widely available information that "could originate from anyone" - like a standard "7 tips for X" article. Non-commodity content provides unique expert or experienced insight that goes beyond common knowledge - like a documented case study of a specific outcome, original research data, or first-hand analysis from someone who has actually done the work.

For GEO, this distinction is critical because AI tools are designed to synthesise and summarise generic information themselves. They do not need to cite a source that says something they can generate. But they cannot fabricate original data, first-hand case study results, or unique expert perspectives - so they cite those instead. Creating content that AI systems must cite because they cannot replicate it is the core content strategy for GEO.

Query fan-out is how Google's AI systems gather comprehensive information for a single user question. When a user asks something like "best SEO agency for SaaS", the AI generates multiple related sub-queries - "SaaS SEO specialists", "SEO for B2B software companies", "SEO with GEO for SaaS" - and retrieves results for all of them to build a comprehensive answer.

For your content strategy, this means topical depth matters as much as individual page optimisation. A brand that has authoritative content covering the full cluster of questions around their category is far more likely to be retrieved across fan-out sub-queries than one with a single well-optimised page. We map your content against the likely fan-out patterns for your highest-value queries and build content to cover the full cluster - which also benefits your Google rankings for related terms simultaneously.

We measure citation frequency - the percentage of your target queries on which your brand is cited in the AI response - across ChatGPT, Perplexity and Gemini separately. This is tracked monthly against a defined query set of 20-50 questions relevant to your category and buyer stage.

Monthly reports include: citation rate per AI tool, which specific queries are generating citations, which content is being sourced, how your citation rate compares to top competitors on the same queries, and how the rate is trending month-over-month. We also track the quality of citations - whether your brand is mentioned as a recommended option, as a source of data or expert insight, or as a general reference - because the context of the citation matters for conversion.

Google's guide introduces agentic AI experiences as an emerging development: AI agents that can autonomously browse your website, analyse its structure and gather data to complete tasks on a user's behalf - such as comparing product specifications, checking availability or even initiating purchases. These agents interact with your site through visual rendering, DOM structure inspection and accessibility tree interpretation rather than just reading text.

For most businesses, this is a forward-looking concern rather than an immediate action item - Google's guide suggests it as an "if you have extra time" consideration. However, the preparation required is largely aligned with best practices you should be following anyway: semantic HTML, clear page structure, good accessibility and clean technical architecture. Our GEO onboarding includes an agentic readiness review so you understand where your site stands before this becomes a mainstream ranking factor.