Google’s AI Optimisation Guide and What It Actually Means for Marketers

On 15 May 2026, Google published a detailed guide to optimising for generative AI features in Search. The document covers AI Overviews, AI Mode, and the broader question of how content gets surfaced in AI-generated responses on Google Search.

For anyone who has spent the past 18 months seeing a flood of conflicting advice about AEO, GEO, llms.txt files, content chunking, and AI-specific schema, the guide is worth reading carefully. It settles several debates directly, names specific tactics that do not work, and makes a clear statement about the relationship between SEO and AI search visibility.

This piece breaks down what the guide says, what it changes, and what marketing teams should do with it.

The Core Message: AI Search Is Still SEO

The most significant statement in the guide is also the most direct. From Google Search’s perspective, optimising for generative AI search is optimising for the search experience, and thus still SEO.

This is not a diplomatic reframing. Google’s AI features rely on retrieval-augmented generation, also called “grounding” as the technique that drives AI Overviews and AI Mode. When a user submits a query, Google’s AI does not generate an answer from training data alone. It retrieves content from the Search index and uses it to construct a response.

The practical consequence of this is significant. The signals that determine traditional search rankings (quality, authority, E-E-A-T, technical health) are the same signals that determine AI Overview inclusion. Strategies that diverge from foundational SEO in the name of AI-specific optimisation are working against the grain of how the system actually functions.

What the Guide Says About Content Quality

Creating content that people find unique, compelling, and useful will likely influence a website’s presence in generative AI search in the long run more than any of the other suggestions in the guide.

The guide introduces a useful distinction between commodity and non-commodity content. Commodity content is the kind of article any competent writer could produce from a search and a basic understanding of the topic. Non-commodity content carries a perspective, experience, or insight that only the author or organisation can provide.

Content with a distinct perspective stands out because AI pulls from many sources. First-hand reviews, personal analysis, and real case studies go beyond common knowledge and beyond what an AI could generate independently.

For B2B marketing teams, this distinction has a direct implication for content planning. Generic category-level articles, guides that summarise publicly available information, and posts that reflect what already ranks are all commodity content by this definition. The content that earns AI visibility is the content that brings something forward the rest of the category has not said.

How AI Overviews and AI Mode Actually Work

Understanding the mechanism helps marketing teams make better decisions about content structure and coverage.

Both AI Overviews and AI Mode may use a “query fan-out” technique, issuing multiple related searches across subtopics and data sources to develop a response. While responses are being generated, advanced models identify more supporting web pages, allowing a wider and more diverse set of helpful links to be displayed than a classic web search result would typically show.

This means comprehensive, topically authoritative content can earn AI visibility for queries it was never specifically written for. A well-structured pillar article that covers a subject in depth can be retrieved in response to several related queries, not just the one it was optimised for.

This is the mechanism behind topical authority as a strategy. It is not simply an SEO concept. It reflects how AI retrieval systems reward depth and coverage.

Technical Fundamentals Still Matter

On the technical side, pages must be indexed and eligible for snippets to appear in generative AI features. Google recommends following crawling best practices, using semantic HTML where possible, following JavaScript SEO best practices, providing good page experience, and reducing duplicate content.

To maximise a site’s visibility in generative AI search features, content must be crawlable, as Google Search generative AI models use publicly accessible, crawlable content to learn patterns and provide relevant, grounded responses.

None of this is new guidance. What the document clarifies is that these fundamentals apply equally to AI search features as they do to standard search results. A page that Google cannot crawl and index is a page that will not appear in AI Overviews regardless of how well it is written.

What the Guide Says to Stop Doing

The guide dedicates specific attention to tactics that have been widely marketed as AI search optimisation strategies. The guidance here is direct.

You do not need to create new machine readable files, AI text files, markup, or Markdown to appear in generative AI search. There is no special schema.org structured data required to appear in these features.

The guide states clearly that standard structured data remains useful for rich results eligibility, but it is not a factor in AI retrieval specifically.

The practical takeaway on structured data is to keep using it for rich results and general SEO, but not to add schema specifically on the belief that it unlocks AI features, because it does not.

The guide also addresses fabricated mentions and inauthentic placements directly. Earning mentions genuinely through quality content and real community participation is the approach that holds up. Inauthentic placements are caught by spam systems and do not improve AI visibility.

The Emerging Agentic Layer

A section of the guide covers a development that most current AI search discussions have not yet reached: agentic experiences.

The guide describes AI agents as autonomous systems that can perform tasks on behalf of people, such as booking a reservation or comparing product specifications. Browser agents may access websites by analysing screenshots, inspecting the DOM, and interpreting the accessibility tree.

Accessibility and clean DOM structure matter more going forward, not just for human users but for AI agents. The guide references the Universal Commerce Protocol as an emerging standard that will allow Search agents to perform more complex transactional operations.

For most marketing teams, this is not an immediate priority. The near-term implication is straightforward: a site with strong fundamentals, clean structure, and good accessibility is already better positioned for agentic interactions than one that is not.

What This Means for Marketing Teams in Practice

The guide does not introduce a new discipline. It confirms that the discipline already exists and that AI search visibility is built through the same sustained investment in content quality, technical health, and editorial credibility that effective SEO has always required.

The specific areas worth acting on based on the guide are: ensuring your highest-value pages are crawlable and indexed, building topical depth rather than isolated articles, ensuring content carries visible E-E-A-T signals, and stopping investment in tactics the guide explicitly identifies as ineffective.

The path to better visibility in AI Overviews, AI Mode, and other AI-assisted search experiences still starts with the same core disciplines: useful content, technical accessibility, clear site architecture, and a strong overall user experience. What has changed is the surface area where that content can now appear, and the increased reward for content that only your brand can produce.