How to get a website into Google’s AI search results: tips from Google

Google has updated its guide on optimizing websites for generative Search features (AI Overviews and AI Mode). We highlighted the key points and prepared a full translation of the text.

Key takeaways

1. SEO is not dead, quite the opposite. Generative features (AI Overviews, AI Mode) work on top of the main search index through RAG and query fan-out. To appear in an AI answer, a page must be indexed and eligible to be shown with a snippet in regular search results. From Google’s point of view, “AEO/GEO” is the same SEO.
2. The main long-term factor is unique, non-commodity content. Personal experience and specific details that an LLM cannot generate and that are not found in the article next door. An example from the text: not “7 tips for home buyers,” but “why we skipped the inspection and saved money: a breakdown of the sewer pipe.”
3. The technical requirements have not changed. Crawlability, indexability, page experience, correct JavaScript operation, minimizing duplicates, Search Console. Semantic HTML is desirable, but “perfect” validity is not required.
4. What you do NOT need to do (explicit mythbusting from Google itself):
   • llms.txt and other “special” files for AI are not needed.
   • Chunking content for AI is not needed; there is no ideal page length.
   • Rewriting content specifically for AI wording is not needed: models understand synonyms.
   • Chasing artificial mentions across the web is useless.
   • Structured data specifically for generative AI is not required, although it is useful for rich results.
   • Creating pages for every fan-out variant of a query falls under the scaled content abuse policy.
5. E-commerce and local business: a separate channel. Merchant Center with feeds, Google Business Profile, and optionally Business Agent. These are the channels through which products and locations get into AI answers.
6. Agentic experiences: a new direction. Browser agents read the DOM, screenshots and the accessibility tree. Protocols such as the Universal Commerce Protocol (UCP) are emerging. If this applies to your business, check the web.dev guide on agent-friendly UX.
7. Google’s meta-test. Before making any decision, ask yourself: “Would this content satisfy my visitor?” If yes, you are on the right track.

Translation

User preferences are changing quickly, and more people are turning to generative AI for information. As we evolve Search to meet these expectations, site owners have opportunities to reach audiences that are more likely to engage with content, spend more time on pages, and even convert: subscribe or make a purchase. This guide is for site owners who need Google’s official recommendations on working with generative AI features in Google Search (AI Overviews and AI Mode).

Is SEO still relevant for generative AI search?

In short, yes. Basic SEO practices remain relevant because generative AI features in Google Search are built on top of the core ranking and quality systems. These features rely on AI methods that highlight content from the search index:

• Retrieval-augmented generation (RAG). A method, also known as grounding, that improves the quality, accuracy and freshness of AI answers by relying on the core ranking systems. The systems retrieve relevant web pages from the index, parse specific information from those pages and generate a reliable, helpful answer with prominent clickable links to the sources.
• Query fan-out. A set of parallel related queries that the model generates to request additional information and pull in relevant results. For example, if the original query is “how to fix a lawn overgrown with weeds,” fan-out queries might include: “best herbicides for lawns,” “remove weeds without chemicals,” “how to prevent weeds from growing.”

What about “AEO” and “GEO”? “AEO” stands for “answer engine optimization,” and “GEO” stands for “generative engine optimization.” Both terms are used to describe work aimed specifically at improving visibility in AI search. From the point of view of Google Search, optimization for generative AI search is optimization for the search experience, meaning the same SEO.

Apply basic SEO practices to generative AI search

In this section, SEO practices are reframed with modern AI systems in mind, including what matters to them and how to apply these practices in the context of generative AI search to improve your site’s visibility both there and in Google Search overall.

Create valuable, non-commodity content

Unique, engaging and helpful content will have a greater long-term impact on your site’s presence in generative AI search than any other recommendation in this guide. Everyone interprets “unique, valuable, good content” in their own way, but such content usually has several common traits:

• A unique point of view. AI systems look across different sources, and this is where your own distinctive position can help you stand out. For example, a first-hand review offers a unique perspective based on personal experience, while a retelling of existing material repeats what is already available elsewhere. Create content yourself based on what you know about the topic, and think about what deep experience you can bring into it. Do not retell what others have already said online and what an AI model can easily generate.
• Non-commodity content that is helpful, reliable and people-first. Commodity content, such as “7 tips for first-time home buyers,” is built on commonly known information that anyone could write and rarely adds anything unique. Non-commodity content, such as “Why we skipped the technical inspection and saved money: what was wrong with the sewer pipe,” offers a unique expert or experience-based perspective that goes beyond common knowledge.
• A structure that helps the reader. Write for a real audience and make sure the text is well written and easy to read. Readers like pages that are divided into paragraphs and sections with headings that provide a clear structure.
• High-quality images and videos. Many people like finding images and videos in Search. Like Google Search overall, generative AI features can pull in relevant images and videos, which gives a site additional opportunities to appear beyond regular text links. Where appropriate, support your text with high-quality, relevant images and videos. If you already follow image SEO and video SEO, you are already optimizing for generative AI search as well.
• Focus on the user, without going too far. There may be a temptation to create separate content for every possible query variant, for example by targeting related user queries or fan-out queries. If this is done to manipulate rankings or AI answers in Google Search, it violates Google’s scaled content abuse policy. The strategy is also ineffective in the long term: a large number of pages does not make a site higher quality or more relevant. Google’s AI systems have learned to understand page relevance even where there is no exact match between the query and the main content of the page.
• If you use generative AI tools to create content, make sure the result complies with Search Essentials and the spam policies. More details are available in the guide on AI-generated content.

The approach can be reduced to one principle: think about what visitors will like, what they will find useful and what will leave them satisfied after visiting your site. If you are unsure about a decision, ask yourself: “Would this content be satisfying for my visitors?” If yes, you are on the right track. Google’s systems are designed specifically to connect people with this kind of helpful information. More details are available in the guide on creating helpful, reliable, people-first content.

Build and maintain a clear technical structure

How Google Search finds and processes your pages remains the foundation of how AI systems access data. Technical clarity ensures that content is ready for discovery and indexing, and all existing SEO practices continue to apply:

• Compliance with Search technical requirements. To be eligible to appear in generative AI features, a page must be indexed and eligible to appear in regular search results with a snippet, meaning it must meet the technical requirements for Search.

Following all requirements, best practices and policies does not guarantee that Google will crawl, index or serve page content. Indexing and serving are not guaranteed. More details: how Search works.

• Crawling best practices. To maximize visibility in generative AI features, content must be available for crawling. Google Search’s generative AI models use publicly available crawlable content to learn patterns and provide relevant grounded answers. For large and frequently updated sites, see the guide on optimizing crawl budget.
• Semantic HTML: focus on human readability, perfect code is not required. The web as a whole is not valid HTML, and Google understands it. Still, using semantic HTML where possible is generally a good idea: it helps screen readers parse and navigate the page.
• JavaScript. If you use it, follow JavaScript SEO best practices. Google can process content inside JavaScript if it is not blocked. SEO on sites with JS frameworks is usually more complex than on other sites. Follow standard SEO practices for JavaScript.
• Good page experience for people who come to the site: proper display on all devices, low latency and clear separation of the main content from other page elements.
• Reducing duplicate content. Duplicates create a poor user experience, and search engines may spend crawling resources on URLs you do not need. If you have time, try to reduce duplication.

To quickly detect and diagnose technical issues, verify your site in Search Console. See also the technical SEO guide and SEO support for your site.

Optimize local business and e-commerce data

Where relevant, generative AI answers may include product cards, product information and local business data. Using products such as Merchant Center, including Merchant Center feeds, and Google Business Profiles helps products and services be visible both in AI answers and in regular search results. More details: how to add and manage business details in Google Search.

Depending on the type of business and your goals, consider other merchant opportunities as well, such as Business Agent, a conversational format in Google Search where customers can interact with your brand.

Mythbusting generative AI search: what you do NOT need to do

As generative AI search evolves, theories, practices and sometimes misconceptions appear around it. Terms such as Answer Engine Optimization (AEO) and Generative Engine Optimization (GEO) are common online, along with “hacks,” many of which are ineffective and do not match how Google Search actually works.

To help you focus on what really affects visibility, we have collected the loudest topics around generative AI and Google Search. Here is what you can ignore when it comes to Google Search:

• llms.txt files and other “special” markup. You do not need to create new machine-readable files, AI-text files, markup or Markdown to appear in generative AI search. Note: Google can discover, crawl and index many file types besides HTML, but that does not mean any file is processed in a special way.
• Content “chunking.” There is no requirement to break content into tiny pieces so that AI can “understand it better.” Google’s systems understand the nuances of multiple topics on one page and show the relevant passage to the user. Sometimes short pages work well, sometimes long pages do, depending on the audience and topic. There is no ideal page length, and ultimately you should create pages for people, not for generative AI search.
• Rewriting content specifically for AI systems. You do not need to write in some special way for generative AI search. AI systems understand synonyms and the general meaning of what the user is looking for, and can connect it with content even if the exact same words are not present. You do not need to worry that you have “not enough long-tail keywords” or that you have not covered every possible query wording.
• Chasing unnatural “mentions.” Like the rest of Google Search, generative AI features can show what is being said about products and services across the web, including blogs, videos and forums. Chasing unnatural mentions is less useful than it may seem. The core ranking systems focus on quality content, while other systems block spam. Generative AI features rely on both.
• Excessive focus on structured data. Structured data is not required for generative AI search, and there is no special schema.org markup that needs to be added. Still, using it as part of a broader SEO strategy is a good idea: it helps obtain rich results in Google Search.

Explore agentic experiences

AI agents can autonomously perform tasks for a person: book a table, compare product specifications. They vary in type. For example, browser agents can visit a site to collect the required data: analyze the visual rendering (screenshots), the DOM structure and the accessibility tree.

If this is relevant to your business and you have time, look at available agentic experiences and the best practices guide for agent-friendly sites. It explains how a site can prepare for today’s browser agents. Protocols such as the Universal Commerce Protocol (UCP) are emerging and will allow search agents to do more.

What to do next

When working on your site, remember: a lot of content performs well in Google Search, including generative AI experiences, without any explicit SEO at all, and you do not need to do everything in this guide to succeed. Key conclusions:

• Apply SEO best practices to generative AI search. Continue focusing on basic SEO practices, building a clear technical structure and creating unique valuable content. This is the foundation of visibility both in generative AI search and in Google Search overall.
• Create non-commodity content that is helpful, reliable and people-first. Focus on unique expert materials that provide value beyond common knowledge.
• Prioritize effective SEO strategies over “AEO/GEO hacks.” For Google Search, you can ignore tactics such as content “chunking,” creating unnecessary AI-text files, for example llms.txt, and chasing unnatural mentions.
• Explore agentic experiences. Keep an eye on technologies that allow AI agents to interact with websites, browser agents and new protocols.

Translation of the official Google Search Central guide “Optimizing your website for generative AI features on Google Search” as of May 15, 2026. The original was published by Google under the CC BY 4.0 license, and this translation is distributed under the same terms. Link to the original.