Google confirms it: SEO remains the key to optimizing for AI overviews

Although Google can sometimes drive us crazy with its rule changes or the SEO tremors caused by its algorithm updates, if there is one thing we have to give it credit for, it is its communication skills. Something essential for any relationship to work… and for SEO professionals to adapt their content to rank better.

Today, we are going to break down one of Google’s latest publications in this area: a guide focused on how to optimize your website for Google Search’s generative AI features (such as AI Mode or AI Overviews).

Talking about GEO sounds very cool, but SEO is still the foundation

As we have explained more than once, SEO is the main mechanism for driving your web positioning strategies, and GEO is one piece within that framework. Generative Engine Optimization practices draw directly from the core principles of Search Engine Optimization, even if they are adapted to the specific nature of AI.

When it comes to Google’s generative AI features, their mechanisms pull directly from content in Google’s search index to extract the information they later shape into answers. For example, Retrieval-Augmented Generation (RAG) relies on Google Search’s core ranking systems to retrieve relevant and up-to-date web pages and, based on those, generate a more reliable and useful response.

On the other hand, its AI’s Query Fan-Out system runs additional searches that complement the user’s original query in order to obtain relevant search results that enrich the final response.

This means that optimizing for AI Overviews or AI Mode still means optimizing for Search and, therefore, we are still talking about SEO. In Google’s own words: “SEO best practices remain relevant because our generative AI features in Google Search are grounded in our core ranking and search quality systems.”

SEO best practices for being relevant in generative search

Now that the importance of SEO when creating web content is clear, let us review the SEO best practices that Google’s AI systems use as a reference when selecting sources.

Let us start with the ones related to the content itself. Above all, it must be valuable:

• Offer a unique perspective: Google’s AI systems analyze several sources, so content that provides its own distinctive point of view is more likely to stand out.
• Create content that is not generic: your website content should be helpful, trustworthy, and people-first.
• Organize content in a way that makes it easy to read: besides being well written, content should follow a clear order, be structured into paragraphs and sections, and include headings that provide a clear navigational structure.
• Include high-quality images and videos: besides complementing written content, these resources can also be powerful allies when it comes to being cited by Google’s AI. Generative AI search features can display relevant images and videos, creating opportunities for your website to appear beyond standard links.
• Focus on what your users want and do not overdo it: do not create different pieces of content on the same topic just to target search variations. Google’s AI systems understand the relevance of web pages even when there is not an exact match between the query and the page’s main content.
• Make sure you comply with Google’s Search Essentials and spam policies if you use generative AI tools to create your content.

The next area where we should apply Google’s advice is technical structure. The goal should always be to build and maintain a clear structure:

• Meet the technical requirements for Search: your page must be indexed and eligible to appear in Google Search.
• Follow crawling best practices: Google Search’s generative AI models use crawlable, publicly accessible content to learn patterns and provide useful answers.
• When using semantic HTML, focus on human readability rather than aiming for perfect code: using semantic HTML will help other types of users, such as screen reader users, analyze and navigate your website more easily.
• If you use JavaScript, make sure to follow JavaScript SEO best practices.
• Provide a good user experience: make sure your site displays correctly on all types of devices, reduce latency, and make it easy for users to distinguish the page’s main content from the rest of its elements.
• Reduce duplicate content.

In addition, Google recommends that, if you have a business, you optimize every detail of your local business and eCommerce presence. “Using tools such as Merchant Center (for example, Merchant Center feeds) and Google Business Profiles can help make your products and services visible both in AI responses and in other Google Search results.”

Likewise, with regard to AI agents, the tech giant encourages users to explore the agent experiences available on Google and to take a look at the best practices guide for agent-friendly websites.

• You may also be interested in: Journey to the center of Googlebot: how Google Search’s web crawler works

Doing this will not help you succeed in generative AI search

Google also took the opportunity to debunk a series of myths that have emerged around how generative AI search works.

“While terms like Answer Engine Optimization (AEO) or Generative Engine Optimization (GEO) are common on the internet, many of the supposed ‘tricks’ are neither effective nor aligned with how Google Search actually works,” the company explains.

• You do not need to create LLMS.txt files or other “special” markup to appear in Google’s generative AI searches.
• You do not have to split content into small chunks: AI understands content nuances and when a page covers several topics, and it can extract only the information it needs. There is no ideal page length.
• Do not rewrite content for AI systems: you do not need to adjust your writing style to “adapt” it for AI. These systems can understand synonyms and broader meanings, allowing them to connect a search with useful content that may not use the exact same words.
• You do not need to chase false mentions: Google’s generative AI features rely both on core ranking systems focused on high-quality content and on systems that block spam content. So do not waste time looking for possible false mentions.
• Avoid over-focusing on structured data: you do not need to add any special schema.org markup to catch AI’s attention. That said, keep using it as part of your overall SEO strategy.

We hope these tips have been useful in helping you better understand the reality of Google’s AI features and put all those claims about the death of SEO out of your mind.

Likewise, it is important to point out that you do not need to implement every single practice listed in Google’s guide in order to rank your content. The key is to apply the ones you actually need thoughtfully and follow this core SEO principle: “create helpful content for the user.”

Photo: generated with GPT Images 2