Google has published a new documentation page to help websites optimize the productivity features of AI in Search, including AI Overview and AI Mode.
The page, “Preparing your website for productive AI features in Google Search,” expands on Google’s previous AI features documentation published in 2025. The front page explains how AI features work, how installations are managed, and how performance is reported. The new guide focuses more on optimization tips and tricks that Google says site owners can ignore.
Two categories deserve special mention. Google specifically mentions popular optimization tactics as unnecessary, and redefines the AEO/GEO discussion as part of general SEO.
Google Says AEO And GEO Are ‘Still SEO’
Google opens by ensuring that basic SEO best practices remain essential for productive AI searches. Its AI features are “embedded in our core Search metrics and quality systems” and rely on retrieval-augmented generation (RAG) and querying followers to surface content from the Search index.
In the discussion of words, Google is specific. It defines “AEO” as “response engine optimization” and “GEO” as “generative engine optimization,” and says:
“From Google Search’s point of view, preparing a productive AI search improves the search experience, and thus SEO.”
This is similar to the positions taken by Google employees at conferences. Gary Illyes and Cherry Prommawin told Search Central Live attendees that GEO and AEO do not need separate entities. The status now appears in Google’s published articles, providing an official citation index.
What Google Says You Don’t Need To Do
The guide includes “Productive AI Search Mythbusting” which is a list of tactics it calls unnecessary for Google Search. The guide is much clearer than Google’s previous AI features page, especially on naming llms.txt, chunking, aliasing, and AEO/GEO specifically.
The guide states that site owners can ignore the following in Google searches.
Opened llms.txt files and other “exclusive” markup, Google says you don’t need to create machine-readable files, AI text files, markup, or Markdown to appear in AI-generated searches. Google may find and index many file types beyond HTML, but that doesn’t mean those files get special treatment.
Opened “to burst” content, the guide says there’s no need to break down content into smaller chunks for AI systems. Google’s systems “are able to understand the differences between multiple topics on a page and show the right chunk to users.” Danny Sullivan made a similar comment in January 2026, saying he spoke with Google engineers who recommended against chunking.
Opened content rewriting for AI systems, Google says that AI systems can understand synonyms and common meanings. Site owners don’t need to capture all kinds of long-tail keywords or write in a specific form for AI search engines to generate.
In search of falsehood”mentions it,” the guide acknowledges that AI features can reveal what’s being said about products and services across blogs, videos, and forums. But it says that looking for fake talk is “not as helpful as it might seem” because basic rating systems focus on quality while other systems block spam.
Opened structured datathe guide states that it is not required for productive AI searches and no special schema.org markup will be added. It recommends continuing to use structured data as part of an overall SEO strategy to optimize for rich results.
Several recommendations conflict with advice from other guidelines for conducting AI searches. Many GEO resources have promoted the collection and structured data as essential to AI search visibility.
What Google Says to Focus on
The optimization advice follows the general SEO landscape, although Google puts it in terms of AI features.
Google puts a lot of emphasis on “non-material content.” Compares commercial content (“7 Tips for First-Time Home Buyers”) and non-commercial content (“What Makes Us Stop Checking and Save Money: A Look Inside the Sewer Line”). The difference is that the content provides unique insight beyond the general knowledge.
On the technical side, pages must be indexed and caption-worthy to appear in AI-generated features. Google recommends following best practices for clarity, using semantic HTML where possible, following JavaScript SEO best practices, providing good page experience, and reducing duplicate content.
Site optimization and ecommerce gets its own category. Google recommends the Merchant Center feed and Google Business Profiles for product and local business visibility in AI responses. It also talks about the business agent, a conversational experience that allows customers to chat with products in Google search.
Agenttic Experience Get First Direction
The new category of agent experience describes AI agents as “autonomous systems that can perform tasks for humans, such as booking a place or comparing product information.”
Google notes that browser agents can access websites by analyzing screenshots, inspecting the DOM, and interpreting the accessibility tree. The guide links to web.dev’s guide to agent-friendly website best practices and refers to the Universal Commerce Protocol (UCP) as an emerging protocol that will “allow Search Agents to do more.”
Google announced UCP earlier this year, and Vidhya Srinivasan’s annual book said it was developed in collaboration with Shopify and more than 20 companies that endorse it.
Why This Matters
This guide provides Google’s clearest guidance yet on the dos and don’ts of the productive aspects of AI in Search. It consolidates previously scattered positions from conference talks, podcast appearances, and blog posts into a single reference.
The mythbusting part carries a lot of weight. Google now tells you in its documentation to skip the tricks that the growing AEO/GEO services industry has been promoting. That doesn’t solve the problem of non-Google AI platforms like ChatGPT or Perplexity, which might weight signals differently. But for Google’s AI features, the guide is now on the record.
The agent experience section lists browser agents and UCP in Google’s official documentation for site owners. The guidance is early, and Google is pitching it as optional for businesses where agent access is relevant.
Looking Forward
Google’s closing paragraph says you don’t need to accomplish everything in the document to be successful. It notes that “a lot of content thrives on Google Search (including artificial intelligence) without any obvious SEO at all.”
The agent’s experience guide is labeled as something to check out “if this is something related to your business and you have some extra time.” That suggests Google sees agent development as more forward-looking than urgent.
Featured Image: Anatolir/Shutterstock
