For many brands, optimizing for AI search has felt less like a clear strategy and more like an evolving set of hacks, gimmicks, and guesses.

These systems are anything but transparent, so it’s understandable why the marketing industry has been running in circles trying to crack the code to AI search optimization.

Recently, Google released its official guidance for optimizing for AI features in Google Search. It confirmed what many of us already assumed, while calling several popular AI search tactics into question.

Here’s what Google says you should do, what you don’t need to do, and where VELOX stands on current GEO strategies.

Much like its Organic Search guidance, Google doesn’t provide a clear instruction manual for improving AI visibility, but its guidance can still help shape your strategy and priorities.

Here’s what Google recommends for AI search optimization.

To succeed in AI features (and really anywhere online), your content must stand out. These systems pull from a variety of sources to provide an answer, and they tend to ignore content that reads like a rehash of what’s already been written.

Instead, provide a unique point of view that can make people think deeper or answer a question nobody else has thought to address. Quality content isn’t about recycling what’s already out there; it’s about leveraging your unique experience to further the conversation.

There are countless articles and web pages targeting broad, common knowledge topics. Think “7 Tips for New Homeowners” (Google’s example). This type of content could originate from anyone, meaning it isn’t necessarily valuable.

Helpful, high-quality content has been a key part of Organic Search success for years, so it’s no surprise this same guidance carries over into the AI era. The goal is to add unique insight and help the user. There should be a clear takeaway that makes your content worth reading, whether by a real-world user or an AI system surfacing information.

Above all else, brands should write content for humans. Agentic AI is evolving, and it may change how users discover, compare, and interact with websites. For now, though, Google’s guidance is clear: write content for human readers first. It should be organized in a way that is easy to follow, with clear headings, structure, distinct sections, and images and graphics to support the text.

Technical SEO is crucial for Organic Search rankings, and remains vital for AI search visibility. It’s the foundation upon which all other on-page optimizations are built. Without a solid technical SEO framework, other optimizations may never get the visibility they deserve, as technical SEO supports discovery, crawling, indexing, and page experience.

Plus, technical SEO is just good for the user experience. E-commerce brands should pay close attention to site speed, navigation, and stability, as these factors can impact the on-site shopping experience.

As far as AI search optimization goes, Google recommends the following:

All of these optimizations help with discovery and indexing in traditional Search and carry over to Google’s AI features.

If you look at these recommendations as a whole, they might seem familiar. That’s because they largely reflect the same Organic Search best practices that top-performing brands have implemented for years.

Google even admits this in the guide: The best practices for SEO continue to be relevant because our generative AI features on Google Search are rooted in our core Search ranking and quality systems.

This essentially confirms SEO is not dead, and optimizing for Organic Search is more important than ever.

However, it’s important to differentiate SEO and GEO, since their goals are vastly different.

SEO aims to improve Organic Search rankings, which can be directly attributed to revenue growth. Meanwhile, GEO is an additional layer that helps improve the likelihood of appearing in AI search features.

In addition to providing recommendations, Google took the opportunity to list optimizations it says are unnecessary for visibility within its AI features, framed as AI search myths.

Some marketers have treated llms.txt files as the key to being featured in AI search results. However, Google clarified in this guide that it doesn’t treat this file (along with others) any differently. Google discovers, crawls, and indexes many types of files in addition to HTML, so it doesn’t play favorites with any type, at least according to this guidance.

However, just because Google says it doesn’t care about llms.txt files doesn’t mean it isn’t important to the broader AI ecosystem (more on this later).

If you’ve noticed that blog content has begun to read less like one cohesive piece and more like a series of articles and FAQ blocks pasted together, you’re not alone. This is called “chunking” content, or separating content into tiny blocks that stand on their own.

Marketers have theorized that chunking makes content more easily discoverable by AI systems. However, Google says this is unnecessary and that its crawlers can understand the nuance of multiple topics on a page. Instead, Google recommends you continue with a human-first approach and organize content based on the audience and subject matter rather than hoping AI picks up just one or two sections.

Similar to chunking content, Google states that you don’t need to write content with AI search in mind. This means overleveraging long-tail conversational keywords, including unnecessary definitions for words and phrases a user could reasonably understand, or adding extra variations of particular terms.

Google claims its AI systems can understand synonyms and general meanings well enough to connect a user query to a piece of content, whether that content uses the exact phrasing or not. Focus on writing human-first, helpful content, and if your technical optimizations are implemented, AI crawlers should find it.

Sometimes Google’s generative AI features don’t display first-party information about a product or service and will instead cite mentions from other sources. These mentions could be from blog posts, videos, or forums.

While it’s natural to want as many mentions as possible, Google explicitly warns against seeking out or purchasing inauthentic mentions. Just like you shouldn’t participate in black-hat link schemes, Google’s systems are trained to prioritize helpful, original content and limit unhelpful, spammy content.

Structured data helps crawlers better understand the meaning, context, and purpose of a page. However, according to Google, it’s not essential to surface in AI search features, and there isn’t an AI-specific markup to add. Instead, they recommend you prioritize crawler-friendly technical optimizations.

So, because Google says you don’t need to worry about these, you can delete them from your strategy, right? If all you care about is appearing in Google’s AI search features, such as AI Overviews and AI Mode, you might be able to cross these off the GEO to-do list.

However, it’s important to remember Google is just one destination in the larger AI ecosystem. It sets the rules for its own AI features, but ChatGPT, Claude, Perplexity, and Copilot also exist and make up a considerable portion of the current AI market share.

That’s why VELOX remains confident in our existing strategy, both to help brands expand their footprint across all surfaces today (including Organic Search) and to help hedge against future changes.

At VELOX, we treat GEO as an executable layer on top of Organic Search, although Organic Search still receives priority.

That’s because Organic Search is a proven channel, even with the rise of AI search features. While AI visibility is valuable and provides additional surfaces for user discovery and engagement, conversion rates remain minimal, while we can directly attribute revenue growth to Organic Search.

Even still, GEO is a core service, and we approach it holistically with an eye toward the future. Instead of relying on what works now or on Google's recommendations for its AI features, we follow the data, iterate based on results, implement proven best practices, and experiment with emerging tactics.

Although this guidance from Google is useful, it doesn’t necessarily change how we approach GEO.

Google doesn’t speak for ChatGPT, Claude, Perplexity, and all other AI systems, so it’s important to view this guidance through that lens.

For example, Google claims llms.txt files are unnecessary, but they might be important for other AI tools. That’s why we still provide them for clients. The cost is minimal, but the optionality is well worth it.

The same can be said for schema markup. Even though Google says structured data is not required for its AI search features, that could change in the future. It’s well worth the effort to hedge against future developments so client sites are well-positioned in the long run.

Beyond these two examples, we continue to prioritize first-hand, expert-written content for client campaigns. AI can be used as an optimization tool, but it can’t replace the expertise of a human content writer. This ensures clients receive helpful, human-first content that provides unique insights and helps achieve their Organic Search goals.

Organic Search rankings are still crucial for revenue growth, and the channel remains the most durable. AI search visibility has its benefits and could become a significant channel in the future, but for now, we emphasize tactics that deliver tangible, dollars-and-cents results for our clients.

With Organic Search, technical SEO, and AI search working together, brands can dominate high-value surfaces across the entire ecosystem.

Ready to strengthen your Organic Search and AI search strategy? Contact us today.