AI Search Content Strategy: A 3-Level Guide

Traditional search engines use many ranking signals to give users the most relevant results. In the AI era, it appears that signals are currently simpler. According to court documents from Google’s antitrust case, AI Overviews depend on fewer documents in the index and lighter signals. However, topical connections and content clarity are more crucial than ever. Content must be discoverable in the age of AI search.

In other words, the same SEO framework that has been in place for years – architecture, optimization, and expert content – remains essential. This article will show how each component can strengthen the most crucial signals in AI search. The court case U.S. v. Google LLC describes how Google’s FastSearch is used to ground the generative AI answers from Google. The technology uses lighter ranking signals and retrieves a smaller subset of search results.

According to the court documents:

This might explain why some AI Overviews have shown low-quality content. The court described RankEmbed as one of Google’s “top-level” deep-learning signals that is capable of “exploiting and finding patterns in vast data sets.” RankEmbed aims to recognize semantic relationships between documents and queries to capture meaning.

Unlike other signals that count backlinks or measure popularity, RankEmbed hones in on how closely content aligns with a user's question. RankEmbed is an important part of FastSearch. This may explain why FastSearch, which grounds AI Overviews, sometimes turns up results that differ from typical search rankings. It focuses on semantic clarity over authority signals.

If RankEmbed is central to FastSearch, then the factors that improve AI visibility are those that strengthen semantic connections. These strategies will be discussed next.

Information Architecture

The way you organize content on a site establishes relationships, making it simpler for AI systems and search engines to understand relevance. A clear information architecture also enables website visitors to engage more fully with a website, giving complete answers to questions during their search. This is not a new idea. Our agency has been using SEO siloing for 25 years. However, in the age of AI, it helps signal semantic clarity and contextual relevance for a site.

So, what is it all about? Let’s quickly examine this strategy in action. While it appears simple, the initial step offers surprising clarity when organizing a website’s information. It involves analyzing a brand’s products, services, and the types of information provided. This data serves as a starting point for further investigation. This research includes identifying the topics and queries for which the site is currently visible in search results, as well as identifying missed opportunities. If the website were a book, what would the story be? What chapters would fit naturally, and which are missing?

This research provides:

• Complete keyword list.
• Gap analysis.
• Site architecture outline.

Silos

The structure is called a “silo”. It keeps topical themes closely linked without mixing them with other topics that weaken relevance. When topics are mixed, it confuses both users and search engines. Organizing them brings clarity. The structure can be created in two different ways: through the physical directory (URL structure) or through virtual connections (internal linking). Let’s examine these two tactics more closely.

Physical Silos

Physical silos structure URLs to create a hierarchy of topical relevance. If a company’s primary focus is CRM software, then the overarching theme would be “CRM Technology.” Categories could be based on the major functions or use cases customers care about, like:
• Sales automation
• Customer support tools
• Lead generation

These are three silos. Each category then gets its own landing page, which is supported by subpages and subcategories that cover specific solutions or features. For example, the sales automation silo might include:
• Email automation
• Contact scoring
• Workflow triggers

For this example site, this structure signals to search engines and LLMs that the site is a valuable source of information about CRM software.

Virtual Silos

Virtual silos connect related pages through internal links, even if the pages are not in the same directory. This works well when the physical directory cannot be changed. It’s also a good hybrid approach for creating relationships between content when it lives outside of the physical directory, like in a blog. As an example, a blog post on a CRM website about improving customer retention might link to the primary landing page for customer support tools, even if the blog is in a separate /blog/ directory. These contextual links create virtual silos. The internal linking connects the blog back to the main theme and strengthens topical authority, even if the blog is not physically nested under /crm/. Website visitors have more opportunities to follow links and engage more deeply with the site as they continue on their journey.

AI models tend to retrieve content that demonstrates clear topic coverage and sites that show semantic consistency across related pages. However, a lot of content about a topic spread across the site will not build authority on its own. It needs to be properly linked and match intent. The siloing strategy might seem simple, but there are many technical factors to consider. And then there's the disruption. Siloing a large website is difficult. In these situations, partnering with an expert can be essential.

Expert Content

We are entering a world where AI content is starting to dominate search results. Some think expertly written, human-generated content will soon be highly valued. If that is the case, it will be easy to stand out from generic, machine-generated content by putting in the effort. When creating content for silos, each page needs to expertly and completely answer a query, as well as provide additional resources to other areas of the site. Consider how to elevate content, not just make it more efficient. This will again become a challenge for many in the age of AI. However, there are still guideposts for doing this. Let’s examine the baseline content strategies that will help a brand remain an expert.

Hire professional writers, even to refine AI content. They should naturally know how to write well and navigate things like:
• Writing in a brand voice
• Ensuring consistent quality
• Adding nuance that machines often miss

Most people are now aware of Google’s guidance on helpful content, which includes helpful self-assessment questions. Marketing leaders should use these as basic guidelines when doing an editorial review of content created by an SEO/GEO program.

E-E-A-T is a holistic framework for content quality. E-E-A-T must be earned over time by consistently providing value and developing genuine trust with an audience. It is most important for “Your Money or Your Life” (YMYL) topics. You can improve E-E-A-T by:

• Citing credible sources.
• Highlighting author expertise.
• Getting reviews.

The bottom line is to write for people and build trust over time. Ultimately, it comes down to creating good results for searchers. Subject matter expertise is essential for ranking high in search. Good SEO is needed for good GEO, according to Google’s Danny Sullivan. There is often overlap between AI Overview inclusions and high-ranking search results, even though the underlying systems use different signals. There is a good chance that if you are trusted in search, you will be trusted for inclusion in Google’s generative AI answers.

Some may argue that poor content written by machines is already visible in AI Overviews, so why bother? First, because of ethics. Ethics should be upheld in marketing. Second, algorithms can change, and content can disappear from search results. If everything has been done correctly, you will be catapulted to the top when that happens.

Schema

Structured data, or schema, is a layer that translates content into signals that machines can better understand. This can make it easier for AI systems to connect the dots. Some research indicates that it may not directly affect AI Overview visibility. However, Google advises using structured data to ensure content performs well in Google’s AI experiences, so it is a practice that requires experimentation.

Google says:

Many in the industry are already implementing or planning to implement schema/structured data as part of their GEO strategies. Structured data/schema was most frequently mentioned as a way to optimize for AI search. The challenge will be to implement the schema methodically. For example, don’t only use schema on the homepage or a few products; add it where it makes sense. However, don’t overuse or abuse it. Structured data should accurately represent the main content of the page, so choosing the most relevant structured data type for the content is crucial. (Google states that structured data issues can trigger a manual action.) Above all, test and validate.

AI Overviews rely on understanding entities like people, places, products, organizations, and concepts. Structured data defines these entities, making it easier for search engines like Google (and its AI-powered technology) to trust the information. While not new, Schema.org has been around for almost 15 years and is helpful for clarifying content, even if its impact on AI visibility is still being tested. Structured data is resurfacing as a way to reduce ambiguity in the AI search era.

What does it take to stand out in AI-powered search? The answer is clarity. Clear structure, signals, and expertise help search technologies and audiences connect the dots. This is the groundwork that AI systems and search engines depend on. The fundamentals that have guided SEO for decades remain the best approach. Focusing on information architecture, creating expert content, and using optimization techniques like schema helps create connections that search engines, AI systems, and people can rely on.