Google reiterates the demand for unique content

The myths Google dispels

The guide (Optimizing your website for generative AI features on Google Search) opens by taking the edge off a claim that has spread through the SEO world. GEO and AEO sound like new disciplines, but Google describes them as SEO under a different name. If you are already working with search optimization, you do not need to start over.

Later in the guide they address some industry myths that have no basis in how Google actually works. To summarize, they essentially say:

• You do not need to create LLMS.txt files. The file, which is supposed to signal to AI systems what they may read, is neither required nor preferred by Google.
• You do not need to split content into short chunks. "Chunking", deliberately breaking content into short sections to make it more AI-friendly, has no effect. Google does not need help navigating a full page.
• You do not need to rewrite existing text for AI systems. Google understands the meaning of what you write, not just the exact words. Tailoring text specifically for AI has no effect.
• You do not need to seek exposure on third-party sites. Chasing mentions purely to appear in AI answers pays off just as little as it does in regular search.
• You do not need to reprioritize structured data. Schema Markup is not required to appear in an AI search, but it is still useful as part of a broader SEO strategy.

What Google means by commodity content

In the guide Google mentions the term non-commodity content, but the meaning is more complex than their own examples suggest.

ALM Corp, who have analyzed Google\'s guidance in depth, formulate the distinction more clearly: commodity content is content that is easy to reproduce. It covers a familiar topic in a familiar way, with the same structure and the same general advice found on hundreds of other sites. It does not have to be incorrect or poorly written, just interchangeable.

As we know by now, AI systems are very good at compressing repeated information. Ten pages saying roughly the same thing can be summarized in a single AI answer. That, in turn, leaves little room for the eleventh page saying exactly the same thing.

Harry Clarkson-Bennett, an SEO specialist who writes a newsletter for Leadership in SEO, points to a Google patent that gives a more concrete picture of how this is measured. The patent describes an information gain score per landing page, meaning how much an individual page contributes compared to what already exists on the same topic. Pages that repeat what is out there receive a low score, while those that add something new are rewarded.

Why content gets stuck in the commodity trap

In many organizations, the people with the most niche knowledge — managers, sales teams and specialists — sit far from those who produce the content. Without a deliberate process to capture that knowledge, the blog fills up with material that could be produced from publicly available information. The result is well-written text containing content anyone could have produced.

ALM Corp also highlights AI language models as a risk factor, because they are trained on the average of what is available online. Used passively, without feeding in specific internal knowledge, they contribute to more generic content.

Think of it this way — if a competitor could have written exactly the same text as you, without being particularly informed or familiar with your company, your text is likely to be classified as commodity content.

What makes a content page unique

According to ALM Corp's review, non-commodity content shares five characteristics:

1. 1

   It is specific and deals with a real situation rather than a general pattern.

2. 2

   It is experience-based and built on observations, tests or decisions.

3. 3

   It is hard to copy, because it relies on internal data, customer knowledge or expertise competitors cannot easily replicate.

4. 4

   It interprets what facts mean, which trade-offs weigh the heaviest and what changes the answer in different situations.

5. 5

   It answers the follow-up question after the follow-up question, not just the most obvious one.

Take Google's own example. A real estate agent who explains a concrete decision about a specific sewer line on a house gives the reader something they will not find on another site. Articles with general buyer tips, like "7 tips for first-time buyers", do not.

Unique content is therefore not a question of length or complexity. Most of all, it is about whose experience the text is built on.

Technology is still the foundation

In the guide Google confirms that technical SEO is still just as important. A page has to be indexed and crawlable to even be considered for AI answers, and sites with poor structure or duplicated content struggle to compete no matter how good the content is.

Something worth knowing is how the AI system handles queries. When someone searches, Google automatically generates a number of related follow-up questions and fetches results for all of them. This is called query fan-out and it means that a page covering a topic in depth can show up in answers to questions the page was never written to address. Breadth and substance pay off in a different way than when you optimize for individual keywords.

If you run an e-commerce site or a local business, it is also worth knowing that Merchant Center and Google Business Profile continue to be relevant. Google's AI answers can include product listings and business information, so keep those channels up to date.

What should you focus on now?

Visibility in AI search largely follows the same logic as traditional search: whoever adds the most value wins the most. The difference is that the bar for what counts as valuable content rises when AI can summarize generic information without linking onward. If you are already working to make your content more specific and experience-based, you are on the right track.

Frequently asked questions