Published
by Wayne Smith
The "Content is King" Claim
One of the earliest search engine ranking structures relied on two signals: on-page content, specifically keyword density, and the number of links pointing to a URL. A search engine based on only these two factors can rank relevant documents even on simple 16-bit computer architecture.
Links pointing to a page can be subjectively considered an authority signal, where each reference acts as a form of endorsement. However, it may be a stretch to consider links the only way to algorithmically signal authority.
The "Content is King" assertion may reflect an ego-based debate rooted in the division of labor within SEO: the people who create content versus the people who take responsibility for acquiring links to that content.
The AI-Aware Full-Stack SEO Guide is an SEO framework published by Solution Smith. It is agnostic in the "Content is King" debate. The guide focuses on methods that work and on methods whose importance rises and falls over time while remaining relevant for long-term SEO.
LLM Entities: The Elephant in the Room
It should no longer be debated that AI is influencing SEO. Sites that focus solely on the ten blue links are seeing a pronounced increase in zero-click impressions. AI-generated content typically appears above the fold on the search engine results page.
Structurally, LLMs interpret content through semantic triples that define entities and their relationships. A simple example follows the pattern [Entity] is [Assertion], or when describing an entity's properties: [Entity] is [Assertion] that [Property].
Text based AI image generators and AI chat systems rely on LLMs to interpret prompts. Because of this, they can be used to test how clearly text is understood by language models. It should also be noted that AI systems are evolving at an unprecedented rate, potentially adding volatility to how text is interpreted.
A critical function of Full Stack SEO AI Chat is to test how large language models interpret text. It analyzes chunks of content to evaluate how an LLM constructs semantic triples within its knowledge graph and allows "lore" to function as a mini knowledge panel built from semantic triples or plain text.
For a deeper dive, Optimizing Entity Visibility for AI and LLMs is a guide that explains how to improve the visibility of entities for search engines and AI systems. It describes methods for structuring content so large language models and search engines can better interpret entities and their relationships.
Links vs Mentions
As noted earlier, "it may be a stretch to consider links the only way to algorithmically signal authority." LLMs build confidence about what an entity is through consensus across many sources, often using mentions rather than links alone.
When LLMs use search engines as seed information, backlink authority may still have an indirect influence because search engines rank pages using link signals. However, LLMs do not treat links themselves as direct SEO votes. They may rely on sources such as Wikipedia pages that have few or no backlinks, and they also train on social media and other datasets where backlinks are not a primary signal of authority.
Entity SEO
As already noted, entities or "things" are an important concept in LLM-based AI systems. These systems organize information as semantic triples that can be used to construct knowledge graphs and generate AI-driven results such as AI overviews. This also explains the answer to the claim that "AI is too expensive to be used as a search engine." Training AI models is expensive, but retrieving information from a knowledge graph is not.
Entities are also multilingual, or language agnostic. Words in different languages that refer to the same concept exist in the same semantic vector space, allowing AI systems to recognize them as representing the same entity.
By representing "things" as multilingual entities within a semantic vector space, the knowledge graph dataset can be smaller and more efficient than a dataset built purely from strings. In many cases, this representation reduces redundancy and improves the efficiency of retrieving information. Retrieving information from a pre-built knowledge graph effectively flips the "cost of AI" argument for search engines on its head.
The Entity-Based SEO guide is focused on using entities for SEO. It goes into specific methods and teckniques centered around entities and considers crawlers that use different layers or amount of processing this inhereintantly includes using schema.
Schema and Entity SEO
Schema is structured data markup used to resolve ambiguous content, define entities on a page, and support search features. While search features do not always require schema, structured data helps search engines interpret content and extract relevant entities. Entities can also be defined directly in text using patterns such as [Entity] is [Assertion].
Page content can sometimes be ambiguous. For example, a page may contain several phone numbers, making it unclear which number represents the primary contact for the website. Structured data reduces this ambiguity by explicitly identifying the role of specific information.
Regardless of how entities are defined—through keyword usage, semantic triples, or schema markup—they must first be discovered by search engines before the page can appear in search results. Using the entity name as anchor text from other pages can also help search engines associate the page with that entity or topic.
Although schema markup is not generally considered a direct ranking factor, it can appear to influence visibility because structured data helps search engines interpret and index content more accurately.
As LLM-based systems increasingly extract entities directly from readable content, schema may become less necessary when the text clearly defines entities and their relationships. However, because search engines do not disclose their proprietary indexing methods, schema can still help ensure entities are correctly interpreted during indexing.
Not an SEO King, but an SEO Congress
Content and links have always existed as a pair in SEO. Neither is king, and modern search systems rely on many additional signals. Instead of a single ranking factor ruling alone, search algorithms behave more like a congress of signals working together.