Why Knowledge Architecture Matters More Than the Next AI Protocol for SEO

Search Engine Journal contributor Bill Hunt has issued a pointed warning to SEO professionals and enterprise marketing teams: publishing the latest AI protocol will not fix a broken content strategy. In a detailed analysis published on September 2, 2026, Hunt argues that the industry is repeating an old cycle of mistaking format for substance.

Hunt, president of Bisan Digital, describes a familiar scenario. A senior executive receives an alert from an AI visibility assessment vendor. The recommendation includes implementing an llms.txt file – a still-debated publishing format designed to help AI models discover an organisation’s content. Suddenly, an emerging technical specification becomes an executive mandate, and teams scramble to assess its validity and impact.

Por que importa

Hunt calls this dynamic part of what he terms the “AI FUD Tax.” The cost of a single recommendation may be small, but the cumulative organisational cost of responding to every new AI audit, vendor pitch, protocol, acronym, or competitive claim can become substantial.

The core problem, according to Hunt, is not the technology itself. Whether it is llms.txt, the Model Context Protocol (MCP), or markdown, each serves a different purpose. Some help machines discover information, some provide alternative representations, and others define how systems exchange data. Lumping them together is technically inaccurate. But strategically, they create the same organisational temptation: treating the latest delivery mechanism as the solution rather than examining the underlying knowledge.

Contexto

Hunt believes the industry is once again focusing too much attention on the format and not enough on the information those formats are supposed to communicate. He advises organisations to ask a more fundamental question before implementing any new protocol: “Do we have the knowledge required to support it?”

This distinction matters because AI creates more ways for machines to consume organisational information. If the underlying knowledge is incomplete, fragmented, inconsistent, or trapped inside individual departments, adding another machine-readable format does not solve the problem. It simply creates another place to publish the same limitations.

Hunt introduces the concept of “Decision Coverage” as a way to measure how completely an organisation has exposed the evidence AI needs to evaluate, compare, qualify, and confidently recommend its products or services. He argues that many organisations have enormous amounts of product information yet lack the evidence AI needs to support an actual customer decision.

Consider a customer asking for the best family-friendly beachfront resort in Cancun. “Best” is not an attribute a hotel can simply add to a page. The recommendation may depend on beachfront access, family suitability, room configuration, amenities, price, availability, reviews, and other criteria. The AI evaluates these conditions collectively before deciding which properties qualify.

Decision Coverage approaches this problem from the organisation’s side. Once the variables influencing a decision are understood, an organisation can determine whether it has authoritative evidence to support them. If a critical criterion cannot be substantiated, the problem may not be that the brand ranked poorly. It may never have provided enough evidence to make the cut.

Hunt argues this gives organisations a more defensible way to diagnose a lack of AI visibility. Rather than observing that a competitor was recommended and immediately responding with more “me too parity content,” more links, or another complex technical implementation, teams should deconstruct the decision, identify the criteria influencing qualification, and determine where supporting evidence is incomplete.

The temptation, Hunt warns, is to put missing information into whatever format is receiving attention at the moment. Daily LinkedIn posts recommend expanding schema, creating a markdown version, building an MCP endpoint, or deploying llms.txt. That may solve an immediate publishing problem, but it does not solve the underlying knowledge problem.

If a customer decision depends upon five meaningful criteria and the organisation can substantiate only four, publishing those same four pieces of evidence through another protocol does not suddenly establish the fifth. The same evidence gap is simply made available in another format.

Hunt’s analysis suggests that the organisations best positioned to adapt will not necessarily be those that implement every new protocol first. They will be those that organise and govern their knowledge well enough that supporting the next useful format becomes a publishing decision rather than another reconstruction project.

Datos clave

Aspect	Key Point
Core argument	Publishing a format is not the same as possessing the knowledge
Recommended metric	Decision Coverage: measuring evidence for AI decision-making
Key risk	Organisational cost of responding to every new AI audit or protocol

For Indian SEO professionals and digital marketing teams, Hunt’s analysis carries a clear implication. The race to adopt llms.txt or MCP should not distract from the harder work of building structured, authoritative knowledge that answers real customer questions. The next AI protocol will not save a strategy built on weak foundations.

Source: Search Engine Journal – The Next AI Protocol Won’t Save Your SEO Strategy via @sejournal, @billhunt (https://www.searchenginejournal.com/the-next-ai-protocol-wont-save-your-seo-strategy/585714/)