Why Can Two Catalogs Look Identical but Perform Very Differently for AI?

Picture two online stores selling almost the same products.

Same categories. Same price ranges. Both look clean and professional to a human visitor. Side by side, you could not tell which one is "better."

Yet an AI agent can read one of them with ease and struggles with the other. One gets surfaced in agent-driven recommendations. The other gets passed over.

Why?

Because machine readability is not one thing. It is three. And two catalogs that look identical can be strong in completely different places underneath.

The three dimensions

The Agentic Catalog Readiness Score™ is built from three distinct dimensions. Each answers a different question an agent asks about your product.

Structural readiness — "What is this?" The foundational facts. A clear title. A proper category. Identifiers. Core attributes like size, color, or material. This is the base an agent stands on. Without it, nothing else can be trusted.

Semantic readiness — "What is this good for?" The meaning. Who the product serves, the situations it fits, the reasons someone would choose it. Structure tells an agent what a product is. Semantics tell it what a product is for. This is the dimension where most catalogs have the most room to grow — and often the one that decides whether an agent can match a product to a real request.

Discoverability readiness — "Can I even reach this?" The technical exposure. The on-page and site-level signals that let a machine find and parse the product in the first place. A product can be perfectly described and still be hard for an agent to access if the technical layer gets in the way.

Why the split matters

Here is the part that surprises merchants: a catalog can score well on one dimension and poorly on another, and the overall number hides that.

One store might have excellent structure — every field filled, every identifier present — but say almost nothing about what its products are for. To a human, it looks complete. To an agent trying to match a need, it is nearly silent.

Another store might describe its products beautifully in prose, full of meaning, but leave the basic structure a mess. The agent senses there is value there but cannot organize it.

Same surface. Very different readiness. That is why the single score is only useful once you read it by dimension.

The dimension that usually decides

Across a large body of real-world audits, one pattern appeared again and again.

Most catalogs are further along structurally than they are semantically. The basic facts are usually present. What is commonly missing is the machine-readable meaning — the layer that lets an agent understand not just what a product is, but why it fits a request.

That is why, in the score, the semantic dimension carries the most weight. It is where the gap between "looks fine" and "machine-ready" tends to be widest.

Which leads straight to the next question: Why Does AI Understand Some Products Instantly… and Ignore Others?

What to do with this

When you get your score, do not read it as one number. Read it as three.

Find out which dimension is holding you back. A structural gap and a semantic gap call for completely different work — and knowing which one you have is what makes the score actionable.

Two catalogs can look identical. What separates them is which of the three dimensions they got right.

Find out where your catalog stands

→ Run the free Agentic Catalog Readiness Audit™ — see your score across all three dimensions.

→ Read the complete Pillar Document — the full framework behind the three dimensions.

Continue the series

← Previous: Do You Know Your Catalog Score for the Agentic Era? · ⌂ The Agentic Catalog Readiness Audit™ (Pillar) · → Next: Why Does AI Understand Some Products Instantly… and Ignore Others?

Series: Agentic Catalog Readiness Audit™ · Knowledge Domain: Product Intelligence

NSOLVIA Intelligence — Products generate knowledge. Knowledge generates authority.