The Agentic Catalog Readiness Audit™

1. Executive Summary

The Agentic Catalog Readiness Audit™ measures one thing: how ready a product — within an e-commerce catalog — is to be understood and chosen by AI systems.

A growing share of buying now happens through AI agents. These agents do not browse the way people do. They read structured product data. When that data is incomplete or unclear, the product becomes hard for an agent to understand, and easy to skip.

The Audit evaluates a product as the merchant published it, and produces a single score — broken into three dimensions — that shows where that product stands today and where it could stand with better preparation. Read across a representative sample of products, it reveals the readiness of the catalog as a whole.

This framework was developed and refined through the analysis of 253 real-world catalog audits across Shopify, WooCommerce, Wix, and Squarespace. The tool is the product of that work. The audits give it legitimacy.

This document explains what the Audit measures, why it exists, how to read the score, and how to act on it. It is written for merchants, agencies, and the teams responsible for product catalogs.

2. Why This Matters

Commerce is shifting from human browsing to agent-mediated buying.

When a person shops, they forgive messy data. They infer. They guess. They click around.

An AI agent does none of that. It selects products based on what the data says, clearly and structurally. It does not reward brand familiarity or advertising spend. It rewards clarity.

This levels the field. A small merchant with clean, well-structured product data can be chosen by an agent over a larger competitor with disorganized data.

It also creates a new risk. A catalog that is not prepared for agents becomes invisible to a growing portion of buyers. Not penalized. Invisible. The agent simply cannot understand the product well enough to recommend it.

The principle is simple: data readiness comes first, protocols come second. A store can connect to every agentic platform available, but if its product data is unclear, those connections lead to products no agent can confidently surface.

The Audit exists because most catalogs were never built with this in mind.

3. Background

For two decades, product discovery was organized around search engines and human readers. The discipline that grew around it optimized for ranking and for clicks.

That era assumed a human at the end of the journey. A human who could interpret a vague title, tolerate a thin description, and fill in the gaps.

Agent-mediated commerce removes that human in the middle. The agent becomes the reader. And the agent needs the catalog to say, explicitly and in machine-readable form, what a product is, what it is for, and why it fits a request.

This is a different requirement from anything that came before. It is not about ranking higher. It is about being understood at all.

NSOLVIA describes the layer that makes this possible as the Semantic Commerce Layer™ — the connective tissue that turns raw product data into something machines can reason about. The Audit is the first point of contact with that idea. It shows a merchant, concretely, how legible their catalog is today.

4. Definitions

Agentic Catalog Readiness Score™ — A score from 0 to 100 that expresses how ready a catalog is for AI agents. Higher means more legible to machines. The score is divided into three dimensions.

Structural readiness — Whether the foundational facts of a product are present and well-formed: clear titles, categories, identifiers, and core attributes. This is the base an agent stands on.

Semantic readiness — Whether the meaning of a product is expressed in a way a machine can use: what it is for, who it serves, the situations it fits. Structure tells an agent what a product is. Semantics tell it what a product is good for. This is where most catalogs have the most room to grow.

Discoverability readiness — Whether the product is technically exposed to machines: the on-page signals and site-level structures that let an agent find and parse the product in the first place.

Machine readability — The umbrella idea behind all three: can a non-human reader extract, understand, and act on this product's data without guessing.

What the Audit Is Not

The Audit does not evaluate SEO rankings. It does not measure advertising performance. It does not score conversion optimization or sales funnels.

It evaluates one thing: the machine-readability of a product catalog.

This matters because the market is full of tools that measure visibility to people. The Agentic Catalog Readiness Audit™ measures readiness for machines. They are not the same problem, and they are not solved the same way.

5. How the Framework Was Developed

The Audit framework was not designed in the abstract. It was developed and refined through the analysis of 253 real-world catalog audits.

Those audits spanned four major commerce platforms — Shopify, WooCommerce, Wix, and Squarespace — and a wide range of product types, from apparel and footwear to food, beauty, and home goods.

Each audit looked at a real product, exactly as a merchant published it, before any enrichment. This is an important discipline. The framework was tuned against the catalog as it actually exists in the wild, not against an idealized version of it.

The evaluation works on a before-and-after basis. It captures the readiness of a product as it stands today, and the readiness it could reach once its meaning is properly prepared for machines. The distance between those two states is where the opportunity lives.

This grounding in real catalogs is what gives the Audit its authority. The dimensions, the weighting, and the score ranges were shaped by what repeatedly appeared across hundreds of live products — not by assumption.

The specific distributions, comparisons, and statistical findings from this body of audits are the subject of forthcoming Research Papers. This document focuses on the framework those audits produced.

6. What We Consistently Observe

Across these audits, a few patterns appear again and again.

Most catalogs are structurally further along than they are semantically. Merchants tend to provide the basic facts of a product. Far fewer express, in machine-readable form, what the product is for.

Category information is a frequent weak point. It is commonly missing, placed inconsistently, or expressed in ways a machine struggles to resolve. When a product's category is unclear, much of its meaning becomes hard to organize.

Product descriptions are often written for people, not machines. They lean on tone and persuasion, and leave the concrete, usable signals — materials, use cases, intended situations — implicit rather than stated.

Identifiers and standardized attributes are commonly incomplete.

And platforms themselves contribute to the problem. The way different commerce platforms structure and expose product data varies, and some make a product's meaning harder to read than others.

These are observations, not statistics. They are the recurring shape of what we see. The precise measurements live in the Research Papers.

7. Analysis

The recurring lesson across these audits is that good structure is not the same as readiness.

A catalog can look complete to a human and still be hard for a machine to understand. A product can have a clean title, a price, and an image, and still leave an agent unsure what the product is actually for.

This is why the Semantic dimension carries the most weight in the score. It is the dimension where the gap between “looks fine” and “machine-ready” is widest, and it is the dimension that most directly determines whether an agent can confidently match a product to a buyer's request.

Structure gets a product into the room. Semantics is what lets the agent choose it.

The implication for merchants is that readiness is rarely about adding more text. It is about making meaning explicit — turning what is implied in a human-written page into signals a machine can act on.

8. Findings

The central finding of this work is that the Audit makes a previously invisible problem visible.

Before an audit, most merchants have no way to know how legible their catalog is to AI systems. The catalog looks finished. Sales may even be healthy. The gap is silent.

The Audit surfaces that gap. It shows, dimension by dimension, where a catalog is strong and where it is not yet machine-ready. It shows the difference between a product's current readiness and the readiness it could reach.

It also reframes the conversation. The question stops being “how do I rank” and becomes “can an AI system understand this product.” For a merchant preparing for agent-mediated commerce, that is the more useful question.

The tool, in other words, does not just produce a number. It gives a merchant a clear, structured view of a problem they could not previously see — and a direction for closing it.

9. Industry Implications

Agent-mediated commerce changes who a catalog is written for.

For merchants, this means catalog readiness becomes a competitive variable in its own right. The merchants who prepare their data early will be the ones agents can confidently surface. The cost of waiting is rising invisibility.

For agencies and platforms, it means a new category of work: preparing catalogs not for human search, but for machine comprehension. This is adjacent to existing services, but it is not the same skill.

And for the ecosystem as a whole, it points toward a healthier dynamic. When agents choose based on clarity rather than spend, well-prepared small merchants can compete on the merit of their data. Readiness becomes something a merchant can earn.

10. Recommendations

Start by running the Audit. It costs nothing to see where a catalog stands.

Read the score by dimension, not just as a single number. A strong structural score with a weak semantic score tells a very different story than the reverse, and points to a different kind of work.

Treat the Semantic dimension as the primary opportunity. For most catalogs, that is where the largest gains live.

Prioritize the products that matter most — the ones that drive the business — and prepare their meaning first.

And treat readiness as ongoing, not one-time. Agentic commerce taxonomies are still forming. A catalog prepared today will benefit from continued attention as the standards mature.

The Audit is the starting point. It tells a merchant where they stand. What comes next is preparation.

11. Limitations

This framework is honest about what it does and does not cover.

The Audit evaluates a representative sample of products, not necessarily an entire catalog in its free form. A sample reveals the shape of a catalog's readiness; it is not a line-by-line inventory.

It is a snapshot in time. A catalog changes, and so does its readiness.

The taxonomies that agentic commerce relies on are themselves emerging. Some product categories are still finding their place in machine-readable standards. A product in an emerging category may score lower today simply because the broader ecosystem has not yet matured around it — and its readiness can improve as that coverage expands.

Stating these limits is part of the method. The Audit is a clear instrument, not an absolute verdict.

12. Future Work

The Audit is the entry point to something larger.

NSOLVIA is building toward a Product Intelligence Platform — an ecosystem for making commerce data legible, usable, and ready for an agent-driven market. The Audit is how a merchant first sees the need. It is not the end of the path.

The next natural step in that ecosystem is AI Feeds™ — the operational layer that takes a catalog from “readable” to “delivered,” preparing product data for the agents and platforms that consume it. A forthcoming Pillar Document will cover AI Feeds™ in full.

Alongside the products, NSOLVIA Intelligence will publish a series of Research Papers grounded in real audit data — beginning with the gaps this work has surfaced in category, identifier, and semantic readiness. Those papers will carry the detailed measurements this Pillar deliberately leaves aside.

Two directions are worth naming. First, deeper, multi-product audits — evaluating a larger share of a catalog in a single pass — are part of the platform's roadmap. Second, the framework's principles apply to any structured product catalog, not only e-commerce storefronts; the same readiness logic can extend to catalogs delivered in other formats as the platform matures.

The Audit answers a question. The platform answers what to do about it.

13. References

McKinsey & Company — projections on agentic commerce and AI-driven retail revenue.

eMarketer — forecasts on AI platform share of retail spending.

Industry reporting on agentic commerce readiness and the primacy of data readiness over protocol adoption.

NSOLVIA Intelligence — The Semantic Commerce Layer™ (Whitepaper).

Full citations to be finalized at publication.

14. Continue Exploring

The Semantic Commerce Layer™ — the foundational concept behind catalog readiness.

Companion articles — a series expanding each dimension of the Audit for practitioners.

AI Feeds™ — the next step: turning a ready catalog into delivered, agent-consumable data (forthcoming).