Catalog quality, defined.

Report terms

Index

The overall 0–100 Catalog Quality Index for a store, with an A–F grade and a confidence band.

Grade

The letter grade derived from the index, from A+ down to F. A starts at 90; the full report uses plus and minus steps, and the preview shows a coarser grade.

Confidence band

How much of your catalog we could read. A wider band means less of the data was machine-readable.

Sub-index

One of the two halves of the index: storefront experience (how shoppers read your pages) and AI discovery (how agents and feeds read them).

Dimension

One scoring area within a sub-index — 13 in total, such as identifiers, structured data or category depth.

Pillar

A grouping of related dimensions within a sub-index.

Check

A single graded test on a product page — for example, whether it has a GTIN.

Problem

A check that failed across a share of your products, reported with how many are affected.

Punch list

The per-product table of every issue found, exportable to CSV. See one in the sample report.

Action plan

The prioritized list of fixes, ranked by reach and effort. See one in the sample report.

Reach

How many products a given fix would improve — one axis of the action plan's ranking.

Agent-reachability gap

Products an AI agent can't fully read — usually because of JS-only rendering, missing structured data, or login-gated content.

Preview

The free, roughly one-minute sample score on up to 20 of your public product pages, with no account and no card.

Profile

The scoring profile applied to your catalog: B2B/technical or consumer. Auto-suggested from your catalog and confirmed by you.

Methodology

The versioned set of weights, thresholds and standards that produce the score.

Family

The product family CatalogIQ detects for a product (for example apparel, HVAC/plumbing, consumer electronics), which decides which checks apply.

Gate

A rule that marks a check not-applicable when it doesn't apply to a product's family, so irrelevant checks never count against you.

Cap

The 5,000-product ceiling for the self-serve report. Larger catalogs are quoted.

Crawl

Reading your public product pages to score them. CatalogIQ crawls public pages only — see Our crawler.

Standard

A published data standard the score measures against: Google, Amazon, GS1, ETIM, UNSPSC and schema.org.

Attribute completeness score

A score Google is adding to Merchant Center to show which of your products are missing the structured product attributes its AI surfaces use (announced 27 May 2026). The CatalogIQ report shows the same kind of gap, product by product, from your live pages.

Product data standards and identifiers

GTIN

Global Trade Item Number: the barcode number that identifies a specific product worldwide. It is 8, 12, 13 or 14 digits long; a UPC is the 12-digit US form and an EAN the 13-digit form. GS1 issues them to brand owners. Google and marketplaces use the GTIN to match your listing to the right product.

MPN

Manufacturer Part Number: the maker's own code for a product. It identifies items that have no GTIN, and it is the main search key in B2B procurement, where buyers search by part number.

GTIN vs MPN

A GTIN is a global number from GS1 that is unique across all brands. An MPN is assigned by one manufacturer and is unique only together with the brand name. Publish the GTIN when the product has one, and the brand plus MPN when it doesn't. CatalogIQ checks for both.

UPC

Universal Product Code: the 12-digit barcode number used on products sold in the US and Canada. A UPC is a GTIN-12.

UNSPSC

United Nations Standard Products and Services Code: an eight-digit, four-level classification (segment, family, class, commodity) used in procurement, purchasing systems and spend analysis. It is owned by the UN Development Programme (UNDP). A UNSPSC code tells a buyer's system what kind of product an item is.

ETIM

An open classification standard for technical products such as electrical, HVAC, plumbing and building supplies. For each product class it lists the features a product should carry, with units and allowed values, so distributors and manufacturers describe products the same way.

GS1 GPC

Global Product Classification: GS1's system for grouping products into segment, family, class and brick, with attributes per brick. Retailers and suppliers use it to classify products when they exchange product data.

Google product category

A category from Google's product taxonomy, which Merchant Center uses to classify each product. Google assigns one automatically, and the category decides which attributes Google expects.

Amazon product type

Amazon's classification for a listing. Each product type comes with its own set of required and recommended attributes.

Attribute

A single, named fact about a product, such as material, voltage, size or color. Standards like ETIM, GS1 and Google's taxonomy define which attributes a type of product should have.

Product structured data

Machine-readable markup on a product page that states the product's name, price, availability, identifiers and reviews, so search engines and AI agents don't have to guess them from the page text. It is usually written as JSON-LD using schema.org types.

schema.org Product

The schema.org type that describes one product. Its price, currency and availability sit in a nested Offer. Google reads it for product rich results and merchant listings.

ProductGroup

The schema.org type for a product sold in variants, such as sizes or colors. It lists each variant as a Product (hasVariant) and names what changes between them (variesBy). Without it, agents may see only one variant, or a price and stock state that belongs to a different one.

JSON-LD

A format for adding structured data to a page as a separate block of JSON that browsers don't display. Google recommends it for structured data.

Product feed

A file or API that sends your product data to a channel such as Google Merchant Center, a marketplace or an AI shopping channel. Google compares the feed with your product pages and can disapprove items whose price doesn't match.

Google Merchant Center

Google's tool where merchants submit and manage product data for Shopping ads, free listings and AI Mode. It flags missing or invalid product data and disapproves items that break its rules.

PIM

Product information management system: software where a business stores and enriches product data before publishing it to the storefront, marketplaces and feeds. CatalogIQ scores what actually gets published, so it works with or without a PIM.

AI shopping agent

An AI assistant, such as ChatGPT, Gemini, Copilot, Perplexity or Google AI Mode, that researches and compares products for a shopper by reading product pages and feeds. Some can also complete the purchase.

Agentic commerce

Shopping where an AI agent finds, compares and sometimes buys products on a shopper's behalf. What the agent can read about a product, from its pages and feeds, decides whether that product gets recommended.

Shopify agentic storefronts

A Shopify feature that makes eligible products available to AI channels such as ChatGPT, Microsoft Copilot, Google AI Mode and Gemini. Shopify turns it on by default for eligible stores.

UCP and ACP

Two protocols for AI agents to shop with merchants: Google's Universal Commerce Protocol (UCP) and the Agentic Commerce Protocol (ACP) from OpenAI and Stripe. Both depend on accurate, complete product data.