# catalogaudit

> Free audit that shows how Google Shopping (Merchant Center) and AI shopping agents (ChatGPT, Perplexity, Google AI Mode) read a Shopify store's product catalog, and rebuilds the five worst products as proof.

catalogaudit reads only public storefront data — /products.json, the storefront Ajax API, each product page's JSON-LD, robots.txt and policy pages — and works with headless Shopify storefronts. It produces two scores out of 100:

- Ad feed health: what gets products disapproved or throttled in Google Shopping and Meta catalogs (invalid or missing GTINs, structured-data prices and stock that don't match the store, missing apparel attributes, image size, promotional titles, zero prices).
- AI agent readability: what makes products invisible to AI shopping agents (no structured data or data only JavaScript can see, undescribed variants, specs buried in prose, robots.txt blocking AI search crawlers, missing policy pages).

Scores of 80–100 are good, 50–79 need work, and under 50 are poor. For the lowest-scoring products, catalogaudit outputs corrected JSON-LD, Shopify metafieldsSet input and Merchant Center feed rows. Running an audit is free at https://catalogaudit.app/.

## Published reports

- [Brooklinen catalog audit](https://catalogaudit.app/reports/brooklinen.com.md): ad feed health 86/100, AI agent readability 56/100, 437 products. Top issue: Product structured data only appears after JavaScript runs.
- [Allbirds catalog audit](https://catalogaudit.app/reports/allbirds.com.md): ad feed health 75/100, AI agent readability 44/100, 691 products. Top issue: No material, measurements or specs in the product data.
- [Death Wish Coffee catalog audit](https://catalogaudit.app/reports/deathwishcoffee.com.md): ad feed health 75/100, AI agent readability 55/100, 147 products. Top issue: No material, measurements or specs in the product data.
- [Rothy's catalog audit](https://catalogaudit.app/reports/rothys.com.md): ad feed health 94/100, AI agent readability 93/100, 749 products. Top issue: Missing description.
- [Taylor Stitch catalog audit](https://catalogaudit.app/reports/taylorstitch.com.md): ad feed health 83/100, AI agent readability 76/100, 1,974+ products. Top issue: No material, measurements or specs in the product data.

## Findings across published audits

- 5 of 5 audited stores: No material, measurements or specs in the product data (https://catalogaudit.app/checks#specs_absent)
- 4 of 5 audited stores: Specs buried in prose, not structured (https://catalogaudit.app/checks#specs_unstructured)
- 3 of 5 audited stores: Missing description (https://catalogaudit.app/checks#description_missing)
- 3 of 5 audited stores: Very thin description (< 150 characters) (https://catalogaudit.app/checks#description_thin)
- 3 of 5 audited stores: No product type (https://catalogaudit.app/checks#product_type_missing)
- 3 of 5 audited stores: Apparel item: color / size / gender / age group not published (https://catalogaudit.app/checks#apparel_attrs_missing)

## Reference

- [Every check](https://catalogaudit.app/checks): the 30 rules behind the scores, with the Merchant Center or agent impact of each
- [Reports library](https://catalogaudit.app/reports): all published audits
- [FAQ](https://catalogaudit.app/faq): what the audit checks, why AI agents skip products, robots.txt and AI crawlers, GTINs, Merchant Center disapprovals
- [Our crawler](https://catalogaudit.app/bot): what the CatalogAudit user agent reads and how to opt out
- [Privacy](https://catalogaudit.app/privacy)