Resources
Learn what your catalog needs.
A free self-test, plain definitions, and what we learn from catalogs.
10-minute AI visibility test
Check your store in ChatGPT yourself.
No tools required. Takes about ten minutes. Tells you whether your products are getting recommended.
Pick five products from your catalog
Choose products that have accessories, compatible items, or fit a specific use case. Avoid hero products with dedicated pages all over the web.
Open ChatGPT (or any AI assistant) in a private window
Private mode avoids personalisation from your previous searches.
Ask three questions per product
Use the shopper's voice: "What's a good [product type] for [use case]?", "What [product type] fits [compatible item]?", "I need a [product type] that [attribute].". That's 15 questions total.
Count how many times your products appear
Note each time your store or a specific product is recommended by name.
Score yourself
Your score is the number of questions, out of 15, where your product was recommended at least once.
Your score
The number of questions, out of 15, where your product was recommended at least once. Under 5 is a gap worth fixing.
Glossary
Plain definitions for catalog and AI-commerce terms.
| Term | Definition |
|---|---|
| Agentic commerce | Shopping done by an AI agent on a shopper's behalf — finding, comparing, and sometimes buying products without the shopper browsing manually. |
| AI shopping assistant | A conversational AI that helps shoppers find products. It reads product data to match recommendations to the shopper's question. |
| Attribute | A single piece of product data stored as a field: color, weight, compatibility, skin type. Attributes are what search, filters, and AI assistants read. |
| Catalog audit | A check of your product records for missing, inconsistent, or unstructured data. ReLUnit's free audit shows you which attributes are absent and what that costs you in search and AI channels. |
| Compatibility data (fitment) | Structured fields that record what a product fits, works with, or requires. Common in auto parts; increasingly important in electronics accessories, outdoor gear, and anything sold as a system. |
| Confidence score | A number ReLUnit assigns to each filled attribute — how certain we are the value is correct. Low-confidence values are flagged for review before import. |
| Controlled vocabulary | A fixed list of allowed values for an attribute. "Black", not "blk", "BLACK", or "Jet Black". Controlled vocabularies make filters, comparisons, and AI matching reliable. |
| Custom fields (metafields) | Extra product fields beyond a platform's standard set. Shopify calls them metafields. They're where compatibility and use-case data usually lives — when it exists at all. |
| Facets (filters) | The filter options shoppers use to narrow a collection: brand, size, colour, use case. Filters only work if the underlying attributes are filled and consistent. |
| Product taxonomy | A classification system that groups products into categories and subcategories. Platforms like Shopify and Google use taxonomies to decide which standard attributes apply to a product. |
| Relationship mapping | Recording how products relate to each other as structured data: what fits what, what's needed alongside it, what replaces a discontinued model. ReLUnit's core output. |
| Shopping feed | A file or API that sends your product data to channels like Google Shopping or Meta. Missing required attributes can limit or disapprove listings. |
| Structured data | Product information stored as discrete, machine-readable fields rather than buried in a description. Search engines and AI assistants read structured data; they struggle to extract it from prose. |
| Value standardization | Merging variant spellings of the same thing into one canonical value across your catalog. Prerequisite for filters that work and comparisons that hold. |
| Zero-result search | A search query that returns no products, usually because the catalog lacks the attribute the shopper named. "Waterproof hiking boots wide fit" returns nothing if width isn't a field. |
Coming soon
What's missing across catalogs.
We're compiling what our audits find across stores: which details go missing most often, by category. Get the first report when it's out.
One email when the report is published. No newsletter unless you ask for it.
Find out what your store could be answering.
Paste your store URL. See which shopper questions your catalog misses, and which details are missing.
Free. Takes about a minute. Works with Shopify stores today.