Definition
Catalog Enrichment
The work of cleaning up and filling in a product catalog — fit, fabric, occasion, compatibility, pairings — so site search, sales agents, and the AI assistants where people now shop can sell from real attributes instead of guessing.
Definition
Catalog enrichment is the work of taking a product catalog as it actually exists — inconsistent titles, thin descriptions, attributes trapped in photos and PDFs — and cleaning it up and filling it in: fit notes, fabric, occasion, dimensions, compatibility, what pairs with what. The result is a catalog where every product carries the facts a good salesperson would know about it.
Why it matters now
Catalogs were written for humans skimming a grid. A growing share of selling is now done by software — the on-site assistant, ChatGPT, Google's shopping surfaces — and software can only sell from the attributes it can read. A dress with no fit data can't be recommended to the shopper asking whether it runs small; a product ChatGPT can't read gets skipped or described wrong. Enriched catalogs get recommended accurately; thin ones lose the sale invisibly.
Where the new attributes come from
From the products themselves — images, descriptions, and specs read carefully. From reviews, where hundreds of shoppers have already said how something fits. From real shopper conversations, because the questions people ask reveal exactly which attributes are missing. And from the brand's own materials. Good enrichment is grounded: every added attribute traces back to evidence, because an invented attribute is worse than a missing one.
Where the enriched catalog goes
Two directions. Inward: the brand's own site, where search, product pages, and the sales agent all answer from the same completed data. Outward: distributed to ChatGPT, Google, and the surfaces where AI shops, so the brand is represented by its own facts. Kinect runs catalog enrichment as one of its six revenue jobs, feeding the same enriched catalog to its own agents and to the outside ones.
Related concepts
Agent Readability
How easily AI shopping agents can parse, trust, and act on a brand's catalog, content, and policies — the agent-era equivalent of crawlability.
Agent-Ready Storefront
A storefront that AI assistants can read and act on — both the ones browsing the store on a shopper's behalf and the ones pulling its data to answer questions elsewhere — so the brand is described by its own facts.
Revenue Agents
AI agents that do revenue work for a brand — selling to shoppers, keeping the catalog complete and accurate, turning conversations into customer research — as distinct from the support bots that deflect tickets.
One Intelligence (One Brain)
The architecture in which a single model of a brand — its catalog, policies, customers, and conversations — powers every AI job the brand runs, so whatever one job learns makes all the others better.
See catalog enrichment in action.
Book a call to talk through your store — live on your catalog the same day.