Proof, not adjectives
Collective Shoes: 1,539 products that were invisible to their own search.
New Zealand footwear retailer
The problem
Collective Shoes, a New Zealand footwear retailer, had a deep catalog and a search bar that could not reach most of it. Shoppers typed what they wanted and the search returned almost nothing, or the wrong thing. 1,539 products existed in the catalog that the store's own search could not surface, because the attributes a shopper searches on, use case, material, width, style, were not structured in a way any agent could read.
What Atomz did
Catalog Genius read the full catalog and mapped every product to the Shopify Standard Product Taxonomy, writing structured attributes to metafields. AI Search then connected real shopper intent to those products, with Discovery Prompts guiding vague queries to the right result. The catalog stopped being a list and became something the search agent could actually answer from.
The result
NZ$624K
in search-driven revenue
2.2x
repeat purchase rate among shoppers who used search
44%
higher lifetime value from those shoppers
1,539
previously unfindable products surfaced and selling
Search went from a cost center to our second-biggest revenue channel.
The same catalog work that powered this on their store is what makes a store findable when shoppers ask off-site. Fix the catalog once, and every agent that reads it, on-store and off, gets smarter at the same time.
Collective Shoes, in short
What did Atomz do for Collective Shoes?+
Catalog Genius mapped their footwear catalog to the Shopify taxonomy and wrote structured attributes to their metafields. AI Search then ran on that readable catalog. It surfaced 1,539 products their old search could not reach and turned the search bar into their second-biggest revenue channel.
How much revenue did search drive?+
NZ$624K in search-driven revenue, with a 2.2x repeat purchase rate and 44% higher lifetime value among shoppers who used search.
Why couldn't their old search find those products?+
The catalog was unstructured. Their products were described in prose, not in fields an engine can match on. Once the attributes existed, the same products became findable. Unfindable is missing.
How long did it take?+
It took days rather than a re-platform, because Atomz reads the catalog you already have and there is no migration. Run the free audit at gpt.atomz.ai to see your own gap first.
Get started
Want results like these on your store?
Drop your store URL for a free readability score, or add the app and Atomz starts making the catalog readable today.