What a term is being searched on. The need a bid is placed against. Whether the page the click lands on answers it.
Campaigns and product pages running on one description of the product.
Retail media optimization
Search term: "deodorant for sensitive skin"
Illustrative term read
What the keyword records
The string typed, and the category node it was filed under
The need under it
Rules an item out
The product page the click lands on carries none of it.
Retail media targeting is built on keywords and on the retailer's category hierarchy. A keyword records the string a shopper typed, and a category node records where the retailer filed the product.
Neither field records why the shopper is searching. The need and the situation behind the term sit outside both of them.
Where it comes from
The need behind an item is read from millions of product pages and the reviews under them, then tied to the UPC that carries its sales.
Refreshed on a schedule agreed with you, and applied the same way to every product in the category.
What a search term is standing in for. Long wear, sensitive skin.
The moment the item is being bought for. Game day, first apartment.
What rules an item out before anything else matters. Fragrance free, nut free.
Which need sits under a term you are bidding on.
Contextual attributes
What the product page is missing before paid traffic lands on it.
Product data enrichment
Whether an agent answering a plain-language shopping question returns your item.
AI and agent readiness
What a rival item carries that makes it the better answer to the same need.
Competitive benchmarking
A Fortune 100 CPG company defined the attributes it needed, and we populated and maintained them after in-house coverage capped near 68%. The file now runs at 98% accuracy against an independent test set, including 26,185 items that arrived with no tags at all.
Core, Contextual and Custom attributes for the categories you advertise in, on a schedule agreed with you.
The same attributes, open against the terms and audiences you are buying.
We check your terms and your product pages against the attributes.
Thirty minutes on your categories. How the attributes get built, and the questions your team could put to them.
