Home / Glossary / P / Product Data
The attributes describing a product: title, description, identifiers, images, specifications. The raw material for every shopping surface and most AI product answers.
We work on this: eCommerce SEO Services
What counts as product data
Identifiers first: part number, SKU, GTIN, manufacturer code. Then the specification: material, dimensions, tolerance, capacity, rating, certification. Then the context a buyer needs but a spec table does not carry: what the product is used for, what it replaces, what it is compatible with, how it fails.
The last group is the one that separates a manufacturer’s page from a reseller’s. The specification is shared. The application knowledge is not.
Why it is usually missing, not badly written
In most product businesses the data exists and is well understood internally, and almost none of it is in a form a website can use. Engineering drawings hold the authoritative specification in a format no crawler reads. Supplier sheets arrive as PDFs, often scanned, inconsistent between suppliers. ERP holds the part number, the unit of measure and the price, and little else. The rest is in the sales team’s heads, because they have answered those questions by phone for years.
That is why adding a writer rarely fixes a thin catalog. There is nothing to write from until the extraction has happened.
Why it decides what can rank
Buyers who know what they need search the specification, not the category. A page can only answer that query if the specification is on it, in crawlable text rather than inside a PDF, an image or a table rendered as a graphic. Product data is therefore not a content-quality question. It is an eligibility question.
How it is measured
Field completeness against a target model, the share of specifications reaching the crawler as text, and impressions on product-level queries. Not word count.
Related: Product Catalog, PIM, Product Taxonomy. Back to the glossary.