Two hundred and twelve products came back empty, and fourteen of twenty were only badly searched

What arrived

A percussion and pro-audio retailer’s catalogue, with gaps to fill from manufacturer and retailer pages. Every value we add carries a link to the page it came from, so a product with no page found gets nothing added.

The first pass came back empty for 212 products.

What we found

An empty result is easy to read as missing data: the product is obscure, discontinued, or never listed anywhere. We tested that reading before accepting it. We took twenty of the 212 and put them straight back through, with no code changes and nothing retyped.

Fourteen of the twenty came back with a real source attached. They were never missing, just badly searched.

What it showed

Fourteen of the twenty products we had marked empty had a source all along. The gap was in the searching, not in the data. That changes what an empty result means: not a dead end, but a question to ask again before anyone types a value in by hand.

How we counted

Twenty products from the 212, put back through the same process, unchanged. A product counts if it came back with a source page for that product. One page that matched but turned out to be a company register entry is left out, which is why the figure is the strict one.

What it means for your file

When a tool, a supplier or a first attempt reports products it cannot find, treat the gap as a question rather than an answer. Run the same products again before anyone fills them in by hand, and keep a link to the page behind every value that comes back, so a page that only looks right can be caught.

What is product data enrichment? covers where the missing values come from, and how to keep each one traceable.