The Resolution Layer · Case Study in BriefEKOM
Gemstone rings.
A cold read of the gemstone rings of a colored-gemstone jewelry retailer — where every ring is described three times on its own page, and the structured specification is the copy most likely to disagree. One of a set of anonymized EKOM analyses — what surfaced, and where it leads.
Lead finding · the ring's name and its record
A ring named and described as 10-karat gold is specified as 14-karat.
The product name and the written description agree on the gold; the structured material field does not. Two more rings show the same shape, one on karat and one on gold color, and a black sapphire is specified as white. Wherever a filter, a marketplace or an assistant reads the specification rather than the copy, it carries the wrong gold and the wrong stone — while the correct value sits on the same page, already written.
Also surfaced
A band thinner than paper
208 rings publish the inch figure for band width with "mm" attached. Of 552 rings that state band width in two panels, 288 disagree with themselves. One ring top is recorded as more than sixty inches wide; its own description gives the width correctly.
Stone records that contradict the stone
A heart-shaped diamond recorded as not faceted. Mined white diamonds recorded as heat-treated. Tiny ruby accents filed as an onyx ring's primary stone. 75 rings give a stone a shape its own dimensions rule out.
A savings claim stored as words
637 reference prices name their basis — an estimated retail value, a competitor comparison, a list price — only as text typed in front of the number. The page shows the label; the data cannot tell one basis from another.
What held
Gemstone against species, primary stone against category, and the savings arithmetic all passed on every ring read. The failures concentrate in units and free text — not in a broken taxonomy.
How EKOM read it. The engine read the retailer's own public product pages cold — no file, no credentials, no schema — and confirmed each category's count against the total its listing page declares. It read every stone group in its own panel, so an accent stone is never mistaken for the center, and then held each ring's specification against its own name and description. Eleven findings were withdrawn when that same comparison showed the record was right; the headline findings were re-checked live two weeks later.
From diagnosis to channel-ready. Most of the 47 findings point to five upstream patterns — three descriptions of one ring written separately, measurements typed as text with the unit inside, stone groups filled by position rather than role, free text where a list belongs, and some product lines entered with fewer fields. Fixing the findings fixes the findings; fixing the patterns fixes every ring that has already passed through them and every one that will. Then the same intelligence holds the line at intake, so the next ring whose specification disagrees with its page never reaches a shopper.