Attribute Coverage Analyzer

Paste or upload a product export to measure product attribute coverage across your catalog: which fields are populated, which are blank, and where the gaps cluster by category or supplier. It is built for anyone preparing data for syndication, AI search, or onboarding — distributors, retail and marketplace teams, and API-first platforms alike.

What it checks

The analyzer reads your column headers and rows, then computes coverage at the field, category, and record level. For each upload it reports:

• Fill rate per attribute — the percentage of records with a non-empty value for every column, so you can see at a glance that, for example, weight is 98% populated but country_of_origin is only 41%.
• Empty, null, and placeholder detection — treats blanks, NULL, N/A, -, 0, and TBD as missing so a column that looks full is not silently hiding gaps.
• Coverage by category or group — pivots fill rate against a category or product-type column, surfacing that your MRO fasteners are well-described while a furniture line is missing dimensions.
• Required vs optional gaps — lets you flag a set of must-have attributes (title, brand, GTIN, primary image, key spec) and scores records against only those.
• Record completeness distribution — a histogram of how many attributes each product has, separating near-complete records from skeletons that need the most enrichment.
• Outlier and uniqueness flags — columns where every value is identical (likely a default) or where a single value dominates, which often signals a mapping error rather than real coverage.

How attribute coverage is measured

Coverage is a simple, transparent ratio: for each attribute, the number of records with a meaningful value divided by the total number of records. The nuance is in what counts as “meaningful.” A naive COUNT of non-blank cells overstates reality because placeholders, zeros, and whitespace pass the test. This analyzer normalizes each cell first — trimming whitespace, lowercasing common sentinels, and treating a configurable list of placeholders as empty — so the number reflects data you could actually publish.

There is no universal standard for which attributes a product “must” have; requirements come from your channels. A Google Merchant feed, an Amazon flat file, an ETIM class, and a distributor’s PIM each define their own mandatory set. The tool therefore separates two questions: how complete is the catalog overall and how complete is it against a target list you define. That distinction matters because a record can be 90% populated and still fail a channel that needs the specific 10% it is missing.

1. 1
   Load your data
   Paste CSV or upload an export. Nothing leaves your browser.
2. 2
   Mark required fields
   Optionally tag the attributes a channel or AI surface needs.
3. 3
   Read the coverage map
   Review fill rates, category breakdowns, and the record-completeness distribution.
4. 4
   Prioritize enrichment
   Export the gap list and work the highest-impact attributes first.

All processing happens client-side. Your file is parsed in the browser with JavaScript; there is no upload, no login, and no file-size limit beyond what your own machine can hold in memory. That makes it safe to run against a raw CPG export or an industrial-distribution catalog with confidential cost or supplier columns.

Related resources

FAQ