Catalog
A practical catalog data audit you can run this week
Nine checks that surface most of what is silently costing you visibility across marketplaces — no tooling required beyond a spreadsheet.
Most catalogs do not fail dramatically. They accumulate small faults — a missing attribute here, a broken variation there — until nobody trusts the data and every channel launch starts by rebuilding it from scratch.
This is the audit we run at the start of a catalog engagement. It needs an export and a spreadsheet, and it will find most of what matters in a day or two.
1. Required attribute coverage
Export your live listings per channel and count the percentage of required and recommended attributes filled, by product class. Sort ascending. The bottom of that list is where your visibility is leaking. On Wayfair and similar platforms, coverage varies enormously by class, and a single under-filled class can drag a whole category.
2. Variation integrity
Look for parents with a single child, children with no parent, and variation themes that do not match the category standard. These are near-invisible in a UI and obvious in an export.
3. Duplicate and near-duplicate listings
Group by normalised title and by any manufacturer identifier. Duplicates split reviews, split ranking signals and confuse buyers. They also multiply every future bulk update.
4. Cross-channel drift
Pick fifty SKUs at random and compare title, key attributes and price across every channel you sell on. Drift is normal and usually unintentional. Systematic drift means nobody owns the catalog.
5. Image compliance
Check main images against each channel's minimum dimensions, background requirements and text-in-image rules. Then count listings with fewer than four images. That second number is often the single highest-return fix available.
6. Suppressed, stranded and inactive
Pull every listing not currently buyable and categorise by root cause rather than fixing one at a time. In our experience roughly half of a suppression backlog shares three or four underlying faults, which means three or four bulk corrections instead of hundreds of individual ones.
7. Logistics data accuracy
Box dimensions, weights, lead times and freight classes. On home and furniture channels these are merchandising data — they affect ranking, conversion and margin — and they are almost always the least maintained fields in the catalog.
8. Taxonomy placement
Sample your best sellers and check they sit in the most specific correct category, not a convenient parent. Miscategorisation quietly excludes products from filtered browse, which is where a large share of marketplace discovery happens.
9. Orphans and end-of-life
Products that are live but unfulfillable, discontinued but not retired, or seasonal and long past. They cost you nothing directly and cost you a great deal in buyer trust and defect metrics.
Root causes before symptoms; classes with the worst coverage before individual SKUs; anything affecting buyability before anything affecting ranking. Resisting the urge to fix the first broken thing you see is most of the discipline.
None of this is difficult. It is simply nobody's job in most teams, which is exactly why it compounds.