In the first workshop with a prospect, one sentence comes up almost every time: “we’re at 72% completion on the catalog”. Two or three meetings in, somebody finally asks the real question: 72% of what?
The denominator nobody looks at
A percentage without a denominator isn’t a number — it’s a feeling. When a catalog reports 72%, what it means is “72% of the attributes we decided to track are filled in”. But those attributes were picked two years ago, by a team that’s no longer around, based on what the ERP required at the time.
The missing 28% isn’t necessarily what’s blocking you. And the filled-in 72% isn’t necessarily moving you forward. We’ve seen 85%-complete records that can’t ship to Amazon Business, and 60%-complete records that work just fine on internal EDI.
The real number is always conditional
Completion only makes sense relative to something. At minimum, three axes:
- Per channel: a record shipping to Mirakl doesn’t have the same requirements as the same record in a print catalog.
- Per family: a bearing and an electrical cabinet have different taxonomies. Required attributes don’t overlap.
- Per market: a DACH record needs weights in grams and an ETIM classification; a French one may not.
A useful completion indicator therefore looks like:
Family “Modular circuit breakers” · Channel “Distributor B” · Market FR → 94%. Same family · Channel “EU marketplace” · Market DE → 61%.
Not this:
Catalog → 72%.
What to track instead
With the teams who fixed this problem, the dashboard isn’t a global gauge. It’s a completion matrix: rows = families, columns = channels, each cell = blocking percentage (required and missing) plus recommended percentage.
And each cell is clickable. You see exactly which records block which channel, with which attributes missing. Announcing “we’re at 82%” to a CEO is useless — what they care about is “we can’t publish 412 records on Amazon Business because the HS code is missing”.
The round-number bias
The reason “72%” sticks around is that it feels reassuring. A number going up looks like progress. But a catalog moving from 72% to 78% without knowing what got filled in may well have made its problem worse — somebody completed the easy fields, not the blocking ones.
What measures real progress isn’t the average percentage, it’s the number of (family × channel × market) combinations that have hit 100% on blocking attributes. When that number goes up, it corresponds to actual channels you can open, actual markets you can serve.
A PIM doesn’t help you check boxes. It helps you tell the boxes that matter from the ones that don’t — on that day, for that channel, for that family.