Category mappings

This is the bridge between the two trees. A vendor sends a category called "Road bikes — carbon"; your shop has "Bikes → Road". A mapping says the first belongs in the second — and from then on all of its products file themselves there, including the ones that arrive tomorrow.

1 — Mappings2 — Rule count3 — Enabled4 — Mapped products5 — Target category6 — The category's rule count7 — Add category8 — How it works

Controls on this screen

1. 1Mappings — The fourth tab of the Categories section. The first three show trees; this one shows the bonds between them.
2. 2Rule count — How many mappings are configured.
3. 3Enabled — How many of them are active. A disabled rule stays written down but files nothing.
4. 4Mapped products — The real impact. Zero with rules present means the filter is catching nothing.
5. 5Target category — Your category, the one the rules file into.
6. 6The category's rule count — Several vendors' categories can point at one of yours — and usually do.
7. 7Add category — When the target category does not exist yet, you create it from here.
8. 8How it works — A summary of the procedure, on the screen.

Why a category is mapped, not a product

Filing products one by one is impossible at tens of thousands, and mostly pointless: the vendor already filed them. A mapping takes their work and translates it into your structure — once, for a whole category. A new product arriving in that same vendor category is filed with no further intervention.

The procedure

1. Let the AI propose

   The AI compares names and structure and proposes where a category belongs. On clear cases its hit rate is high.

2. Review the proposals

   Watch for categories that sound alike but are not the same — "Accessories" from a bicycle vendor and from an electronics vendor.

3. Finish the rest by hand

   Categories with unhelpful names have to be placed by what is actually in them. Looking at a few products in By vendor helps.