Memo 01 · Catalogue architecture
A catalogue is a decision system
The useful question is not whether each listing is acceptable. It is whether the catalogue helps buyers, operators and advertising make the same distinctions.
Most catalogues are histories disguised as structures. They record the order in which products were launched, exceptions were granted and marketplace problems were solved. That history can be commercially sensible and still produce a catalogue no one would design from scratch.
A catalogue architecture starts with decisions. Which differences matter to a buyer? Which products should share demand signals? Which variants genuinely belong together? Which distinctions must remain visible to the team managing margin, inventory and advertising? Taxonomy is useful only when it makes those decisions easier.
The practical work begins with a map. Put every product, parent, child, variation theme and category assignment in one view. Then mark where the same buyer intent appears more than once, where unlike products have been grouped, and where reporting boundaries disagree with buying choices.
The target structure should reduce ambiguity rather than maximise consolidation. Fewer parent listings are not automatically better. More detail is not automatically clearer. The right structure is the smallest set of distinctions needed to support purchase, control and measurement.