War-game a catalog taxonomy before category trees, facet rules, and merchandising talk tracks harden into how every browse path gets built.
Taxonomies fail when categories invent products shoppers never find, when SKUs dual-count across sibling nodes, when search still ranks retired labels above live ones, and when ops cannot show who owns a rename. A neat tree slide is not evidence.
Freeze the package
State the proposed tree depth, which facets and synonyms ship with it, who owns migrations and redirects, what success looks like after two quarters of findability and support ticket quality, and the kill criteria if duplicate SKUs or dead-end paths past a named threshold. Attach the current tree, sample search logs, redirect plan, and the measured lag from rename to live index. If product, merchandising, and support disagree on whether the tree buys conversion or just a cleaner admin screen, reconcile before seating.
Name the decision you will make if the war game finds nothing new, and the hold criteria if any top category still lacks a named owner or a measured redirect path.
Seats
- Shopper. Fear that the label they used last month now dumps into an empty aisle.
- Merchandising. Facet math that looks clean while inventory still sits under two parents.
- Search. Synonyms that invent relevance the ranking logs will not show.
- Support. Ticket macros that still teach the old path after cutover.
- Skeptic. The claim that looks strongest and is least sourced.
Optional data seat if feed partners bind the form. Same pack for every seat.
Run the loop
Feed Pingpong the taxonomy memo and exhibits. First pass steelmans the tree. Later passes attack from the seats above. Final pass turns surviving objections into clearer ownership, a published redirect SLA, or a hold. Delete invented "shoppers already know where things live" claims and dual-counted category traffic.
Force a day-after narrative: what happens if a journalist quotes a broken browse path against a launch, if redirects never log, or if a large account cannot reorder under the new labels. If those stories are stronger than your mitigation plan, fix the package before you publish. Separate a catalog taxonomy from a product naming system or a SKU rationalization so each decision stays testable.
Ask every seat to mark which steps are optional in practice. Optional synonym reviews become skipped within a month if nobody names the owner and the time box. Related: war-game a product naming system, war-game a SKU rationalization, war-game a packaging change, pretend you are the CPO, and the war-game decisions hub. Process: how to run a Pingpong.