A taxonomy cannot be judged only by its structure at launch. Its real value appears when users search, browse and interact with content in production. This session focuses on how to measure whether a taxonomy is genuinely improving findability after it has been deployed.
The session will move beyond coverage and depth metrics to examine runtime signals that reveal how users experience search. Attendees will learn how to use click-through rate, zero-result rate, reformulation rate, ranking position and relevance-judgment benchmarks to assess whether search quality is improving. Query logs will be treated as a diagnostic tool for identifying failing searches and turning them into actionable taxonomy improvements.
The session will also show how explicit and implicit feedback can be interpreted, how root causes can be traced from a failing query to a taxonomy change and how results can be remeasured after each update. Participants will leave with a practical measurement framework, a findability scorecard and a repeatable improvement loop for proving taxonomy value over time.
