Reviewed July 2026. This article was substantially updated to reflect current web standards and practices.
A taxonomy is a controlled way to classify content. It can power navigation, filters, related content, search refinement, personalization, and reporting, but only when the terms match real user and editorial needs.
Start with content and tasks
Inventory representative content and study how users look for it. Define a small set of useful dimensions such as topic, audience, format, location, service, or lifecycle stage.
Control the vocabulary
Give each term a clear definition, scope, preferred label, synonyms, owner, and retirement rule. Avoid duplicate labels, overlapping categories, and tags created casually during publishing.
Choose relationships deliberately
Hierarchies work when broader and narrower concepts are real. Facets work when users need to combine independent attributes. Do not force every classification into a deep category tree.
Connect taxonomy to governance
Use structured fields where consistency matters, train editors, review unused or overloaded terms, and test how classification affects URLs, canonicals, indexation, analytics, and content migration.
Design a taxonomy workshop
Bring representative content, search terms, analytics, support questions, and editorial stakeholders. Ask participants to group content and name the groups, then compare their language. Test the proposed terms with people who were not part of the design discussion.
Taxonomy record
- Preferred label and plain-language definition.
- Broader, narrower, and related terms where needed.
- Synonyms and terms that should not be used.
- Applicable content types and required or optional status.
- Owner, creation rule, review date, and retirement process.
- Effect on navigation, filters, URLs, metadata, and indexation.
Common failures
Taxonomies become noisy when every editor can invent a tag, categories mix unrelated dimensions, singular and plural terms duplicate each other, or internal jargon replaces audience language. Deep hierarchies can force content into one branch even when users need to filter across several attributes.
Measure usefulness
Review filter use, zero-result searches, abandoned result pages, overloaded terms, and terms assigned to almost nothing. Interview editors about classification uncertainty. Merge or retire terms carefully, preserving URLs and updating relationships when public archive pages move.