The main purpose of a taxonomy in information architecture is to describe content using consistent concepts and terminology. This helps systems find related information and connect content across different areas. A taxonomy can also support filtering by multiple criteria, offer search suggestions, and help users narrow their searches. It does more than divide a website menu into levels: it can describe articles, products, documentation, and files. Many of these features work behind the scenes, so users can find content through related articles or filters without seeing the full taxonomy.
Standardize how content is described rather than simply defining menus
Under the definition used in NN/g’s 2022 article, a formal taxonomy is a planned controlled vocabulary: a limited set of approved, hierarchically related terms used to describe content. Content creators can choose appropriate terms from this set, but cannot invent category names on the fly whenever they add content.
In practice, opinions differ on whether the tree structure of an information architecture also counts as a taxonomy. Here, we distinguish formal metadata taxonomies from website structure: a taxonomy describes the concepts that content covers and the relationships between those concepts, while website structure determines how content is organized. Taxonomies prioritize logical precision, while the navigation users see needs to reflect how they understand things. The full taxonomy should therefore not be turned directly into a menu.
Connect scattered content to recommend related material
Content about the same concept may be spread across different navigation categories. A taxonomy provides a shared way to describe it. Systems can then find content based on topical relationships as well as a page’s location, and create links to related content.
In its 2022 article, NN/g uses its own website as an example: readers may see the topic “Information Architecture,” while the system uses the more detailed hierarchy “Information Architecture > Metadata > Controlled Vocabularies > Taxonomies” behind the scenes. This detailed classification can help select related articles, connecting them even when they sit in different navigation categories. The purpose of a taxonomy is not to confine each piece of content to a single folder, but to let the system use the conceptual relationships between pieces of content.
Support filtering by multiple criteria and search beyond matching query terms
Taxonomies can also support two specific ways of finding information:
- Faceted filtering: Users apply multiple criteria at once to narrow down content without working through each level of navigation. The NN/g article uses the US Library of Congress as an example, showing how criteria such as subject, collection, and original format help users find materials.
- Search suggestions and narrower searches: Search systems can use a taxonomy to provide results related to the query terms but expressed differently. They can also list categories to help users limit the scope of a search.
These features require a well-defined and well-maintained taxonomy. Although a controlled vocabulary is intended for long-term use, that does not mean it never changes. Teams still need to review whether terms should be added, renamed, merged, or removed, and spot-check whether content is tagged correctly.
Taxonomies: perspectives from across the web
Should website menus be organized by topic or content format?
If most tasks on a website involve finding answers about a particular topic, top-level navigation is generally better organized by topic rather than first separating content into formats such as articles and videos. Users assess the options to decide which category is most likely to solve their problem; at that point, the content’s format may not be their primary concern. If users have different tasks, the design should still reflect their actual needs.
Can the same piece of content appear in multiple categories?
Yes, but this should be supported by user data. When a substantial proportion of users look for the same content under different parent categories, cross-listing can accommodate different ways of understanding it. If the content merely fits a category in theory but no one looks for it there, an additional entry point may not be necessary. Excessive cross-listing can actually make the information architecture less clear.
How can you tell whether category labels need to be adjusted?
Tree testing can be used to examine users’ first clicks and final selections. If their first clicks lead to the correct broad category but overall success rates are low, lower-level categories may overlap too much. You can examine where users end up and make subcategory names easier to distinguish. If first clicks themselves are scattered, you can also reconsider the overall categorization approach.
Can category tags support internal work as well as help readers find content?
Yes. Besides determining where content appears on a website, tags may also support internal marketing. In the BuzzFeed example, a piece of content can have multiple tags for different purposes, including website content categorization and internal marketing. Their actual uses still depend on how the team designs and uses them.
How can taxonomies help search engines?
A taxonomy can show a website’s topic hierarchy and make it easier to plan internal links between related pages. This involves creating categories based on relationships between products or articles and linking subtopics back to main categories. One view is that this clear topic structure helps communicate the website’s topical expertise to search engines. However, the available material does not establish that doing so will necessarily improve rankings.
References
- Information Architecture vs. Sitemapswww.nngroup.com · 2023-09-03T16:00:00+0000 — Source used in the article above.
- SEO Taxonomy Best Practices: Complete Guideneilpatel.com · 2023-01-03T20:00:00+00:00 — Discusses: Can category tags support internal work as well as help readers find content?
- Taxonomy 101: Definition, Best Practices, and How It Complements Other IA Workwww.nngroup.com · 2022-07-03T16:00:00+0000 — Source used in the article above.
- How Can You Implement Entity Optimization Without Relying On Schema Markup? – Ask An SEOwww.searchenginejournal.com · 2026-06-04T08:30:30+00:00 — Discusses: How can taxonomies help search engines?
- Avoid Format-Based Primary Navigationwww.nngroup.com · 2014-06-08T16:00:00+0000 — Discusses: Should website menus be organized by topic or content format?
- Tree Testing Part 2: Interpreting the Resultswww.nngroup.com · 2024-01-19T17:00:00+0000 — Discusses: Can the same piece of content appear in multiple categories?; How can you tell whether category labels need to be adjusted?

