Taxonomies define the structure that underpins document and records management systems, knowledge management projects and more.

Considerable effort goes into developing these taxonomies, with the goal of creating a common structure that will benefit the whole organisation.

The challenge, however, is to ensure that these taxonomies work well for staff, beyond any organisational benefits that are sought. It is here that taxonomies often fail.

If not designed well, taxonomies can become ‘white elephants’, too hard to understand and too complex to use. At their worst, poorly designed taxonomies are the direct causes of project and system failure.

The field of information architecture (IA) has much to offer those creating taxonomies, including a range of structured techniques for building and testing their effectiveness.

This briefing outlines some of these approaches, and encourages creators of taxonomies to retain a clear focus on usability throughout the design process.

Building taxonomies

Taxonomies are typically drawn from a number of sources, including existing industry-wide classification schemes, business functions and structures already in place within sections of the organisation.

These are pulled together to create a larger or more complete taxonomy. Testing of this taxonomy usually relies on internal review, discussing the taxonomy with staff, and gaining input on areas of strength and weakness.

While effective for gaining broad user and stakeholder input, this kind of review is very shallow, and is not sufficient to ensure that the taxonomy can be used in practice.

Instead, structured techniques must be used, getting beyond staff and expert opinions.

Three key purposes of a taxonomy

As outlined in the earlier article Rolling out a records management system, there are three clear purposes of a taxonomy:

• knowing where to file information correctly
• retrieving information easily when needed
• meeting legislative, compliance or business objectives

It is easy to create a taxonomy that meets the third goal, at least superficially. If the first two goals are not achieved, however, the taxonomy will certainly fail.

Information architecture

Information architecture (IA) is a discipline that focuses on creating effective structures and navigation for websites and intranets.

Through this work, the field of IA has built up a toolbox of techniques that can be applied equally well to taxonomy creation. Three of these techniques are outlined below, with pointers to further information.

Card sorting

A very simple technique for building an understanding of how staff think about information, used as an early input when creating a taxonomy. For more on this technique, see the article Card sorting: a definitive guide on Boxes and Arrows.

Tree testing

Provides a rapid way of testing a taxonomy to ensure that staff can correctly store information and find it again later. For more, see the article Tree testing for effective navigation.

Usability testing

Designed to test the overall ease of use and effectiveness of not just the taxonomy, but the system used to implement it. Should be used throughout the design and implementation process.

Best-practice approaches

It is no longer sufficient to simply gather staff input to assess the effectiveness of taxonomies. Instead, practical IA techniques should be used to ensure that a taxonomy works in practice.