6 - User Testing and Validation

Summary

Editor's note: Tom's passion for helping users shines through in this exploration of the ways you can ensure your taxonomy will work in the real world. Sometimes this means disregarding what seems like the ‘correct’ way and doing something that better matches users’ mental models. Through examples such as mushrooms and marathons, and a case study from one of the UK's largest charities, Tom explains not just the ‘why’ of taxonomy validation, but also the ‘how’.

Introduction

It's an oft-used phrase that knowledge is knowing that a tomato is a fruit, but that wisdom is knowing not to put it in a fruit salad. Something similar applies when developing a taxonomy: terms should be organised in a way that makes sense in real life. In this chapter, we’ll investigate ignoring your taxonomic impulse to create the perfect classification and instead focusing on classifying in a way that makes sense to your user. That could mean arranging your taxonomy in a different way to how you first imagined.

There is a cultural hump you may need to get over when making your taxonomy work in the real world. If your user research has given you evidence that your ‘tomato’ content should live under ‘vegetable’, the purist in you may wrestle with that idea if you have always considered it a ‘fruit’.

Go and listen to your users. Implement a taxonomy structure based on what they tell you. You are not your user.

What follows are examples of where I have taken a user-centred approach in my work at two well known organisations: Getty Images and Cancer Research UK.

Pizza in a search for ‘fungus’?

Getty Images’ collection of rights-managed and royalty-free photography, creative and editorial content, video and music content, numbers well into the tens of millions. When I started, I was in a team of image data experts that was split three ways: image classification (things like the orientation of the image, whether it was colour or black and white, etc.); a search data team (that applied the taxonomy tags to images); and the search vocabulary team (that managed the vocabulary tree – and the team I was part of).

I went from taxonomy rookie to classification expert in the fascinating world of controlled vocabularies. Terms like ‘Boolean’, ‘nested search query’, ‘keyword’, ‘synonym’, ‘exact’ (a user-generated tag) and ‘parent-andchild hierarchy’ became the bread and butter of my professional vernacular.