IA Testing
IA performance testing quickly identifies how well a particular IA structure is meeting key customer and business goals, and flags up the parts that need improving. It can be used to optimise your current IA, test and refine a draft version derived from card sorting, or compare different versions to see which works best.
We use simple html mock-ups for speed of turnaround, and to ensure issues of structure and labelling are isolated and dealt with before you move on to wire framing or overlaying visual elements. This step by step approach means you can optimise different elements of your website easily and in a focused way.
The html mock-up is a clickable version of the core IA structure, reduced to the headings/link titles at each level, with a dummy “content” page at the lowest level and a simple breadcrumb trail navigation at the top to allow users to go back a level easily.
Performance
In the performance testing, each user undertakes a series of quick fire tasks clicking down through the hierarchy to where they think the content would be located. The actual paths they take are recorded. The key to accurate testing is recruitment of users who truly match the audience profile, and the development of tasks that reflect user goals for the website.
Analysis
“Success paths” for each task are defined before the test. Different levels of success are also pre-defined and form the basis of the quantitative analysis. This enables quick identification of which elements are causing problems for users.
Full analysis also requires understanding why users took particular paths – for example, are there common misunderstandings or expectations about certain links? For this reason we recommend always including some one to one moderated sessions in the sample. Remote testing can be a very useful way to boost the quantitative data, but the qualitative feedback from face to face sessions provides a wealth of insights to ensure the quality of recommendations.
Based on the analysis, we will recommend necessary changes to:
- The location of items
- The labelling of items
- Suggested cross linking
- Suggested new content
IA performance testing provides clear metrics for benchmarking and comparison, pinpoints problems areas and delivers actionable recommendations to optimise your IA.
If you want to be sure you’re not hiding your best content, it’s a vital tool.
