So, who is Julian?
Julian Apatu is the Content Management Leader at PHARMAC, the New Zealand Government’s Pharmaceutical Management Agency. He has a strategic advisory role providing commentary on web strategy, governance, web standards and compliance. PHARMAC manages the Pharmaceutical Schedule, the list of pharmaceuticals and medical devices subsidised by the Government. PHARMAC’s role is double-edged; managing the available funding and expanding the range of funded treatments to ensure New Zealanders’ health continues to improve through the use of pharmaceuticals. Rising healthcare costs present a worldwide challenge. Everyone working in healthcare wants the available money spent where it does the greatest social good. For Julian and the team at PHARMAC the same attitude applies to their work. PHARMAC’s stated mission, consistent with its role as defined in legislation, is: “To secure for eligible people in need of pharmaceuticals, the best health outcomes that can reasonably be achieved, and from within the amount of funding provided.” The organisational tenet is to secure the best tools and services for the available budget.
Why tree testing?
Julian considers IA and UX testing are an important part of his role, and he’s a strong advocate for the value of tree testing. PHARMAC serves a diverse audience of consumers, healthcare professionals, administrators and Government agencies. Julian says Treejack provides an excellent tool for evaluating structural and architectural web decisions and provides the means to evaluate and test the IA with stakeholders and users who engage through the online channel. Julian uses tree testing to:
- Explore multi-level information architecture options before committing to the interface design
- Identify and examine nuances in quite complex lexicons
- Interpret the implications of changes to user click paths and understand the impact of changes on users
“We have discovered issues, found breakthroughs and revised our architecture accordingly before re-testing. I like to move forward with the confidence we’ve considered and understood where and, perhaps more importantly, how improvements can be made.”
Julian also counsels that as with any tool, it doesn’t do your job for you. The tester is in control of the inputs and the calibre of the outcomes. “Poor preparation equates to poor outcomes” is a mantra that Julian subscribes to. Most UX designers have also been testers at some stage. Julian feels we’ve probably all seen poorly constructed tests, where not enough thought has gone into in the preparation stages of important interactions like the testing process. In these cases, the user is not valued and can quickly become removed from and disillusioned with the process. Accordingly Julian’s five tips are focussed on how better to prepare and improve the test process.
5 tips for tree testing:
Clarify test objectives
Project objectives sometimes evolve. You’ll discover there’s another question to be answered, or your team expands. Having clarity about your objectives and documenting them makes it much easier and expedient to brief and report against. Return and review frequently.
Invest time choosing truly meaningful titles
Taking time to make sure you’ve carefully considered the titles for navigation and content can generate incremental improvements throughout the architecture and allow users to intuitively find the information they need. Sloppy naming leads to inconclusive test results and poor architecture.
Explore nuances in your architecture
When testing be aware there often may be subtle distinctions in vocabularies that are highly relevant to a professional audiences, but almost imperceptible to non-professionals. In our case medical professionals may draw far more information from our titles than non-medical professionals. Do your research and ask for professional help where required.
Show respect for your testers
We’ve probably all felt we weren’t being asked the right questions - empathise with your testers by reflecting how you could better frame a context for them to consider.
Find the right people
It’s easy to negate your good preparation. Recruiting a representative cross-section of your audience ensures that you can have a reasonable level of certainty and confidence in your results. Keep pushing to acquire the testing team your project deserves.
What do you think? How has tree testing worked for you, and do you have any more advice for effectively validating information structures?