Tree testing

Tree testing helps you understand how easily people can navigate your Information Architecture (IA). By removing design and layout from the equation, it focuses purely on structure. It’s a fast, evidence-based way to spot where users get lost, which labels confuse them, and where improvements are needed.

Whether you’re creating a new website or reworking an existing one (take a look at card sorting), tree testing shows how well your IA supports your users’ goals.

Why do tree testing?

Tree testing helps you:

• See if users can find key content easily
• Identify points of confusion or dead ends in your IA
• Compare versions of your structure to see which performs better
• Give your team confidence in the IA before design and build
• Reduce the risk of expensive rework later

While card sorting shows how users group information, tree testing checks if your structure actually works for real tasks. Together, they form a complete IA research process.

How we do it

Every tree testing project is tailored to your needs, but typically includes:

Prototype preparation

We work with you to create a simple text-based version of your site structure (no design or layout distractions) and prepare realistic tasks based on typical user goals. These might include:

• “Where would you go to renew your membership?”
• “Where would you find information on delivery options?”

Each task is designed to reflect real-world use of your site.

Tree testing sessions

We use an online tool to test this prototype with 30-50 users. Participants complete the previously discussed tasks and we capture:

• Success rates (did they reach the right place?)
• Number of attempts
• Alternative paths taken
• Drop-off points

This provides statistical confidence and helps highlight widespread issues.

What you’ll get

Tree testing doesn’t just highlight problems – it delivers a clear path to improvement. You’ll receive:

• A prioritised list of tasks users struggled with
• Detailed insight into success rates, click paths and errors
• A revised IA structure based on findings
• Evidence to explain and support IA decisions to stakeholders
• Clear next steps to refine and retest if needed

We can also train your team to carry out future rounds of tree testing, helping you build long-term capability and confidence.

Ready to test your IA and make it work better for users?

Book a discovery call – we’ll help you uncover how your site structure performs, where it falls short, and how to fix it.