Tree testing: A valuable UX research method

Believe it or not, the practice of tree testing is incorporated into our daily lives. Whether you’re on a website shopping, navigating a grocery store, or using public transportation, this technique allows there to be a sense of organization while going about these tasks.

Along with staying organized, the results of this user research technique enable people to find what they need in a quick and manageable way.

What is tree testing?

Tree testing evaluates hierarchies. It’s called a tree test because of the branch-like structure participants receive that enables them to have quick and easy access to the information they’re looking at. This method begins with a main category and follows with subcategories underneath so the participants won’t have difficulty navigating.

In reality, a tree test often gets used to test out a navigation menu on a website. For example, when you are shopping on a website for clothes, furniture, appliances, etc., depending on the company’s website, they usually have categories across the top of the screen. When your mouse hovers over a category, a list drops down and it breaks down into specifics. Sometimes those subcategories have another subcategory to make the search even more particular. This allows the user to get straight to the point when they browse.

Card sorting vs. tree testing

Card sorting is another method of evaluating the organization of information. However, the difference between the two is that card sorting focuses on generating ideas by creating general categories.

With card sorting, the researcher provides the participant with cards to sort. There are different types of card sorting techniques that we won’t get into here, but you want to use card sorting at an earlier stage in your product development than you would tree testing.

For instance, if you’re a university researcher and you want to understand how people group different topics together, you’d give them cards with the components you want them to look at. A participant would then group those cards into specific categories, but they wouldn’t necessarily create a branching hierarchy.

The results of a card sort create a starting point for designers to make a navigation structure that you would then evaluate using tree testing.

Both techniques encourage organization and make a simpler platform for the audience. They are useful for a variety of matters. Card sorting is the foundation for tree testing because it allows you to get a glimpse of the vision you have in mind while tree testing is when you begin to arrange and focus on being more specific so the public can understand the concept you’re aiming for.

How do I know tree testing will be successful?

If you are someone who wants to fix an existing website’s navigation or create new website navigation, tree testing would be the suitable method. It is where you can test your ideas because tree testing will uncover where the issue is occurring. There are a few steps to complete to figure out if it will end up becoming successful.

Step 1: Figure out your goal

Preparation is important because this is where you figure out the topic/question you want to test. If you’re an ecommerce website, you might want to find out the ease with which participants can navigate your website and do important tasks such as: finding a product, buying a product, checking out, and finding out the status of an order. In that case, you’d write goals for each of those tasks and make them your research goals.

Step 2: Plan your tasks

While planning your tasks, always go back to your research goals. You would rewrite your goals into a prompt that you would give to your participant. You should always analyze and be precise with what you think the public would click on, but don’t lead them to the answer.

As a general best practice, it’s important to make the subcategories easy to follow. For example, if your category is Accessories, the subcategories can be Sunglasses, Bags, Belts, etc. when the subcategory Sunglasses is clicked on, those are the only items that should be shown.

It’s good to try to give participants a certain section to find and the direction they took to find it. You can also see how long it would take a participant to complete the task.

Step 3: Identify the right audience

You need to understand who your target user is for your website. If your main target user cannot navigate your website, then you’re in trouble.

You might want to look outside of the main demographic as well, but you want to make sure you’re meeting the needs of the people you’re trying to serve. You should be looking to your buyer personas for direction.

Step 4: Recruit participants

Once you identify your target user, you will need to recruit them. Recruitment is a big challenge. There are a number of recruiting and user-testing tools out there that can help with this. Alternatively, if you have the contact info of your users, you could send out an email blast inviting them to participate in your tree testing project.

Step 5: Run your study

We’ll talk about a few options further down in this article for different ways you can conduct tree testing, whether it’s analog or digital.

Before running your full tree test, you might want to do a pilot test to get the data you’ll need in order to continue or fix any potential issues with your test.

Once you run your study, you will see if it is easy to navigate through your website.

Step 6: Analyze your data

Once the results are in, you will be able to see if it was successful or not. Whether you conducted an online study, an in-person study, or both, your research will show you the routes your participants took to accomplish a task. This will help figure out how simple or difficult it was for the participants to find the information they were searching for. You’ll also be able to find how long it took for each participant to complete.

Step 7: Share your findings

Typically, you’ll want to share your findings internally with your various stakeholders so that you can get buy-in on your research. There’s no point in doing a research project that no one acts upon. Collaborating with others about your findings will allow you to receive feedback and bring them into the research process.

Strategies to conduct a tree test

There are a few different considerations when it comes to conducting a tree test. Should it be online or in person? Should it be digital or physical? Should it be moderated or unmoderated? We’ll explore these different options in this section.

Online vs. in-person

Conducting a test online or in person will depend on your resources and time. If your audience is local and it makes a big difference to have them in person, you might want to conduct your study in person. Perhaps you’re a grocery store and you want to improve how people navigate through your store. You might have them walk through your store and then bring them to do a tree test with you.

If you’re doing it online, it widens your pool of potential participants and saves you time and money.

Digital vs. Physical

Even if you’re doing your test in-person, you could use a digital tool to have the person do your test. That way, you’d be able to track your data more closely. Two of the leading digital tools for tree testing are UserZoom and Treejack through Optimal Workshop. Both of these are able to provide you with various methods to conduct your research to your preference. Also, they quickly offer and deliver participants worldwide to partake in your study.

A physical test would require you to manually create the groupings and simulate the test for them. It would take a lot more effort to analyze the data. However, physical tests can open the door for creativity that a digital test may not. For instance, you could have your participant make notes or draw on the paper you’ve given them.

Moderated vs. unmoderated tree tests

You can complete a tree test either moderated or unmoderated. Moderated involves having a dedicated moderator to run a session with participants. This can take a longer time to conduct, but you get the opportunity to ask follow-up questions. Moderated can also be extremely helpful because you are able to communicate with the person and see their way of thinking and why they chose what they chose.

Unmoderated would involve giving participants step-by-step instructions to navigate the website to see if it is clear to follow on their own without someone guiding them. There are pros and cons to both, and it often depends on your time and resources.

Final thoughts

Overall, tree testing is a beneficial and valuable method to use when you want to evaluate the organization of information. It’s best to start with card sorting to see the mental model of your participants and then use tree testing to further refine how you’re structuring your data.

Tree testing can save companies a lot of money because it may discover a critical issue that prevents users from completing a key task. An incorrectly categorized item could be the key barrier preventing the user from finding the product they want to buy.

Featured image source: IconScout