A digital product’s navigation is its skeleton. It provides structure and guides users through the different features and content within the product. An app or website with a poorly designed navigation system can frustrate, confuse, and ultimately, drive users away. 

Tree testing is a UX research method designed to address this issue head-on. Unlike traditional usability testing methods that involve interacting with live websites or assessing an app prototype’s visual design elements, tree testing focuses solely on the navigation structure.

In tree testing, users are presented with a hierarchical representation of an app/website’s content and asked to find specific items within the structure. The hierarchical representation typically consists of text-based outlines or visual diagrams that only display the labels of categories and subcategories of the product’s information architecture. There are no design elements or interactive features in this representation.

So, participants are not asked to evaluate any designs or interact with the interface; instead, they are simply asked to locate specific pieces of content within the presented structure. This basic approach allows researchers to pinpoint all areas of confusion users encounter as they locate specific categories of content within their information architecture.

By validating your website or app’s proposed information architecture with this method you can head into the design phase with much more confidence, knowing that your product’s navigation is intuitive and aligned with user expectations.

The foundational understanding of your product’s navigability will also reduce the likelihood of costly design mistakes later on in the process. Hence, tree testing is an extremely revered UX research method. Go through this article to understand how tree testing helps UX designers validate the usability and efficiency of their information architecture structures. Also, understand how to perform tree testing to evaluate your website or app’s navigability.

What is Tree Testing?

Tree testing is a task-based UX research method where test participants are asked to look for specific items inside a simplified version of the app or website’s information hierarchy. This ‘hierarchy,’ usually presented as text-based outlines or simple diagrams, is what UX researchers call ‘trees.’ A tree is basically a text-only version of an app or website’s information architecture (IA).

Information architecture in UX is a term that describes the organizational structure of all content inside an app or website. So, a ‘tree’ is similar to a ‘sitemap’ – it only shows how content is organized in a product, without any actual content or design elements.

To perform tree testing, you can create a unified tree that includes all categories and subcategories of your entire site/app’s content. Or, you can focus on specific sections of your app or site, such as the blog section or the product pages. The latter – a more focused approach – is preferable among researchers because it makes the tests shorter and simpler for the users and easier to manage for the researchers.

For example, you can create a separate tree for your website’s blog section that includes only the main blog topics and subtopics. Ask participants to navigate through the content section and locate specific blogs. Just after a few, short-duration tests, you’ll find out how navigable your website’s blog section is.

For the next tree test, you can create a tree focused exclusively on the product section. By conducting multiple tree tests like this, you can uncover all potential navigation issues in your app or website. You can also gauge the effectiveness of all the labels used in your app. Do all the names, labels, and content categories make sense to users?

More importantly, you can peek inside users’ minds and discover what mental models they use when they are searching for information. Tree testing reveals how users expect to find information inside an app or website and how they subconsciously categorize different pieces of content in their minds. That is why tree testing is also known as ‘reverse card sorting.

Tree Testing vs Card Sorting

Tree Testing vs Card Sorting is a hotly contested topic in the UX research world so let us clarify it really quick. Card sorting is another UX research method that is used to evaluate the navigability of digital products. In this test, users are asked to group pieces of information into different categories, based on what feels right and natural to them.

Users are either given paper cards or digital cards and asked to organize them into groups/categories that make the most sense to them. For example, let us say you are given cards that represent different types of clothing products on an eCommerce website, such as “Leather Shoes,” “Leather Boots,” “Jackets’, and “Socks.”

You might group “Leather Shoes,” and “Leather Boots” under a category called “Travel Gear” and place “Jackets’ and “Socks” under a category called “Winter Wear.” These test results will help researchers understand how you perceive relationships between different pieces of information on clothing products and how you would naturally categorize them.

Now you see why tree testing is called reverse card sorting. Think about it – in card sorting you build your own content labels navigation system from the ground up using your own intuition. In tree testing, you critique or validate an existing information hierarchy. Another key difference between these two methods is how the test results are analyzed and used.

In tree testing, researchers analyze how users locate specific pieces of info within a predefined hierarchy. The goal is to identify what routes users choose intrinsically as they look for specific items. Once researchers identify where users struggle to find content and which navigation routes are the most confusing, they can revise the product’s IA for better navigability.

In card sorting, the focus is on understanding how users’ minds work as they naturally group and label different pieces of content. Researchers analyze the labels, categories, and sub-categories created by the test participants. They are looking for common patterns that reflect the users’ mental models. Once they find out how most users group and categorize information they create intuitive navigation structures for their apps/sites based on those mental models.

Overall, tree testing is better at evaluating the findability of items in a pre-defined navigation structure whereas card sorting is better at understanding how users intrinsically categorize specific types of information in their minds. Both UX research methods provide a ton of actionable insights on improving the product’s navigability and overall UX.

Plus, both are extremely easy and cheap to perform. For example, to perform tree testing, you don’t need to create any clickable prototypes or design layouts; nor do you have to write long pieces of content. You only need a text-based ‘tree’ that displays a series of navigation categories, without any graphic components.  

The next thing you need is a set of strategic tasks that instruct participants to look for specific pieces of information within the tree. With these two basic components, you can find out exactly where participants feel confused, take wrong turns, or experience delays while exploring your app or website’s information hierarchy.

Common Use Cases for Tree Testing

Tree testing is the most effective in the earliest stages of the product development/design phase. Right after you have defined your IA and, right before you start designing the visual elements or creating a clickable wireframe, is the perfect time to validate your planned information structure’s usability. That means you get to pre-emptively avoid any major navigation issues before you design or code anything.

Here are some other scenarios where tree testing is particularly effective at saving time and resources:

Website Redesigns and Migrations

If you are redesigning or rebranding your site, or migrating to a new address, performing tree testing will help you in many ways. You get to improve your old website’s navigation and findability and ensure that your new site is more intuitive than ever.

Testing Different IAs

Most design teams create multiple app layouts, user journeys, sitemaps, and IAs. If your team has done the same, tree test each IA with real users to find out which one’s worth implementing.

Testing Different Information-Grouping Methods

Let us say you are launching a new app section for local users. Tree testing can help you pick the most effective IA for this new section. Use it to compare different navigation and organization schemes and check which ones your target user base prefers navigating before actually designing or developing the new section.

Most UX design teams or IA engineers perform tree testing every time they perform card sorting. It helps them validate navigation and IA structures that the users themselves have organized during the card sorting exercise, using their own mental models.

Benefits of Tree Testing

The most apparent advantage of tree testing is its ability to improve website or app navigation. The better the navigation, the better the customer experience. Tree testing separates the IA from other design elements, enabling testers to evaluate the IA on its own. Therefore, it is very effective at exposing navigability or ‘findability issues. Here are some other benefits of tree testing:

Resource-Efficient

You can set up and perform a traditional tree test within a week. This includes analyzing the test results! That is how fast and effective tree testing is! And we are talking about traditional tree testing with paper, notes, and in-person sessions. With online tree testing tools like UXtweak, PlaybookUX, or Treejack, you can speed up the process even further.

These tools allow you to recruit participants online. You can create and distribute tree tests directly and gather data in real time. Conduct the tests remotely and then use the automated analysis features on these platforms to swiftly interpret the results and identify the most pressing navigability/usability issues.

No Need for a Prototype

You do not need to create a prototype demonstrating your app/website’s structure (although you can if you want) to perform tree testing. You only need a basic ‘tree’ and some willing participants.

Time-Efficient

Unlike hours-long surveys or usability tests, tree tests are usually no longer than 15-20 minutes. This means you shall have no trouble recruiting willing participants and collecting their valuable feedback.

Quantitative Feedback

The insights you glean from tree tests are ‘quantitative’ not qualitative. You get to know exactly how many times users successfully locate specific items, how long it takes them to find those items, and how they navigate through the headings and categories. This data provides clear metrics on user performance and helps identify patterns in navigation behavior.

Combine this quantitative feedback with qualitative insights from early user interviews, and boom – you have a comprehensive understanding of your product’s UX long before you have made any major design decisions.

Limitations of Tree Testing

Before we teach you how to perform tree testing, let’s discuss some glaring limitations of this UX research method. We do not want you to go into it with unrealistic expectations or draw wrong conclusions about your product’s navigability or usability. Here are the key limitations of tree testing you need to know about:

Limited Scope

Isolating the site structure to purely test an app or website’s IA and navigability is a highly valuable exercise. However, this hyper-focused approach only provides a limited understanding of the site or app’s overall navigability and usability. That’s because there are several other factors, other than the main navigation tree, that influence an app/site’s navigability and findability. Factors like the search bar, the design’s visual hierarchy, the layout of content, and the presence of CTAs all play critical roles in guiding users to the info they need.

Tree testing does not evaluate any of these other crucial elements. It looks at the navigation structure in a vacuum, without considering the real-world context of the actual site or app. The narrow scope means you cannot conclude overall usability from tree testing alone. It is best used as one piece of the larger UX research puzzle.

Reliance on Quantitative Data

Most tree tests today are done online, remotely, and unmoderated. Researchers do not stand next to users as they navigate through trees. That is why tree testing is so cheap. But it is also why it only produces figures, statistics, and other forms of quantitative data. Quantitative data will show you exactly what users did, but not why they did it. Once again, tree testing falls short on its own and it needs to be combined with other qualitative UX research methods.

Task-Phrasing

The term ‘task-phrasing’ refers to the process of creating ‘tasks’ for a tree test. ‘Tasks’ like “Find the Service Option,” or “Locate the Jewelry Section,” guide participants through the navigation structure during tree tests. Creating such tasks is not easy. Researchers have to carefully consider how users think and what language they use when searching for information.

They must ensure that the tasks are clear, relevant, and reflective of real user scenarios. All of this is hard work, and if you do not do it diligently, it can significantly impact the quality of the results. If the tasks are poorly phrased, they can confuse users or add bias to their answers.

So, in the next section, we are going to teach you how to perform tree testing and task-phrasing.

How to Conduct a Tree Test

Follow these steps to conduct a successful tree test and validate your website/app’s navigational structure:

Tree testing process in 2024

Step 1: Define Your Objectives

Before diving into the tree test, define what outcomes you want. Do you want a list of all the areas in your site’s navigation that confuse users? Or, do you want to discover what types of decisions users make when they’re presented with multiple sub-categories? A list of well-defined objectives like this will keep your research team focused. Knowing what you want to achieve upfront will also help them evaluate the test results more objectively.

Step 2: Finalize the Research Plan and Tree Testing Questions (Task-Phrasing)

Finalize the research plan and timeline with stakeholders. Then, start defining the research questions. You need to give the test participants four types of tasks, at least.

  • Essential tasks that mimic real-world user queries. For example, “Find the Return Policy and the FAQ section.” List all potential real-world user queries and present them as essential tasks. These types of tasks will reveal the most pressing navigability issues in your tree.
  • Softball tasks help identify if participants are actually paying attention. Slip a few of these in there (for example, “Find all 25 sub-headers in the main homepage menu.”) to keep the test participants in check.
  • Thesaurus tasks that test the effectiveness of category labels. Ask users whether all the labels make sense. For example, if a label says “Return Policy FAQ section,” do users understand what it means? Ask these types of probing questions for all labels/categories within the structure.
  • Diverging roads tasks that evaluate how users differentiate between two or more similar-sounding categories. For example, if an app has categories like “Product Description” and “Product Features” the diverging roads task should ask participants to select the label that conveys information about “how the product works.” Whichever one the users choose the most should be used as the official app page label.

Before finalizing these tasks, make sure that everyone in the design team understands what they imply.

Step 3: Build Your Tree/s

Once you have created tasks for the participants and defined the key metrics you’ll record to analyze the data set to be gathered – create a text-only blueprint of your site/app’s structure. All main and sub-categories of content should be included in this blueprint.

Outline the tree structure with all the categories, subcategories, and pages in your site/app. Use clear and concise labels to maximize clarity. If your site/app’s structure is too complicated, create multiple trees for different sections of the structure.

Step 4: Choose a Tree Testing Tool

Use tools like UXtweak, PlaybookUX, or Treejack for this step. Enter your tree structure into the tool, in the right format. Do not miss any category or subcategory. Review the tree for accuracy before you enter the tasks.

Step 5: Enter the Tasks tasks

Enter all the tasks you prepared in Step 2 into the tool. Include anywhere between 10-30 tasks. For every task, define the correct answer/s as well. By entering this info, you make the tree testing tool automatically calculate the success/failure of each task.

Step 6: Choose Your Participants

Choose test participants who have different levels of familiarity with your site/app. Pick some who have used it before, some who’ve only heard about it, and some who have no clue about it. Pick at least 15 users for the test.

Step 7: Pilot Run

Do a pilot test with your team before you hire any test participants. This pilot run will ensure that the test is ready to be given to users.

Step 8: Run the Test

Use your chosen tree testing tool, share it with your registered test participants, and then launch the test. These tools will automatically monitor how users navigate the tree and complete each task.

Step 9: Analyze Results

Review the collected data. Some tools will automatically explain to you what each data point means in relation to your app/website’s navigability. If the tool you are using does not offer that feature, no worries. Here are the key metrics you need to look for and analyze: 

  • Success Rate: % of participants who completed all the tasks. 
  • Directness: Measure of how many users found the correct answer for each task on their first attempt.  
  • Time: Average time taken by the participants to complete each task. 
  • Paths: Visuals depicting all users’ routes throughout the tree.

Understand their navigation choices by assessing these details.

Step 10: Refine Your IA

Use the insights from the last step to create a list of all the changes you need to make to your IA. For example, if “Success Rate” and “Directness” are low for a particular category, you may have to rename/reorder all the category labels in that section. If “Time” is significantly higher than expected in some sections, simplify their navigation and provide clearer pathways. If “Paths” shows that most users are taking circuitous routes to reach specific sections, you’ll have to rethink the structure of that section entirely.

Tree testing is an iterative process and the more you perform it, the more accurate results you get. So, do not quit after one test. Keep repeating the process for every section in your website/app’s structure. Regular testing and refinement will gradually perfect its navigability.

Using Tree Testing with Other Methods

We have already discussed tree testing’s limitations as a UX research tool. Now, let us learn how to overcome these shortcomings by strategically combining tree testing with other research methods:

  • Begin with card sorting to uncover your target users’ ‘mental models’ when it comes to grouping and labeling information.
  • Next, tree testing to validate whether those user-defined groupings translate well into a functional navigation structure.
  • Next, give users a clickable prototype and ask them to find specific information.
  • Use the click-testing results to refine your tree test. Perform tree testing again.
  • Now, A/B tests different variations of your navigation labels to identify which label names and navigation formats offer the highest degree of ‘findability’.

Finally, end the test by conducting usability testing on clickable, high-fidelity prototypes.

Conclusion

At Design Studio, we always perform tree testing but we do not give it precedence over other UX research techniques. Behind each of our successful web app designs, lies hundreds and hundreds of such tests. That is what it takes to deliver highly refined design experiences in 2024. To learn more about our tree testing or other UX testing methods, contact us now!