Tree Testing: What It Is, How to Run It and the Best Free Tools

A website may be nicely organized, but that doesn’t imply visitors can always find what they need. If users struggle to navigate around your website or find specific information, the fault may be in its information architecture. 

Tree testing is a usability testing method that can help you uncover these navigation issues by measuring how readily users are able to discover information presented in a website’s menu layout. It’s not a test of the visual design but of the organizing and labeling of sections and subcategories. 

In this article, we explain what tree testing is, why it’s important, how to do tree testing step by step, and the best free tools you can use to optimize your website’s navigation.

What Is Tree Testing?

Tree testing is a UX research method used to evaluate how easily users can find specific information within a website or app’s navigation structure.

Participants complete tasks by navigating a text-only hierarchy of pages, categories, or features without seeing the actual interface. This helps UX designers identify confusing labels and navigation structures before implementing the design.

This method allows teams to test their information architecture and identify navigation problems before building the final design, according to the Nielsen Norman Group’s guide to tree testing.

For example, let’s say you’re building a SaaS project management app. Users want to locate settings for invoices, team access, and subscription details. Tree testing can tell if they know what groupings these features are in.

The results help you improve navigation labels, category links and content layout before production begins.

Why Is Tree Testing Important in UX Design?

Tree testing is important in UX design because it helps identify navigation problems that may prevent users from finding the information they need. A structure that makes sense to your product team may confuse new users who are unfamiliar with the product’s terminology and features. 

Tree testing reveals these issues early, allowing UX designers to improve navigation and create a more intuitive user experience.

Tree testing lets you test your information architecture (IA) from the user’s perspective. This is a test to see if people can guess where content belongs from your category names and structure. This is important since navigation difficulties can make it harder to locate essential product features. 

Tree testing allows you to:

1. Detecting fake labels: Learn examples of categories that people get wrong or misinterpret.
2. Organize content better: Make sure that the pages are in the categories visitors expect to see.
3. Reduce friction in navigation: Identify superfluous steps, wrong pathways and unclear hierarchy.
4. Early design decisions: See navigation ideas before you start designing or developing the interface.
5. Consider other structures: Compare several navigation hierarchies for the most typical tasks.
6. Support product usability: Simplify the process for consumers to locate key features, options and resources.

You may also check out our approach to usability testing for SaaS apps for more general usability issues around layouts, interactions and whole user journeys. 

How Does Tree Testing Work?

In tree testing, participants receive specific tasks and select categories or subcategories they believe will lead them to the correct destination. 

Researchers then analyze their choices, success rates, and navigation paths to identify areas where the information structure may be confusing.

The test eliminates visual design elements, page content and interactive interface components. This allows you to focus on the structure and the phrasing of your navigation. 

Tree Testing Example: SaaS Settings

Task: You wish to change your subscription. Where would you go looking?

1. Dashboard
2. Account Preferences
3. Profile Security
4. Billing and Subscriptions
5. Team Admin Support
6. Expected destination: Account Settings > Billing and Subscription.

If most people follow the intended path, your hierarchy may be communicating the destination clearly.” 

If people continue to look under different categories, you have evidence that your labeling or structure needs to be improved.

A successful result doesn’t immediately show your entire navigation works. You should test several important tasks across different parts of the hierarchy.

Tree Testing vs. Card Sorting: What's the Difference?

The main difference between tree testing and card sorting is their purpose in UX research. Card sorting helps you understand how users expect information to be grouped and organized. 

Tree testing evaluates how easily they can find information within an existing navigation structure. Both methods complement each other and can be used together when designing a new product or redesigning a website.

The Nielsen Norman Group's information architecture study guide explains how these methods support different stages of information architecture research. The following table outlines the key methods and when to use each one.

When Should You Use Card Sorting?

Use card sorting when you need to understand how users expect information to be organized or how important sections should be labeled. 

For example, you can ask users to sort 30 SaaS features into groups that make sense to them. In an open card sort, participants create and name their own categories, while in a closed card sort.

After that, they organize the cards into predefined categories. You can then use these insights to design your recommended navigation structure. 

When Should You Use Tree Testing?

Use tree testing when you already have a proposed navigation structure and want to evaluate how easily users can find specific information. 

For example, if you organize SaaS settings into billing, account, security, and team management, you can ask users to complete realistic tasks and observe where they look for each option. This helps you identify confusing categories and improve the navigation structure before launch.

If you’re still unsure of your framework, start with card sorting. A structure, if you have one, tree testing helps to validate. 

Can You Use Both Methods Together?

Yes. Combining the methods gives you a practical way to build and validate information architecture.

1. Card Sorting: Discover how users group and label content.

2. Create Your Information Architecture: Build categories, subcategories, labels for navigation.

3. Tree Test: You can always find out what's available to you on the website.

4. Refine and Retest: Clarify misleading labels and navigation paths.

This process allows you to see structural issues before you start generating detailed page layouts or developing the product. 

How to Run a Tree Test in UX Research

To run a tree test in UX research, start by defining clear research questions, creating realistic tasks, and preparing a navigation hierarchy that reflects your planned website or app structure. 

Here are 5 steps to follow when you run a tree testing study.

Step 1: Define Your Research Goals

Start by identifying what you actually want to learn. A broad objective such as “improve navigation” is not specific enough to guide meaningful research.

Instead, identify the questions you want the study to answer. These questions will help you create relevant tasks and interpret the findings.

For example:

• Can users access billing options without navigating through unnecessary pages?
• Do users understand the difference between Account Settings and Workspace Settings?
• How easily can new clients find the cancellation policy?
• Which navigation structure helps users find team permissions more directly?

Try to establish some specific research objectives instead of relying on a single goal. This will help you design more meaningful tasks and draw actionable conclusions from the results.

If you’re overhauling an existing product, review your support requests, consumer feedback and analytics in order to uncover common navigation problems. A UX audit might uncover sites and user experiences that require additional research.

Step 2: Build Your Information Architecture Tree

After defining your research goals, design a text-only version of your proposed structure for a website or app.

Top-level categories Subcategories Individual destination pages Keep the hierarchy how you want to test the navigation.

The structure of an e-commerce website could be:

• Primary Navigation
• Store
• Apparel
• footwear
• Add-ons
• Orders & Returns Track order
• Return Item Or
• Refund Policy
• Customer Care
• Get in touch
• Delivery Details

Keep the tree focused on the pages and categories that are relevant to your inquiry. Refrain from adding visual styling, promotional banners, or interface features that could influence participants’ decisions.

You should also keep the separation between categories and destination pages. Participants need to recognize which elements they can extend and which are the final answers to their objectives. 

Step 3: Create Realistic Tree Testing Tasks

Your tasks should describe realistic goals that users want to accomplish rather than tell them which category or page to select.

For example, avoid asking: “Where would you find the Refund Policy page?” This gives away the destination by mentioning its name.

Instead, describe a real-life scenario: “You have returned a product and want to know when you will receive your money back. Where would you look for this information?” Here are samples you can tweak for different products. 

The task should describe the user's goal without repeating the exact wording of the correct category.

Step 4: Recruit Participants and Run the Study

Hire people who look like your intended users. For instance, if you’re testing a SaaS application for finance teams, your participants should reflect the people who would actually use that product.

Tree testing can be done remotely, either via a specific research platform or manually with a simple text-based hierarchy.

Pilot your tasks with a few people before launching the whole research . This helps you identify confusing directions, ambiguous destinations, and issues with your test set up.

In the course of the research:

• Provide each participant with one job at a time.
• Let them find their way on their own without telling them the right way.
• Do not tell them what category to choose.
• Record final answers and their navigation paths.
• Ask for optional feedback when you need to know why they choose one thing over another.

If you wish to collect responses without hosting each session, remote unmoderated tree testing is a good option. Moderated sessions are useful if you want to ask follow-up questions on participants’ reasoning. 

Step 5: Analyze the Results and Improve the Tree

When you have the responses, determine where the participants did well, where they went off track and where they chose the wrong destinations.

Don’t look at the great success rate. A task can have a high success rate and still contain superfluous stages in the navigation. Another task might reliably send participants into the same incorrect category.

Structure findings as actionable problems: confusing labels, pages in the wrong place, categories that overlap, hierarchy that is too deep. Then change the navigation and test again. This gives you evidence to drive design decisions rather than depending on internal preconceptions. 

Best Free Tree Testing Tools in 2026

The ideal free testing tool will depend on your study size, participant recruiting needs and the type of analysis you desire.

Tree testing is available for free on some systems, whereas on others it is a premium or business feature. Check the current limits before you pick a platform, especially if you need more than a handful of responses.

Here are 5 tools to consider for your next UX research project: 

1. Lyssna

Lyssna also offers tree testing, card sorting, questionnaires and other UX research methods. The free plan includes 3 collaborator seats and up to 15 responses to review from self-recruited participants.

You can measure task success, task directness, time to completion and navigation pathways. Further participant-panel recruitment.

Best for: Small teams and studies that need swift navigation. 

2. UXtweak

UXtweak provides research tools for information architecture, navigation, usability, and user feedback. Consider it if you expect to combine tree testing with other methods during a redesign.

Before you commit to a research, check the current free plan response constraints and the expected allowances and tree testing availability.

Best for: Teams looking for a lot of research methodologies on a single platform.

3. Useberry

Useberry Free Plan 10 responses/month 1 project 5 blocks. These constraints do allow some testing, but you should check that the present plan can support the tree-testing technique you have in mind.

Best for: Designers that want to blend navigation research with prototype-testing.

4. Maze

Maze has tree testing to check category labels and navigation topologies. But it’s worth noting that its official documentation features tree testing in its Enterprise plan, so don’t assume that the basic free offering has this way.

Best for: Product teams who wish to combine navigation research with broader UX testing.

5. Optimal Workshop

Optimal Workshop is known for Treejack, a specialist tree-testing tool. It supports evaluating findability, navigation paths, and category labels.

It's worth considering when information architecture research is a regular part of your workflow. Before you get the idea there is a free plan check the latest pricing and access specifics.

Best for: More systematic study and info architecture research.

Which Free Tree Testing Tool Should You Choose?

Use your research requirements to narrow down the options rather than choosing based on the number of advertised features.

Here is the list of tools you can use based on your requirements:

For a small startup validating its first navigation structure, begin with Lyssna or investigate UXtweak's current free offering. If you only need to validate a few categories, a manual test may be sufficient.

Can I Do Tree Testing for Free Without a Tool?

Yes, you can conduct tree testing for free without using a dedicated tool. Simply create a text-based navigation hierarchy, give participants realistic tasks, and manually record their navigation choices and success rates. 

This approach works well for startups, student projects, and teams that want to identify navigation problems and gather early user feedback before investing in research software.

Here's an easy method to accomplish it:

1. Sketch out a website hierarchy on paper or in a spreadsheet
2. Create 5-8 realistic challenges that highlight user goals.
3. Show sub categories as participants browse Share hierarchy with participants
4. For each participant, record the first decision, final location, backtracking, and time to completion.
5. Compare data, uncover groups that people get confused about consistently

You may calculate basic success rates and aggregate popular paths of navigation using a spreadsheet.

Manual testing requires more effort to maintain consistency in the process. You need to control the participant’s perception of the hierarchy and correctly record their selections. There's a lot of this automated by dedicated tree testing software. 

How Many Participants Do You Need for a Tree Test?

The number of participants needed for a tree test depends on your research goals. For early directional feedback, a small group of users can help you identify common navigation problems. 

However, if you want to make quantitative comparisons or draw more reliable conclusions, you’ll need a larger sample size based on your study design and the level of confidence you need.

For an exploratory study, a small group can help you find unclear labels and navigation patterns. Quantitative research requires more participants, to be able to accurately predict rates of success. 

The Nielsen Norman Group's tree-testing guidance distinguishes between qualitative studies for identifying problems and quantitative studies for benchmarking navigation structures.

Use the following as a practical planning guide, not a guarantee of statistical reliability.

These ranges are planning suggestions rather than universal requirements. Your sample size depends on the expected success rate, acceptable uncertainty, number of trees being compared, and whether participants test one or multiple versions.

For formal quantitative studies, use an appropriate sample-size calculation. The Nielsen Norman Group's guide to quantitative research sample sizes explains why the required number varies with the research design and precision you need.

Recruit participants that look like your actual users. A larger sample of people who don’t represent your audience can be less valuable for early discovery than a smaller, relevant group.

How Many Tasks Should a Tree Test Have?

A typical tree test should include five to eight well-designed tasks for an initial study. Each task should reflect a realistic user goal and test a specific part of the navigation structure. 

However, this is only a starting point, as the ideal number of tasks depends on the complexity of your website or app and the goals of your research.

Each assignment needs a meaningful destination or navigational decision. Including too few tasks may mean that key parts of your hierarchy are not being adequately tested; including too many may tire your respondents out and reduce the quality of later responses.

Select jobs covering:

• Key user flows like upgrading subscriptions, or managing permissions.
• Various areas of your hierarchy.
• Confusing or overlapping category labels
• And features that are important to users and they need to quickly find.
• Content that historically creates support tickets

Guide the duties before the whole study. If participants are misunderstanding what the work is, you may be judging faults of language rather than navigation quality.

How to Analyze Tree Testing Results

To analyze tree testing results, examine whether participants reached the correct destinations and how they navigated the hierarchy. Focus on key metrics such as success rate, directness, time on task, and navigation paths to understand where users struggle. These insights help you identify confusing categories, improve navigation labels, and make information easier to find.

The Nielsen Norman Group has a guide on analyzing tree test results, which will explain how these metrics help to discover information architecture difficulties. The following steps will help you analyze these metrics and identify areas for improvement. 

1. Success Rate

Success rate The percentage of participants who selected the correct task final location.

With this formula:

If 24 out of 30 people discover the correct destination, you have an 80% success rate.

Better findability usually means a higher success rate. But it should be taken in the light of work significance and other metrics.

2. Directness

Directness assesses whether participants went to the right place without backtracking or attempting other paths.

For example, a participant selecting Account Settings, then Billing, has taken a direct path. If someone looks at Profile first, then goes back to the parent category and chooses Billing, they’ve taken a bit of a roundabout way.

A high success rate with low directness may imply that people do get the answer ultimately, but nonetheless struggle along the route. 

3. Time on Task

Time on task is the time taken by participants to complete a navigation assignment .

Long timeframes to complete may imply misleading labels, too many levels of navigation, or doubt about where information fits.

Compare how long tasks take to complete. Longer time does not necessarily indicate that the structure is bad as the complexity and familiarity of the activity may also influence the results. 

4. First-Click Data

Initially-click data indicates the top-level category participants clicked initially.

This measure indicates their first assumptions. If many of the participants start in the wrong category, your top-level labeling or grouping could not fit with the participants’ mental model.

For example, if users are searching Team Management often for billing credentials, you might need to clarify how team positions connect to account billing. 

5. Navigation Paths and Final Destinations

Path analysis shows where individuals travel through the hierarchy, and what wrong destinations they choose.

Don’t just accept every wrong answer as a mistake. Look for recurring patterns:

• Participants consistently choose the same incorrect category: Revisit the category labels or grouping to make them clearer.
• Participants start correctly but lose their way toward the end: Review the subcategories and how they connect to the main categories.
• Participants explore multiple unrelated branches: Reconsider the overall information hierarchy and navigation structure.
• Participants reach the correct destination directly: Keep the existing structure unless other findings suggest a problem.

What Is a Good Success Rate for a Tree Test?

A good success rate for a tree test depends on the task complexity, target audience, and the consequences of users failing to find the right information. 

There is no universal benchmark that proves a navigation structure is effective. Instead, compare success rates across tasks and look for patterns that reveal confusing categories or navigation problems that need improvement.

However, the Nielsen Norman Group’s study of tree-testing data shows that the median success rate for a task is 62% among the 98 tree-testing experiments that Bill Albert and Tom Tullis analyzed. Their rubric is 61-80% good, 81-90% very good, and over 90% exceptional.

Think of these numbers as a guide, not a general rule. For instance, a task like account security or subscription management can require a substantially greater success rate than a task containing company information that is accessed less often.

Also your past findings are useful. If you change the hierarchy and now do the same task with 75 % success rather than 55 % success , that might be a big improvement . At least you should think about sample size and uncertainty before you say that it is a significant improvement . 

Tree Testing vs. Usability Testing

Both tree testing and usability testing examine user behavior but they are examining different aspects of the experience.

Tree testing tests navigation hierarchy without the entire UI. Usability testing is about how people use a website, an app or a prototype to do things. Take a look at the following table to understand the main differences

Use tree testing when you want to know where information should live. Use usability testing when you need to understand whether users can complete tasks in the actual interface.

These approaches can be mutually reinforcing. Tree test can test the navigation structure and a follow-up usability research can clarify if the designed menus and pages help users to achieve their goals.

For a broader understanding of navigation and discoverability, you can also review the World Wide Web Consortium's guidance on navigable web content.

Common Tree Testing Mistakes to Avoid

Tree testing produces useful results only when the study measures the problem you intend to investigate. 

Poor task wording, irrelevant participants, and misleading interpretation can undermine otherwise useful research. Watch out for these common mistakes:

Writing Leading Tasks

If your task includes the exact name of the correct category, participants may choose it because you supplied the answer.

Fix: Describe the user’s intent in natural language without repeating the destination label.

Testing Too Much at Once

With a big hierarchy of dozens of unconnected branches, it's hard to manage and understand your initial investigation.

Fix: Keep the relationships needed to traverse the structure realistically, while including the structure relevant to your research topic.

Recruiting the Wrong Participants

Internal team members often know product language that new customers don’t know. Familiarity may make it more difficult to see navigation problems.

Fix: Identify people matching your target audience, considering any important differences in experience, if relevant.

Treating Success Rate as the Only Metric

Participants may take several false turns to arrive at the right place.

Fix: Review task success + directness + time to complete + first clicks + common paths.

Changing the Hierarchy Without Retesting

The improvement could generate more ambiguity or shift the problem to another class.

Fix: Re-test critical modifications using similar tasks and participants. When you compare the outcomes, look carefully at differences between the groups that did the research.

Assuming Tree Testing Replaces Usability Testing

A successful tree test doesn't guarantee users will understand site content, identify interactive elements, or complete a workflow once they come to the page.

Fix: Want to evaluate a comprehensive user experience? Follow up with user testing.

When Should You Use Tree Testing?

Tree testing is a fantastic way to learn how to structure material and how people will find that item.

Use tree testing in these circumstances:

• Website Redesigns: Testing new category concepts for navigation prior to designing the pages.
• SaaS product development: Make it easy for users to find their settings, integrations, rights and paying options.
• Large content libraries: Help users find documentation, policies, and support articles.
• E-commerce Navigation: Test if your clients can find products, delivery information and return policies.
• Rewrite the navigation: There are several hierarchies, so look at them before you choose one.
• Post-launch enhancements: Log any client-reported or support-ticketed findability issues.

If the main topic is about visual hierarchy, where to place buttons, how to give feedback about interactions, or how usable a complex workflow is, then tree testing is not as useful. You could use different ways to address such queries like prototype testing or usability testing. 

How to Create a Tree Testing Report

A tree testing report should describe the story of what you tested, what participants did and the design decisions that the data supports. It should not only display your graphs and not say what they signify.

This report is short and action-oriented. These ought to include:

1. Research Objectives: What navigation challenges did you aim to investigate?
2. Study Design: Explain the study structure, including the information hierarchy, task wording, sample size, testing method, and participant selection criteria.
3. Key Metrics: Report task completion rates, directness, time taken to complete tasks, and common navigation pathways.
4. Key Findings: Identify confusing labels, misplaced pages, and frequently selected incorrect destinations.
5. Recommendations: What changes would you suggest based on the findings, and why?
6. Next Steps: Explain what you plan to change, how you will retest the structure, and which metrics you will use to measure improvements.

Example Tree Testing Report

Picture a SaaS dashboard with 30 users trying to use it. As you can see in the next table, you can convert hypothetical findings into recommendations.

Use the report to prioritize changes based on frequency, severity and business impact. Fix the navigation issues that prevent people from doing the primary jobs first. Then clear up the less important labels.

Frequently Asked Questions 

What Is Tree Testing in UX?

Tree testing is a UX research tool that allows designers and product teams to try out information architecture before a website or app’s navigation is finalized. Participants choose categories for realistic missions to reach the target destination.

What Is the Best Free Tree Testing Tool?

Lyssna is a great starting point for small teams – its free plan includes tree testing and allows you to observe up to 15 self-recruited responses. UX tweak is another to look out for. Which one is best for you depends on what features you have now, what your participants need, and how complex your question is.

Is Tree Testing Similar to Card Sorting?

Both can help improve information architecture, but for different reasons. Card sorting provides insight into how users group and name content . Tree testing indicates if users can discover information in a suggested hierarchy.

How Many Participants Do You Need for a Tree Test?

5-8 participants will help you find deceptive labels and navigation patterns in early exploratory study. You tend to need bigger sample sizes quantitatively. 

If you compare many hierarchies, you may require 50+ per version. The final sample size will be decided by the kind of investigation and the accuracy necessary.

How Many Tasks Should a Tree Test Have?

For a small study, you can begin with five or eight tasks. Pick jobs that span critical user journeys and areas of the hierarchy. Try them out to see whether the wording is clear and does not give away the correct answers.

Is Tree Testing Qualitative or Quantitative?

Tree testing may be qualitative or quantitative. Moderation can be useful to researchers to discover what makes people follow certain paths exploratory investigations.

Final Through

Confusing navigation can make it difficult for users to find useful content and product features. Tree testing helps identify these issues early by setting clear study goals, creating a logical hierarchy, and observing how representative users complete realistic tasks. 

Whether you're building a SaaS product or redesigning an existing website, use these insights to improve your information architecture before investing in professional design and development.

Orbix Studio works with startups and emerging businesses to design and build digital solutions around real customer needs. Discuss your product’s navigation, information architecture and broader user experience with our UI/UX design services.