90% of users stop using an application just because of it’s poor performance. 

Usability is a key metric to track for a successful website or app design. It helps ensure that users can seamlessly navigate and find the information they seek. Tree Testing is great way to test your Information Architecture (IA) and optimize your website or app. 

Tree Testing helps organize the content in a logical and intuitive way. Especially when combined with card sorting, it can yield great results. It can enable you to get answers to questions like-

• Can my users find the information they are looking for? If not, where are the friction points?
• Does the categorization of the content make sense to the users? 
• Do users find the labels on my website or app logical?

In this article, we do a deep dive into tree testing and how it can be a complete game-changer for you. 

What is Tree Testing?

Tree testing is a popular method used in UX research that is used to assess the discoverability and hierarchy of content in a website or app. 

‍

Tree testing is an important step in your research journey. The respondent needs to navigate the text and category labels without any navigational aids or visual design which are present on the actual site or app. This helps you get key UX insights and create a more user-friendly design. 

What is Tree Testing Used For?

Tree testing can be used to test your hypotheses before producing a product or making any big modifications, from redesigns to site migrations.

What Questions Can Tree Testing Answer?

1. Objective: Validating concepts

Question to ask: Do users find the labels on my website or app logical?

2. Objective: Usability Testing

Question to ask: Is the organized content easy to find?

3. Objective: Ensuring logical and clear hierarchy 

Question to ask: Are there any existing navigational structural hierarchy problems?

4. Objective: Create a base for the design that will be layered on top of the framework 

Question to ask: Is the available information easy to find?

5. Objective: Eliminate redundant paths 

Question to ask: Is the navigation redundant in any way?

6. Objective: Identify areas where crucial details or elements are absent

Question to ask: Are there any gaps in the navigation?

7. Objective: Enhance the readability and clarity of menu choices and labels

Question to ask: Did the users find any of the choices and labels misleading?

8. Objective: Identify and reduce any confusing paths

Question to ask: Is the user facing any issues with the flow?

Tree Testing Use Cases

Tree-testing is a great way to gather valuable insights about your product. Some areas where tree-testing can be used are- 

1. Development of a New Product

Your product's information architecture is a crucial component, whether you're creating a new website, mobile app, software suite, or digital platform. Before spending time and money on development, tree testing enables you to assess and validate the usability of your proposed information architecture.

2. Redesigning and Migration of a Website

When it's time to upgrade, relocate, or rebrand your website, you need to make sure that the new or reorganized version is still intuitive and doesn't lose key functionality. Test your changes using a tree to make sure they improve navigation and findability while keeping your users in mind.

3. A/B testing various Information Architectures (IAs)

Suppose you have a software suite, but users keep complaining that they can't find what they're looking for as new features are added. With tree testing, you can assess each suggested IA without having to fully implement it first. Your team has diverse ideas for how to reorganise the features.

Examples of Tree Testing

Tree-testing is an extremely versatile UX research method. Here are a few examples of how you can make use of tree testing-

• A retail store can ask the respondents to locate where they check their rewards or special discounts
• A music app can ask the respondents to locate the payment page to upgrade the existing account plan
• A travel app can ask respondents to navigate to the previous and upcoming bookings page
• A video-on-demand service can ask respondents to find the top new releases of the month
• A social media app can ask respondents to find the new location of the explore page

And so on. 

What Does a Tree Test Look Like

‍

In tree testing, there are mainly two elements involved- the tree and your tasks.

Before we understand what a tree test looks like, you need to know what is a tree. 

What is a Tree? 

A tree is a text-only representation of the organization of your app or website (like a sitemap).

What Does a Tree Test Look Like? 

You design a series of challenges that require your participants to nominate places in the tree where they anticipate finding the task's solution (e.g., "I'd find it here"). To find any navigational or structural problems, the participant's success rate and the time it takes to get there are recorded and analyzed.

Key learnings you’ll take away from a tree test are:

• Whether the respondents were able to reach the correct page 
• How many respondents were able to complete the task
• Which paths did the respondents take to reach the goal screen
• How long did the respondents take to complete the tasks
• How many respondents dropped off or got lost along the way 

The Advantages of Tree Testing

Early in the design process, tree testing is a great way to assess how well your website's or app's navigation hierarchy works. You can better organize your content by using the information from a tree test to better understand how users go through your website. Let's examine each benefit in more detail-

1. Usability & Ease of Navigation Testing

Tree testing can be used to identify and eliminate issues with navigation. Problems like misleading or unclear labels, pages or confusing paths that as an obstacle for users can be easily optimized. 

By analyzing whether your goals align with the test results and understanding if users take a longer time than expected to locate certain information, you can change the way your content has been categorized in the information architecture. 

2. Evaluate Labels and Language Used

Labels help users navigate through your site or app. Since tree tests are text-only versions of your site or app, you can understand whether your labels work or do the design, interactive elements and content play a role in influencing the user’s interpretation. 

3. Assess User’s Mental Models 

Your site or app will not be the first site or app your users have used. This makes understanding their inherent expectation of how something works important.

This massively impacts the way the user approaches and interacts with your product. Tree Testing can helping you get answers to questions like- 

“What do the users expect to find?”

“Where do the users expect to find it?”

“How do the users expect that particular item to behave?”

4. Prioritizing Issues

Tree testing gives you pin-point actionability about the areas that need to optimized. 

You can analyze and identify the areas in your tree that are causing- 1) confusion, 2) misunderstanding, 3) vagueness or 4) difficulty. But solving all issues simultaneously is as arduous task. Tree testing highlights the areas that need care the most. 

Other advantages of tree testing are -

• Tree tests are quick, simple and easy to answer
• Recruitment for respondents is easier 
• Tree tests can be conducted remotely, thereby reducing cost and researcher bias
• Data analysis is fast and actionable

Limitations of Tree Testing

Tree testing like any other UX research method has its own set of limitations, some being- 

1. It is difficult to visualize as the tree contains only text and no visual elements.

2. Since tree tests are conducted remotely, there is no further probing from researchers. 

How to Conduct Tree Testing

Here is step-wise guide to help you conduct effective tree tests- 

Step 1: Design your Tree Test

The first step to conducting a tree test is designing it. The design acts as the foundation by helping you keep your study on track. Let’s see how you can effectively design your study-

• Create a research plan 
• Define the objectives of your research
• Choose the questions you want to ask in the test
• Communicate a tentative timeline to the team
• Create a site map
• Map the tasks you want respondents to attempt
• Identify the metrics you want to measure

Step 2: Define the Structure of Your Tree

Outline the categories, subcategories, and pages that make up the tree structure of your website or app. It's crucial to be detailed while creating your subcategories because doing so will encourage actual user behaviour. 

For instance, Bookings could be the name of a navigational category. The menu structure for the many subcategories may include Upcoming, Completed, and Cancelled.

Make sure your target audience is aware of how that portion of your product relates to the product as a whole, even if you wish to test a particular aspect of it. You can then use this information to take appropriate action after examining your results.

Step 3: Set Goal-Based Tasks

Create challenges for participants that require them to use a top-down technique to locate a page in a tree. Assigning effective and insightful tasks are key while conducting a tree test. 

For instance, to uncover insights about the findability of the accounts page, you might design a task that asks participants to figure out the best approach to update their profile on the app. An example of a tree test problem is given below: 

“You recently changed your mobile number. You want to update your personal details on your profile. Find out how to do that.”

Step 4: Recruit Respondents

While preparing to conduct the tree test, keep in mind that it is critical to select the right respondents because they represent the preferences, opinions and feedback of target audience. The number of respondents depends on various factors like -

• The type of test you are conducting
• The objectives of the the UX study
• The target group, etc. 

You can also choose a research platform like Qatalyst to conduct tree tests remotely, get access to an global online panel and seamlessly gather and store insights. This way, you can reach out to a wider audience and get a better understanding of your target audience. It is important to select the right target audience and understand how your changes will impact them. 

Step 5: Choose the Right Method

You can conduct tree testing in-person or remotely. In-person testing is moderated and it lets you ask relevant questions about the respondent’s choices in the moment. But, remote testing will yield you better results overall as it is easy to set up and quick. Respondents can take the test from anywhere at any time, without you being present. Qatalyst’s templates enable you to create your tree easily by adding and removing blocks, making tree-testing a breeze.

Step 6: Conduct a Pilot Test

Conducting a pilot test is a great way to ensure that your final test has no hiccups. It helps you understand if your tree test makes sense to the respondents and works the way you expect it to.

Doing this will reduce the chances of- 1) missing out on important details, 2) any logistical issues, 3) misleading or confusing instructions and help you gather valuable and actionable feedback. It also helps bring in fresh perspective and new ideas.

Step 7: Conduct the Tree Test

Depending on the method chosen, the testing tool will generate a link that will allow your respondents to take the tree test. Add survey questions before and after the tasks to gain valuable data about your respondents.

Once the respondents are done taking the test, you can start analyzing the data to make data-driven decisions about your UX. 

Step 8: Analyzing your Tree Test Results

The collected data will get recorded in the research platform of your choice. Qatalyst gives you insights about the usability, average time taken, misclicks, drop-offs, bounce rate, areas of interest, the different paths taken, etc. 

{{cta-case}}

These insights help you answer critical questions like- 

• “Where and why did respondents drop-off?”
• “Were the labels confusing?”
• “Why did the respondents feel that they would find certain information under a particular category?”
• “Why did they take a longer path?”
• “Why did they take a long time to complete the task?”

And so on. 

You can validate your concepts and make data-driven designs by analyzing this data.

When Should You Conduct Tree Testing?

A short answer to this is regular and often. An IA might get "stale" over time. This is especially true when new links are introduced, or old links are changed. But it can also happen when users' understanding shifts in response to new information and ideas. Testing your benchmark trees regularly is a 'non-invasive' and straightforward way to determine the state of your IA.

How to Use Tree Testing with Other Usability Testing Methods

Tree testing being a versatile research method, it is ideal for a multitude of use cases. It helps researchers get insights into the information architecture of their site or app.

By combining tree testing with usability testing methods, you can understand your users in-depth and make your site or app more efficient and user-friendly. 

Let us look at some of the other UX research methods you can use and how you can combine tree testing with them-

Card Sorting

What is Card Sorting?

Card sorting is used to get an understanding of how users categorize information. The idea is simple, respondents are asked to categorize topics and information into buckets.

Card sorting is of three types- open (respondents are allowed to create and name groups), closed (respondents categorize the cards into pre-defined buckets) and hybrid (respondents can mix both open and close together)

Card Sorting Example 

For example, there is a new gaming app based on Harry Potter, and it wants to see how users categorize the following characters-

Here you can see that Harry has been categorized as a “Hero” by 88% of users and as a villain by 12% of users. 

Similarly, under the “super-smart people” users have added cards associated with McGonagall, Dumbledore and Snape. 

How to Use Card Sorting with Tree Testing

Once you are done with card-sorting, you will start to see your product through the user’s lens. It will help form the foundation of your information architecture and design it. Once that is done, you can conduct a tree test to identify gaps or issues with the IA or the labels. This way, you can get to know user expectations (using card sorting) and validate their behavior (using tree-testing).

A/B testing

What is A/B Testing?

Also known as ‘split testing’, A/B tests help you to evaluate two versions of the same webpage or app. 

How to Use A/B Testing with Tree Testing

By using A/B testing and tree testing in tandem, you can evaluate the effectiveness of both versions of the Information Architecture. 

These variations may differ in the category names, item grouping, tree depth, or any other component of the IA you want to evaluate and improve. Run a tree test by grouping your participants and giving them a variety of tasks. Then, compare the results to see which IA performed better.

You can conduct an A/B test to validate the findings after implementing them in your live environment. Display the updated version of your IA to some users while showing the old version to others (as a control). The effectiveness of the new IA may then be verified by comparing user behaviour and key performance indicators (KPIs) between the two versions.

First-Click Testing

What is First Click Testing?

The first click testing is used to find where the respondent clicked first to complete their intended task, on the interface. 

How to Use First-Click Testing with Tree Testing

Once you conduct a first-click test, use tree testing to evaluate if the number of correct first-clicks increases. Misleading labels or categories tend to confuse the users and respondents who start with the correct first-click tend to successfully complete the task.

Bottom Line 

Tree testing is a great way to refine your information architecture and create a seamless and user-friendly site or app. However, using it in tandem with othe UX research methods will yield the best results and help you gather reliable and actionable insights.

{{cta-button}}