Sep 25, 2019
When learning your way around a new app, you might like to dive straight into the product to understand how it works by means of trial and error, trial and… aha!.
Or you might go through the welcome package and read the product's documentation, with a "study first, make second" approach. Most people are a combination of both.
We're always publishing how-to articles on our Help Center, but we also wanted to share with you a complete guide to using Maze that you can access when you're first getting started or when you need to brush up on your Maze skills.
Here, we'll go through everything you need to know about using Maze to test your designs with users. Navigate the table of contents below to jump to a section that interests you, or get started by reading the chapters in order.
Maze allows you to test your prototype with users and get valuable data that helps you to design better. In this chapter, we’ll look at how to get your prototype ready for testing.
Maze works seamlessly with a prototype you've already created. To begin testing your prototype with Maze, you only need the prototype URL—it’s that simple!
Maze supports Figma, InVision, Marvel, and Sketch prototypes.
Testing on Maze takes place with missions and questions, all built together in a maze test for users to complete. Missions are tasks users normally do in a live product, e.g., signing up or adding a new profile picture. These missions are completed by clicking or tapping on the screens in a prototype.
That's why it’s very important the screens and hotspots in your prototype are specifically defined for testing. If you’ve already built your prototype in one of the tools Maze supports, you can start testing right away by importing your prototype to Maze. Section two covers how to import your prototype to Maze from each design tool.
With Maze, you can test a prototype at any level of fidelity—all you need is a flow you can test with users. For example, for an e-commerce website, the user flow for the check-out process can be divided into these categories:
With the user flow determined, you can create a prototype with your designs, and submit it to test with Maze.
You can test the full flow of your product from beginning to end to learn if users can complete tasks with your website or app. Or, you can dive into a particular section of your product, such as the check-out process, and test copy and language, element placement, visual design, and more.
Make sure to test at least once before you handoff the designs to development using a mid- to high-fidelity prototype, and whenever possible, test earlier and more often during the design process. This way, you'll make sure your customers are able to easily learn and use your product.
Import your prototype to Maze to start testing. If you don’t have a Maze account yet, now is the time to sign up to Maze. After verifying your account, you’ll be able to import your prototype and begin creating your test.
Here are step-by-step instructions on importing a prototype from each design tool Maze supports:
When you import your prototype, Maze automatically creates a new project with a draft test (maze) inside based on the prototype.
💡Good to know! Within a Maze project, you can run as many tests as you need to fully assess your prototype.
It’s time to create your maze test! Build your maze out of blocks from the left-side panel in the maze creation dashboard. There are three main types of blocks you can add to your maze:
💡 Good to know! When you edit and create a maze, it's in draft mode. After you’re done creating your maze, you’ll need to set it live to start testing. A live maze is one that records and stores interactions and responses.
There are some best practices to follow when you create missions to get accurate results and provide a good testing experience to users. When you build your maze, aim to:
For example: “You want to start tracking expenses for your business. Sign up and set up your business details to get started.”
💡 Read more tips and practices for creating great missions for your maze tests.
A mission is made up of a title, a description, and the expected path users are likely to take. Let's look at these three elements in detail.
💡Tip! You can format the text in your maze with Markdown.
1. The title. A short, actionable sentence that defines the goal users have to complete.
Examples of mission titles:
Aim to keep your title concise and give users a clear task to complete. Take a look at more best practices for writing mission titles.
2. The description. A longer sentence that gives a general instruction and sets the scene of the task.
💡Tip! Refrain from giving users instructions on where to click or tap in the mission description.
Example of a mission description: “You’ve forgotten the password to your account. Go through the necessary steps to change your password.”
The mission description should be clearly written and avoid leading instructions that give users hints on where to click, what to do, or how to navigate the interface.
Check out more best practices for writing mission descriptions.
3. The expected path. The flow you expect users to take to complete the mission. You can set more than one expected path for a mission if there are different valid ways to complete it. For example, your product allows users to sign up from a landing page and from the main website.
💡 Good to know! Maze records every path users take to complete the missions in your maze. The expected paths you set will be used to measure the direct and indirect success metrics in your results.
To set the expected path, click on the hotspots in the prototype on the right side of the maze creation dashboard.
You can see how missions look like—and your maze overall—by clicking on Preview in the top corner. Note that no interactions are recorded when you preview your maze.
Your maze test can include missions only or you can add usability testing questions to survey users during testing. Missions allow you to record quantitative metrics and usability data, while questions can help you to gauge user opinion and gather qualitative feedback.
There are four types of questions you can add to your maze:
The phrasing of your questions matters: aim to write clearly and avoid leading users towards an answer. For example, asking "How great was your experience?" leads the respondent towards a positive reply. Instead, ask users "How was your experience?" to get more accurate insights.
To add questions to your maze, choose and click on the question type from the list of blocks in the maze creation panel. When you start typing, you can preview how respondents will see your question on the right. Write a description if you need to add details or specify something (optional).
💡 Good to know! For Open Questions, you’ll be able to choose which answers go into your maze report. For Multiple Choice, Opinion Scale, and Yes/No questions, you can read individual replies and view the aggregate results for each question. Read more about your maze results in section 12.
Planning the setup of your tests is important no matter the usability testing method you use: remote or in-person, unmoderated or moderated. For in-person testing, think about setting up a dedicated device and space where you'll meet with users comfortably.
For remote unmoderated user testing, it's important to consider the context.
When you share the link to your maze with users—especially with new or external testers—it's vital to introduce them to what's about to happen. To help them get started, introduce yourself and the company, and explain what and why you're testing.
As they go through your maze, their attention might be disrupted and they might switch to another task. Consider these factors when you test remotely, and provide necessary information for each step of the process to help your users complete your maze.
Use a Context Screen in between missions and questions, or before a new part of your maze to communicate information.
💡 With a Context Screen, you can place contextual messages throughout the test when you need to. Think of it as your chance to keep users engaged during their sessions.
To add a Context Screen to your maze, choose the Context Screen block from the left-side panel in the maze creation dashboard, type a title and a description, and you're good to go!
You've built your maze with questions and missions, you've added context messages where needed, and now you’re almost ready to start testing with users. Before you send your maze live, customize its welcome and thank you screens, and set up maze options.
💡 Customizing Welcome & Thank You screens is available on all paid plans.
Besides the blocks that you created, your maze has two more adjacent blocks by default: the welcome and thank you screens. You can view the default messages on these screens in Preview.
If you've purchased a paid plan, you're able to customize these screens with your own messages and images. This allows you to welcome your users with a custom message and conclude the testing sessions with your own parting words. Here’s how to customize the Welcome Screen and the Thank You Screen.
The options panel includes a few settings for working with your maze test. Here are the available options for your maze:
You will find these options in the top right corner menu under the wheel sign.
💡 Good to know! After you send your maze live, you'll also be able to stop and restart recording from the Options menu.
Hooray! You've built your maze with missions and questions, customized it, and now you're ready to start testing. To begin recording interactions, send your maze live by clicking on the Start Testing button in the top right corner.
If you still need to make changes to your maze, continue editing in draft mode. Use Preview to see how users will experience the test, and only click on Start Testing when you've finished creating your maze. It's not possible to edit a live maze to avoid interfering with results as they come in.
💡 Important: You won’t be able to make any changes to your maze when it's live. Please make sure you’re happy with your maze before you start testing.
Once your maze is live, you can start testing with users. There are a couple of ways to get your maze tested, and in this section, we’ll cover remote or guerilla testing with your own pool of participants.
If you already know who you're going to test with, and have access to a pool of participants from your own users, click on the Copy your link button and share it with your participants.
When users begin testing, your results will start appearing in Maze. Move to section 12 to learn how to analyze your results.
If you want to test with participants in a public space—known as guerilla or hallway testing—you can do so with one simple feature. When you click on Copy your link, you’ll get a modal with options for your maze URL.
Switch the Unique session toggle off to test with different users on the same device. Copy the new link, and open the maze on the device you’ll use for testing. When a user finishes their session, just refresh the page: your maze will be ready to be tested again.
💡Tip! If you’re testing a mobile or tablet prototype, you can use our PWA to offer an immersive experience to your users. This is optional: you can also test via a browser with your maze URL. Learn how to install the Maze PWA on your device.
Optionally, you can also share your maze on social media. To do so, click on Social Networks and choose the platform you want to share your maze on. A new modal will open up and you’ll be able to edit the custom message before sharing your maze on your social profiles.
This option may be useful for personal projects, when you don’t have access to users, or can’t hire external participants. Check out our guide to finding user testers for your maze.
When you share your maze URL with multiple users, you might want to track their sessions by name, group, or even source. Maze uses a simple nomenclature to record metadata with your URL. Here's how that works.
The second URL will record Name: elena in the session info.
You can add as many tags as you want to your URL like so: t.maze.design/XXXXXXXname=elena&source=twitter&gender=female before you share it with users.
The information you record with meta tags will be available for every tester under +more info button.
Note that the tags id, createdAt, and userId aren't available, and won’t record your information.
This feature can help you differentiate users by the information you already know about them, or help you understand where the testers came from when you send your link to multiple batches of users. However, if you don’t know these types of details, simply use a question in your maze to gather this information from users.
In case you can’t test with your own participants, you can hire testers from the built-in Maze panel. You can do this in two simple ways:
To order testers from the Maze panel, choose the Hire Testers option on the live maze page, select the number of participants you need (orders start at five testers), and define criteria (sex and age).
Once you place your order, your results will appear in a few hours depending on the length of your maze, day of the week (longer on the weekend), and the criteria you’ve defined: the more specific the criteria, the longer it might take for results to appear.
Read more about recruiting testers from Maze.
Similarly, you can purchase Maze credits to add to your account and use them across all your projects and mazes whenever you need them. The process of using credits to order testers is fairly straightforward. Access the Maze credits page and choose the credits package that suits your needs.
The credits you buy will be added to your account. When you want to test with users, choose the Hire Testers option, and use the credits in your account to order testers. Learn more about Maze credits.
If you want to compare how two or more variations of your design perform with users, you can conduct an A/B test with your prototype in Maze.
Follow the guidelines below depending on whether or not you have created missions and questions for your maze yet.
💡 For both options below, you don’t need to create two separate prototypes. Make changes to the same prototype after importing your first version to Maze.
After you imported version A of your prototype to Maze, go back to your design tool, and make the changes you need for testing. When you’re done, return to Maze and click the Import new prototype version button. This will create a new draft maze with the changes you made to your prototype.
Now that you have two draft mazes—version A and B—define missions and create questions to send both tests live. When testing is done, analyze your results and determine the winning version.
If you already defined missions for version A of your test, go back to your design tool, and the make the changes you need in the design for testing.
Back in Maze, use the Duplicate feature to create a new draft maze with the changes you made in the prototype while keeping the defined blocks.
Now that you have two draft mazes—version A and B—set them live and start testing with users. When testing is complete, analyze your results to determine the winning variation of your design.
💡 Tip! You can change the name of your maze to differentiate between multiple tests in the same project.
Hooray! You’ve finished testing with at least twenty recommended users. Let’s now look at how to access and read your results from your users.
💡 Your results are available on the live maze page. To get the latest data, simply refresh the page.
There are two high-level ways to read results:
For each mission you tested with users, you get heatmaps at the path, user, and screen level. The paths users take to complete the missions are aggregated based on the mission outcome:
Aggregated results are best for making informed decisions and answering high-level questions about your designs. It can help you understand if users are able to complete tasks with your designs, where they get lost in the flow, and how you can optimize your designs to improve the user experience.
If you included a Multiple Choice, Opinion Scale, or Yes/No question in your maze, you’ll get aggregated results and individual responses from users.
For Open Questions, you can read individual replies and choose to add comments to your maze report. More on the maze report in section 13.
Another way to read your results is by looking at the full session from each user. This way, you can dive into a particular issue a user has experienced, and understand what’s causing the problem. Coupled with question replies, this information can help you improve your design based on the feedback provided by users.
You can find individual session results from beginning to end in the Testers tab.
💡Good to know! You can also download CSV files and export all your maze results.
In addition to your results, you also get an instantly generated report for your maze that you can analyze and share with colleagues and clients. You will find this report by clicking on the Go to report button.
You maze report includes a usability score for every screen, each mission, and your maze. This score is an indication of the usability of your designs based on the average metrics for time on screen, mis-click rate, success rate, etc.
The score is intended as a benchmark you can use to improve upon over time, and it’s not an assessment of your designs. When you share the report with stakeholders, you can display or hide the score as you need.
To do this, click on the Share report button in the top right corner and switch the toggle for Usability Score off. Your no-score link will be copied to dashboard instantly, and you can proceed to share the report with your team, executives, or clients.
Congrats! You're now fully equipped to test your product's design with real users, gather actionable insights and data, and only implement new designs that work.
Testing the designs you create is a memory muscle: the more you test and collect valuable learnings from users, the more prone you're to research and test the next time around. Now you have a full guide to testing with Maze when that time comes.