User testing is a fundamental core of the UX design process for a great and pleasant user experience. User testing provides critical insights in guiding and validating the design process, and with most things in life, conducting user testing takes time and patience.
Which is why we have compiled a list of top 10 tips to follow to help you make the most of your user testing sessions:
1. Test Early, Test Often
The earlier you start testing, the easier it will be to make changes. Instead of testing with a large group of people all at once, consider testing frequently with smaller groups. This will provide sufficient insights into improving the design to test again for a quick iteration.
You may be thinking “the product isn’t done yet, we should wait to test”, but there is a danger of having to make significant changes, especially after working on a design for so long without any feedback. It’s a classic mistake in thinking that we know what the user wants. But there is no way of knowing that with certainty unless you really test with your users.
By investing energy and time to test early, it can help to prevent problems from snowballing further down the line, saving a tremendous amount of time that would otherwise be wasted. Plus, there is no need for a high-fidelity prototype or completed product in order to start testing, design mockups or low-fidelity prototypes would be enough. There are many tools available to convert the design into a clickable prototype (e.g. Adobe XD, Axure, Invision etc)
With low-fidelity prototypes and mockups, all you really need to do is set the context for the user test and explain clearly what is required of them.
2. Outline Your Goals
Before going into user testing, it’s important to be clear on your goals. Going into user testing without outlining your goal will lead to an aimless excercise, so think about why you want to test the product. What are you trying to learn? What do you need to know from the research?
Once you identify your goals, identify which specific features and areas you want to focus on for feedback. Here are a few common objectives to get you started:
- Explore whether users are able to complete specific tasks successfully, e.g. purchase a product.
- Find out how long it takes to complete specific tasks and what the completion rate is.
- Identify users’ pain points, thoughts and opinions, whether they are satisfied with the product and identify changes required for improvement.
3. Prepare Questions and Tasks Carefully
Once you have a goal, define which tasks to test in order to answer the questions you have outlined or to validate your assumptions. Bear in mind, the objective shouldn’t be to test the functionality itself, but to test the user experience of that functionality.
Create realistic and actionable tasks, these could be specific parts of the product you want users to test, such as completing a checkout, or getting started with a product.
Testing and analysis will take a lot of time, list down which tasks to focus on and order them by priority. You can first test with five users, come back to the drawing board, and test again with another five users instead of testing with 10 users at once.
Describe Tasks Clearly
Users need to know what to do, this doesn’t mean giving them step-by-step instructions, but you have to be clear on what the task is. Users will tend to be discouraged when tasks are unclear. For example, if you would like to find out whether users are able to go through the entire checkout journey, instead of asking vaguely “Select any item to buy.”, the task can be clearer: “You are looking for a new pair of shoes to buy, could you show me how you would do this?”
Set A Goal For Each Task
As a moderator, the goal of a task should be clear, for example: the user should be able to complete the checkout within three minutes.
This information should not be shared with the user, but it is useful for the moderator to know how long each task should approximately take in order to keep the session moving.
Limit the Number of Tasks
Considering the time limit of each session, which is usually 60 minutes, be mindful not to overload too many tasks, and remember you will also need time for follow-up questions.
Conduct A Pilot Test
This is a great way to iron out any glitches, revise any instructions, tasks or questions that need to be fine-tuned to avoid any potential misunderstandings.
If the pilot test takes more than 60 minutes, it is an important red flag to cut down on the number of tasks.
Create A Scenario
People tend to perform more naturally when a scenario is provided to them as opposed to instructions. So instead of asking “Download an ebook with recipes” you could create a scenario “You are looking for dishes to make at home, how would you download an ebook for this?”
A scenario provides more context for the user and so feels more natural when the user performs tasks, deriving better insights as a result.
4. Recruit Wisely
Asking the right questions is important, and recruiting the right people is as important. It’s a waste of time and effort to test with people who are not your target audience, their feedback won’t be representative of your actual users.
Recruit people based on your goals and target users criteria, it is helpful to write down the demographic criteria. In addition, you can make use of respondent recruitment services that can help to find users that represent your target audience.
Define The Recruitment Criteria
If you are testing for a food delivery service, it would be better to recruit those who order food regularly as opposed to a person who cooks most days.
Create a precise and measurable criteria, this will be helpful when it comes to screening prospective users for testing; for example: people who order food at least twice a week and have used at least two different delivery services in the past six months.
Don’t Recruit Those Who Are Friends and Family
Feedback from friends and family is better than nothing but for unbiased results, you need independent and unbiased users who have no relations to you in order to obtain realistic results.
Analyze Existing Data
If you already have a customer base, you can then do a quick analysis of the available information, such as data analytics, surveys, previous usability tests, customer support tickets, and customer feedback. As a starting point, this will help to assess what you already know and don’t know about your users.
5. Getting The Most Out Of In-Person Testing
The best way to learn about your product and how to improve is to hear it directly from your users, and by observing how they use your product, you can identify problem areas. To do that you need to:
Build A Good Rapport
It’s normal for users to feel nervous and unsure when they come into a user testing session, but the quality of the session is related to the rapport with the users. If the user feels at ease and that they can trust the moderator, they will likely provide more open and honest feedback. Conduct the sessions in the best way you can to help users feel comfortable in giving you feedback:
- People tend to blame themselves, instead of the flaw in the design so it’s important to emphasize to the user that “we’re not testing you, we are testing the design. There are no wrong answers or feedback.”
- Start easy, this will get the user talking and to feel relaxed, opening up as the session goes on.
Listen, Do Not Lead
It is so important not to ask any leading questions or to guide them, the moderator is there to facilitate the session, not to give the users the answers to their tasks. The goal is to understand how users will use the product, so if the users go down a ‘wrong’ route, hold back on correcting them. Waiting to see what happens can provide valuable insights.
The moderator should not emotionally affect the user, and most definitely not judge the user. The goal is to gain understanding from the users’ perspectives and to gain as much information as possible in the allocated time.
Avoid saying “That was quite obvious, wasn’t it?” and avoid negative body language like raising eyebrows. Instead ask in a neutral tone “How easy or difficult would you say it was to complete this task?” and “Why do you think that?”.
Don’t Explain, Don’t Interrupt
If the moderator explains the functionality, it defeats the purpose of the study and introduces bias to the session. In the real world, there is no explanation or guidance from a third-party and so it should be the same for the user testing session. Users should figure things out independently.
Moreover, try your best not to interrupt them. The more interruptions, the less confidence the user will have in completing their task. Be patient, and let the users find their flow to capture their natural behaviour.
The think-aloud technique is critical in getting inside the user’s head, the moderator should ask the user to continuously think out loud as they use the product – to verbalize their thoughts.
This technique helps you discover what users really think about in their thought process, turning the findings into actionable design recommendations, such as “this is loading too slowly” or “I expected to see something different”.
Users may need a reminder now and then to keep the think aloud going, the moderator can ask “what are your thoughts right now?” to encourage users to keep talking.
People want to appear smart, which is why when they struggle through a task, they will still say it’s easy. It’s important to study their body language, seeing how they react. Listening is vital, so is observing. It can uncover a lot of information that goes unsaid.
Clarify When In Doubt
If you’re not sure what the user is talking about, it’s okay to ask for clarification. Simply ask “when you said …, what did you mean?” to make things clear.
Don’t leave this to the end of the session either, it will be too late to go back and figure out what they are talking about.
Follow Up Questions
The moderator should have an eager and curious attitude, and to dig deeper by asking follow-up questions. It will give a lot of insights into what really happened, as people don’t often state their motivations clearly without being prompted. A well-timed follow-up question will help to uncover more thorough and valuable information.
Answer Their Questions With Questions
Users will usually ask questions, such as “Should I use this?” or “Is this the right icon to click?”
It’s tempting to give them the right answer but asking them politely “What do you think?” can keep the conversation going and keep the user engaged.
6. Design As An Iterative Process
A lot of product teams think that the design process is linear, starting with user research and ending with testing. This mindset can hold back the team because there is so much pressure to getting it right the first time.
Instead, by thinking of it as an iterative process, it removes the unnecessary pressure. There should be an understanding that what isn’t being tested now, can be tested later. Regular user feedback is crucial for continuous improvement, and should be at the heart of the UX design process. It is common for usability testing to be run every 2-4 weeks to support continuous design improvement.
7. Consider Remote User Testing
There is a huge benefit when it comes to testing in-person to derive more qualitative data such as body language. But in-person testing is more time-consuming and expensive compared to remote user testing.
If there is a need to test only a small feature, or if the target audience is geographically dispersed, or if there’s a very strict timeline, remote user testing is a great alternative.
To find out more about remote usability testing, along with information for tools and platforms, check out this article: Best Practices for Remote Usability Testing
8. Include The Whole Team
By involving the entire team in the testing process, it gives them the opportunity to observe, understand and empathize with the users. The whole team can then be the same page by building a shared understanding.
Direct involvement in preparing the test will help draw interest from team members and by communicating the test strategy, it helps them understand how the findings will aid their work.
Encourage team members to observe the session, they don’t need to observe all the sessions but try to get them involved in at least one of the sessions. This helps them to understand first-hand how user testing can provide valuable insights.
9. Test Before, During, And After Redesign
“When should we test?” is a common question, and the answer is quite simple: test before, during and after.
- Before: If you’re planning to redesign an existing product, testing the current product can help to identify existing pain points to guide the design process moving forwards. Consider testing competitors’ products to compare results and know what the users like and dislike.
- During: In the time it takes to build and launch a new product/feature, several small testing sessions can be conducted to fine-tune and improve the prototype.
- After: The knowledge and feedback from real users will help to improve the product even more.
10. Don’t Try to Solve Everything At Once
It is impossible to solve everything at once. Instead, prioritize your findings and fix the most important problems first. Test again to see if the problems have been fixed properly.
Investment in constant user testing is about the only way to generate a consistent rich stream of data on user behaviour, even a simple round of testing could make or break your product.