How to improve usability testing for your site

In the initial usability test, I met an amiable old aunt who never used a mouse. During the test, she always gestures with her hands, gesturing to the screen in the air, and talking to the cursor. At the end of the test, I certainly didn't get any results, but she insisted that I was a "cute boy" and should know her granddaughter. Soon, I learned how to set the core criteria for the recruitment of experimenters.

If you've done a usability test before, you'll know it's not as easy as it looks. Although this is not rocket science, it also contains some important complexities. In this article, I'll share some of the lessons I've learned about how to help you avoid your usability testers from finally getting into a frustrating outcome during the test.

In the first year of my career there are so many valuable experiences that I could have learned, and I think that learning by mistake is the best way to learn, and we don't always have the chance to fail. Here are some tips that I can take to help you quickly improve usability testing techniques and avoid pitfalls.

Design your usability test script to answer specific research questions

When we start a new usability test, don't hesitate, all you have to do is pick out the main area of the site and have the user complete those tests. You'll find some useful insights in the process, but maybe when you present these findings, you won't be surprised if you can't answer the questions posed by the website's profit-makers.

　　Important Readings

Ask the people you need to report last, the key questions they need most in the survey, and then answer them. If you get a lot of questions at the end of the test, sort them by priority, and then do what you can to solve them. If these questions look too vague, or you're not sure why they're asking, find out why. The more you understand the reasons behind these problems, the better you will be able to answer the questions you are asked by the test and the tester during the mid-term test.

Give participants confidence and make them behave naturally

When participants are doing a test, they usually do not know how to behave. They get a little nervous when a camera is on them or someone is always looking at their shoulders. So don't be surprised at the beginning if they see you seeking guidance. If you are busy directing the whole field at the beginning of the test, you need to strengthen the idea that they need your permission to move.

　　Important Readings

At the start of a series of tests, encourage users to show you their most natural behavior by allowing them to leave and develop any direction they like. I'm going to make a prediction for them, which contains a problem they're going to have in real life, and then I'll let them get out of the way and answer the question as naturally as possible. For example, I would like to test an online property law website, so in the first phase of the test I would ask testers to look for a house that they might buy in a specific budget. This gives us an idea of how they have a realistic perspective when using the network, and also sets the background of the next task in the test.

Leave the room for testers to be free enough to complete the test in their own way

In the early days, I set up a task in the test script, and once the user began to deviate from the task I would control them and ask them to try again. Not just because I have control, but sometimes when I lose contact with the tester, I think I've wasted the opportunity to learn something this time.

Important Readings

Always leave the tester room to allow them to roam freely to the site, or you can deviate a little bit. You may feel that you are out of control, the participants misunderstand the task, but relax a little, because you will be fascinated to see where they go and why. You will often find something valuable in this process, so be bold enough to let it happen. If you need to track it, you can pull it back, but it has to wait until you're sure they can't find their way back on their own.

Relax, keep quiet and see what you can find.

It's easy to get tough with the testers and keep them focused on the tasks you want them to do. When they do something interesting or unexpected, it is useful to ask them what they are thinking at this time. But ask them too often or prematurely, as I mentioned above, that you may not be able to observe their natural manifestations.

　Important Readings

In this process, try not to interrupt the participant's thinking process. The more times you interrupt, the less information they have to accomplish these tasks independently. If you ask them every 30 seconds, they will lose the flow of thought and you will not see any natural manifestation. You can bring their thinking back and ask what happened. I've seen a lot of beginners make such a mistake when doing usability testing, so I remind them that it's impossible to ask questions and watch what the user is doing.

Tailor different tasks to different testers

I think it might be that when you do something new, you want to control the variables and lock in the unknowns, so until now I've been more strict. But in your experience, you have to believe that you can handle anything that happens, and that this confidence should slow you down and weaken your control.

In my early days, I liked to write the task of setting up a scene for the user. But I quickly realized that when my settings didn't match their usual performance, they were less likely to be involved. I remember once, in order to complete a test, I asked a 19-year-old boy to think of himself as the mother of three children. Needless to say, he looked at me strangely and did not really take part in the test, knowing that at last he could not finish the task and give up.

　　Important Readings

You can set up all the test tasks that you want users to do, but it's best to tailor the tasks to different users. While this is not always the case, it is valuable to spend a little time at the beginning of the test to study who the tester is and what kind of products or services they are currently using. If you can use it to build a test environment that suits them to solve a real problem or situation, when someone really pretends to be in this test environment, you can find far more than that.

Always include test tasks on the peer or competitor Web site

For you and your testers, spending hours on a Web page can be very boring. However, boredom is not the only problem. All of your findings are based on an isolated event. Do you really understand whether the person is always going to the search box, or are they just being confused by the navigation options? Just looking at a single page test doesn't give you a real idea of how people use the site.

　　Important Readings

The time to use your site and the time to use the peer site are also included as one of the tests. The best way to do this is to ask the tester what they are currently using and show them to you when you start the test. Then you can introduce a competitor or peer site that they have not used. You can learn more about their behavior patterns, why they choose to use this site instead of using another, and more importantly, you will understand what works on other sites. This can be a great source of inspiration when you need to address a thorny usability issue on your site.

Don't let the tester immediately know which website you are testing

In the past I would have made such a mistake and obviously let them know which website I was testing. Although sometimes it's hard to avoid, if you can, I would advise you not to let them know which website you are testing. The main reason for this is that it can be hard for anyone to be honest with their experience, especially if you are an employee or an agent working for the company.

　　Important Readings

If I didn't have previous experience in designing a usability test for a website, I would always emphasize my independence. Another trick is to get participants to use a competitor's website and give you feedback about their real experience before you actually test your site. If you can do this when they don't know which Web site they're testing, it's much more likely that you'll get their most honest initial ideas. At the end of the test, it's clear that you've spent most of your time on the task of setting up usability tests for your site, but at that point you should have been able to get a good understanding of their intuitive first reaction.

Summarize

If you want to improve your usability testing techniques, there is no better way to increase the number of tests. However, as I have highlighted here, you can look at how you set up tasks and how interacting with testers affects your research results. It's a good starting point to focus the test on the key issues, rather than being too strict. In addition, using a competitor's website as part of the test and encouraging users to behave as naturally as possible can produce better results.