I am working on the full circle of Usability Research, and you can also call it User Experience research, of the websites and mobile apps.
Currently, at least twice a year, in April and October, I prepare and conduct the usability tests with the visitors of the Global Sources Trade Shows in Hong Kong. But I started conducting the remote online usability tests in 2016. I would say, sometimes, testing online is easier and more convenient. However, if you want the users to focus on your study, you’d better prepare for the laboratory test (but make sure the respondents will feel comfortable, give them snacks and drinks, allow charging their phone, sit comfortably).
The usability test can be conducted at any stage of your developing process, but the earlier you will start – the better.
Think about usability research in this way. If there is nothing extremely unique, what else the users will have to rely on? Aesthetic, usefulness, emotions, trust to your brand and your product and engagement. And proper usability research will help you to make improvements in all these areas.
As the ISO describes usability as “the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency, and satisfaction in a specified context of use.” this means that you can measure your product’s usability. This means you can know the ROI of your research!
What normally as a usability researcher will I do?
- Staring at your interface for a while, trying to identify any possible issue your users might have. As well as identifying the users. If you don’t know your users, we can go through the Customer Research first.
- Grouping all possible issues that I found and rank them. Ranking should be confirmed with the client.
- Identifying a goal and objectives of the usability study. This is a good way to see whether the usability study is actually what you were looking for.
- Identifying hypotheses for the study.
- Look for the respondents.
- Contact them and invite to the test.
- Conduct the study 1:1. With at least 5 users.
- Collect usability metrics (UMUX, SUS, Reaction Cards, etc).
- Prepare the report with the results of the study, whether the goals were reached, whether the hypotheses were confirmed, what kind of issues were discovered and what kind of ideas the users suggested. As well as the design solutions for the issues.
- But we can also try something abnormal!
What can be tested?
- Sketch
- Paper mockups
- Wireframes
- MVP
- Finished product
- Readability of your texts
- Accessibility
- Anything you would consider giving your clients joy!
How many respondents are required?
From 5 and it depends on the scenarios and areas you want to test. For example, if you are testing a website with the registration, you may test the registration with 5 users and login with the other 5. But remember, the usability research is the qualitative research which means that you don’t need to test 100 people, but test a few that represent your targeted group the most. This is also why the usability research can be conducted quickly and with a low budget.
How does the usability test process look like?
- Identify how many respondents do you need. Ideally, 5-7 respondents may be more than enough.
- Identify who are your respondents, how old are they, any specific group you want to target.
- If you require the test with kids, make sure the test is ethical for the age group you are targeting.
- Start to look for those people. Either online, offline, or both.
- You can consider inviting respondents via EDMs clarifying what will happen and when.
- Prepare the incentives. For a 30-minutes test, around 30 US Dollars should be enough.
- Prepare the introduction. I usually use this text:
Hello and thank you for your time. My name is Marina, I am a usability researcher. Today we are planning to test the product ABC, it should take around XYZ time. During the test I will ask you to do some simple tasks using the product ABC, it should be easy to do it however if you feel that you don’t want to continue, let me know. I would like to emphasize that we are not testing you but our product, so all your actions will be considered proper by us.
During the test, please comment on your action, and express your opinion. Try to do a little narrative where possible. This will be helpful.
I also have a camera, so I would like to record you, your voice and the screen of your device, but these records will only be used inside our company and only for the test purpose. All your details will be depersonalized and will not be published anywhere. Do you mind?
Okay, before we start, do you have any questions?
Proceed with your tasks.
Make sure your questions are not direct. You should not ask a user to click the button, you should guide a user there.
A good task should be phrased like this:
As you are interested in fashion accessories, I would like you to show me what would you do to buy the earrings for coming birthday of your best friend. You have a budget of XXX USD and may consider gift wrapping and/or engravings. Please proceed.
As well as you should ask before the task will be completed what a respondent would expect to happen, and after the task – whether what happened was what a respondent expected. And if not – what should we improve to make it more useful to them.
During your test, you may invite observers to observe the test silently and to make notes. The best idea is to ask your team to follow on the test and place sticky notes to the important areas, highlighting positive comments, negative comments and issues, bugs, or any future ideas a user will mention. I normally use Miro for this, for example:
After the test session with a respondent, I will ask them to fill in some additional questionnaires, such as System Usability Scale (SUS) and Usability Metric for User Experience (UMUX). These metrics help to clarify the issues and address them to your client.
What should happen next?
After the test results are combined and shared, it is your turn to make a decision. You can update all the issues that were discovered, and test further with other groups of users.
You test the same interface with the same group of users in a few months, 3-5. This way you can check whether the interface is dummy-proof.
If you want to learn more, SAY HI