Why It's Needed
People are very opinionated creatures. If you are a UI & UX expert, or a designer/developer, it is so easy to get caught in the cycle of making ultimate decisions based solely on past knowledge and experiences. The problem is, every scenario is different so past knowledge and experience won't always give you the correct answer. All it can do is help guide you into the right direction. This is where user testing comes into play. A construction company for example is not going to have the same "answers" as a web design firm. A company who is trying to appeal to consumers is going to be completely different from those focusing on B2B. Just because something works for one client doesn't mean it will work for another client. There are simply too many different factors involved.
Types of Usability Testing
There are so many different tests available. Some are fairly expensive while others are very affordable or even free for basic parameters. A few types of tests are:
Below are real world examples of these tests in action that I used to improve my company's website - www.cyberbytesinc.com
User Video Tests
In this test, a user records themselves using your site for anywhere from 5-20 minutes, and discusses their experience. They’re then given a variety of simple and difficult tasks to perform, allowing you to see where the problematic areas of your site lie. As expected, a test that requires such a high level of user testing is expensive and time-consuming, but it does provide a wealth of knowledge for you to analyze and learn from.
A great site we use is Peek by UserTesting. They give you three short user video tests per month per email address, for free! The only problem with this is that you cannot choose the demographic or tell your users what specific tasks you want them to complete. It is still very useful to get an insight into the minds of people who will be using the site and finding the problematic areas. On the other hand, if you have the resources available, their full-service allows you to actually customize the tests, choosing a demographic that matches your target market while also selecting the specific tasks that you want completed.
Five Seconds Tests
5 Second Tests are another test commonly used. It consists of showing the user a site design, logo, or graphic for 5 seconds, then asking them questions about what they just saw. Common questions, such as:
- What do you think this page was about?
- What services does this company offer?
- Does this brand appear trustworthy?
This test format allows you to identify if what you made properly gets across your intended message.
Preference tests allow you to compare two interfaces to see which one performs better. This can be used to compare UI changes in a website to see which layout is preferred. These tests also work well if you’re trying to compare subtle graphic/logo changes. When dealing with a 50-50 test like a preference test, statistical significance plays an important role when trying to determine which alternative truly performs best.
For those unfamiliar, statistical significance can be defined as:
A result that is not likely to occur randomly, but rather is likely to be attributable to a specific cause. Statistical significance can be strong or weak, and is important to research in many math and science-related fields, including medicine, sociology, psychology and biology. Statistical significance does not always indicate practical significance. In addition, it can be misinterpreted when researchers do not use language carefully in reporting their results. Source
I have taken my fair share of stats courses throughout college to know that statistically significant sample sizes will always vary. There are equations that can be ran to determine how many tests will need to be run in order to create statistically significant data though which depend on many factors. It is although pretty safe to assume that on most things, a sample size of 30 will give you fairly high statistically significant results. Depending on how important the element/layout is that you are testing, you might want to run more tests to be safe. I usually run with a sample size of between 30-50 before considering the results to be conclusive.
In the below example, I ran 34 tests and the program actually ran the proper tests to figure out the statistical significance, placing it at 99.99%. But also keep in mind that there are many exceptions to this rule. If you are testing something in the real world, a new medical drug for example; a sample size of only 30 people out of a population of 7+ billion (assuming it is a drug that can help/affect everyone in the entire world) then a sample size of 30 is in no way near 99.99% statistically significant. It would be more like a fraction of a percent if anything. For the most part though, you will be pretty safe running between 30-50 tests so long as the tester demographic matches similarly to the target demographic of your site.
Click tests allow you to find out how your users interact with your interfaces. For example, this is the test you’d use if you want to see if your main call to action is working properly. There are many other ways to use this test though. Unlike preference tests which only allow you to compare two images, you can use click tests to compare multiple images or pieces of content at one time, as shown by the below example. This works exceptionally well when comparing 3 or 4 logo concepts.
Nav Flow Tests
Nav flow tests help you find out how users navigate around your websites and applications. You can use these tests by asking the user to find a certain piece of information or to complete an action and see how they funnel through the app or site. This allows you to figure out where users are having difficulties so you can fix the problem areas.
If you are the designer/developer of a site, graphic, or logo, it is often difficult to stand back and take honest criticism. Especially when users cannot achieve what you may see as a simple task. It's very easy to start getting defensive and blaming the problems on the user instead of the site. The thing is, you are not necessarily the user-base that the site or graphic is being made for. The only way to improve upon what you're making is to take the problem areas and improve them until your design fits the actual target markets’ needs.
Do you usability test? If so, what tests have you found the most success with? Let me know in the comments below.
- UsabilityHub - Five Second Tests, Click Tests, Preference Tests, and Nav Flow Tests
- Peek by UserTesting - User Video Tests
- Desinion - Preference Tests
*The only resource listed that has a referral code is UsabilityHub as it credits both of our accounts with 10 free credits to use on testing
If you would like to evolve your business with usability testing and other online marketing solutions, please contact my company, CyberBytes Inc. at (716) 876-1824 and ask for Alex or visit our website to get a free quote!