Through the years, the design concepts applied to user interfaces and user experience have evolved to become an integral part of the creative process. Almost every designer has to face the problem of having beautiful ideas but terrible execution. Even if what you think as a designer is the right approach, maybe it is not for your final user(s).
For that reason, testing your work before going to market is something crucial if you want to create appealing products and incredible experiences. So, let’s talk precisely about that: User Testing in the UX world.
According to Wikipedia, user testing is: “A technique used in user-centered interaction design to evaluate a product by testing it on users”. This is true in the interaction process, but the real thing here is that we can test whatever we need to test, not only the product itself: a concept, an architecture of information, the color palette used on the design, even the sound we put on a page intro. You can develop your own types of test according to what you need in your design process.
We can use testing in the early stage of the design process. For example, in the discovery phase, which we use to catch the biggest amount of relevant information for the project, testing could be used in an existing product that we need to optimize and understand what problems with the user experience that digital product could have before we move to other phases. That way, we avoid possible issues and strange behaviors that could harm the overall experience in the future.
In the developing phase, when the product already has some structure, we can use different types of testing, like card sorting (we’ll talk about it later in the post) as a way to organize and prove some content and architecture that will be delivered to the final user. Finally, in the delivery phase of the design process, we can use the testing to prove the final solutions for our digital product in terms of usability.
Any kind of user testing falls into one of two categories: moderated and unmoderated usability testing. Let’s explore its differences:
This kind of testing requires the active participation of a moderator or facilitator. The moderator’s mission is to take action on administering tasks, guiding the participants, recording behaviors and so on, all in real-time. This testing could be done remotely or in-person. In the first case, the participants need to share their screens with the moderator to be able to record the actions performed by them.
Moderate testing is the majority used in the early stages of the design process and has an impact on the next steps because it throws a kind of controlled metric that indicates how the final user could see the final digital product.
Unlike the previous one, unmoderated testing doesn’t require a moderator or facilitator to perform this task. The participants are not guided in this one, so this is recommended only to test specific features or minor elements of your product. Unmoderated testing is an aid to UX projects that have a tight budget and in which the time is a concern.
Let’s see now the types of testing that we can perform in a design project:
Card sorting is quite simple: it is just a pile or group of sort cards that help to enhance the Information Architecture (IA) of a digital product. To structure the IA better, your user finds your site/application more intuitive, less confusing and they’ll enjoy visiting it.
Card sorting uses the mental model of your participants to structure the content in the most obvious ways and make it easily accessible. Card sorting is a quick, simple and cost-effective method of user testing that’s also quite reliable, depending on the number of users you perform it with.
Put it simply, the card sorting is a way in which your participants understand the information in front of them and how they organize and evaluate that information.
Similar to its approach to the previous card sorting technique, tree testing is a way in which your participants test an existing navigation concept for a particular website/app. The user is let only with the most basic form of navigation, clutter the unnecessary elements and eliminate the necessity for layout and design. Here the most important is the way in which the participants measure how easy it is to navigate through your site.
To measure the effectiveness of the tree testing technique, you need to analyze the results and the data which were submitted by the participants, for example, the percentage of users that performed the task without hesitation, the number of participants that completed it and the time it took to them.
The tree testing could be done with online tools like Treejack, an application that performs testing on this kind without a hassle.
This technique is the most popular type of test in the UX field. Why? Because it is simple and well structured. In basic terms, you make two versions of the same page and show that product to different groups of users. After measuring the results, the page with more conversion rate is the winner.
The conversion rate depends on the product that is going to be created; could be different in certain types of digital products, for example, the ones that need people to buy items from it or the other one, in which its conversion rate is measured by the number of users that subscribed to the newsletter.
But before you start thinking about how to perform your A/B testing, you need to research your product and gather the data you think you will need: buttons that are hiding, call to actions (CTAs) with no proper results, pages that make your users feel disconcert, etc. You need to work on something that must be enhanced.
Finally, when you have enough information, you can start your AB testing and try to apply the solutions you found in the analysis of that data.
Surprisingly like the A/B testing, Multivariate testing (Mvt) is thought to test many things at the same time. It is used to create a combination of elements and to test which one could deliver better results. As a result of testing different components at once, the data that the test throws could be massive and difficult to read.
Supposed that you have a small website, where traffic is very low. Mvt could not be the best option to choose, due to the difficulty of the process and the analysis. Instead, A/B testing offers you a better approach to your actual needs.
The importance of Testing in UX design is that with it you can search for possible issues in your digital product even in the early stages. Also, performing these kinds of tests, you ensure your product reaches the market with the quality that your users deserve.
Maybe you’re wondering: what kind of testing is better for my project? Well, there’s no correct answer for that, but it seems that some tests perform better for some projects than others. For example, card sorting and tree testing could be used if your goal is to analyze the organization of the information (navigation). Multivariate testing is for a more extensive project and if you are going to decide on alternative design options.
So you have it. A little introduction into the user testing on UX that could leverage your next digital product to the top.
Design always must be center on the user and for the user, so if you want to know a little more about it, check my post of User Interface for Humans.