Design, research, and improve: usability testing as the path to an ideal portal
User research is an indispensable element of UX design (i.e. designing to ensure a positive user experience with a website, app, or other user-focused interface). It should be included in the design process at an early stage so that the actual needs, expectations, and problems faced by users are clear.
Nobody can tell us more about these issues than the users themselves. Without this knowledge, it is very difficult to design a good product or verify its utility. Frequently talking with users throughout the process allows the design team to detect potential problems early on.
At e-point, we treat all UX testing as an integral part of the design process. Additionally, our team supports research aimed at verifying the usability of existing solutions.
Why does your company need usability testing?
User research itself is no guarantee of a project’s success, but it can certainly increase your chances. This is well understood by big names like Amazon and Google, who regularly perform this research to see if they’re really understanding and meeting users’ needs.
Usability testing is a basic source of information about how easily users can find information or complete some task. It also offers an opportunity to see how they consume online content and navigate through an app or website.
Usability testing saves time and money
One of the advantages of user research is that it offers an opportunity to significantly reduce costs. Correcting issues detected at an early stage of the design process is usually faster and cheaper than modifying an almost finished solution. Usability tests prevent investment in faulty conceptions.
According to author Robert S. Pressman, each dollar spent on solving an issue detected at the design stage saves between ten dollars (to fix the same error during development) and 100+ dollars (to fix the problem after implementation).
When teams are tasked with creating interactive solutions, the benefits of catching problems early are strongly felt. Making changes to a mock-up or prototype is usually a matter of a few to a dozen hours. When the project swells with more people, more teams, or additional technologies and licenses, it’s a different story. Even the slightest change in such a puzzle takes a lot of time and generates substantial costs.
Usability testing is also a chance to eliminate factors that cause customer churn or customer journey interruption. A well-designed, easy-to-use portal leads to fewer queries by customers who are unable to find relevant information on their own. It also enhances the efficiency of the customer service team and can even lessen the demand for new customer service staff.
What is usability testing?
It seems reasonable to start with the meaning of usability itself. Jakob Nielsen states that “usability” expresses how easy it is to operate a user interface. If it is difficult or confusing, people will simply stop using it.
On the other hand, the ISO 9241-210 norm defines this concept in a slightly more precise manner. According to them, usability is “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”.
That is why running tests that allow us to spot potential interface issues are so important. In this case, employing qualitative research – through which we can get to know how average users navigate the website while performing predefined tasks – is highly recommended.
When and how to start usability testing?
Usability tests are best started at an early stage of the design process and should be repeated after each major change. These tests require us to have an existing product, a prototype, or at least a mock-up demonstrating the concept being developed.
Naturally, if we do not know the target audience of something we’re designing from scratch, user research should be started at the very outset of the project. It will then be used to identify the target group and establish its needs.
If we already have a solution that we want to verify, our first step should be to determine research goals and areas. What we want to learn from the research and what elements or processes of our solution will be tested should be defined from the start of any research. Only when we’re clear on these points should we decide on the actual research methods.
Most of the research at e-point is run in the form of task tests with elements of in-depth interviews added in. Depending on the project, some other methods may be applied, such as:
- Card sorting
- Tree testing
- 5-second tests
Based on our defined goals and selected research methods, a research scenario is created. This is simply a detailed description of the research procedure and a list of tasks that will be given to research participants.
The next step is setting the size of the research sample group, the requirements for participants, and the recruitment methods we’ll use. The right selection of participants is key in usability testing, as it can affect the quality of findings.
Qualitative research is conducted on small samples. Sometimes a few participants are all that is needed to complete a usability test. As Jakob Nielsen and Tom Landauer proved, testing a portal on five users allows designers to detect 85% of functionality problems; the participation of a larger group does not guarantee a linear rise in the number of problems identified.
The number of participants will depend on how diverse (in terms of the goals pursued) the target group of the designed solution is. For each segment of users, there should be several segment representatives recruited – this is extremely important. Then, participant recruitment criteria should be established. These criteria serve as guidelines for team members tasked with recruiting participants or for an outside company that has been commissioned to recruit participants.
So, if we aim to study a banking application, the fundamental requirement for participants is that they are users of the relevant banking products and services. They should also have some experience in operating the device used in the test, e.g. an Android smartphone.
Apart from that, demographic characteristics such as gender or age are taken into account during the recruitment process. It may seem obvious, but persons who regularly develop similar solutions or whose professional activity relates to launching new products (such as people who work in marketing or advertising) should not participate in this kind of research.
The next step in research preparation is choosing a location. Most of the research run at e-point is conducted in the lab or another controlled environment; less-formal research is conducted in other areas of our company's premises. The preparation stage ends in running a pilot test, which serves to verify the scenario and to detect possible prototype problems (if the research subject matter is a clickable prototype).
Usability testing procedure
Apart from the participants, there are always two people participating in the survey – a moderator who runs the meeting and an "observer" who monitors and takes notes. On top of that, the test is conducted in a research lab equipped with a one-way mirror or online streaming capabilities, which means the test can be watched by other project team members. These team members observe how the participant navigates the interface and how they react to what they see. Facial expressions and other reactions are valuable sources of knowledge on how users perceive particular products.
During the test, a participant is presented with a computer or another device (e.g. a smartphone or tablet) and is asked to perform tasks on it, as instructed by the moderator. These tasks are determined according to the previously-created scenario. While performing tasks, the participant, if necessary, may be asked to answer additional questions. A very popular method for running a user test is the “think-aloud protocol”: participants are encouraged to continuously verbalize their thoughts and comment on the actions they perform.
Having completed the task part, the participant is requested to provide a general assessment of their interaction with the interface. They will be asked to answer the moderator's questions or complete a post-survey questionnaire. This is the declarative part of the research, which complements the knowledge gained from other research techniques; however, during the findings analysis, it is weighted significantly less than other techniques.
For the purpose of further analysis, the whole test (i.e. the audio and video recording of the participant as well as the interaction path) is recorded by dedicated software. After the test is completed, there will be a recording generated in a video format for each research session.
Analyzing and presenting findings
Research analysis starts with replaying the recorded sessions and writing down or supplementing observations and conclusions. During analysis, the research team prepares a list of all interface problems in the solution studied; these are then ranked in terms of their weight, from critical problems (which either disabled or seriously affected the completion of the task) down to negligible quibbles. We can often formulate interesting findings or tips for further product development from the information gleaned in this type of analysis.
Based on the data processed, a report is prepared that includes recommendations for further actions to be taken by the development team. The complete document is usually presented to the client during a workshop session. This allows the recommendations proposed by the research team to be used as a starting point for joint decisions on the scope and form of the changes required.
The advantages of Including UX testing in the development process
Thanks to usability testing, many businesses have learned how future users will perceive their services or solutions. And they gain this knowledge at an early stage of the development process. This has spurred the creation or redesign of various concepts, making them better suited to the needs and expectations of the target group. It’s also prevented costly mistakes and miscalculations, saving a significant amount of money.