A novice designer’s guide to usability testing
You cannot make a product or application truly user centered if you do not involve user in the design process. It is important to involve target users in many phases of design. Validation of the solution is one of them. Usability testing is industry wide accepted and standard techique to validate the solution designers has come up with. Many companies and products have been hugely benefited by the usability testing and further changing their design based on the findings and make it more user-centric. A good example is, Macafee decreases support calls by 90% and Mozilla’s support site was able to improve 233% by doing extensive usability testing and making it truly user centric. by doing extensive usability testing and making it truly user-centric. Here I, a fledgeling designer, want to share my learnings from my practice of Usability Testing. Feel free to give feedback as well.
What is Usability Testing
As defined by Interaction Design Foundation
Usability testing is the practice of testing how easy a design is to use on a group of representative users. It usually involves observing users as they attempt to complete tasks and can be done for different types of designs, from user interfaces to physical products. It is often conducted repeatedly, from early development until a product’s release.
Types of Usability Testing
There are many ways to classify usability testing like formal & agile, Guerrilla Usability testing, Summative and Formative, Hi-fi and Lo-fi usability testing.
Here I am going to elaborate the types based on Fidelity of the prototype which is High Fidelity and Low fidelity Usability Testing.
In high fidelity usability testing you use finished and more mature designs, neatly designed detailed prototype and sometimes even actual product. High Fi testing gives you something to test which is close to the actual product and uncovers the loopholes or design mistakes there are before making a huge investment in building the actual product. In case of redesigning of the product or website, it uncovers the issues with the old solution and provides the ground for future improvement.
Low-Fi or Low Fidelity Usability Testing is done on initial design solutions with paper prototypes or clickable wireframes. This helps to get initial feedback on the design at the very early stage of the process hence. It helps to know whether we are going in the right direction with the solution or not.
It is adviced that one should not wait until the design is mature, start doing usability testing in the initial stage itself because it is relatively easy to alter the solution and iterate at the early stage.
Steps of Usability Testing
There are 3 Key steps of Usability Testing
- Planning the Usability Test
- Conduct The Test
- Analysis/Assesment and Test Report
Planning the Usablitiy Test
Before you conduct usability testing you have to plan out that what you want to test, which assumptions you want to check or what you are trying to measure, how many people you want to bring in for the task, where do you want to test like in lab, remote or infield, so on and so forth. So, you have to plan all these things as per your need for the effective testing.
Before we jump to planning in detail let me give the brief info about what do we measure.
What are we measuring?
There are two types of metrics to measure the interface. Qualitative and Quantitative.
Quantitative metric includes time taken on the task, number of clicks, success rate etc. The success rate is the best key metric. This gives the factual data. While Qualitative metric includes their satisfaction and behavioral response.
Usability testing is more of qualitative research and testing but you need quantitative data so you can form the facts take the decision based on that. For example, If you are testing effectiveness of your sign up process and all you gather is some qualitative data like their satisfaction, emotional response and video recording that can be helpful to you as designer but you need data like 8 out of 10 users finished the form or 4 out 10 did not go past the landing page to present in your report in front of stack holders or product owners for them to take decision. Hence it is advised to use both metrics in your testing.
What does a test plan look like?
Test plan includes objective — Assumptions or design patterns that you want to test, Task — Scenario that you give to your user to perform on using your prototype (make sure they are well formed and picture the actual situation), Answer — correct steps to perform that task successfully, Assets — Things that are needed to conduct the test like device to be used in test, camera, eye tracking tools, things to create environment for the test, etc.
On side note, recruiting right and right number of people for the test is important. You have to be careful that people you bring in for the usability test are matching to your persona or have same mindset as your future users because you don’t want to end up with dirty data or inconclusive usability test.
Conduct The Test
While conducting the test you can print out the task part of the test plan and hand over to your user. Duration of the testing can be of any length but make sure your user does not get exhausted.
You may want to create the realistic environment around them for example if you are testing the mobile app, you should add the distractions same as they get while they are actually using the app like getting called in between the use or using the app while walking down the hallway. This kind of realistic environment gives you better data. It is advised to conduct usability in the actual environment of the user on the field if possible so.
While interacting with your recruited user, you must make sure they feel comfortable and friendly also make sure you remind them that this is not their test but is a test of your design solution. Moderating the user is important but make sure you are not guiding or biasing their views. While conducting the usability testing you are a mere observer. Follow the user in their steps while verifying your mental model like, what you assumed that how will user do the task compare to what they actually do. Remember to ask them to think out loud which gives you more insights into their mental model.
Don’t ask the users whether they like something or not because they will say “Yeah, I love it totally !” but might also get confused while using, so instead of asking such questions put them in scenarios and observe.
In Margaret Mead’s words
What people say, what people do, and what they say they do are entirely different things.
This is the crux of the usability testing.
Analysis/Assesment & Test Report
Now once you are done with the testing and gathered all the data it comes the phase from where you analyze the data and draw conclusions. You will need to adopt the usability attitude in this phase. You will have to shift your focus from your design’s goals to user’s goals, their satisfaction, in other words from what you want to what your user want. You will have to be more empathetic towards your user. This assessment is something you use to negotiate the design changes and to guide your team to more of the user-centered design this practice is called User advocacy and it is very important to have in order to design which are truly user-centered.So, you note down all the points say success rate, user satisfaction and user frustration by each task and figure out what that all means and where design changes are necessary. You can also include fixes or idea if you have it is not that necessary.
In the report, you have to prioritize the finding. Include the observed behavior instead of remarks about what do you think. If you are suggesting design change you have to consult development team as well to see its feasible or not and negotiate on common grounds.
Keep is short, dont write long descriptions. Add summary at the start. Include positive findings as well. Acknowledge feelings of your audience, don’t be just negative and point out just flaws, you have to be empathetic to your audience as well. Having concrete quantitative data with a mixture of qualitative data make the report more effective. Include quotes which user said. It is okay if you just measure and point out success metrics it is still very effective.
So, That was it. This is just basic summary of the usability testing and this is just one way of doing it. You can use any technique based on the requirements of your project.
Remember, Usability testing is not something you do for this one special project. Usability testing is inseparable part of truly user centric design.
Thank you.
