12 helpful UX methods and when to use each

UX methods are tools to guide user-centered design. The right tool can get you the same results faster or even better results. Most of the time, UX Designers need to solve real-world problems with definitive recourses and deadlines. Ultimately, you want to get to the solution as soon as possible. In which case, picking the right tool matters.

This article is a list of 12 typical methods that I found helpful in my work and when to use each. Hopefully, they can be helpful to you too.

*feel free to learn more methods extensively from usability.gov or Nielson Normal Group.

BEFORE WE DIVE IN

The British Design Council coined the Double Diamond process in 2005. This is the typical process I follow for design, and this article is dividing the methods up into the 4 stages defined in the Double Diamond. If you are familiar with IDEO’s design thinking, it is somewhat similar. Feel free to learn more about these concepts from Wikipedia and IDEO.

Fig 1: Double Diamond Process, source:wikipedia.

Fig 2: Divergent and convergent thinking diagram from IDEO.

EARLY RESEARCH

If you are in the early stage of a project, esp. for an industry that you do not have much experience in, generative researches will help. Some methods to use are:

• User Interviews: this is when you talk to users following a carefully drafted interview script. This allows a lot of open-ended questions and could be more conversational. You have room to build rapport with the user while listening to their stories. Very useful for initial researches.
• User Observation/Contextual Inquiry/Ethnographical Interview: these are the ones that you need to visit the user in their actual context of use. e.g. an operation room if the design is something physicians will use during operations (realistically, there might be HIPPA restrictions but this is just an example). If you have time to travel to the site and permissions to observe on-site, this will be good for projects where the context of use influences user behavior heavily.
• Qualitative Survey: this is where you design a structured questionnaire and send it out to users. In a general sense, surveys tend to be more quantitative, but here I am talking about using it as a qualitative tool, where the questions are carefully written, concise, and targeted (i.e. gather quality text answers rather than stuff like Net Promoter Score). A qualitative Survey is good when you don’t have time to sit down with the user, or the research question is a follow-up to previous researches that don’t necessarily worth another entire afternoon of meetings. Ideally, you would need to know the subject matter at least a little bit AND have some trust built with the users before pulling this off because it’s easier for users to lie (intentionally or unintentionally) in surveys. In summary, this method works better for small follow-ups and/or in-house projects, and please use it with caution.

ANALYSIS

Raw research data has limited meaning to design until they are synthesized into insights. This stage is when the designer turns the convergent second quarter in the Double Diamond. Here are some methods for surfacing designable insights:

• Inductive Coding: where the researcher/designer goes through raw notes and tagging similar information. e.g. if two users talked about how slow the landing page is, you can tag both with “landing page performance issue.” The goal is to eventually group similar tags and quotes to reveal user patterns.
• Google Rainbow Spreadsheet: a spreadsheet where insights are listed on the left and participants are tallied against each insight. This method can be applied after the research meeting or during. However, doing this while talking to the user requires a lot more cognitive energy, esp. if you are a one-man team. So use it based on your specific situation. The benefit is since more “work” was done during the meeting, the turnaround time from research to design is much faster.
• Affinity Diagramming: this is when highlights in notes are turned into stickies, and researchers form clusters to see trends. This method is the most collaboration-friendly. You can bring in teammates, group stickies together, and address questions all in one go. What’s sweeter is, in the end, there will be a sharable visual artifact — the affinity diagram — for stakeholders.

These three above are pretty similar to each other. Sometimes I use either Inductive Coding or Rainbow Spreadsheet with Affinity Diagramming to make insights grouping easier.

The Rainbow Spreadsheet is much faster. However, the cognitive load is real. Additionally, it requires the researcher to be a lot more self-conscious about his/her biases. To counter, your team can document individually, then combine results. This counteraction is also applicable for inductive coding.

If you are ok with processing information during research, I recommend the Rainbow Spreadsheet. If you tend to analyze afterward, then there’s not much of a difference among these, so feel free to pick whatever works.

Fig 3: an example for affinity diagram. Source: UserTesting.com.

GENERATE SOLUTIONS

Once insights are revealed, it’s time to generate potential solutions. This is the 3rd quarter on the Double Diamond. And here are some methods to help with ideation:

• Whiteboarding: sketch out ideas quickly on a whiteboard with your teammates and ask for feedback. This allows quick solution generating, and usually is very low cost.
• Pen and Paper Sketching: sketching out ideas on a piece of paper. Again, fast and low cost. I tend to use this when I am going through potential solutions myself, it helps clear up my own thoughts before even going into a whiteboarding session with others.
• Rapid prototyping/Wireframing: quickly put together a prototype or wireframe to validate or invalidate design ideas. If the project has an existing design system, rapid prototyping would work nicely, esp. for a team that’s used to seeing the visual detail. If prototyping would take much more effort and your audience won't care about the fidelity, wireframing is better. Leverage existing low-fidelity tools like Balsamiq or Figma kits. It will be more of a greyscale design, but rough spacing and quick labels should be sufficient to communicate ideas and allow teammate feedback.

Worth noticing is that you can use these methods whenever in the design process when you need to generate ideas. e.g. if you are in the middle of a project but got stuck on how to design a certain interaction, you can run a whiteboarding section with the team. Design is easier when informed teammates chime in. More brains are usually (USUALLY) better than one.

Fig 4: example whiteboarding a sitemap. Source: UI Stencils Blog.

EVALUATION

Finishing the design does not mean it’s successful, and that’s why we need evaluation methods. I consider this as an extended 4th quarter in the Double Diamond. Here are some helpful methods:

• Usability Testing: during which you observe the user performing key tasks on an interactive prototype. This is usually more qualitative. Depends on how many resources there are, you can either do moderated testing or unmoderated (e.g. UserTesting.com). However, unmoderated testing usually entails more risk (similar to a survey, you can’t be there with the user, it could be hard to ask follow-up questions or figure out WHY).
• or Usability Feedback Interview: instead of asking users to click through the prototype, the researcher/designer walks it through like a presentation and ask for feedback (more of a stakeholder review format). This works better if you have limited resources to design high-fidelity interactive prototypes, which could happen due to large workload, project juggling, or simply an inherently complex project. The tricky part is that you will have to give a little more emphasis on your openness for anything that does NOT work because it’s easier for the user to think the design is done and just be reserved on their negative feedback.
• Focus Group: where a group of participants is brought in and openly discuss the research topic. These things can be VERY TRICKY. Some professionals will even recommend not using this method at all. This is where having a script and level-setting participants at the beginning of the meeting is VERY IMPORTANT. Professionally inform them to be aware of their thoughts and don’t just agree with others to agree. Use Focus Group cautiously and stay alert during the meeting to make sure it is not turning into a one-man show. If you have users who are the same persona, bringing them into one meeting rather than one-on-ones will save a lot of time.
• Web Analytics: with which defined interactions are tracked and corresponding metrics are presented on a data dashboard. These are good and fairly low cost to set up for quantitative metrics, e.g. number of clicks on the landing page CTA. Once the tracking is set it up, it will collect data on its own. As long as the metrics and thresholds are clearly defined, it can gather a lot of performance data. Abnormal behavior in the data can guide where to allocate usability testing resources. e.g. if there’s a spike in the number of clicks on the “refund” button and conversion is crucial to the project, then it’s easy to make a case for spending more time for usability testings to drill deeper into WHY users are clicking “refund.”

Fig 5: an example of a web analysis tool from Google.

All above are just a subset of all UX Methods. If you are interested, feel free to learn more on NNG or Usability.gov. At the end of the day, the spirit is never to know all of the methods but to know when to use what. That’s how both the research quality and project turnaround time can be improved.

*link to usability.gov and Nielson Normal Group.