UX Research is on the rise, and it has never been more important to make sure it is conducted properly. Bad UX Research is more damaging than no UX Research – you build your house on unstable foundations. It is therefore important to structure your research properly and use well-crafted questions.
Structuring the UX Research interview
It’s just asking a few questions, isn’t it? No, it isn’t. You need to think about the objectives and how you will achieve them through the interview structure. If you don’t you will almost certainly lead the participant and ask them to predict the future. You need to avoid both.
There is no one structure that fits all UX Research types. There are also choices about methodology you can make. To begin with, think about the objectives and where you are in the product development process. Your methodology at the discovery stage will be very different to when you are evaluating prototypes.
There are many methodologies to choose from so I have highlighted a few examples that we use regularly at UX24/7. If you don’t have the time or inclination to explore these, you can use a basic pyramid approach as follows:
Consider the topic you are interested in (there will be more than one in the research) as the top of the pyramid. As you work down there are increasing numbers of paths available that lead from the top. This is the same with your questioning. Plan these before the interview as they will provide the structure.
Here are some other methods based on two stages of the product or service development process, Discovery and Generative.
Discovery research methods
At the discovery stage of a project, we want to be as divergent as possible. We are learning about unmet needs, unknown needs, behaviours and more. We need an interview structure that opens the participants frame of mind but also answers our objectives.
You will no doubt have heard colleagues say, “we need to understand the users’ mental model”. You may have even said it yourself. It is an ambitious objective but needed if we are going to create products and services that thrive.
The mental models approach uses the following key steps:
Task based audience segments
Different to market research segmentation, mental model methodology looks to group people based on the tasks they perform. This is like JTBD theory in some ways.
As Young says, task-based audience segments are, quite simply, groups of people who do similar things. The benefit of this approach is that it allows us to structure our interview based on the activity rather than the demographic segmentation.
You should have clear research goals. If you don’t ask yourself why you are conducting the UX Research. When you have the goals, work on them to create clear prioritized business objectives.
Young provides the following example:
Goal: Simplify access to information
Prioritised key business objectives:
- Simplify web navigation
- Have one site
- Make things consistent
- Base navigation on audience needs
- Improve look & feel
- Resolve access to large information stores
Do this for each of your goals so you have a few sets of objectives.
From these business objectives we can then assemble topics that we want to explore with the participant. The topics will guide our questions.
Create interview prompts
If you are an experienced researcher, you can create prompts that remind you to explore each of the areas related to your objectives. This is something Young does herself to avoid the interview being too scripted. It allows the interview to flow more naturally and to pursue subjects brought up by the participant.
For inexperienced research moderators it is better to start with a more scripted interview where your questions are written down. Open questions, those beginning with who, what, why, where, how, when etc., can be difficult to come up with under pressure. You may find yourself resorting to closed questions that close off the conversation.
A closed question is one that results in a yes or no answer. For example:
Have you ever bought shoes online before? [CLOSED]
When was the last time you bought shoes online? [OPEN]
There is a lot more to the mental model’s methodology than this. I have taken a few of the structural elements to help guide your creation of the discussion guide. By thinking in terms of goals > objectives > topics > and prompts you will be able to structure your interview.
Jobs to be Done
Jobs to be done was popularized by Clayton Christensen in the late 90’s and much has been written about it since. We tend to look to Jim Kalbach because his book is a very practical guide that addresses the strategic aspects we tend to look for.
What I really like about the JTBD framework, is that it shifts the focus from the organisation and the product or service to the customer and their needs. That is very helpful when structuring your UX Research session. There can be a tendency to focus on the subject matter, rather than the customer.
There are five elements of JTBD that helps tom focus the research:
- Job performer – who in the research
- Jobs – what they do
- Process – how they do it
- Needs – why they act the way they do
- Circumstances – when and where this activity takes place
You can already see that this model provides a useful framework for structuring our research sessions.
In Kalbach’s book, there is a section on how to prepare for and conduct the interview. There are three sets of guidance that are helpful as follows:
Interviewing guidelines and tips:
- Create rapport: this is about using eye contact, being empathetic
- Listen: he focuses on avoiding leading and actually listening to them. This is crucial as researchers can be focussed on the next question rather than listening and reacting to the current answer
- Avoid yes-or-no questions: see above about open questions
- Dig deep: linked to listening, if you hear something of interest, pick up on it. I am often asking my research colleagues to be curious.
- Minimise distractions: make sure the environment is right and that you won’t be interrupted or distracted.
- Go with the flow: sometimes you must work with what you have got. Go with it.
- Don’t interrogate: the interview should be like a conversation, not an endless stream of questions
- Use pauses: You will be surprised at the power of silence. It gives the participant time to think.
- Research in pairs: This is less important in my view now. We use tools like Condens that mean a second researcher is not always needed.
These are valuable for any interview.
Critical incident technique:
The critical incident technique is presented as a way of encouraging the participant to be more specific. In research, when we ask a participant an open question, they tend to respond in broad terms. We don’t want that, and these three steps provide a useful tool.
- Recall a specific incident. Ask them to recount a situation where doing the job went wrong for them.
- Describe the experience. Ask them to describe this in detail in terms of what went wrong, why, and how they felt about it at the time.
- Discuss the ideal state. Ask them to describe what should have happened. This reveals their underlying needs in specific terms.
The third technique provided in the book is a structure that ensures the interview is about the participant. There are 5 aspects:
- Get background about the participant and the job
- Understand the main job and related jobs
- Understand the process of executing the job
- Find needs
- Probe on circumstances
As researcher, you construct questions under each heading. Each build on the last section creating an in-depth picture. It is important to understand what is meant by “job” here and so some knowledge of JTBD theory is needed.
Generative research methods
The final approach I want to talk about is “The path of expression”. This technique was created by Liz Sanders and Pieter Jan Stappers and is set out in their book “Convivial Toolbox: Generative Research for the Front End of Design”. In UX24/7 it is heavily sponsored by John Dumas our Consultancy Director, and with good reason.
One of John’s pet peeves is tendency in research to create a situation where the participant is being asked to predict the future. This happens when the interview dives straight into the detail before the participant has time to acclimatize. A short warm up interview is OK for usability testing. But for discovery and generative research, we need more structure.
The path of expression provides the structure for generative research interviews. It uses “Make tools” to support the leap into the future by basing the discussion first on deep interpretations of the past.
The four stages to path of expression
Step 1: Using the topic of the research, ask participants to observe, reflect and describe their current experiences. This is done in the session as the first stage of the interview.
Step 2: Ask the participant to select and reflect on memories from earlier experiences. It is suggested that this is achieved using photographs and other “evocative triggers”. This should form the basis of a homework task.
Step 3: The participant is then asked to reflect on their memories and consider the possibilities for the future. Sharing the current and past allows access to the underlying needs and values and provides a basis to explore the future.
Step 4: In generative research this final step is where we conduct co-creation or use exercises to create artefacts that underpin the experience design. For research conducted when there is already a high-level proposition, this is where it is introduced.
UX Research structure summary
These are just three methods that you can use to base the structure of your UX Research interview. There are many more. The important takeaway here is that you need to invest time in preparing the structure to achieve your goals.
Without a structure, your research will deliver answers, but they may not be valid. Your design decisions could be built on shaky foundations.