← Blog from Guindo Design, Strategic Digital Product Design
From hypotheses to actionable insights, a guide to user interviews
Beyond answers: how to get deep insights in UX interviews
User interviews are a fundamental tool in design research. Getting answers to fill in the report is not enough, the real value lies in uncovering patterns, motivations and deep needs that drive transformative design decisions.
In our experience, a single powerful discovery can completely rethink a product or service. For example, when we start StepAlong, The main motivation for defining the project was kill crappy PDFs But it was the first customer we had that gave us a slap in the face of reality: «If I could generate a PDF from a StepAlong, I wouldn't have to keep producing and maintaining PDFs, but I could still offer them to my customers if they needed them».»
While most people are fine with using their smartphone to follow instructions, some people still like to have a PDF they can print or use on their tablet, and in some cases, such as complex products, it could be really useful to have a PDF to complement StepAlong, i.e. give the customer a choice. So we developed the functionality of dynamic PDFs in any language, freeing customers from their maintenance and giving end users a complementary format to follow instructions.
This article explores how to structure effective interviews, from hypothesis formulation to implementation of findings, to achieve actionable insights that inform successful design decisions.
Formulate hypotheses before creating questions
A particularly useful method for defining powerful questions is to first formulate hypotheses about the behaviours or needs that we want to validate or refute.
Having prior hypotheses allows us a posteriori to contrast them with the users' real answers, avoiding biased interpretations. It also helps to generate more precise and relevant questions.
It is crucial to list even the most obvious assumptions, as they can often be incorrect. A common mistake is to assume that we already know how users think, which leads to biased questions and missed opportunities for discovery.
Examples of scenarios for a digital tutorial platform:
- Users prefer interactive, step-by-step tutorials over static documents.
- People drop out of tutorials mainly because of confusing instructions.
- Users value the ability to navigate quickly between sections more than detailed explanations.
- Most prefer purely digital content and do not see value in downloadable formats
By listing these hypotheses in a spreadsheet, we can structure our research and connect each hypothesis to specific questions. Then, by cross-checking with user responses, we can confirm, refute or adjust our initial assumptions.
Structure of the interview script
The process starts with setting clear learning objectives - what do we really want to find out? Before defining specific questions, it is essential to write research objectives aligned with the central question we seek to answer.
Interview typologies: when to use each one
There are three main types of design interviews:
- Structured: They follow fixed questions in a specific order, ideal when you are looking to compare responses across many participants. They are especially useful when you need consistent data for quantitative analysis or when you are working with a large team of interviewers and need to maintain consistency. However, they offer less flexibility for exploring emerging themes.
- Semi-structured: They combine fixed questions with the flexibility to explore emerging themes, providing depth and consistency. This format is the most versatile and commonly used in design research, as it allows you to follow a script while adapting to user responses. It is ideal for teams with some interviewing experience and when seeking a balance between consistency and discovery.
- Unstructured: They are based on a free conversation, allowing unexpected insights to be discovered. They are useful in exploratory phases or when complex contexts need to be understood. This format requires experienced interviewers who can guide the conversation effectively without a rigid script. It is particularly valuable when exploring unfamiliar domains or when seeking to deeply understand the subjective experiences of users.
The choice of interview type will depend on the research objectives and the degree of exploration desired. An effective interview script addresses five key dimensions:
- TargetWhat does the user hope to achieve, what are their tasks to be accomplished, and what are their outstanding tasks?
- RealityWhat is the user's current situation, how do they currently achieve their tasks, how often, how much execution time do they spend, what steps do they perform?
- AlternativesHow did you always have this objective and how did you achieve it in the past? If they are not satisfied, have they considered or found an alternative?
- MeaningWhat impact does this problem have on your life, what do you gain by achieving your goal, and what would you lose if you don't achieve it?
- SolutionList the main problems. This reveals what they are looking for solutions to.
It is not necessary to follow this order strictly, but make sure to cover each aspect to get a complete picture of the user experience. Questions related to solutions are best addressed at the end of the interview, when the user feels more comfortable to share deeper insights.
At the end you should have a list of questions that tries to address everything that stands between the user's problem and the potential product, with the intention of understanding the user's objective and especially inquiring into the motives of the user. why it is important to them.
You can start by probing into the user's reality: «Can you describe how you usually look for and follow instructions when you need to learn something new?» This provides context and makes the conversation natural. Start with broad questions and refine as you go along. Group questions in a coherent way so that the conversation flows naturally.
It is also important to plan the preamble to the interview:
- How you will present yourself
- How you will explain the dynamics and duration
- How you will ask for permission to record the session
Design open-ended, neutral, unguided questions.
Well-worded questions are key to obtaining genuine information. They should be open-ended and avoid suggesting an expected answer.
A bad question suggests or directs towards an answer, while a good question explores without influencing the interviewee.
Example:
- HypothesisUsers prefer interactive digital tutorials over static documents.
- Bad questionDo you prefer interactive tutorials instead of boring PDFs?
- Good question: How do you like to consume instructions when you need to learn something new?
Open-ended questions allow the conversation to flow and reveal unexpected insights. Supplementing with «Tell me more about that» is a powerful technique to deepen initial responses.
Plan questions for a one-hour conversation, but stay flexible. As you go along, you can adjust, change and create new questions based on the discoveries you make.
The role of the interviewer: uncovering the truth, not selling.
Empathetic interviewing is not just about being friendly, but about creating a space where users feel comfortable sharing their genuine experiences. Rather than applying a rigid questionnaire, it is about having a fluid conversation, actively listening and subtly guiding to uncover real motivations and frustrations.
Listening and observing attentively is vital to uncover authentic information, tuning in to the user both verbally and non-verbally.
A common mistake is to try to validate our preconceived ideas rather than obtain genuine information. If we seek only answers that fit our expectations, we will ignore key signals and reinforce possibly erroneous ideas.
“We tend to project our own rationalisation and beliefs onto the actions and beliefs of others”.”
Donald Norman, The design of everyday things
Encourages detailed answers and avoids problematic questions
For richer and deeper answers, use phrases such as:
- «Tell me about the last time you you followed a complex tutorial...»
- «Tell me about how decide if a tutorial is worth your time...».»
- «Describe how you feel when an instruction is not clear...»
- «Explain to me more about that moment of frustration when you you were trying to follow in those footsteps...»
When you want to go deeper, or clarify something, use probing questions:
- «Could you elaborate on that?»
- «Why was it important for you to be able to keep that information?»
- «I'm interested to better understand that point of frustration... could you explain it again?»
A valuable resource is to ask both positive and negative questions. For example:
- «What do you hope to achieve with better interactive tutorials?»
- «If you continue to use traditional tutorials, what opportunities do you think you would lose?»
Types of questions to avoid
- Questions forcing a specific answer
- Questions that do not allow you to explore different options
- Closed questions (yes/no) with no possibility of expansion
Active listening and user relations
Active listening helps to uncover new and unexpected information, rather than confirming preconceived ideas. It creates an atmosphere of trust to elicit honest answers.
Verbal signals
- «I understand», «Of course», «I see what you mean».»
- Little sounds like «mmm», «uh-huh», "uh-huh", "uh-huh", "uh-huh", "uh-huh".»
- Repeating keywords to extend the user's reach
Non-verbal signals
- Frequent eye contact.
- Nodding.
- Mirroring the user's body posture.
Observing what is not there
Sometimes it is not what the user says that is most revealing, but what they do not mention. Users may not be aware of latent needs or underlying issues that affect their behaviour.
A landmark case in online banking revealed that many users avoided using the app in public places because they feared others would see their balance and financial status. This insight led to the implementation of the «hide balance» functionality, which increased the app's usage significantly.
Often, users cannot articulate needs that they do not even know they have. A good interviewer goes beyond direct answers and looks for patterns in user behaviour.
As Henry Ford said:
«If I had asked my customers what they wanted, they would have said ‘a faster horse’.»
The challenge for designers and researchers is to discern when incremental improvement is needed and when radical innovation is needed. Interviews alone cannot answer this question, but they can provide invaluable clues about the underlying problems and deep motivations of users.
What you can learn by talking to users is what their life is like today, what frustrates them, what they value, how they solve their problems today, and this will eventually lead you to solutions that have a good product/market fit.Sometimes, the most revealing thing is not what the user says, but what they don't mention. Users may not be aware of latent needs or underlying problems that affect their behaviour.
Consider price as part of the conversation
Price is not just a number, but a representation of perceived value. Asking about willingness to pay can reveal valuable expectations:
- «What features would make a tutorial platform worth a monthly subscription?»
- «How would you assess whether an instruction system justifies its cost to your company?»
Understanding the perception of value helps to prioritise functionalities and establish appropriate monetisation models.
Observe patterns and segment users
During interviews, you might notice that different groups of users have different needs and behaviours. For example, in a research on tutorial consumption, you might identify:
- Visual learners: They prefer image-rich tutorials and videos.
- Quick consultants: They are looking for specific references, they don't want to read the whole content.
- Methodical followers: Need detailed step-by-step instructions
- Archive-dependentRead more : Value being able to download and save content for offline access
Identifying these emerging segmentations is key to designing solutions adapted to different user profiles.
The post-interview analysis process
Once the interviews have been conducted, systematic analysis is essential to extract actionable insights:
- Transcription and initial reviewReview recordings to identify relevant quotes and initial patterns.
- Thematic codingGroup answers into conceptual categories (motivations, frustrations, behaviours).
- Pattern identification: Searches for significant trends and anomalies between different interviews.
- Contrast with hypothesesEvaluates each initial hypothesis against the evidence gathered.
- Extracting insightsFormulates key findings that will inform design decisions
- PrioritisationEvaluates each insight according to its potential impact and feasibility of implementation.
For example, a recurring pattern in interviews about tutorials might reveal that, contrary to our initial hypothesis, many users highly valued being able to download PDF versions of tutorials for future reference or to share with colleagues. This insight led our startup to develop a «rich PDF» functionality that combined the navigability of the digital format with the portability of downloadable documents - a feature that we initially considered a low priority but that turned out to be critical to business adoption of our platform.
Adaptations according to the research context
Interview techniques should be adjusted according to the type of project:
For new products (exploratory research):
- Focus on existing problems and current solutions
- Delves into contexts of use and points of frustration
- Avoid describing your solution too early
For iterative improvements:
- Focus on specific experiences with the current version
- Explores unfulfilled expectations and latent desires
- Use prototypes or mockups to get concrete feedback.
Remote vs. face-to-face interviews:
- In remote sessions, consider using screen sharing tools.
- Pay more attention to verbal cues when non-verbal cues are limited.
- Allows more time for rapport building in digital environments
Practical advice
- Record the interview (with user's permission)
- Do not take excessive notes in order to maintain eye contact.
- Ask for references from other interviewees with similar profiles.
- Pilot the interview to fine-tune the guide.
- Set aside time immediately after each interview to document initial impressions.
Turning insights into UX artefacts
Data can be transformed into:
- Empathy maps: Representing thoughts, feelings, actions and pains
- PersonsUser profiles based on actual behaviours and needs
- User journeysVisualising the complete user journey, including pain points and opportunities
- List of needsPrioritised according to frequency and impact
- Opportunity matrices: Crossing technical feasibility with user value
For example, an insight into the frustration of having to switch between seeing instructions and carrying them out could be transformed into a «busy hands mode» function with voice navigation on the tutorial platform.
Triangulation with other methods
Interviews work best when they are complemented by other methods:
- Behavioural analytics: To validate what users actually do vs. what they say
- Usability testing: To observe concrete interactions with existing solutions
- SurveysTo quantify trends identified in qualitative interviews.
- Daily studiesTo capture behaviour in the moment, minimising memory biases.
Limitations of the interviews
Interviews have important limitations to consider:
- They depend on what the user remembers and is willing to share.
- May generate incomplete or biased data
- The quality of the information depends to a large extent on the skill of the interviewer.
- People sometimes rationalise instinctive behaviours, offering inaccurate explanations
To mitigate these limitations, it is advisable to cross-check findings with direct observation where possible.
The ideal interviewer: key qualities
Skills that distinguish a great interviewer include:
- Genuine empathy: Ability to connect authentically with different people
- Insatiable curiosityWillingness to question one's assumptions
- Analytical thinkingAbility to identify non-obvious patterns and connections.
- AdaptabilityFlexibility to adjust the approach according to the personality of the interviewee.
- Active listeningAbility to process information while planning the course of the conversation.
Developing these qualities requires constant practice and honest self-evaluation.
Scalability of research
When resources are limited or you need to interview many users:
- Group sessionsConsider interviews with 2-3 users simultaneously for non-sensitive topics.
- Strategic prioritisationFocus on higher value or higher risk user segments.
- Partial automation: Use recruitment and scheduling tools to streamline logistics
- Compact interviewsDesign shorter sessions (30 min) focused on specific aspects of the project.
Conclusion: from research to action
UX interviews are more than a collection of answers: they are a tool for uncovering deep truths that drive meaningful innovation. The best insights combine surprise («we hadn't thought of that») with retrospective obviousness («of course it was like that»).
The real value of research materialises when we transform these findings into concrete design and product decisions. In the case of our tutorial startup, the insight about the need for enhanced PDFs not only changed our product strategy, but opened up a whole new market segment: corporate customers who needed accessible, navigable and archivable documentation.
Every interview is an opportunity to question our assumptions and approach solutions that truly resonate with fundamental human needs.
Do you have cases where interviews have transformed your design decisions?
Share them - I promise to quote you.
Good luck and enjoy the conversations!