User Research Methods

This deck is a friendly introduction to the methods UX researchers use to understand the people who use their products. It starts with foundational concepts like what UX research is and the difference between qualitative and quantitative approaches, then moves into the differences between attitudinal and behavioural research. From there, it walks through specific techniques including user interviews, contextual inquiries, diary studies, and surveys, along with practical details like the three types of interviews, best practices for asking good questions, and how to design survey scales such as Likert scales and Net Promoter Scores.

The cards are well suited to anyone who is new to user research or who wants a refresher on the core vocabulary of the field. UX and product design students, junior designers, product managers, marketers, and developers who collaborate with research teams will all find this a useful grounding in how insights are gathered from real users. If you are preparing for an interview, a course, or simply want to speak more confidently about research methods with your team, this deck gives you the essential terms and distinctions in one place.

Because the topics build on one another, it helps to review the earlier concepts like qualitative versus quantitative research before moving into the specific methods. Try spacing your study sessions across a few days rather than cramming everything at once, since flashcards work best when your brain has time to forget and re-remember. As you go, pause to think about when you would actually use each method in a real project, which will make the distinctions between interviews, diary studies, and surveys stick much more naturally.

UX research is the systematic study of target users and their requirements, designed to add realistic contexts and insights to the design process and ensure that products meet real needs. It sits at the heart of human-centred design, helping teams move beyond assumptions and gut feelings toward decisions grounded in evidence about how people actually behave, think, and feel when interacting with a product or service.

Researchers typically distinguish between two broad approaches. Qualitative research explores the "why" behind user behaviour through methods such as interviews, observations, and open-ended data collection, while quantitative research measures the "what" and "how much" through numbers and statistics. A related distinction is between attitudinal research, which captures what users say (their opinions and preferences), and behavioural research, which captures what users actually do (their observed actions and click patterns). Because people often say one thing and do another, combining both kinds of evidence yields a more complete picture.

UX research is also commonly split into generative and evaluative work. Generative research explores problems and opportunities early in a project to define what should be designed, while evaluative research tests existing designs to assess how well they work. To strengthen the validity of insights, practitioners use triangulation, which means combining multiple research methods such as interviews together with analytics and usability tests, so that findings confirmed across different approaches can be trusted with greater confidence.

These research activities fit naturally inside a broader design thinking process with five phases: Empathise, Define, Ideate, Prototype, and Test. In the Empathise phase, researchers build deep understanding through observation, interviews, and immersion, setting aside assumptions to see the world through users' eyes. The Define phase synthesises findings into a clear problem statement, often expressed as a "How Might We" question that reframes the issue as an opportunity for design. From there, Ideate generates a wide range of creative solutions through brainstorming, sketching, and other divergent techniques; Prototype builds quick, inexpensive representations of those solutions; and Test puts prototypes in front of real users to learn what works, iterating back to earlier phases as needed.

Qualitative research relies on direct conversations and observations to uncover the motivations, behaviours, and pain points that numbers alone cannot reveal. The most common conversational method is the user interview, a one-on-one discussion with a current or potential user to understand their needs, behaviours, and context. Interviews come in three forms: structured, where every participant is asked the same fixed questions; semi-structured, which follows a guide but allows flexible probing; and unstructured, which takes the form of an open exploratory conversation. Best practice is to ask open-ended questions, avoid leading questions that suggest a preferred answer, and listen more than you talk, aiming for roughly an 80/20 listen-to-talk ratio so the participant does most of the talking.

A useful rule of thumb is that five to eight interviews per user segment will typically reveal around 80% of usability issues, based on Nielsen's research on diminishing returns. Leading questions such as "Don't you think this feature is easy?" bias responses, so a researcher would instead ask neutrally, "How was your experience with this feature?" To bring research even closer to reality, contextual inquiry involves observing and interviewing users in their natural environment while they perform tasks, combining observation with dialogue to surface tacit knowledge that interviews alone miss. When researchers need to understand behaviours over time, habits, or infrequent events, they turn to diary studies, in which participants log their experiences, activities, and thoughts over days or weeks using a journal, app, or photographs.

Several synthesis tools help teams make sense of qualitative data. An empathy map is a collaborative canvas that captures what a user says, thinks, does, and feels, building shared understanding across a team. An affinity diagram organises qualitative data by grouping related ideas, quotes, and observations on sticky notes into clusters that reveal themes, while thematic analysis is the broader qualitative method for identifying recurring patterns across interview transcripts and diary studies. Researchers also conduct task analyses, breaking down user goals into discrete steps and sub-tasks to understand the actions, decisions, and information needed to complete a workflow. Throughout this work, recruiting the right people is essential: a screener survey is a short questionnaire used to filter and select participants who match the target user profile, and participant recruitment more broadly is the process of finding users who represent the target audience so that findings remain valid and reliable. Finally, a research repository acts as a centralised collection of all research findings, insights, and artefacts, making past work discoverable and avoiding duplicate studies.

Quantitative methods translate user experience into measurable data. A survey is a structured questionnaire distributed to many users to collect both quantitative and qualitative data about attitudes, preferences, and demographics. Survey questions come in two flavours: closed-ended questions offer predefined answer options such as multiple choice or rating scales and produce quantitative data, while open-ended questions invite free-text responses and yield qualitative insights. Among rating scales, the Likert scale typically offers 1–5 or 1–7 points measuring agreement or satisfaction, ranging from "Strongly disagree" to "Strongly agree".

Two widely used standardised questionnaires help benchmark and compare experiences. The Net Promoter Score is a loyalty metric based on the question "How likely are you to recommend us?", scored from 0 to 10, with the score calculated as the percentage of Promoters (those answering 9 or 10) minus the percentage of Detractors (those answering 0 to 6). The System Usability Scale, or SUS, is a 10-question questionnaire that produces a score from 0 to 100 measuring perceived usability, with a score above 68 considered above average. These instruments allow teams to track changes over time and compare against industry norms.

Behavioural analytics complement survey data by capturing what users actually do. Tools like Google Analytics, Hotjar, and Mixpanel provide quantitative data on page views, click paths, drop-offs, and conversion rates. A heatmap is a visual representation of where users click, move, or scroll on a page, with warmer colours indicating more activity, and tools such as Hotjar and Crazy Egg help teams spot attention patterns. Session recordings replay actual user sessions, including mouse movements, clicks, and scrolls, so researchers can observe confusion and friction in context. Funnel analysis tracks the percentage of users who complete each step in a multi-step process such as a checkout flow, identifying exactly where drop-offs occur.

For more precise attention data, eye tracking uses specialised hardware to measure where users look, for how long, and in what order, revealing visual hierarchy and elements that get overlooked. Research from the Nielsen Norman Group has shown that users tend to scan web content in an F-shaped pattern: reading across the top, then down the left side, with progressively shorter horizontal scans as they move down the page. Recognising this pattern helps designers place key messages, calls to action, and headings where they will be seen.

Usability testing is a method in which representative users attempt tasks with a product while researchers observe to identify usability problems and measure performance. In moderated sessions, a facilitator guides the participant through tasks in real time, asking follow-up questions and probing for insights, whereas unmoderated testing has participants complete tasks independently, often remotely through platforms such as UserTesting or Maze, making it more scalable and cheaper. A common technique in both settings is the think-aloud protocol, in which participants are asked to verbalise their thoughts continuously while performing tasks, revealing their mental model, expectations, and points of confusion.

Three key usability metrics anchor most tests. Task success rate is the percentage of participants who complete a given task correctly and serves as a primary measure of effectiveness. Time on task measures how long it takes to complete a task, while error rate captures how often mistakes occur; together with satisfaction ratings, these provide a robust view of usability. Beyond single-session tests, teams often compare designs through controlled experiments. A/B testing, also called split testing, compares two versions of a page or feature, with the existing version as the control (A) and a modified version as the variant (B), to determine which performs better on a specific metric. Statistical significance, typically reported at a 95% confidence level or \( p < 0.05 \), indicates that the observed difference is real and not due to random chance, and the required sample size depends on the baseline conversion rate and the minimum detectable effect, often running into hundreds or thousands of users per variant. When several variables need to be considered together, multivariate testing examines multiple variables simultaneously, such as headline, image, and button colour, to find the optimal combination, though this approach requires much larger sample sizes than A/B tests.

Two complementary methods address how information is structured and found. Card sorting has participants organise topics into categories that make sense to them: in an open card sort, they create their own category names to discover structure, while in a closed card sort they sort cards into predefined categories to validate an existing structure. Tree testing evaluates the findability of topics in a site's hierarchy by asking participants to locate items within a text-only tree, without visual distractions. The two methods work well together: card sorting helps create the information architecture, and tree testing validates whether users can actually find content within that structure.

Several lighter-weight techniques offer quick insight. A usability benchmark is a baseline measurement of task success, time, and satisfaction that can be compared against after design changes. The 5-second test shows a design for five seconds and then asks what the user remembers, testing whether the key message and purpose come across at a glance. A first-click test measures where users click first to complete a task, and research shows that users who get that first click right succeed around 87% of the time. Remote usability testing conducts sessions with participants in a different location, either moderated via video call or unmoderated via a testing platform, while guerrilla usability testing takes an even more informal approach, recruiting random participants in public places like coffee shops for fast, low-cost early feedback. Throughout a project, evaluation can be formative, conducted during design to identify and fix issues, or summative, conducted after design to measure overall usability against benchmarks.

Information architecture, often abbreviated as IA, is the structural design of shared information environments: organising, labelling, and structuring content so that users can find and understand it. It rests on four components: organisation systems that group content meaningfully, labelling systems that name things consistently, navigation systems that let users move through content, and search systems that help users locate specific items. A sitemap is a visual diagram showing the hierarchical structure of a website's pages and their relationships, used as a planning tool during IA work, while navigation design creates the menus, breadcrumbs, links, and search affordances that let users actually traverse content and find what they need efficiently.

Wireframes and prototypes bring IA to life at different levels of fidelity. A wireframe is a low-fidelity visual guide representing the skeletal framework of a page, showing layout, content placement, and basic functionality without visual styling. Low-fidelity wireframes are simple sketches or grey-box layouts that are quick to create and focus on structure, while high-fidelity wireframes are more detailed and use real content and closer-to-final proportions. A prototype goes further: it is an interactive simulation that allows users to click through screens and experience the flow, making it suitable for testing before any code is written. In short, a wireframe is a static layout, whereas a prototype is interactive and clickable, simulating the actual user experience.

A range of tools supports this work, including Figma, Sketch (often paired with InVision or Principle), and Adobe XD, all of which support interactive prototyping and collaboration across teams. To ensure visual and behavioural consistency across a product, organisations maintain design systems: collections of reusable components, patterns, and guidelines covering buttons, typography, colour, spacing, and other foundational elements. Together, IA, wireframes, prototypes, and design systems form the connective tissue that turns research insights into testable, buildable product designs.

Personas are fictional but data-driven representations of key user segments that summarise demographics, goals, frustrations, behaviours, and context of use. A good persona includes a name and photo, demographic details, goals and motivations, pain points, behaviours and context, and a quote that captures the user's mindset. Most projects benefit from three to five primary personas, enough to cover key segments without overwhelming the design process. The risk of creating personas without underlying research is that they become fictional stereotypes, leading to design decisions based on assumptions rather than real user needs.

A customer journey map is a visual representation of the end-to-end experience a user has with a product or service. Its key components include stages or phases of the experience, user actions, thoughts and feelings, touchpoints where the user interacts with the brand, pain points, and opportunities for improvement. A related artefact, the experience map, captures the broader experience independent of any particular product, while the journey map focuses specifically on a single user's interaction with that product. Extending this further, a service blueprint adds behind-the-scenes detail to the journey, showing front-stage actions, back-stage actions, support processes, and physical evidence so that teams can design not just what the user sees but also the operations that make it possible.

Competitive analysis is a complementary input to all of this work. It involves evaluating competitor products to understand their strengths, weaknesses, features, and user experiences, and identifying gaps and opportunities for your own design. A useful competitive analysis typically includes a feature comparison matrix, an assessment of UX strengths and weaknesses, screenshots or flows of competitor products, a summary of user reviews, and a clear articulation of opportunities for differentiation. Together, personas, journey maps, experience maps, service blueprints, and competitive analyses help teams anchor their work in a shared, evidence-based understanding of both their users and the market.

Expert-led evaluation complements user testing. Heuristic evaluation is an expert review method in which evaluators assess a user interface against a set of established usability principles, known as heuristics, to identify problems. The most widely used set was developed by Jakob Nielsen in 1994 and remains the industry standard. These ten heuristics include visibility of system status, meaning the system should always keep users informed about what is happening through appropriate feedback such as loading indicators and success messages; match between system and the real world, meaning the interface should use language, concepts, and conventions familiar to users rather than internal jargon; user control and freedom; consistency and standards; error prevention; recognition rather than recall; flexibility and efficiency of use; aesthetic and minimalist design; help users recognise and recover from errors; and help and documentation. Typically, three to five evaluators working independently will find around 75 to 80% of usability problems. A related expert method, the cognitive walkthrough, has experts step through tasks from the user's perspective, asking at each step whether the user will know what to do and whether they will understand the feedback.

Several classic UX principles and laws help explain why certain designs feel intuitive. Fitts's Law states that the time to reach a target is a function of the distance to it and its size, so larger and closer targets are faster to click, implying that important buttons should be made large and accessible. Hick's Law states that the time to make a decision increases logarithmically with the number of choices, so reducing options speeds up decisions. A mental model is a user's internal representation of how something works, and designs that match users' mental models feel intuitive while mismatches cause confusion. Progressive disclosure is a design pattern that shows only essential information first and reveals more detail on demand, reducing cognitive load, which is the total amount of mental effort required to use an interface. Finally, the aesthetic-usability effect describes the observation that users perceive aesthetically pleasing designs as more usable, so good visual design creates positive first impressions and increases tolerance for minor issues.

Accessibility is a non-negotiable dimension of usability. The Web Content Accessibility Guidelines, or WCAG, are international standards published by the W3C for making web content accessible to people with disabilities. They are organised around four principles known as POUR: content must be Perceivable, Operable, Understandable, and Robust enough to work with assistive technologies. WCAG defines three conformance levels: Level A as the minimum, Level AA as the recommended standard for most sites, and Level AAA as the highest, which is generally not required for entire sites. For Level AA, normal text must meet a colour contrast ratio of at least 4.5:1, while large text (18pt or larger, or 14pt bold or larger) requires at least 3:1.

Alt text, or alternative text, describes an image for screen readers and should convey the image's purpose rather than just announcing "image of". Common accessibility issues include missing alt text on images, insufficient colour contrast, and keyboard-inaccessible interactive elements. Screen readers such as JAWS, NVDA, and VoiceOver are assistive technologies that read aloud the content of a screen for people with visual impairments, and ensuring compatibility with them is central to creating inclusive experiences for all users.