Design thinking research methods: The research toolkit for each phase

Design thinking is a problem-solving framework built on a core premise: understand users deeply before generating and testing solutions. Each of its five phases, Empathize, Define, Ideate, Prototype, and Test, calls for different types of research activity. The methods appropriate for building user understanding at the start of the process are fundamentally different from the methods appropriate for evaluating a prototype at the end.

Getting this right matters practically. Teams that apply evaluative research methods in the Empathize phase test solutions before they understand the problem. Teams that use generative methods in the Test phase produce insight but no usability data. This article maps the right research methods to each phase and explains how they connect across the full process.

Research in the Empathize phase

The Empathize phase exists to build genuine understanding of the user: their goals, their context, their frustrations, and the way they currently navigate the problem space. Research here is generative and exploratory. You are not testing a hypothesis. You are trying to understand a domain well enough to define the right problem.

User interviews are the primary tool for empathize-phase research. These are open-ended conversations that explore a user’s goals, workflows, and challenges without reference to any specific design or product. The questions are deliberately broad at first, narrowing as patterns emerge. The goal is to understand their world on their own terms, not to confirm what your team thinks their problems are. See how to conduct effective user interviews for session structure guidance and the customer discovery interview guide for question frameworks.

Contextual inquiry adds a layer that interviews alone cannot provide. Observing users in their actual work or life environment as they complete relevant tasks reveals what they actually do, not just what they say they do. Environmental factors, workarounds, interruptions, and context-switching behaviors that users never mention in interviews become visible when you observe them in context. This is particularly valuable for complex professional workflows where the gap between self-reported behavior and actual behavior is wide. See how to do ethnographic research online for remote contextual approaches.

Diary studies fill a gap that both interviews and contextual inquiry leave open: they capture experience over time. Participants document their behaviors and experiences across days or weeks through structured entries, capturing the longitudinal dimension of a problem that point-in-time sessions miss. Diary studies are especially useful for workflows that span multiple sessions, behaviors that happen sporadically, or experiences where the cumulative effect matters more than any single interaction. See how to run a diary study for the methodology.

Secondary research provides context before primary research begins. Existing studies, market reports, support ticket analysis, customer feedback data, and behavioral analytics reveal what is already known and where the genuine unknowns are. Teams that skip secondary research often spend primary research budget rediscovering things that were already documented.

For empathize-phase recruitment, participants should be current users of the existing approach to the problem, not necessarily your intended customer for the proposed solution. You want to understand how people currently navigate the problem, including those using workarounds, manual processes, or competitor products. For professional or B2B research requiring specific job functions, industries, or company sizes, CleverX’s panel of 8 million verified professionals gives you the filtering to reach the exact practitioner profiles relevant to your domain. See best user research tools for B2B for platform options.

Synthesis in the Define phase

The Define phase is primarily analytical. Its purpose is to synthesize what you learned in the Empathize phase into a clear, evidence-grounded problem statement that can guide the Ideate phase productively.

Affinity mapping is the foundational synthesis method. All observations from empathize-phase research are written on individual notes and organized into thematic clusters. The clusters reveal patterns that were not visible at the individual session level. Which problems appeared across multiple participants? Which behaviors were consistent? Which contexts produced the most friction? See user research synthesis methods for affinity mapping and other synthesis approaches.

Journey mapping translates synthesized research into a visual representation of the current-state user experience across a relevant workflow. A journey map shows the steps users take, the actions they perform, the touchpoints they encounter, and their emotional state at each stage. Journey maps make patterns accessible to cross-functional teams who were not present for the research and help the full team see where the most significant friction and unmet needs are concentrated.

Jobs-to-be-done analysis reframes the problem in terms of the underlying job users are trying to accomplish rather than the surface features they are using or requesting. A job-based problem definition produces solutions that address the root need rather than the surface symptom. See jobs to be done research guide for the framework. The output of this analysis feeds directly into problem statement formulation: a clear articulation of who you are designing for, what they are trying to accomplish, and what constraints or obstacles currently prevent them from doing it well.

A well-formulated problem statement is not a product requirement. It is a user need: “How might we help [specific user] accomplish [specific goal] given [specific constraint]?” That framing keeps the Ideate phase grounded in the research rather than drifting toward the team’s existing preferences.

Research in the Ideate phase

The Ideate phase generates potential solutions to the defined problem. Research activities here are supporting rather than central: they provide inputs that keep generated ideas grounded and informed.

Competitive analysis reviews how existing solutions in the market address the problem you have defined. Understanding what has been tried, what works, and where current solutions fall short generates informed starting points for ideation. See how to do competitive UX analysis for a structured approach to evaluating competitor experiences.

Stakeholder workshops bring cross-functional participants together for structured ideation. Researchers play a facilitating role in these sessions, grounding generated ideas in the user evidence from the Empathize phase. When ideation drifts toward the team’s preferences rather than user needs, the researcher is positioned to redirect by referencing specific observations from the research.

Analogous inspiration research examines how adjacent industries or domains solve similar problems. A healthcare team studying patient scheduling might look at how other complex appointment systems operate. A financial services team studying onboarding might study how other high-stakes, document-heavy processes work in different contexts. Solutions from unrelated domains often open up approaches that are not visible from within the target domain.

The Ideate phase ends with a prioritized set of ideas to carry into prototyping. Research informs the prioritization by connecting ideas to the frequency and severity of the user problems they address.

Research in the Prototype phase

The Prototype phase creates low-fidelity representations of proposed solutions that can be tested quickly with users. Research activities here focus on creating testable artifacts that are grounded in what was learned in the Empathize phase.

Concept validation happens before full prototype development. Showing early concepts, paper sketches, or static mockups to five users in informal sessions surfaces major directional issues before the team invests in building a more complete prototype. This is not formal usability testing. It is a quick directional check: does this concept make sense to users? Does it address the problem they actually have? A few hours of informal concept review can prevent weeks of work on a prototype that was headed in the wrong direction.

Prototype specification uses empathize-phase insights to make specific design decisions: the vocabulary users actually use (not the vocabulary the product team uses), task flows that match the mental models users revealed in interviews, information hierarchies that match user priorities rather than organizational priorities. Prototypes built without this grounding tend to reflect the team’s mental model rather than the user’s.

Research in the Test phase

The Test phase evaluates prototypes against real user behavior. This is evaluative research: you have a proposed solution and you are assessing whether it works. See what is evaluative research for the full methodological context.

Moderated usability testing is the core Test phase method. A researcher observes participants completing specific tasks on the prototype, noting where they succeed, where they struggle, and what the think-aloud protocol reveals about their reasoning. Moderated sessions allow real-time probing of unexpected behavior, which is particularly valuable for prototype testing where surprising participant responses may indicate deeper issues than surface usability failures. See what is moderated usability testing for a full explanation.

Unmoderated usability testing scales participant count when the prototype is sufficiently developed for participants to navigate independently. It is faster and less expensive than moderated testing and appropriate for validating specific flows at larger sample sizes. See what is unmoderated usability testing.

Five-second testing evaluates first impressions and visual hierarchy before participants engage with interactive flows. Showing a prototype screen for five seconds and asking what participants remember reveals whether the primary message and key actions register at first glance. See five second test guide.

Preference testing is useful when the team is evaluating two distinct design directions rather than testing a single prototype. Showing two variations to participants and measuring which better serves specific criteria generates evidence for design decisions that would otherwise default to team opinion. See how to do preference testing.

For Test phase recruitment, participants should match the actual target user profile for the product. In B2B contexts, this means recruiting by job function, company size, industry, and product usage rather than general demographics. CleverX’s professional panel with its screening and scheduling infrastructure is well suited for this type of specialized recruitment, handling participant sourcing, consent, and incentive payment so the research team can focus on running sessions rather than logistics. See how to recruit participants for product research for sourcing strategies.

How research evolves across iterations

Design thinking is iterative. Most projects cycle through these five phases multiple times as understanding deepens and solutions are refined. Research in each subsequent cycle is more focused than the previous one.

The first cycle’s Empathize phase is broad: you are building foundational understanding of a domain. The second cycle’s empathize work is narrower: you are investigating specific questions raised by the first cycle’s test findings. Third and subsequent cycles are focused even more precisely on the most critical unresolved assumptions.

The research toolkit stays consistent across cycles. What changes is the scope of the questions being asked. Broad exploratory research at the beginning narrows progressively into focused evaluative research as the solution takes shape. Good design thinking research programs have a clear sense of which questions remain open at any given cycle and design research activities specifically to close those questions.

Frequently asked questions

How much research is needed in a design thinking process?

The right amount is the minimum needed to reduce the risk of the decisions being made at each phase. Eight to twelve interviews in the Empathize phase typically builds sufficient understanding to define the problem for most scopes of work. Five to eight moderated sessions in the Test phase surfaces the major usability issues in a prototype. Research investment should scale with the stakes of the decision: a strategic product pivot warrants more research than a component-level UI decision. See how to calculate research sample size for sample size guidance across methods.

Can design thinking be applied to B2B product development?

Yes. The core premise applies equally to B2B and consumer contexts. B2B design thinking adds complexity because multiple stakeholder roles, including buyers, end users, administrators, and executives, have different needs that the product must serve. Research in B2B design thinking requires interviews and testing across these roles, not just with end users. Platforms with professional participant panels and role-based filtering make this multi-stakeholder research practical without requiring separate recruitment for each role.

What is the most commonly skipped research step in design thinking?

Synthesis in the Define phase is the most frequently underinvested step. Teams complete empathize-phase research and then jump directly to ideation without doing the analytical work to identify patterns, build journey maps, or formulate a problem statement grounded in evidence. The consequence is an ideation phase driven by individual team members’ interpretations of what they heard rather than shared evidence about what the research actually showed. The synthesis step is not optional; it is what makes the empathize-phase research usable for the rest of the process.