UX research methods, step by step

Plan

• Write a project briefAgree what the project is trying to learn and for whom.
• Frame a decisionWrite down what you are deciding, what you believe and what would change your mind, before the research starts.
• Frame research questionsTurn the brief or decision into open questions that research can answer.
• Write an interview scriptPlan the conversation: warm-up, core questions, wrap-up.
• Write a usability test scriptPlan tasks for participants to attempt with the prototype.
• Build a surveyTurn findings into a survey that checks how common they are, or a brief or decision into one that explores.

Research

• Run an interviewTalk to one participant and capture what they said and did.
• Compare competitorsCompare existing products feature by feature, from their own websites.
• Scan the webFind what peer-reviewed studies, official statistics and other sources already say about your questions.

Synthesize

• Affinity mappingGroup observations from several sources into themes.
• Synthesize insightsCombine themes and data into a few statements you can act on.
• Build a personaDescribe a representative user, grounded in research.
• Map a journeyLay out what a persona does, step by step, and how it feels.
• Write jobs to be doneDescribe the progress people are trying to make, and what they hire to make it today.
• Map empathyLay out what someone says, thinks, does and feels, from the evidence.
• Map opportunitiesLink one outcome to the opportunities that could move it, and the ideas for each.
• Find contradictionsFind where the evidence disagrees: between people, between what someone says and does, and between the numbers and the talk.
• Code the dataTag every excerpt with a code and count how often each comes up, across all your sources.
• Ask a persona panelPut one question to several personas and see their answers side by side.

Ideate

• Write how-might-we statementsReframe pain points as open questions to design for.
• Ideate conceptsGenerate ideas for a how-might-we and shape the best into a concept.
• Map assumptionsFind what has to be true for the concept to work, and what your research already says about it.
• Run a pre-mortemImagine the concept failed a year from now, and work out why.
• Write a scenarioTell a concrete story about the persona, grounded in real quotes, for workshops and storyboards.
• Vary the conceptPush one concept in several directions and get a card for each variant.
• Score conceptsScore concepts against the research, with a cited reason for every score.

Prototype & test

• Build a prototypePlan the lowest-fidelity prototype that tests the concept, build it in your prototyping tool, and link it back.
• Run usability testsWatch people use the prototype and report what happened.
• Write a handoff specDescribe what to build, informed by what testing showed.
• Review as stakeholdersHear the concerns engineering, legal, support or a sceptical exec would raise, before the real review.
• Write copy optionsWrite names, value propositions or microcopy in the words participants actually use.

Share

• Write a readoutTurn findings into a deck for people who weren’t in the research, with every claim traced to its source.

Checks

• Audit the evidenceCheck every claim in an artifact against the research behind it.
• Find coverage gapsSee which research questions the evidence answers, which it barely touches, and who is missing from the sample.
• Check for biasCatch leading questions before research runs, and confirmation bias after it.
• Check what new evidence changesSee which insights and personas new evidence strengthens, weakens or contradicts.

Check that something works

• Evaluative studyDoes this design work for the people who will use it? 1–2 weeks.
• Assumption-led (Lean UX)You already have an idea. What could sink it? 1–2 weeks.
• Design SprintFive days from a big question to a prototype tested with real people. 5 days.

Understand people and the problem

• Exploratory studyWe don’t know the problem well yet: why do people do what they do? 2–4 weeks.
• Jobs to be doneWhy do people start using, switch to or leave a product like yours? 2–3 weeks.

Run a whole project

• Double DiamondA whole project, from understanding a problem to a tested answer. 6–10 weeks.
• Design ThinkingEmpathize, define, ideate, prototype and test, in the classic five modes. 4–6 weeks.

Keep learning every week

• Continuous discoveryA product team learning every week towards one outcome. every week.