Conduct a User Research Study

Use this when a product decision needs evidence from the people who do the work. Each step produces an artifact the next one can use, from the research question through the stakeholder readout.

Use this

  • A product decision needs evidence from the people who do the work
  • You do not yet know the jobs, language, or constraints well enough to spec
  • Stakeholders keep arguing from opinion and you need quotes with coverage

Skip this

  • You already have a flow and need to see if people can finish a task. Use usability testing
  • You need counts from many people on questions you already know. Use a survey
  • You have one week to pick a concept and prototype. That is a design sprint

Steps

Each step hands the next one a concrete artifact, so the readout can point back to what people said.

02Recruitment Planning

04Conduct Interviews

Run the sessions. After each one, turn the transcript into a debrief so synthesis is clustering evidence, not reconstructing memory.

This is time with participants. Record with permission and capture the lines you will want later.

05Research Synthesis

Frequently asked questions

I already have research questions. Do I start over?

Keep them. Go to recruitment if you still need participants, or to the interview guide if people are already booked.

Does this have to take weeks?

A narrow question with people already lined up can be five interviews and a synthesis the next morning. Give synthesis its own block of time so the readout has themes, not only notes.

How do I keep the study together across ChatGPT or Claude threads?

Put the research objectives at the top of every new thread, then the last step’s output. The model stays on the same study.

When should I use usability testing instead?

Use this workflow while you are still forming the problem. Use usability testing when you have a flow and need to see whether people can complete a task.