User research process: step-by-step framework for teams

A user research process is the sequence of steps a team follows to turn a business or product question into a decision backed by real evidence. Skipping or rushing any single step is usually where a study's findings stop being trustworthy, regardless of how good the interviews themselves were.

Step 1: Define the decision the research needs to inform

Before recruiting a single participant, get specific about what decision this research will actually change. A vague goal like understand our users produces a vague study. A specific goal like decide which of two onboarding flows to ship produces a study that can actually answer the question.

Step 2: Choose the right method for the stage

Generative, exploratory questions call for open-ended interviews. Evaluative questions about a specific prototype or feature call for usability testing. Choosing the wrong method for the stage is one of the most common process failures, and it happens before a single interview is conducted.

Step 3: Recruit the participants who actually matter

The right participants are the people whose behavior the decision depends on, not simply whoever is easiest to reach. A study recruited from the wrong segment can produce clean, confident-sounding findings that do not actually apply to the population the decision affects.

Step 4: Run the study with adaptive, not scripted, conversations

A rigid script gets through more questions. An adaptive conversation, one that follows up when a participant says something unexpected, gets to more truth. This is true whether the interview is run by a human moderator or an AI moderator trained to probe.

Step 5: Synthesize toward a recommendation, not a summary

The output of a good research process is a clear point of view on what the team should do next, not a stack of transcripts or a list of quotes. Synthesis should identify the themes that matter most to the original decision and state plainly what they imply.

Step 6: Close the loop

Document what was decided because of the research and share it back with the team, and informally with participants where possible. This is what turns a single study into a research practice a team trusts and keeps investing in.

Key takeaways

A user research process starts with a specific decision to inform, not a general topic to explore.
Matching the method, generative or evaluative, to the actual stage of the decision is one of the most common points of failure.
Adaptive, follow-up-driven conversations produce more useful findings than a rigid script, whether run by a human or an AI moderator.
The process ends with a clear recommendation and a closed loop, not a transcript archive.

Frequently asked questions

What are the main steps in a user research process?

Define the decision the research needs to inform, choose the right method for the stage, recruit the participants who matter, run adaptive rather than scripted interviews, synthesize toward a recommendation, and close the loop by documenting what changed.

What is the most common mistake in a user research process?

Starting with a vague goal instead of a specific decision to inform. A vague goal produces a study that cannot clearly answer any one question, however well the interviews themselves are run.

How long should a user research study take?

It depends on the method and scale, but AI-moderated interviews running in parallel can typically compress a study that would take weeks with a single human moderator down to 24 to 48 hours of fieldwork.

Who should be involved in defining the research question?

The people who will actually use the findings to make a decision, whether that is a product manager, a designer, or a research lead, should help define the question before the study is designed.

Generative vs evaluative research: what UX teams get wrong
User interview questions: best examples for UX research
What is data saturation in qualitative research?