UX Research

When to use

Use when a design decision depends on uncertain user needs, confusing flows, or an unexplained drop-off and appropriate evidence can be gathered.

When not to use

Do not present guesses as research findings. Do not recruit users, collect personal data, or run external studies without authorization.

Inputs and assumptions

Define the decision to make, target users, current evidence, constraints, privacy considerations, and whether research is exploratory or evaluative.

Workflow

1. Turn the uncertainty into a neutral research question.
2. Choose a proportionate method: support-ticket review, analytics, usability walkthrough, interview, or prototype test.
3. Define recruitment and consent needs; minimize collection of personal information.
4. Ask about recent behavior and concrete tasks rather than leading users toward a preferred solution.
5. Separate observations, participant statements, interpretation, and recommendations.
6. Identify sample limitations and confidence; do not generalize from a small or biased sample.
7. Connect findings to a testable design decision and follow-up measure.

Decision rules

• Research should change a decision; avoid research theater.
• Protect participant confidentiality and follow applicable policies.
• Report contradictory evidence and uncertainty.
• Use analytics as behavioral evidence, not as a substitute for user understanding.

Example

Instead of asking, “Would you use this new dashboard?”, ask the participant to find a specific overdue invoice in a prototype, then observe where they look and what they expect to happen.

Common failures

• Leading questions and hypothetical intent claims.
• Treating stakeholder opinion or one participant as universal truth.
• Recording sensitive information that is not needed.

Validation

Check that the research question maps to a decision, the method is proportionate, evidence is clearly distinguished from inference, and the recommendation includes limitations.