Tremoloo logo
English
English
Tremoloo logo
English
Person using a laptop to review a rewards program

Survey vs Interview vs Usability Test: Picking the Right Research Method

UX Research

UX research methods, survey vs interview, usability testing vs survey, research method selection, user research guide, when to use surveys

Summary

Each research method answers a different kind of question.

  • Surveys measure how many people think or do something, at scale, but only for things users can self-report.

  • Interviews explore why people behave as they do, revealing motivation and context, but with small samples.

  • Usability tests show where an interface fails by watching real behavior, but say nothing about market size.

The selection rule: match the method to the question. How many means survey; why means interview; where does it break means usability test.

Most research disappointments are method mismatches. A team surveys users to learn why they churn, or interviews five people to estimate market demand, and ends up with confident answers to questions the method cannot address. Surveys That Work spends real energy on this boundary, and the discipline generalizes.

What can each method actually tell you?

Method

Answers

Cannot answer

Scale

Survey

How many think or do X

Why, or what they actually do

Hundreds to thousands

Interview

Why people act, what they value

How many, or usability

Five to twenty

Usability test

Where the interface fails

Market size, or motivation

Five to eight per round

The classic error: surveys can only capture what users can self-report accurately. Users cannot reliably report why they did something, what they would do in a hypothetical future, or how a design made them fail. They can report what they did recently and what they prefer right now. Everything else is fiction with a response rate.

How do you choose in practice?

Start from the decision. “Should we build this feature?” needs why-users-struggle data: interviews. “Which of these two flows converts better?” needs behavior: usability tests or analytics. “How widespread is this pain?” needs scale: a survey, with questions built from the interview findings.

Mature research programs chain the methods: interviews to discover, surveys to size, tests to fix.

Why does Tremoloo share this book?

Because method selection is the first conversation in every research engagement we run, and this book gives clients the framework to have that conversation well. Teams that understand what each method can prove stop wasting budget on the wrong one.

Not sure which method your question needs? Our strategy team will help you design the right study.

Ready to Launch
Your Next Project?

If you’re tired of iterating in circles, we can help. We research,
design, and refine products with focused teams.