How to Choose the Right Sample Size in UX Research

By Philip Burgess – UX Research Leader

In UX research, one of the most common—and often misunderstood—questions is:

"How many participants do I need?"

The answer depends on your research goals, methodology, and whether you’re aiming for qualitative insight or quantitative confidence. Choosing the wrong sample size can either waste time and budget or, worse, lead to misleading conclusions.

This guide breaks down how to determine the right sample size for different UX research methods, when to prioritize depth over breadth, and the key trade-offs to consider.

1. The Purpose Drives the Number

The first step is clarifying why you’re running the study.

• Exploratory (understanding problems, generating hypotheses): smaller, targeted samples.

• Evaluative (testing solutions, validating changes): larger, more representative samples.

2. Qualitative Research Sample Size

Goal: Depth, context, and understanding the “why” behind behaviors.

Rule of Thumb

• Usability testing: 5 participants per iteration can uncover ~80% of the most critical issues (Nielsen Norman Group).

• Interviews: 5–20 participants, depending on complexity and audience diversity.

• Diary studies / contextual inquiry: 5–15 participants, focusing on rich, longitudinal data.

Why Smaller Works

Qualitative research looks for patterns (thematic saturation), not statistical generalization. Once you stop hearing new insights, you’ve reached saturation.

Risks of Too Few

If you have highly diverse user groups, five participants may miss unique needs—run separate small tests for each segment.

3. Quantitative Research Sample Size

Goal: Measure and generalize results with statistical confidence.

Key Factors

• Desired confidence level (commonly 95%)

• Margin of error (often 5% for surveys)

• Population size (if finite and known)

• Expected effect size (smaller differences need larger samples)

Examples

• Surveys:

  • Large population: 385 responses = ~95% confidence, ±5% margin of error.

  • Smaller populations require adjusted calculations.

• A/B tests:

  • Often require hundreds to thousands of participants to detect meaningful changes.

  • Use online calculators to determine the needed sample size based on baseline metrics and desired improvement.

Risks of Too Few

Underpowered studies may miss true effects or produce results that look significant but aren’t reliable.

4. Mixed-Methods Research

Combining qualitative and quantitative often means running:

• Small qualitative sessions first to explore the “why.”

• Larger quantitative surveys or tests to validate at scale.

This approach ensures both insight and confidence.

5. Method-Specific Guidance

Method	Typical Sample Size	Notes
Usability Testing (Qual)	5–10 per user segment	Run multiple rounds for iteration
Interviews	5–20	Stop at thematic saturation
Surveys (Quant)	100+ (ideally 385+)	Depends on confidence/margin of error
A/B Testing	200+ per variant (often more)	Based on baseline & expected lift
Card Sorting	15–30	Enough to see clear grouping patterns
Tree Testing	20–50	Balance between confidence & cost
Diary Studies	5–15	Manageable for analysis
Eye Tracking	20–30	Identify clear attention patterns

6. Budget, Time, and Practical Constraints

• Recruitment cost: The more specific your audience, the harder (and more expensive) they are to recruit.

• Researcher capacity: Large datasets require more time to analyze.

• Iteration cycles: Smaller, faster tests can lead to better outcomes than one large study.

7. The “Goldilocks” Approach

Instead of aiming for the largest possible sample, aim for the right sample:

• Big enough to answer the question with confidence.

• Small enough to stay within budget and timeline.

In many cases, you’re better off running several small, focused studies than a single massive one.

Final Thoughts

Choosing the right sample size in UX research isn’t about blindly following a magic number—it’s about aligning your sample size with your research objectives, methodology, and constraints. Start with clear goals, select the method, then use established guidelines and calculators to land on a number that gives you credible, actionable insights.