How many research participants do you need?

Your sample size affects the quality of your findings and the time and budget needed to complete the research—so getting it right is a priority. Recruit too few participants, and you may miss important patterns. Recruit too many, and you can end up collecting more data than your study needs.

In this guide, we look at how participant numbers vary across common UX research methods, what to consider before choosing a sample size, and how to recruit the right participants for your research goals. We also include a handy sample size calculator to help you make sure you’re recruiting the right number of users for your study.

The number of research participants you need depends on the question your study looks to answer.

Qualitative research helps you understand user behavior in context. It helps you explore how people complete tasks, what they expect, and why they make certain decisions. Since each session can include detailed feedback, observations, recordings, and follow-up questions, you usually need fewer participants than you would for quantitative research, but enough to see where the same issues, expectations, or behaviors appear more than once.

Quantitative research measures user behavior or attitudes at scale. It helps you answer questions about frequency, performance, preference, or confidence. For example, you might want to know how many users completed a task, which version performed better, or how satisfied users were after using a feature. Since the goal is to measure trends across a larger group, quantitative studies usually need a larger sample size.

Qualitative and quantitative research need different sample sizes because they answer different kinds of questions. Usability testing sits right in the middle of that conversation. Some usability tests are qualitative, helping teams find where users struggle and why. Others are quantitative, measuring task success, time on task, error rates, or completion rates across a larger group.

That’s why blanket sample-size rules can be misleading.

The five-user rule comes from Jakob Nielsen’s guidance at Nielsen Norman Group. It says that testing with five users can uncover around 85% of usability issues in the first round of a qualitative usability test. But that finding depends on context. This means that if you’re observing a relatively similar group of users completing the same tasks in one design, prototype, or flow, you can use what you learn to improve the experience and test again.

But the five-user rule isn’t a universal sample size recommendation. It’s mainly useful for early, formative usability testing when you’re trying to identify obvious issues in a single design, prototype, or flow.

In practice, testing with more than five users should depend on your study goal, product risk, audience complexity, and how specific your participant recruitment criteria need to be.

Here’s when you should consider using more testers to uncover key insights:

The five-user rule is useful, but only in a narrow usability testing context. For most research, sample size depends on the research method, question, and the level of confidence you need.

Next, let’s look at how many participants you may need across common research methods.

There’s no single sample size that works for every study. The right number of participants depends on your research method, the audience you’re studying, and how complex or high-stakes the research is.

Based on our experience at Maze, however, there are useful starting points you can use when planning user research. These aren’t fixed rules, but can be considered as benchmarks you can scale up or down depending on the scope, risk, and goals of your study.

Here’s a good starting sample size for each research method, along with when you may need to scale that number up or down:

Stakeholder interviews help you understand the priorities, constraints, assumptions, and expectations of the people involved in or affected by a project. They’re usually conducted early in the discovery stage.

For a standard round, plan for 5–10 key stakeholders across the relevant teams. The goal is to only include the people who can provide different perspectives on the problem, users, business goals, and project constraints.

You can scale down when the organization is small and stakeholders are already well aligned. For a simple project or a small team, 3–5 stakeholder interviews may be enough to capture the main priorities and requirements.

You’ll need more interviews when the organization is more complex. Plan for 10–15+ stakeholders if the project spans several departments or teams have different or conflicting priorities.

For in-depth user interviews, plan for 5-8 participants for a standard, focused round. This can work well when you’re speaking to a relatively consistent user group and exploring a specific set of needs, behaviors, or experiences.

Increase the sample to 10–20 participants when the study is more strategic or exploratory, or when your audience is broader or more varied. For example, you may need more interviews if you’re speaking to people across different roles, markets, experience levels, customer segments, or use cases.

For very broad or varied audiences, you may need to scale further to 20–30 participants. In these cases, the larger sample gives you more opportunity to understand how experiences differ across the audience rather than assuming insights from one group apply to everyone.

User surveys can help you gather qualitative insights from users like their needs, motivations, and behaviors. The number of participants needed for a survey depends on project complexity, which reflects the potential impact of UX design decisions on the user experience and business outcomes.

For a standard, medium-risk directional survey, start with around 50 respondents. This can work when you need a useful read on patterns or preferences without treating the findings as statistically representative of a much larger audience.

For a low-risk quick-pulse survey, as few as 20 respondents can be enough to get an early directional signal before deciding whether more research is needed.

Aim for 100–200+ respondents when you need more generalizable results for a broader or more strategic decision. If you’re comparing different user segments, plan the sample separately for each segment so that every group has enough responses to analyze meaningfully.

For research across a broad audience, 300–400+ respondents may be appropriate when you need a specific confidence level and margin of error. In these cases, rather than relying on a fixed rule of thumb, use a sample size calculator that accounts for your population size, confidence level, and acceptable margin of error.

A focus group is a moderated group discussion used to explore participants’ reactions, attitudes, opinions, and language around a topic.

For focus groups, plan for 5–8 participants per group. That’s usually small enough for each person to contribute effectively, but large enough to create discussion between participants.

You can scale up slightly when the topic is relatively simple, and more voices are likely to add value. In these cases, up to 10 participants per group can work well.

If you’re researching multiple audience segments, you’ll usually need more than one group. In such a case, plan for around 3–6 focus groups, or roughly 18–60 participants in total, depending on how many segments you need to cover and whether each group needs to be analyzed separately.

You may also want to scale down for complex or sensitive topics, or discussions that may be heavily influenced by group dynamics. Here, smaller groups of 4–5 participants can work better. This makes it easy for each participant to explain their views and for the moderator to explore important points in more depth.

Diary studies track user behavior, routines, and experiences over time, so participants submit multiple entries across days, weeks, or months. That’s why diary studies usually work best with a smaller, focused sample. As a starting point, plan for 10 participants. This is often enough to identify patterns over time.

You can also scale down for narrower research. For a short, simple, low-risk, or tactical diary study, as few as 5 participants can work when you only need an early directional read.

You may need a larger sample when the research is broader or more strategic. Aim for 15–20 participants when you’re exploring a wider range of behaviors, comparing different types of users, or running a longer study where participant drop-off is more likely.

If you’re comparing distinct audience segments, plan for around 5–10 participants per segment. This gives each group enough representation to compare patterns without combining very different user experiences into one sample.

First-click testing helps you understand where users click first when trying to complete a task. It’s especially useful for checking whether navigation, labels, and page structure point users in the right direction before you invest in deeper usability testing.

For a standard directional first-click test, start with around 15–30 participants. This is usually enough to identify recurring patterns in where people click first and spot obvious navigation or information architecture problems.

Aim for around 50–100 participants when you’re testing a major redesign, a conversion-critical flow, or an audience that spans different devices, experience levels, or user segments. A larger sample gives you more confidence that the patterns you see aren’t limited to one small group.

Card sorting is a generative UX research method used to develop or evaluate the information architecture of a product or site. Unlike usability testing, card sorting requires a larger number of users to achieve reliable results due to the variability in how people categorize information.

For a standard card sort, start with around 20 participants. This gives you enough responses to identify recurring grouping patterns and see where participants agree or differ in how they organize the content.

You can also scale down when you only need an early, directional read. Around 10–15 participants can work for a tactical or low-risk card sort where the goal is to explore how people categorize content before making more significant information architecture decisions.

You’ll need a larger sample when the results will influence navigation, taxonomy, or content hierarchy more broadly. For a strategic card sort, or when you need a lower margin of error, plan for around 30–150 participants. The right number within that range depends on how much confidence you need in the patterns you find.

Often done after card sorting, tree testing helps evaluate the navigability of a system structure.

For a simple tree test, 50 participants can work as a minimum starting point. This can help you spot whether users are choosing the right paths, where they hesitate, and which labels cause confusion.

You can scale down when you only need an early directional read. Around 15–30 participants can work when you’re testing an initial structure before moving into deeper validation. At this stage, the goal is to spot obvious problems rather than make a high-confidence information architecture decision.

You’ll need a larger or more carefully divided sample when the study is complex. If you’re comparing different tree versions or testing multiple user segments, plan for around 30 participants per segment or version. This ensures each group has enough responses to analyze independently.

For a high-stakes information architecture decision, you may also want to increase the overall sample beyond the standard 50 participants. A larger sample can give you more confidence in the patterns you see before making significant changes to navigation or content structure.

Concept testing helps you understand how users respond to a product, design, or marketing idea before you invest significant time and resources into developing it. It can be used early in discovery or later when you need to validate a particular direction before moving forward.

For a standard concept check, start with around 5 participants. This can be enough to understand whether people grasp the idea, how they react to it, and what questions or concerns come up repeatedly.

You can scale down when the concept is still rough and the decision carries little risk. For an early, low-risk, or tactical gut-check, as few as 2–3 participants can work. At this stage, the goal is to identify obvious problems or reactions before spending more time refining the idea.

For a strategic or higher-stakes concept decision, plan for around 8–10 participants. A larger group gives you more perspectives before deciding whether to develop, change, or discard the concept. If you’re running a product concept test, this can be particularly useful before committing significant resources to a proposed solution.

Co-design, or collaborative design, brings users or other relevant participants into the design process to help shape a solution. Participants might work with the research or product team to explore ideas, map out an experience, prioritize needs, or develop possible solutions together.

For a standard collaborative design session, start with around 3–5 participants. Co-design is hands-on, so a small group gives everyone enough opportunity to contribute, discuss ideas, and work directly with the team.

You may want to scale up when the project is more strategic or exploratory and would benefit from a wider range of perspectives. In these cases, you can include up to around 8 participants. Beyond that, collaborative sessions can become harder to manage and give each person less opportunity to contribute meaningfully.

For an unmoderated usability test of a mid-to-high-fidelity prototype, a good starting point is 10 participants. This works well when you’re testing one user group, one prototype, or one flow and want enough responses to identify recurring usability issues.

Remember, the number of participants you need for every study changes depending on how the test is run, the stage of product development, and what you need to learn.

Moderated and unmoderated usability tests also have different sample size needs. In a moderated test, the researcher can ask follow-up questions, observe behavior in real time, and dig into why a participant struggled. So, 5–10 participants can often be enough for a moderated round.

You can go even smaller when you’re testing an early, low-fidelity prototype. In this case, 2–5 participants can give you enough feedback to identify obvious problems before investing more time in the design.

The same applies when you’re running rapid, iterative rounds of prototype testing. Around 3–5 participants per round can work well because the goal is to identify usability issues, make improvements, and test again with a new group.

You may need 20+ participants when testing a finished or live product, comparing multiple design variants, or working with multiple user segments. Larger samples can also be useful for strategic or exploratory usability studies where you need to understand a wider range of behaviors and experiences.

Content testing helps you understand whether the words, information, and other content within a product or website are clear, useful, and easy for users to understand.

For a standard qualitative content test, start with around 10 participants per round. This gives you enough feedback to spot recurring issues with clarity or comprehension without making the study unnecessarily large.

The right number can change depending on what you’re testing. There are several content testing methods, including task-based usability testing, cloze tests, card sorting, highlighting tests, and five-second tests. If you’re testing multiple content variations or comparing different user segments, increase the sample to around 15–20 participants so you have enough feedback to assess the differences between them.

For preference testing across a broader audience, plan for 30+ participants. At that point, the goal is less about exploring why individual users respond a certain way and more about identifying patterns across a larger group.

For live website testing, start with around 20–30 real visitor sessions. This can be enough to spot recurring navigation or path patterns, like where visitors repeatedly take an unexpected route or struggle to reach an important part of the site.

You’ll need more data when you want to move beyond identifying patterns and make statistically reliable comparisons. Aim for 100+ sessions when you need stronger behavioral evidence across your live audience or want to compare performance before and after a change.

Continuous feedback works differently from a one-time research study. There’s no fixed number of participants to recruit upfront because feedback is collected as an ongoing stream while users interact with your product. You ideally want to identify patterns that emerge over time rather than reach a single sample-size target.

At lower volumes, individual responses can still highlight useful problems, requests, or opportunities. But if you want to identify more reliable trends or compare feedback across user segments, look for a steady 50–100+ responses within a defined period, such as a month.

Unlike methods focused on user behavior and preferences, A/B testing relies on quantitative data to inform decisions. With A/B testing, your design team presents two variations of a design or prototype to measure which one resonates better with users. There are also several testing methods to choose from for A/B tests.

For a standard A/B test, 100+ conversions per variant is a useful rough minimum. This can give you enough data to start evaluating whether one version performs differently from another, although the exact sample you need will depend on your baseline conversion rate.

If you expect a smaller uplift, plan for 300+ conversions per variant. Smaller differences are harder to detect reliably, so you need more data to separate a genuine improvement from normal variation.

The sample size increases further as experiments become more complex. Multivariate tests can require thousands of users across all combinations because traffic is divided between several variations rather than just an A and B version.

With Maze, you can run variant comparison and create custom expected paths for each version using testing templates. You can also directly test a prototype for multiple variables, then simply share the test with recruited users via a link, collect their answers, and get an automated report with success rates and alternative paths.

While we've explored the need for more testers in specific user testing scenarios, let's take a look at some general factors that can help you decide how many testers to recruit for well-rounded, inclusive UX research:

The ranges above are a useful starting point, but your final sample size should reflect your study setup. Maze’s sample size calculator helps you estimate how many participants you need based on your methodology, population size, and confidence level.

Use it to check your ideal participant count before recruitment, align stakeholders on the sample size, and plan whether you’ll recruit from your own users, an external panel, or both.

Then, once you know the number of participants you need, Maze can help you recruit them.

Once you know how many participants you need, Maze helps you recruit the right audience and hit your target sample size without managing the process across separate tools.

Once you know how many participants you need, the Maze panel gives you access to millions of vetted B2B and B2C participants across 150+ countries, using 400+ filters to target by demographics, work, lifestyle, and more.

With in-product prompts, you can collect feedback or recruit participants while they’re actively using your product, with targeting based on specific URLs or cohorts through integrations like Amplitude.

Maze helps you recruit the right participants for your study.