One of the most common questions when planning a market research project is how many people you actually need to speak to. But the right number depends on what you’re trying to find out, who you need to speak to, and how you intend to use the results.
For example, a survey designed to understand the views of UK consumers will have very different requirements from a series of interviews with senior decision-makers in a specialist industry. While sample size matters, good market research is ultimately about speaking to enough of the right people to answer your research questions with an appropriate level of confidence.
So, how do you decide what that looks like?
How many people do you need for quantitative market research?
Quantitative research is designed to measure opinions, behaviours, or attitudes across a larger group of people, usually through surveys. Because the findings are often expressed as percentages and used to compare different groups, sample size is particularly important.
For a broad consumer survey, a sample of around 1,000 respondents is commonly used to provide a robust overall picture of a national population. However, that doesn’t mean every quantitative research project needs 1,000 people.
A smaller sample may be perfectly appropriate if you’re researching a more specific audience, working with an existing customer database, or looking for directional rather than nationally representative findings. Equally, you may need a much larger sample if you want to compare lots of different groups within the data.
For example, a sample of 500 people might provide a useful overall result, but if you then want to understand differences across five customer groups, the number of people within each group becomes much smaller. This is why it’s important to think about the analysis you want to conduct before deciding on the overall sample size.
Is a bigger market research sample always better?
Not necessarily. Increasing your sample can improve statistical confidence, but only if you’re speaking to the right people in the first place.
A survey of 5,000 poorly targeted respondents won’t necessarily tell you more than a carefully designed survey of 500 people who accurately represent the audience you’re trying to understand.
There is also a point at which increasing the sample delivers diminishing returns. Depending on the research question, the additional confidence gained by moving from 1,000 to 2,000 respondents may not justify doubling the fieldwork cost.
A good research design therefore balances statistical reliability, the level of analysis required, and the available budget rather than simply aiming for the largest possible sample.
How many people do you need for qualitative research?
Qualitative research works differently because the objective isn’t to measure how many people think something; it’s to understand why they think or behave in a particular way.
This means qualitative studies typically involve far fewer participants. Depending on the project, this might mean conducting 10–20 depth interviews, several focus groups or a combination of different qualitative methods.
The exact number depends heavily on the audience and the range of perspectives you need to explore. If you’re researching a relatively similar group of customers, you may start hearing the same themes after a fairly small number of conversations. If you’re exploring several different customer types, markets, or stakeholder groups, you’ll usually need more participants to make sure those different perspectives are represented.
The aim isn’t statistical representation; it’s to reach enough depth and diversity to understand the themes that matter.
What about focus groups?
A typical focus group might include around six to eight participants, although the ideal number will depend on the subject and the type of discussion you want to encourage.
Most projects involve more than one group because relying on a single discussion makes it difficult to know whether an opinion reflects a broader theme or simply the dynamics of that particular group.
You might therefore run separate groups for different audiences, such as existing customers and prospective customers, or different age groups, markets or customer types. As with other forms of qualitative research, the number of groups should be determined by the range of perspectives you need to explore rather than by a fixed rule.
How many people do you need for B2B market research?
B2B research often requires a different approach because the potential audience can be much smaller and harder to reach.
If you’re researching IT decision-makers at large UK businesses, specialist healthcare professionals or senior procurement directors, for example, there may simply not be thousands of suitable respondents available. Reaching 100 highly relevant decision-makers could be far more valuable than reaching 1,000 people with only limited involvement in the decision you’re researching.
The same applies to qualitative B2B research, where a relatively small number of well-recruited interviews can provide extremely valuable insight, particularly when participants have specialist knowledge or significant influence over purchasing decisions.
For niche audiences, participant quality and relevance should generally take priority over achieving an arbitrary sample size.
Do you need to speak to existing customers or the wider market?
Sample size isn’t the only consideration; you also need to think carefully about who makes up your sample.
Your existing customer database can be an excellent source of research participants, particularly if you’re trying to improve customer experience, understand satisfaction or develop your existing offer. However, existing customers can only tell you about people who have already chosen your business.
If you’re trying to understand your growth opportunity, test a new proposition or explore why people choose competitors, you may also need to include prospective customers, lapsed customers or people who have never considered your brand.
In some cases, the best approach is to combine your own customer data with an independently recruited external sample, allowing you to understand both your current customers and the wider market.
What determines the right market research sample size?
There isn’t one sample size that works for every project. When deciding how many participants you need, some of the most important considerations are:
- Your research objective: What decision will the findings be used to make?
- Your target audience: How large, diverse and accessible is the group you need to understand?
- The type of research: Quantitative and qualitative studies have very different sample requirements.
- The analysis required: Do you need an overall answer, or do you want to compare different customer groups?
- The level of confidence needed: A major investment decision may require more robust evidence than an early-stage exploration.
- Your budget: Larger samples cost more, so the additional insight needs to justify the investment.
These decisions are best made together because changing one element can affect the others. For example, reducing a quantitative sample may save money but could leave too few respondents to make meaningful comparisons between important customer groups.
How do you know if your sample is large enough?
The best way to answer this is to work backwards from the decisions you want to make with the research.
If you need to measure something accurately across a large population, your sample needs to be sufficiently robust to provide confidence in the findings. If you need to understand the motivations of a very specific group of people, a much smaller but carefully recruited sample may be more appropriate.
This is why determining sample size should be part of the research design process rather than a number decided in isolation at the beginning of a project.
Frequently asked questions
Is 100 people enough for market research?
It can be. A sample of 100 may be appropriate for research among a niche audience or for generating directional quantitative findings, but it is unlikely to provide enough depth for extensive subgroup analysis. Whether it’s sufficient depends on the audience, objectives and how the findings will be used.
Is 1,000 respondents enough for a survey?
For many general consumer studies, 1,000 respondents can provide a robust overall sample and is commonly used for national research. However, you may need more respondents if you want reliable analysis across numerous subgroups or particularly small sections of the population.
How many interviews are enough for qualitative research?
There is no fixed number, but many qualitative projects involve around 10–20 depth interviews. The appropriate number depends on the diversity of the audience and whether additional interviews continue to uncover meaningfully different perspectives.
How many focus groups should you run?
Most research projects involve multiple focus groups rather than relying on one group alone. The number required will depend on how many different audiences or perspectives need to be represented, with separate groups often used to explore differences between important customer types.
Getting the sample right
There is no exact number when it comes to market research sample size. A larger sample can provide greater statistical confidence, but relevance, representation and research design are just as important.
The right approach is to start with the business question you need to answer, understand who can answer it, and then design a sample that gives you the right level of evidence to make that decision confidently. In some cases that might mean surveying thousands of consumers; in others, a dozen carefully selected conversations could provide exactly the insight you need.