Sample size in qualitative research is one of those questions that makes experienced researchers wince slightly. Not because it's unanswerable, but because the honest answer is always "it depends" - and clients want a number.
The good news is that "it depends" is genuinely a useful starting point, because the factors it depends on are well-defined and worth working through carefully at the planning stage.
Why qual sample size is different to quant
In quantitative research, sample size is mathematically determined by the confidence level you need, the margin of error you can accept, and the variability in your population. It's a formula, and a statistician can give you a precise answer.
Qualitative research doesn't work that way. The goal isn't statistical representativeness - it's depth of understanding. You're not trying to prove that 47% of people feel a certain way. You're trying to understand why people feel the way they do, and what that means for the brand, product, or decision in question.
This means qual sample size is determined by different factors - primarily the complexity of the research question, the diversity of the target audience, and the type of methodology being used.
The key factors
What are you trying to discover?
A focused, specific research question - understanding the barriers to first purchase for a single product among a defined audience - typically requires fewer participants to reach saturation than a broad exploratory question about brand relationships across multiple segments. Clarity at the briefing stage directly shapes how many participants you need.
How diverse is your target audience?
If your research covers a single, well-defined audience segment, you need fewer participants. If you're researching across multiple age groups, geographies, usage levels, or life stages, each sub-group needs enough representation for patterns to emerge. This is where sample sizes can grow quickly - not because you need more people to hear the same thing, but because you need enough people across different groups to understand how experiences vary.
What methodology are you using?
Different methodologies operate at different scales:
- In-depth interviews: Typically 8-20 participants per audience segment, depending on diversity and complexity. Fewer is often fine when the question is focused and the audience is homogeneous.
- Focus groups: Usually 4-8 groups, with 6-10 participants each. Group dynamics mean you need enough sessions to see patterns emerge across different group compositions.
- Online communities and bulletin boards: These accommodate larger numbers - often 30-100 participants - because the async format and structured activities make it practical to manage and analyse at greater scale. Long-term always-on communities can run into hundreds of participants across ongoing waves of activity.
How long does the project run?
Longer fieldwork with repeated engagement naturally builds depth that can compensate for smaller initial numbers. A 14-day diary study with 30 participants across multiple activities may generate more useful insight than a single-session focus group with 50 people.
Practical approaches to setting sample size
Start large, then reduce
One effective approach is to begin with a larger quantitative screen - surveying a wide group to identify and segment participants - then reduce to a targeted qualitative cohort of around 30-40. This ensures your qual participants are genuinely the right people, rather than the first available.
Use saturation as your guide
In qualitative research, you've found enough participants when new conversations stop producing new insights - when you're hearing the same themes, the same barriers, the same language. This is called theoretical saturation. For many well-defined research questions and homogeneous audiences, saturation arrives earlier than researchers (and clients) typically expect.
The most important factor is the quality of insight you are obtaining. At each planning stage, consider what data you're gathering and what you'll actually do with it.
What about cost?
Sample size has a direct relationship with cost - but the relationship is more nuanced than "more participants equals more money." The activities participants complete, the level of moderation required, and the analysis intensity all affect cost at least as much as raw numbers.
High video transcription and AI analysis is more intensive - and more valuable - than a simple text survey. The right platform should make larger samples more manageable through automated processing, so cost per insight goes down even as participant numbers go up. Design your project so data-intensive elements like video activities are concentrated in the refined cohort, not the full sample.
The underlying principle
Sample size in qualitative research is an organic decision, not a formula. Start by being clear about what you're trying to understand. Consider who you need to hear from, and in what depth. Choose methodology that suits the question. And be honest with clients that the right number of participants is the number that produces reliable, saturated insight - not the largest number that fits the budget, or the smallest that seems defensible.
