What sample size do you need for qualitative research?
A practical way to think about qualitative sample size today
“How many people do we need?” is one of the first questions clients ask about qualitative research.
It sounds like a question that should have a precise answer. In reality, it never has.
Historically, qualitative sample size was about *coverage*, not confidence — enough interviews to ensure all the important themes had surfaced, before moving on to a quantitative study.
Today, that logic no longer fully applies.
With well-designed, guided qualitative surveys, it’s now possible to collect high-quality open-ended insight at scale — and use qualitative research directly to support decisions.
That changes how sample size should be thought about.
There is no single “right” number.
Instead, the right sample depends on what decisions you need to make, how granular you want the output to be, and how confident you need to feel.
Here’s how to think about it.
1. How qualitative sample size used to work
Historically, qualitative research had a very specific role. It was not designed to support decisions on its own. Instead, it existed to *inform* quantitative research.
The objective was simple: Speak to enough people until all the important themes and ideas have surfaced.
Once those themes appeared, the qualitative phase was effectively complete. Those insights would then be:
- Turned into statements
- Converted into closed questions
- Tested properly in a large quantitative study
This idea is often referred to as **thematic saturation**.
In practice, that meant:
- 6–8 depth interviews per audience
- Or 2–3 focus groups
- Or “until nothing new comes up”
At that point, more interviews were seen as unnecessary.
Why? Because qualitative research was never meant to stand alone. It was a stepping stone, not the final evidence.
2. The traditional qual–quant model updated
The traditional approach relies on three assumptions:
1. Qualitative data being difficult to scale
2. Open-ended responses being hard to analyse consistently
3. Decision-makers needing numbers to feel confident
Today, all three assumptions are outdated. Guided qualitative surveys change the economics and usefulness of open-ended data completely.
When:
- Questions are carefully structured
- Respondents are guided to give meaningful answers
- Analysis is systematic rather than anecdotal
Qualitative research stops being purely exploratory.
It becomes **decision-grade**. And that fundamentally changes how sample size should be approached.
3. Qualitative research can now support decisions directly
With guided surveys, qualitative research can now:
- Show how *many* people express a view
- Explain *why* they hold it
- Capture nuance at scale
- Compare responses across segments
The old boundary between qualitative and quantitative research becomes much less rigid. You’re no longer just asking:
> “What themes exist?”
You’re asking:
> “Which themes matter most, to whom, and in what context?”
That requires a different way of thinking about sample size.
4. What does sample size depend on now?
There is still no single “right” answer. But there *are* clear factors that influence what makes sense.
A. How granular you want the results to be
This is usually the biggest driver. Ask yourself:
- Do you need a high-level directional answer?
- Or do you want to compare sub-groups with confidence?
For example:
For overall understanding your sample size might be 100–150.
If you want to compare two user groups it might be 200–300.
Or if you have multiple segments or audiences the total sample size might be 400+
The more cuts you want to make, the more respondents you need.
B. How diverse the audience is
Homogeneous audiences reach clarity quickly. Fragmented markets do not.
If people:
- Use the product in very different ways
- Have different needs or motivations
- Sit at different stages of the journey
You’ll want more respondents to ensure important minority views are properly represented.
C. How confident you need to be in the decision
Some decisions are directional. Others carry real commercial risk.
Examples include:
- Launching a new pack design
- Changing a brand position
- Dropping or reformulating a product
The higher the risk, the more confidence stakeholders usually want.
Today, that confidence often comes from:
- Volume of verbatim evidence
- Consistency of patterns
- Clear signals repeated across many respondents
Not just percentages.
5. The cost question
Historically, increasing qualitative sample size was expensive. More interviews meant:
- More moderation time
- More transcripts
- More manual analysis
So samples stayed small. With guided qualitative surveys, that cost curve changes dramatically.
Once the survey is live:
- Adding respondents is inexpensive
- Fieldwork scales easily
- Analysis effort grows much more slowly than sample size
That’s why most clients now choose **a few hundred respondents** as standard. Not because they have to. But because they can.
6. The bottom line
There has never been a single “right” sample size for qualitative research.
Historically:
- Qualitative work was small
- Exploratory
- Designed to inform quantitative studies
Today:
- Qualitative research can stand on its own
- Scale increases confidence, not noise
- Sample size is a design choice, not a constraint
The best answer remains:
> **Enough people to support the decision you need to make.**
The difference now is that “enough” no longer has to mean “small”.
7. Finding out more
If you would like to discuss sample size for a particular project, please just get in touch.
Or email directly: richard[at]mrqual.com
Read more articles