What Is Saturation in Qualitative Research? A Guide
Saturation is the point where new interviews stop teaching you anything. What it means, why you cannot predict it, and how to run a study toward it.
By GhostClick / Method / Published 16 Jul 2026 / 6 min read
Saturation is the moment in qualitative research when new interviews stop teaching you anything new. It is the only principled answer to “how many participants is enough?”, and understanding it properly is what separates a study that stops for the right reason from one that stops because the calendar ran out. The most cited experiment on it, Guest, Bunce and Johnson (2006), coded 60 interviews and found that saturation occurred within the first twelve, with the basic thematic structure visible after just six.
This guide explains what saturation actually means, why it cannot be predicted precisely in advance, how to run a study toward it in practice, and what changes the number. It is the concept underneath our guide to how many participants qualitative research needs.
What saturation actually means
The Nielsen Norman Group defines it as the point where “conducting more interviews won’t provide new insights that would alter your themes.” Guest and colleagues put it as the point at which “no new information or themes are observed in the data.” Both descriptions point at the same thing: you keep interviewing until you start hearing the same ideas come back, and then a little longer to be sure.
Saturation is not a fixed number of people. It is the point where the next interview stops surprising you.
Why you cannot set it in advance
This is the part teams get wrong. Guest and colleagues are explicit that saturation “provides little practical guidance for estimating sample sizes, prior to data collection.” You cannot know the number before you start, because it depends on how varied your participants are and how broad your question is. That is a feature, not a flaw: it forces you to watch the data rather than commit to an arbitrary target and stop there whether or not you have learned enough.
How to run a study toward saturation
The practical method NN/G recommends is to start small and analyse as you go, rather than booking a fixed number of sessions up front.
- Start with about five or six representative participants. That is enough to see the basic themes take shape, as Guest and colleagues found.
- Analyse between sessions, not just at the end. Saturation is something you notice while collecting data, so code as you go and track which themes are new.
- Keep adding interviews until the new ones stop adding themes. When two or three consecutive interviews teach you nothing new, you are at or near saturation.
- Reset the count for each distinct segment. Saturation is measured against a specific population and question; a genuinely different audience starts its own count.
What changes the number
Two things move saturation more than anything else, and NN/G names both: the breadth of the question and the diversity of the population. A “research project that is very exploratory and targets a diverse study population will require more interviews to reach saturation than a study with a small-scope research goal and a homogeneous study population.” So a focused question asked of one tight segment might saturate in well under twelve; a sprawling question across many kinds of user can run far past it. Narrow the question or narrow the audience, and the number falls.
Reaching saturation without the cost
The honest constraint on saturation was never curiosity, it was cost. Teams stop at five participants because each additional interview means another moderator hour and another scheduling round, not because the themes had stabilised. That is the specific bottleneck GhostClick removes. You bring your own participants; it interviews each of them asynchronously by text or voice in seven languages, so adding the next five interviews to check whether you have truly saturated costs almost nothing, and its cross-participant analysis helps you see when the themes stop changing.
It does not decide saturation for you, that judgement is still yours, and it is not the right tool for deeply generative or emotional work where a live moderator is needed. But for the common case, running enough interviews to actually reach saturation instead of guessing, cheap extra interviews are exactly what the method has always wanted. Once you know roughly how many that is, the practical arithmetic is in how many participants qualitative research needs.
Run enough interviews to actually saturate
GhostClick interviews every participant you recruit, by text or voice in seven languages, on their own schedule, so adding a few more to confirm saturation costs almost nothing. Bring your people; it runs the sessions and the analysis.
Start a studyFrequently asked questions
- What is saturation in qualitative research?
- Saturation is the point at which new interviews stop revealing new themes, so conducting more would not change your findings. It is the standard stopping rule for qualitative research: you keep interviewing until the data becomes repetitive, then a little longer to be sure.
- How many interviews does it take to reach saturation?
- It depends on the study, but the most cited experiment, Guest, Bunce and Johnson in 2006, coded 60 interviews and found saturation occurred within the first twelve, with the basic thematic structure visible after about six. A narrow question on a homogeneous group can saturate sooner; a broad, diverse study takes more.
- Can you predict saturation before starting a study?
- Not precisely. Guest and colleagues note that saturation gives little guidance for estimating sample size in advance, because it depends on how varied your participants are and how broad your question is. The practical approach is to start with a small sample and add interviews until new ones stop adding themes.
- Does saturation reset for different participant segments?
- Yes. Saturation is measured against a specific population and a specific question, so a genuinely different audience starts its own count. Two distinct segments each need to reach saturation on their own terms; reaching it in one does not mean you have reached it in the other.
Related guides