"But you only interviewed 11 buyers - is that even a statistically significant sample?" This question used to make me sweat. Now I see it as prime opportunity to educate and flex true expertise. Here's what I tell my clients to shake the doubt: For quant folks it's all about making sure the sample is 'statistically significant.' But qualitative research follows a different standard: Thematic Saturation. Thematic saturation happens when you start hearing the same core insights from the same buyer types again and again. This inflection point means new interviews won't reveal new information—just more of the same. It's the difference between productive discovery and wasting time. In most cases, thematic saturation hits around 10 interviews (assuming you're talking to the same buyer type and focusing on the same research goal). Beyond that, you're not uncovering new insights—only adding noise. More conversations won't change what you already know. So instead of chasing more data, focus on making sense of the patterns you've already found.
Friends - there is a LOT of peer-reviewed studies, on the ideal sample sizes for qualitative research. They have some variation on what that number is, but the TLDR aligns with what Hannah is saying. You can keep talking to people, but if you aren't learning new things, what is your time worth?
Powerful post Hannah, The law of diminishing returns kicks in after a while. I always say: Get insight from high-quality research then validate that insight with mass feedback (quantitative research). One caveat: of course you want your high-quality research to be as significant as possible, but this costs time money and effort. —And the more time, money and effort you invest in that, the less the relative impact will be. At one point you need to act on your data. Even if that data is insignificant.
When it comes to trying places to unlock growth in #lifesciencemarketing, I'm a huge fan of #ethnography and similar deep-dive qualitative research done well as you describe above. Even though it's a small number of customers, every time I've done a study like that we got such rich insight--beyond anything "big data" ever offered us--that helped my team and I make a quantum leap in subsequent strategies and plans.
The infatuation of volume is deeply implanted in most GTM teams and leaders. If 100 leads is good, 350 must be much better. The same fallacy clogs up sales funnels everywhere. For me, the "right lead or deal" meant it progressed through our process (velocity). I always shared with my teams that I "didn't care whether we won or lost a deal, but that it flowed through our process". That freaks some people out. I should add that I care for Velocity first as a sign of PMF, then turn concern to fine-grained improvement of the sales win rate.
Hannah, it's true indeed! I've seen enough times (especially with B2B clients' interviews) that seven interviews may be more than enough. Thematic saturation is a real thing and can happen even BEFORE you hit the 11-customer mark.
Can't count how many times I've been asked this question. 🫤
I was today years old when I learned about thematic saturation. I've seen it in practice when helping clients conduct customer interviews in order to uncover insights and patterns that can be leveraged into Sales email copywriting....but never knew what it was called. Great tips, Hannah!
Love this
This is great, Hannah! And the flip side: to clients who wonder why we want to have more interviews, I say, we're not just looking for what they say but how many of them say it. The proof is in the pattern!
Lynda with a Why 😁 Relentlessly curious 🤔 The research I do for organisations informs change in #construction #fuelpoverty and #climate. Both freelance and part time employed.
1wThe thing I always tell clients too (on small projects) is that if I'm doing all the interviews I am the one constant across them all. I can make connections, see the themes. I can tell when I've got what we need, and any more interviews will waste time and money (unless we start to explore things that have been revealed as we went, that we didn't know to question when we started out - this is why discussion guides need some fluidity).