Research Flow makes it quick to run AI-moderated interviews, so you can reach the number you need faster. This article helps you work out how many that is.
There's no single right number for every study. This article explains what affects how many interviews you need, gives some starting points, and shows you how to tell when you have enough.
If you're interested in Typeform's Research Flow for creating research studies, please reach out to our Sales team.
Quick guidelines
Use these numbers as a starting point. You can add more interviews if your results show you need them.
- Easy-to-reach audiences (for example, general consumers or your own active customers): aim for around 30 interviews at a minimum.
- Niche or hard-to-reach audiences (for example, senior decision-makers or specialists in a small field): even 5 to 10 interviews can give you valuable insights.
- Comparing groups (for example, monthly vs. annual subscribers): plan your numbers for each group. If you’re aiming for 30 interviews for one group, then aim for around 30 per group when comparing two.
What affects how many interviews you need
Who you're talking to
The more alike your respondents are, the sooner their answers start to repeat the same themes. A broad or mixed audience needs more interviews to cover the full range of experiences. A small, specialist audience often needs fewer, because each person has a lot of relevant experience to share, and there are fewer people to find in the first place.
How many groups you want to compare
If you want to compare segments, like different regions, plans, or customer types, aim for a similar number of interviews from each group. For example, if you’re comparing free, monthly, and annual subscribers, 30 interviews from each (90 in total) will give you a more balanced and reliable picture than 90 interviews where 80 come from free users and only 5 each from monthly and annual subscribers. With only a handful of interviews from a group, you won’t have enough evidence to draw conclusions about it or compare it with the others.
What you want to learn
Open, exploratory questions ("Why do customers leave?") usually need more interviews to uncover the full picture. Focused questions ("What do people think of this new onboarding screen?") usually need fewer, because the range of possible answers is smaller.
Collect interviews in rounds
If you already have a target number from the guidelines above, you can launch straight to it. If you’re not sure how many you need, or you want to check you’re hearing a wide enough range of experiences, collect interviews in rounds instead. If you’re recruiting from a panel, each round can be a new recruitment group in the same study.
- Launch your study to a first batch of respondents, for example, 15 to 20, or about half your target. If you're using a panel, you set the number of people when you create a recruitment group. See Recruit respondents to take your research study with a panel.
- Review your results and highlights. Note the main themes and anything that surprised you. See How to analyze your Research Flow results Insights.
- If you're still hearing new ideas, collect another batch. With a panel, create a new recruitment group. You can't change a group after it's launched, but you can add more groups to the same study. Sessions from all groups appear together on your Results page.
- Repeat until new interviews mostly confirm what you already know. If a recruitment group is still collecting responses, click Stop recruitment on the Share page. See What are panel recruitment groups?
Tip! Your first batch is also a good way to check your study works as expected. For example, answers are detailed and relevant to your research questions.
Signs you have enough interviews
- Several interviews in a row haven't brought up new themes and you've heard from the full range of people you wanted to reach.
- You can answer your original research questions with confidence, and back up your answers with examples.
- Each group you’re comparing has enough interviews for you to draw conclusions about it, even if the groups turn out to be similar.
Researchers often call this point saturation: when more data stops giving you new information.
Common mistakes to avoid
- Collecting too many interviews "just in case." More responses take longer to review and cost more, but often don't change your conclusions.
- Stopping too early. If your first few interviews all say different things, keep going until patterns appear.
- Focusing on numbers over quality. A smaller group of the right respondents is more useful than a large group that doesn't match your target audience. Use screening questions to make sure you're talking to the right people. See How to add tasks to your research study.
- Reading too much into percentages. With 10 interviews, one person changing their answer moves a Multiple choice result by 10%. Use small-sample percentages to spot patterns, and back them up with what people said in their own words.
FAQ
Is there a minimum number of interviews I should run?
There's no fixed minimum. For most audiences, around 30 interviews is a good starting point. For niche audiences, a handful of interviews can still give you useful insights.
Can I add more interviews after my study is live?
Yes. If you're recruiting from a panel, create a new recruitment group in the same study. All sessions appear together on your Results page. How many panel participants you can recruit per study depends on your plan. See What are panel recruitment groups?
Do I need more interviews if I'm using Multiple choice questions?
It depends on how you'll use the results. If you're using Multiple choice or Dropdown answers to spot patterns alongside open answers, the guidelines above still apply. If the percentages themselves are your main result, for example to compare groups or report numbers to stakeholders, you'll need more interviews for them to be reliable.