Qualitative market research: methods, sample sizes and how to run a study

Qualitative market research finds out why and how people buy, use, switch or give up on something, by talking to them or watching them directly. The usual methods are one-to-one interviews, focus groups, observation, diary studies and interviews with people who work in the industry. Samples are small, questions are open, and the analysis is done by reading and comparing what people said and did, not by calculating percentages.

Use it when you do not yet know the options, the reasons or the steps involved. It cannot tell you how common something is. That is a job for quantitative research, which works best after the qualitative work has shown you what to count.

A worked example

Suppose you run a regional food distributor and you notice that many independent restaurants buy from you and from one or two other distributors at the same time. You want to know why they split their orders, and whether you could win a bigger share. This example is hypothetical.

Sales data tells you that restaurants split orders. It does not tell you whether they do it for price, for specialty items, for delivery times or because a chef likes a particular rep. Those are qualitative questions: you need people to explain their choices.

Pick the method that fits the question

Nielsen Norman Group's overview of research methods says qualitative methods are better suited to questions about why or how to fix a problem, because the researcher observes or hears from people directly and can ask follow-up questions. Within qualitative research, the methods answer different questions.

NN/g describes focus groups as sessions of 3 to 12 participants and notes they can give a top-of-mind view of a brand or product concept but tend to be less useful for detailed usability questions. It describes diary studies as typically longitudinal and suited only to data participants can easily record.

For the distributor, one-to-one interviews fit best. The decision to split orders is made by one or two people in each restaurant, and you need them to walk through their last few orders. A visit during a delivery or an ordering session would add detail. A focus group would not: chefs and owners discussing suppliers in front of each other would tell you more about the group than about each kitchen's choices.

Decide how many people to include

There is no single correct number. A 2022 systematic review in Social Science & Medicine looked at studies that tested empirically when interviews stopped producing new findings. Those studies reached that point within 9 to 17 interviews, or 4 to 8 focus groups. The authors note that most of the studies had a fairly similar group of participants and narrowly defined objectives; broader studies needed more.

So plan per group, not per project. If the distributor's customers include fine-dining restaurants, fast-casual chains with a few locations and family-run diners, those may split orders for different reasons. GOV.UK's research planning guidance suggests 4 to 8 participants per round for in-depth interviews, and more rounds rather than one big one, so you can adjust between them. How many customer interviews to plan covers this in more detail.

Recruit people with firsthand experience

Talk to the person who actually makes the decision you are studying. In a restaurant, that might be the owner, the chef or a kitchen manager who places orders. Ask each one early in the call whether they personally choose suppliers, since titles vary a lot between kitchens.

Also consider people on the other side of the decision. A sales rep at another distributor sees dozens of restaurants and can describe patterns no single chef can. The procurement professionals in restaurants and sales professionals in food distribution directory pages show the kinds of people you might invite. Research screener questions helps you confirm someone fits before the call.

Write a discussion guide, not a questionnaire

A qualitative interview follows a list of topics, not a fixed script. GOV.UK's guidance on in-depth interviews suggests writing the topics you want to cover in a logical order, with a starter question and a few possible follow-ups for each, and testing the guide on a colleague first. It also suggests focusing on stories and real examples rather than how things should happen, and not sticking rigidly to the guide when something unexpected comes up. Interviews can run from 30 minutes to 2 hours; longer ones give more detail but are harder to recruit for.

For the distributor, a starter question might be: "Walk me through the last time you placed an order with someone other than your main distributor." Qualitative research interview questions explains how to write and order these.

Analyze by comparing accounts

After each round, write down what each person did and why, in the same format. Then compare: which reasons come up across different types of restaurant, and which appear only once? How to analyze customer interviews walks through this step.

Report what you found as counts of people, not percentages of a market. "Five of the nine chefs we spoke to buy specialty produce elsewhere because their main distributor's minimum order is too high" is honest. "56% of restaurants" is not, because nine people chosen for a conversation are not a sample of the market.

Know what it cannot tell you

Qualitative research tells you what reasons exist and how they work. It does not tell you how many restaurants share each reason, or how much revenue is at stake. Once the interviews show you the three or four main reasons for splitting orders, a short survey of your customer list or a look at order data can size them. Interviews versus surveys and primary market research cover how to sequence the two.

If the people you need are outside your own customer list, you can search for people by describing the experience you need. Instant Expert finds people whose work matches your question, you review them, it sends your invitations, and you pay for each call that gets booked.