With a limited budget, should I do only a focus group or only a survey?

It depends on whether you already know the right question. If you are not yet sure what the real problem is, invest in qualitative research first, because knowing the right question is worth more than getting a precise number for the wrong one. But if you need numbers to request budget approval and you already know clearly what to measure, a small, well-designed survey is enough.

How many focus groups, and how many in-depth interviews, are enough?

The usual principle is data saturation: the point at which additional interviews no longer produce new issues. In practice, this is often 8 to 15 in-depth interviews per target group, and 3 to 4 focus groups per target group. If you have several target groups, multiply by the number of groups.

Is mixed-method research much more expensive than using one method?

It costs more as a project, but it is often cheaper relative to the quality of the decision, because it helps avoid two far more expensive mistakes: measuring the wrong question and misinterpreting the numbers. The way to control the budget is to shrink each step rather than cut steps out, for example 8 interviews instead of 15, while keeping all four steps in sequence.

Can AI replace people in market research?

AI helps a great deal with synthesizing data, analyzing interview transcripts, and searching secondary data, and it genuinely reduces time. But in 2026 AI still cannot replace four things: observing what customers do but do not say, asking follow-up questions when an informant says something unexpected, taking responsibility for the conclusions that decisions are based on, and understanding Thai context that is not in the training data.

Can qualitative findings be presented to the board, given the small sample size?

Yes, if presented properly. Qualitative research does not claim statistical representativeness; it claims to uncover mechanisms and reasons. Good presentation states clearly how many participants a finding came from, which groups, and which issues recurred in almost every case. That carries more explanatory weight than numbers without an explanation.

Focus Group or Survey? The Answer Is a Sequence, Not a Choice (A 2026 Mixed-Method Guide for Executives)

September 10, 2026

Updated September 21, 2026

By Piyamin Trithipcharoenchai

Short answer: If you are not yet sure what the right question is, start with qualitative research (focus groups or in-depth interviews). If you already know the question and need to know how big it is, use quantitative research (a survey). And if the decision carries high stakes, use both in sequence, because numbers tell you what is happening while interviews tell you why. For a decision at this level, using only one method still leaves the other half of the question to guesswork.

This article explains what kinds of questions each method can really answer, how much time and budget each takes, and how to sequence them so you end up with answers you can act on.

What Is a Focus Group, and What Is a Survey?

A focus group is a qualitative research method that brings together 6 to 8 participants who match the target audience to discuss a single topic for about 90 to 120 minutes, with a moderator who asks questions and encourages the group to exchange views. Its strength is that it captures reasons, the language customers actually use, and how people react to one another, none of which a questionnaire can pick up.

A survey is a quantitative research method that asks the same set of questions to a large number of people, from the hundreds to the thousands, and summarizes the results as proportions and figures that can be projected to the market. Its strength is that it tells you what percentage of the market holds a given view, but it cannot tell you why people think that way.

How Do Qualitative and Quantitative Research Differ?

The key difference is not the number of people but the type of question each can answer.

Qualitative research answers "why" and "how." It collects data from a small number of people, but in depth. The main tools are in-depth interviews, focus groups, and observation of real behavior (ethnography). The output is an understanding of customers' reasons, motivations, and the language they actually use. What it cannot do is tell you what percentage of the market a given opinion represents.

Quantitative research answers "how much" and "what percentage." It collects data from many people, but at a shallower level. The main tools are online questionnaires, telephone interviews (CATI), and analysis of existing data. The output is figures that can be projected to the market. What it cannot do is explain why people answered the way they did.

Where executives often go wrong is assuming quantitative research is more credible because it has numbers. In reality, a survey that asks the wrong question from the start will give you a very precise number for a question nobody cares about.

Which Questions Should a Focus Group or a Survey Answer?

Choose based on your business question, not on the budget you have.

Use qualitative research when:

—You do not yet know what the customer's real problem is and need to develop hypotheses first
—You want to understand the purchase decision process: who has influence and at which step customers hesitate
—You are testing a new product concept at an early stage to see whether people understand it the way you intended
—You need customers' real language to write survey questions or marketing copy
—Your quantitative results came out strange and you need to know what happened

Use quantitative research when:

—You already know what to ask and need to know the proportion or size
—You need to estimate market size or share of interest
—You want to test pricing, for example measuring price sensitivity with the van Westendorp method or conjoint analysis
—You need a baseline to measure again in the future, such as NPS or brand awareness
—You need numbers to secure budget approval from the board or investors

Use both when the decision carries stakes high enough that a wrong guess would cost more than the research. In practice, that means almost every decision about investment, new product launches, and entering new markets.

The NXT Insight Loop: The Sequence That Makes Research Usable

Our framework has four steps, and the heart of it is the order.

Step 1 Frame: get the question right with qualitative research

A small number of in-depth interviews or focus groups to find out what the real problem is, what words customers use, and which factors you have never considered. This step very often changes the research question, and that is its purpose. If you skip it, you will end up with a questionnaire that reflects management's assumptions, not the reality of the market.

Step 2 Measure: size it with quantitative research

Take the hypotheses from the first step, design the questionnaire, and collect data from a representative sample to learn how much weight each issue carries, which groups think differently, and what the numbers needed for the decision are.

Step 3 Explain: account for the surprises with another round of qualitative research

Every quantitative study produces numbers that do not fit the hypotheses, groups that answer against the trend, or relationships that cannot be explained. This step means going back to interview people in those groups to find the explanation. In our experience, the most valuable findings usually emerge in this step, not in the second.

Step 4 Decide: turn findings into decision-ready options

Deliver a decision memo that sets out the options, pros and cons, risks, and recommendations, not a report that ends with tables and leaves executives to interpret them.

Approximate Time and Budget for Each Method

The following figures are estimated ranges for the Thai market in 2026, for decision-grade work.

In-depth interviews

Time: 2 to 4 weeks for a project of 10 to 20 interviews.

Cost: about 5,000 to 25,000 baht per interview for general consumers, and about 20,000 to 60,000 baht per interview for hard-to-reach groups such as executives or medical specialists.

Focus groups

Time: 3 to 5 weeks for 4 to 6 groups.

Cost: about 80,000 to 180,000 baht per group for decision-grade work with an observation room, rigorous screening and a professional moderator, including participant incentives and venue. Entry-level work runs about 25,000 to 60,000 baht per group.

Online survey

Time: 3 to 6 weeks.

Cost: about 150 to 600 baht per completed response for general consumers. Specialist groups such as B2B audiences or patients with specific conditions may cost into the thousands of baht per response.

Observation of real behavior (ethnography)

Time: 4 to 8 weeks.

Cost: about 30,000 to 80,000 baht per participant. The most expensive per head, but it yields findings other methods cannot reach.

Full mixed-method project

Time: 8 to 16 weeks.

Total cost: about 350,000 to 1,000,000 baht for a single market, and 1 to 2.5 million baht with several target groups or several areas, depending on sample size and how hard the target groups are to reach.

A Common Case: Great Research Results, but the Product Doesn't Sell

There are two common failure patterns. The examples below are composites drawn from several situations, not the project of any one client.

Pattern one: skipping Frame. A company was confident it already understood the problem, so it went straight to a survey. The results showed 78 percent of respondents were interested in the product, yet once it launched, it did not sell. The cause was that the survey asked, "Are you interested in this kind of product?", which measures respondents' politeness more than their purchase intent. What should have happened was in-depth interviews first to find out what really drives customers' purchase decisions, followed by questions designed to measure behavior rather than opinion.

Pattern two: skipping Explain. A survey showed that the 35 to 44 age group had unusually low satisfaction. The team concluded it needed more marketing aimed at this group. But when they went back and actually interviewed this group, they found the problem was not brand awareness but one step in the service process that this group encountered more than others. Fixing the right thing cost many times less than running a marketing campaign.

Both cases had data, numbers, and reports. What was missing was the sequence that gives the numbers meaning.

How to Choose a Market Research Consultant

Questions to ask before you hire:

—If the results contradict our hypotheses, what will you do next? A good answer includes a plan to go back and find the explanation, not just delivering the agreed report.
—Who actually conducts the interviews and the analysis? The quality of qualitative research depends directly on the person asking the questions.
—Is the deliverable a report or recommendations? If it is only a report, someone in your organization will have to turn it into a decision.
—Have you ever advised a client not to do research? Consultants who say yes are usually the ones who genuinely think about the client's interests.

NXT does market research as a strategy consultant, not a data-collection agency. NXT's consulting team designs the study, runs the interviews and focus groups, and does the analysis itself, while survey fieldwork runs through a vetted external panel under NXT's quality control. The deliverable is a set of strategic options that executives can decide on right away.

Talk to us about your market research

Frequently asked questions

With a limited budget, should I do only a focus group or only a survey?: It depends on whether you already know the right question. If you are not yet sure what the real problem is, invest in qualitative research first, because knowing the right question is worth more than getting a precise number for the wrong one. But if you need numbers to request budget approval and you already know clearly what to measure, a small, well-designed survey is enough.
How many focus groups, and how many in-depth interviews, are enough?: The usual principle is data saturation: the point at which additional interviews no longer produce new issues. In practice, this is often 8 to 15 in-depth interviews per target group, and 3 to 4 focus groups per target group. If you have several target groups, multiply by the number of groups.
Is mixed-method research much more expensive than using one method?: It costs more as a project, but it is often cheaper relative to the quality of the decision, because it helps avoid two far more expensive mistakes: measuring the wrong question and misinterpreting the numbers. The way to control the budget is to shrink each step rather than cut steps out, for example 8 interviews instead of 15, while keeping all four steps in sequence.
Can AI replace people in market research?: AI helps a great deal with synthesizing data, analyzing interview transcripts, and searching secondary data, and it genuinely reduces time. But in 2026 AI still cannot replace four things: observing what customers do but do not say, asking follow-up questions when an informant says something unexpected, taking responsibility for the conclusions that decisions are based on, and understanding Thai context that is not in the training data.
Can qualitative findings be presented to the board, given the small sample size?: Yes, if presented properly. Qualitative research does not claim statistical representativeness; it claims to uncover mechanisms and reasons. Good presentation states clearly how many participants a finding came from, which groups, and which issues recurred in almost every case. That carries more explanatory weight than numbers without an explanation.

The price and timeline ranges in this article are broad estimates for the Thai market as of 2026, intended only to support initial budgeting. Actual prices vary with the scope of work, how hard the target groups are to reach, and the number of markets studied. This article was prepared by NXT Consulting Group.