Learn · fundamentals

Qualitative vs quantitative research: when to use each

Qualitative research investigates why people behave as they do: their motivations, reasoning, and lived experience. Quantitative research measures how many people behave a certain way, how often, and with what consistency across a population. They are not competing methods. They answer different questions, and choosing between them should depend entirely on what you need to know, not on which method your team happens to prefer or has budget for.

The core difference

Quantitative research produces numbers. Qualitative research produces understanding.

A quantitative study might tell you that 34% of users abandon your checkout at the payment step. That number is precise, replicable, and statistically meaningful. It tells you the scale of the problem with confidence.

It does not tell you whether users are abandoning because they don't trust the payment provider, because the form is too long, because they want to check a delivery date first, or because they received a phone call and never came back. Those are four different problems with four different solutions. A number cannot distinguish between them.

Qualitative research goes into the space a number cannot reach. It surfaces the reasoning, the hesitation, the context, and the meaning behind the behaviour. A well-designed qualitative study on checkout abandonment would tell you which of those four explanations is actually driving the pattern, in the words of the people experiencing it.

When to use qualitative research

Qualitative research is the right choice when you need to understand something you don't already know how to measure.

Use it when:

You're in discovery. You don't yet know what the problem is, only that something isn't working. You need to understand the problem space before you can design a solution or a measurement framework for it.

You need to understand reasoning. A behaviour is visible in your data, but the motivation behind it isn't. You need to know why people do what they do, not just that they do it.

You're testing an idea before building it. Committing design or engineering resource to a concept that hasn't been validated with the people who will use it is expensive. Qualitative research, done early, is cheap by comparison.

Your sample is too small for statistics. You're designing for a specific, narrow user group. You can recruit 10 of them and learn more from those 10 conversations than from a survey sent to 200 who barely fit the profile.

You need language. You're writing copy, building messaging, or designing content and you want to use the words your users actually use rather than the words your team uses when talking about users.

When to use quantitative research

Quantitative research is the right choice when you need to measure the scale or frequency of something you already understand.

Use it when:

You need to know how many. A qualitative study told you that users are confused by a specific step in your onboarding flow. A quantitative study tells you how many users hit that step, how many abandon there, and what the conversion rate impact of fixing it would be.

You need statistical confidence. You've made a change and you need to know whether the improvement is real or noise. A/B testing and controlled experiments require quantitative methodology to produce defensible results.

You need to generalise to a population. You want to know whether a finding from 12 qualitative interviews holds across your entire customer base. A representative survey gives you that generalisation in a way qualitative research cannot.

You're tracking a metric over time. Satisfaction scores, NPS, feature adoption rates: these require consistent measurement at scale over time. That's quantitative territory.

The mistake most teams make

The most common research mistake isn't choosing the wrong method. It's using quantitative research to answer qualitative questions, or expecting qualitative research to produce quantitative certainty.

Running a survey to understand why users churn produces answers that are superficially measurable but methodologically shallow. Survey respondents pick from options you gave them, not from the full range of their actual experience. You get percentages, but the percentages describe the options you imagined, not necessarily the reality.

Conversely, running qualitative interviews and then treating the pattern from 12 participants as a statistically significant finding overextends what the method can support. If 9 of 12 participants mention price as a concern, that's a strong signal worth investigating further. It is not the same as "75% of our users are price-sensitive."

Each method has a scope. Staying within that scope is what produces trustworthy findings.

How they work together

The most effective research programmes use both methods in sequence, with each informing the other.

The standard pattern: qualitative research first to understand the problem space and generate hypotheses, then quantitative research to test those hypotheses at scale and measure the magnitude of what was found.

A product team studying onboarding might run 12 qualitative interviews to understand where and why confidence drops. Those interviews surface three specific friction points. The team then runs a quantitative study: a targeted survey sent to 500 recent users, designed around those three specific friction points, to measure which is most prevalent and which correlates most strongly with long-term retention.

The qualitative study made the quantitative study worth running. The quantitative study told them where to invest first.

Running either one without the other leaves a gap. Quantitative alone produces numbers without explanations. Qualitative alone produces explanations without scale.

What this looks like in practice

A research lead at a B2B software company is trying to understand why enterprise customers aren't expanding their seat count after the initial purchase. The CRM shows that 60% of enterprise accounts are still on their initial seat count 12 months in. The number is alarming. It doesn't explain itself.

She runs a qualitative study: 10 interviews with decision-makers at accounts that haven't expanded. The brief is specific: what would need to be true for them to add seats, and what is currently stopping them? Three sessions in, a pattern emerges that wasn't anticipated: the bottleneck isn't satisfaction or budget, it's internal IT approval. The accounts want to expand. They're stuck in a procurement queue.

Armed with that finding, she designs a quantitative survey to validate it at scale: 200 enterprise contacts, three specific questions about procurement friction. 71% identify IT approval as the primary barrier to expansion. That number, grounded in the qualitative insight that made the question worth asking, gives the sales and product teams something concrete to act on.

Neither study alone would have produced that outcome.

Frequently asked questions

Can you use both qualitative and quantitative research in the same study?

Yes. Mixed methods studies combine both in a single research programme, often running qualitative and quantitative phases in sequence. The qualitative phase informs the design of the quantitative phase, which then validates and scales the qualitative findings. This is more resource-intensive but produces richer, more defensible findings than either method alone.

Is qualitative research less rigorous than quantitative research?

No, when conducted systematically. Rigour in qualitative research comes from consistent execution, systematic design, and traceable analysis. It doesn't produce statistical significance, but that's not what it's designed to produce. The two methods have different standards of rigour appropriate to their different purposes.

How do you know which method to use if you're not sure what you need?

Start with the question: do you need to understand something, or do you need to measure something? If you're still trying to understand the problem, start qualitative. If you already understand the problem and need to know its scale or test a solution, go quantitative. If you're genuinely unsure, a small qualitative study is almost always the faster, cheaper way to get oriented.

Can AI tools support both qualitative and quantitative research?

AI interview tools like Fieldwork support qualitative research: structured, adaptive conversations that produce transcripts and thematic analysis. They don't replace quantitative survey tools. The two serve different purposes and work well in combination, with AI-conducted qualitative interviews generating the hypotheses that quantitative surveys then test at scale.

Why do some teams default to surveys when qualitative research would serve them better?

Surveys feel faster and more controllable: you write the questions, you send them, you get numbers. The output looks scientific. The problem is that surveys are only as good as the questions you thought to ask. If you don't yet understand the problem well enough to write the right questions, a survey will measure the wrong things with impressive precision. Qualitative research is the right first step when the problem space is still unclear, which is more often than most teams acknowledge.

What sample size do I need for qualitative vs quantitative research?

Qualitative research typically reaches data saturation between 6 and 20 participants for a well-scoped question. Quantitative research requires larger samples to produce statistically significant findings: typically 100 or more for general patterns, 200 to 500 for segmented analysis. The qualitative sample is smaller because depth per participant is higher. The quantitative sample is larger because the goal is generalisability across a population.

Related on Fieldwork

Last updated: 2026-04-17