Qualitative and quantitative UX research are the two main ways teams learn how people use a product. One digs into motivation, the other counts behavior, and most sound product decisions need both.
Teams that lean on a single method end up with half the story. Analytics can show where users leave, but only conversation and observation reveal why.
In this guide, we’ll break down what each approach does, when to combine them, and how to measure a mixed study without falling into common traps.
What are qualitative and quantitative UX research?
Qualitative UX research explains why users act as they do. Quantitative UX research measures how many of them act that way.
Qualitative UX research
Qualitative UX research is the study of non-numerical evidence, such as observations, interview quotes, and session recordings, to understand why users think and behave the way they do. It works with small groups and goes deep on each person. Common qualitative research methods include user interviews, diary studies (participants log experiences over days or weeks), focus groups, and moderated usability tests, where a facilitator guides participants live.
Quantitative UX research
Quantitative UX research is the collection of numerical data, such as task success rates and survey scores, to measure how many users behave a certain way and how often. It relies on larger samples and produces figures you can compare over time.
Typical sources include:
- Product analytics that track paths and drop-offs
- A/B tests that compare two designs
- Structured questionnaires built in survey software
Qualitative vs quantitative UX research: Key differences
The two approaches differ in the question they answer, the data they produce, and the sample they need. This table puts the differences side by side.
| Dimension | Qualitative UX research | Quantitative UX research |
|---|---|---|
| Core question | Why do users do this? | How many users do this, and how often? |
| Data produced | Quotes, observations, video, themes | Rates, scores, counts, timings |
| Sample | Small groups studied in depth | Larger samples needed for reliable numbers |
| Typical methods | Interviews, diary studies, focus groups | Analytics, A/B tests, structured surveys |
| Best for | Finding problems and motivations | Sizing problems and validating fixes |
| Main risk | Findings may not generalize | Numbers may miss the reason behind behavior |
Neither column is better than the other. Each one covers the other’s blind spot.
Terms people confuse with qualitative and quantitative
Qualitative and quantitative describe the type of data, not whether you study opinions or actions. The most common mix-up is treating qualitative as a synonym for attitudinal, and quantitative as a synonym for behavioral.
- Attitudinal research captures what people say, such as opinions, expectations, and feelings.
- Behavioral research captures what people do, such as clicks, navigation paths, and task attempts.
- Mixed methods studies deliberately combine qualitative and quantitative data to answer one question, as this overview of mixed methods research explains.
- Usability tests and surveys can be either. A think-aloud session, where participants voice their thoughts while completing tasks, is qualitative. A benchmark test with success rates is quantitative.
Nielsen Norman Group treats qualitative versus quantitative and attitudinal versus behavioral as separate dimensions, which is why one method can sit in different places on each.
Which UX research methods are qualitative, quantitative, or both?
Most UX research methods lean one way, but several produce both kinds of data depending on how you run them. This table maps common UX research methods to the data they usually produce.
| Method | Usually produces | Typical use |
|---|---|---|
| User interviews | Qualitative, attitudinal | Understand needs, motivations, and vocabulary |
| Diary studies | Qualitative | Track behavior in context over days or weeks |
| Focus groups | Qualitative, attitudinal | Hear how people react to each other’s views |
| Moderated usability tests | Qualitative, behavioral | Watch users struggle and ask why in the moment |
| Unmoderated usability tests | Both | Add volume to a known flow through self-guided sessions with success rates and video |
| Surveys | Both | Closed questions give numbers, open-ended ones give text |
| Analytics | Quantitative, behavioral | Find where users drop off |
| A/B tests | Quantitative, behavioral | Compare two designs at scale |
Nielsen Norman Group notes that diary studies, contextual inquiry, and qualitative usability testing can gather both behavioral and attitudinal data. Contextual inquiry means observing people in their own setting. The table shows the usual case, not a rule.
When should you use qualitative, quantitative, or both?
Use qualitative research to understand a problem, quantitative research to size or prove it, and both when a decision is costly or the numbers surprise you. The right order depends on what you already know.
- You do not yet know what to measure.
Start with qualitative research. A handful of interviews or diary entries surface the themes that shape your hypotheses, survey questions, and prototypes.
- You see a problem in the numbers.
Start with quantitative research, then follow with interviews or moderated tests to learn the reason.
- You need to move fast on a known flow.
Run both in one session, such as a usability test that captures think-aloud comments and a standardized score.
- A decision is costly.
Validate what a small qualitative study finds with a larger quantitative sample before committing budget.
One method is often enough. A narrow, well-defined question may need only a survey, and early discovery may need only a few interviews. Blended studies cost more and take longer, so weigh that against the decision they support. This piece on blending qualitative and quantitative research makes the same point for market research teams: budget, timeline, and audience decide.
A worked example: Fixing a checkout drop-off
Quantitative research finds where a problem sits, qualitative research explains it, and quantitative research confirms the fix. The scenario below is illustrative rather than a customer case study.
- Signal. Analytics shows that many shoppers on a US online store leave at the shipping step. The team knows where the problem is, but not why.
- Explore. Moderated usability tests with recent buyers reveal that shipping costs appear too late and the delivery estimate feels unreliable.
- Fix. The team shows costs earlier and rewrites the delivery message.
- Validate. An A/B test compares the old and new flow with a large sample and tracks completion rate.
- Follow up. A short survey with one open-ended question asks recent buyers what almost stopped them, which catches any new friction.
A guide to scheduling focus groups, in-depth interviews (IDIs), and usability tests walks through a similar sequence for a bank redesigning a mortgage application. It starts with four interviews with recent applicants. A focus group of six then pressure tests the affordability language, and an unmoderated mobile test checks whether the fixes held.
How to measure mixed methods UX research
Measure a mixed study on two tracks: numbers that show the size of an effect and evidence that explains it. Pair each metric with the qualitative signal that gives it meaning.
| Quantitative metric | Qualitative partner | What you learn together |
|---|---|---|
| Task success rate | Session recordings | How often users finish and exactly where they stall |
| Time on task | Think-aloud comments | Whether long times mean confusion or careful reading |
| System Usability Scale (SUS) score | Interview themes | How perceived usability moves across releases, and why |
| Net Promoter Score (NPS) | Open-ended comments | Which way loyalty is heading and what drives it |
| Conversion rate | Diary entries or interviews | How outcomes connect to real-life context |
The System Usability Scale is a standardized ten-question survey that turns perceived usability into a single score. Net Promoter Score asks how likely users are to recommend a product, on a zero-to-ten scale.
Read the two tracks together. When they agree, confidence in the finding rises. When they conflict, treat the gap as a lead, because Nielsen Norman Group points out that mismatches between what users say and do often produce the best insights. You can also code open-ended answers into categories and count them, which turns themes into numbers.
Where QuestionPro UX fits in a mixed methods workflow
In February 2026, QuestionPro announced that it would merge Trymata, its long-time sister company and a usability testing platform, into a combined UX offering. The stated aim was to put quantitative, qualitative, and UX insights in one place. The asynchronous qualitative tools QuestionPro gained through Digsite, including diaries, Quests, and discussions, joined the same platform.
Qualitative and usability research in one workspace
QuestionPro UX covers moderated and unmoderated usability tests, one-to-one interviews, virtual focus groups, and diary studies built from Quests, which are sets of tasks participants complete. The platform records each session and adds automatic transcription, video annotations, and AI highlight reels for analysis. Standardized scales such as SUS, NPS, and SUPR-Q, a website-experience questionnaire, sit alongside the recordings, so each number stays tied to the moment it came from.
Feedback on the QuestionPro UX page includes a Dannon strategy and insights manager who said the team understood consumer reactions to packaging, product statements, messaging, and ingredients within two weeks.
Participants and the quantitative side
Teams can recruit from the QuestionPro UX panel and filter by demographics such as age, country, and household income, or invite their own customers instead. QuestionPro Audience is the company’s participant offering, and the merger announcement said QuestionPro would integrate Trymata’s 25,000-strong usability testing panel into it.
For validation at scale, a structured survey can follow a qualitative finding with a larger sample. That is the qualitative-then-quantitative sequence described earlier, run inside one connected toolset.
Mistakes that weaken a mixed methods study
Most failures in mixed studies come from asking one method to do another method’s job. These five mistakes show up most often.
- Treating small-sample numbers as statistics.
Nielsen Norman Group warns that numbers from small qualitative sessions rarely generalize, so check confidence intervals the range a true value likely falls within before reporting any percentage.
- Asking a survey to explain why.
Multiple-choice questions size a problem but rarely explain it. Add an open-ended prompt or a follow-up interview.
- Starting from different questions.
If the two strands answer different questions, they cannot confirm or challenge each other.
- Reporting the strands separately.
Stakeholders then hear two stories. Pair each number with a quote or clip instead.
- Trusting AI summaries without review.
AI speeds up transcription and theme detection, but a researcher should check the themes against raw sessions for context and nuance.
One question, two kinds of evidence
The strongest UX research rarely picks a side. It starts with one clear question, borrows whichever method answers the next piece of it, and keeps the numbers and the stories in the same room.
Teams that work this way spend less time arguing over whose data is right and more time fixing what users actually struggle with.
Frequently Asked Questions (FAQs)
Nielsen Norman Group recommends about five users per audience group for qualitative usability testing and roughly 40 participants for most quantitative studies. Distinct user groups need their own sessions, so budget for each segment separately.
An average System Usability Scale score is 68, based on MeasuringU’s analysis of about 500 studies, so anything above 68 is above average. Scores above roughly 80 land in the top 10 percent. A score is not a percentage, so compare it to benchmarks.
UX research studies how people use a specific product or interface so teams can improve its design. Market research studies broader consumer needs, demand, and competitors to guide business strategy. Both use qualitative and quantitative methods, and they often share participants and tools.
Get explicit consent to record from every participant before you start. Recording laws differ by state, and some require all parties to agree, so put consent language in your screener and session script. Confirm specifics with your legal team.
Yes. Unmoderated tests, remote moderated sessions, diary studies, and surveys all run on participants’ own devices, which lets US teams reach people across time zones without a lab. Test on both iPhone and Android when mobile matters.
