Marketing research often begins with completely the wrong question: “Should we do a survey or some focus groups?”
That is a bit like asking whether you need a thermometer or a conversation.
It rather depends on what you are trying to find out.
If sales have fallen by 15%, quantitative data can tell you the size of the problem, when it started, which customers are disappearing and which products are being affected. What it may not tell you is why customers have changed their behaviour.
For that, you might need to talk to them.
Equally, six customers telling you that your packaging is impossible to open might reveal a genuinely important problem. But those six conversations cannot tell you that 63% of the market agrees with them.
Quantitative research helps you measure the market.
Qualitative research helps you understand it.
What Do Quantitative and Qualitative Actually Mean?
Quantitative Research
Quantitative research collects or analyses numerical data. It measures things such as how many, how much, how frequently, how strongly or to what extent something occurs.
Think surveys, experiments, sales data, customer analytics, market-share data, conversion rates and A/B tests.
Qualitative Research
Qualitative research explores experiences, motivations, attitudes, language, context and meaning.
Think depth interviews, focus groups, observation, ethnography, diary studies, usability sessions and online research communities.
The Market Research Society’s qualitative research guidance provides UK-specific professional guidance on planning, participant treatment, recruitment, recording and conducting qualitative research.
Quantitative vs Qualitative Research: The Quick Comparison
The distinction becomes much easier to remember when the methods sit next to each other.
| Question | Quantitative | Qualitative |
|---|---|---|
| Primary purpose | Measure, compare and test | Explore, understand and explain |
| Typical question | How many? How much? How often? Which performs better? | Why? How? What does this mean to people? |
| Type of evidence | Numerical and structured | Words, observations, experiences and themes |
| Common methods | Surveys, experiments, A/B tests, analytics and sales data | Interviews, focus groups, observation, ethnography and diaries |
| Sample approach | Often larger and designed around estimation, comparison or statistical testing | Usually smaller and purposively selected for relevance, variation and depth |
| Typical output | Percentages, averages, distributions, relationships and effect estimates | Themes, explanations, language, tensions, journeys and hypotheses |
| Best used for | Sizing, tracking, prioritising, comparing and testing | Discovery, proposition development, barriers, context and creative exploration |
| Biggest danger | False confidence in bad measurements or biased samples | Treating interesting stories as if they represent the whole market |
Don’t Choose the Method. Choose the Question.
Instead of beginning with the research technique you happen to know, start with the decision the business needs to make.
What Quantitative Research Is Particularly Good At
1. Measuring the size of an opportunity
Imagine ten customer interviews reveal frustration with part of your service.
Useful. But before spending £500,000 rebuilding it, you may want to know whether the problem affects 5% of customers or 55%.
Quantitative research moves you from “this appears to be happening” towards “how widespread is it?”
2. Comparing groups
Quantitative evidence can compare customers with non-customers, geographical markets, behavioural segments, demographics or different stages in a customer journey.
This is particularly useful in research before segmentation, targeting and positioning, where marketers need evidence that meaningful differences between groups actually exist.
3. Tracking change
Repeated measurement can show whether awareness, consideration, satisfaction, retention, conversion or other indicators are rising or falling.
The word repeated matters. If you change the question wording, sample or collection method every quarter, your trend line may start measuring changes in your methodology rather than changes in the market.
4. Testing alternatives
If you have two landing pages, three prices or four proposition territories, quantitative testing can help compare performance.
This includes controlled experiments such as A/B testing used in conversion rate optimisation.
AAPOR’s Best Practices for Survey Research is a useful practical guide to defining objectives, choosing samples, writing questions, pretesting surveys and interpreting results.
What Qualitative Research Is Particularly Good At
1. Discovering the questions you should be asking
This is one of qualitative research’s most underrated strengths.
A survey can only ask about the things somebody thought to put into the questionnaire.
If your entire management team assumes that customers choose between products based on price, quality and convenience, those may become your survey options.
An interview may reveal that customers are actually worried about whether buying the product will make them look inexperienced in front of their colleagues.
Good luck finding that in a drop-down menu nobody thought to include.
2. Finding the customer’s language
Marketers spend a surprising amount of time inventing sophisticated language for problems customers describe perfectly well themselves.
Qualitative research can reveal the words, metaphors and phrases people naturally use — useful for positioning, advertising, website copy, product descriptions and search strategy.
3. Understanding context
People do not make decisions inside questionnaires.
They make them while rushing to work, scrolling on the sofa, comparing six browser tabs, standing in supermarkets or trying to remember a password they definitely changed three days ago.
Observation and contextual research can reveal how behaviour takes place rather than relying entirely on somebody’s reconstruction of that behaviour afterwards.
4. Finding contradictions
Consumers are not lying machines, but neither are they perfectly accurate narrators of their own behaviour.
What people say, remember, believe and subsequently do can be different things.
This is explored in more depth in Psychology, Research and Marketing: How to Turn What People Say into What Brands Should Do.
Behavioural Data Is Not the Same as Understanding Behaviour
Modern marketers can see which advert somebody clicked, which page they visited, what they added to a basket, when they abandoned checkout and whether they returned three weeks later.
That is extremely useful evidence.
But a dashboard showing that 37% of users abandon a page still does not necessarily explain why.
Perhaps the delivery charge surprises them. Perhaps the page feels untrustworthy. Perhaps they are comparing prices. Perhaps the form is annoying.
Analytics tells you what happened.
Research helps you work out what it means.
Why the Best Research Often Uses Both
The phrase mixed methods research sounds terribly academic. The principle is actually very practical.
Different forms of evidence answer different questions, so combining them can give marketers a more complete picture.
A methodological review of mixed-methods research describes sequential, concurrent and parallel approaches, and shows why integrating evidence can reveal complementary, convergent and even conflicting findings rather than forcing everything into a single method.
Four Practical Marketing Research Designs
This is where the distinction becomes useful rather than merely academic. Here are four common marketing problems and how the methods can work together.
Sample Size: Bigger Does Not Automatically Mean Better
One of the easiest ways to sound authoritative in a meeting is to announce the sample size.
“We asked 3,000 consumers.”
Excellent.
Which consumers?
If those people are systematically different from the population you want to understand, adding another thousand similar respondents does not magically remove the underlying problem.
This is why AAPOR’s survey best-practice guidance emphasises defining the population and sampling frame rather than judging a survey simply by how many people answered it.
Qualitative sample sizes work differently. The objective is usually not to estimate exactly what percentage of a population believes something, but to develop sufficient depth and variation to understand the issue being studied.
Hennink, Kaiser and Marconi’s study of qualitative interview saturation found an important distinction between identifying the main codes in a study and reaching the deeper meaning required for rich understanding. Their wider point is more useful than any magic interview number: sample adequacy depends on the research purpose, population and depth required.
Read the study.
The Research Can Go Wrong Before Anybody Answers
Survey software has made it extraordinarily easy to create quantitative research.
Unfortunately, it has not made it extraordinarily easy to create good quantitative research.
- Leading questions
- Double-barrelled questions
- Ambiguous wording
- Poor response scales
- Missing answer options
- Question-order effects
- Unrealistic recall demands
- Questionnaires that simply refuse to end
- Leading moderation
- Dominant participants
- Group conformity
- Poor recruitment
- Confirmation bias
- Ignoring contradictory evidence
- Turning memorable quotes into market statistics
The MRS Guidelines for Questionnaire Design provide a useful UK reference, while Pew Research Center’s guide to writing survey questions includes practical examples of how wording, response options and question order can influence answers.
Neither quantitative nor qualitative research is automatically rigorous. Rigour comes from how the research is designed, conducted and interpreted.
Seven Ways Marketers Misuse Research
Research is supposed to reduce uncertainty. Unfortunately, marketers occasionally find ingenious ways of using it to manufacture certainty instead.
Research Ethics Matter Too
Research depends on people being willing to share information, opinions and experiences.
That makes trust rather important.
In the UK, researchers should be familiar with the Market Research Society Code of Conduct and its supporting professional guidance.
There is an important distinction between research and selling. Research should collect and analyse information rather than act as disguised lead generation or a mechanism for persuading the people taking part.
Marketers also need to think carefully about privacy and personal data. If research involves identifiable individuals, data-protection responsibilities do not disappear because the spreadsheet has RESEARCH written across the top.
The Information Commissioner’s Office provides detailed guidance on research-related processing. Where the specific UK GDPR research provisions are being relied upon, the ICO also emphasises appropriate safeguards, data minimisation and anonymisation or pseudonymisation where appropriate.
Further Reading & References
MRS
Guidelines for Qualitative Research
MRS
Guidelines for Questionnaire Design
AAPOR
Best Practices for Survey Research
Pew Research Center
Writing Survey Questions
Hennink, Kaiser & Marconi
Code Saturation Versus Meaning Saturation
Mixed Methods Review
Combining Qualitative and Quantitative Research
Quantitative vs Qualitative Research FAQs
Is a survey quantitative or qualitative?
A survey is normally associated with quantitative research when respondents answer structured questions that can be counted and statistically analysed. However, surveys can also contain open-ended questions that produce qualitative data.
Are focus groups qualitative research?
Yes. Focus groups are qualitative. They help explore reactions, language, perceptions and interactions in depth. They should not normally be treated as a statistically representative vote.
Which is more reliable?
Neither is automatically more reliable. They answer different types of questions. Rigorous qualitative research can be much more useful than a badly designed survey, while qualitative interviews cannot estimate a population percentage in the way well-designed quantitative research can.
Should qualitative or quantitative research come first?
If you do not understand the problem well, qualitative research often comes first to discover themes and hypotheses. If numerical evidence has already exposed an unexplained pattern, quantitative analysis may come first and qualitative work can investigate why it is happening.
Can marketers use both together?
Yes — and for many important decisions that is the stronger design. Qualitative research can discover what matters, quantitative research can establish how widespread or important it is, and experiments or behavioural data can test what actually changes behaviour.
So, Quantitative or Qualitative?
The answer is frustratingly simple:
Use the one that answers the question you actually have.
Use quantitative research when you need to measure, size, compare or test.
Use qualitative research when you need to explore, understand, explain or discover.
And when the decision matters enough, use both.
Because a percentage without context can mislead you, while a brilliant customer quote can seduce you into believing one person’s experience represents an entire market.
The real skill in marketing research is not knowing how to launch a survey or book a focus group.
It is knowing what evidence you need before making the decision — and being prepared for that evidence to tell you that your original assumption was wrong.
