Quantitative Research

Quantitative research is the numbers side of market research. It's about measurement and attaching numbers to a market - for instance market size, market share, penetration, installed base and market growth rates.

Quantitative research can also be used to measure attitudes, satisfaction, commitment and a range of other useful market data and market metrics that can be tracked over time and used to generate insights as part of a wider business planning and business strategy process.

Most quantitative market research is now conducted online via web-based surveys, but we also do research by phone, post and face-to-face as tailoring the research to the audience remains vitally important.

"Our key research objective was to obtain a hierarchy of consumer needs for our NPD programme, but dobney.com exceeded our expectations by also building us an excel-based model to test consumer preference for different product scenarios - we got much more than we expected!"

Technology Manager, Reckitt Benckiser

The mainstay of business planning is the use of numbers such as market size, share and usage. This form of numerical data, or market metrics, is gathered through the use of questionnaire-based statistical surveys. Advanced use of this quantitative data looks for correlations and relationships within the data. This can provide key insights into the structure or underlying concepts or drivers of behaviour.

The basis of all quantitative research comes from the design of the sample and survey type, the design of the questionnaire and the quality of the analysis and reporting. A good design comes from understanding not just how to do research, but also the business context for that research and knowledge of the decisions that may be taken once the results are in.

For more information see the specific pages on:

A good starting point is our guide to Market Research Basics. We also provide our Questionnaire Wizard software to help businesses better manage and speed up the process of creating questionnaires.

Sample and survey type

The sample and survey type are the statistical bedrock on which quantitative research is based. Survey design relies on properly defining the target universe or population, finding means to make contact with this population and stratifying or dividing the population into a known classification scheme so that the sample can be drawn properly.

The type of survey to be carried out will depend almost entirely on the target population and the subject under investigation. Options range from postal, to telephone, to face-to-face intercept surveys (street interviewing), to house-to-house and on-line research.

Understanding the likely response rates, biases and properly defining the interviewing task will determine the true statistical quality of the final data. We carry out quantitative research in all forms whether postal, telephone, on-line or using face-to-face interviewers which makes us well placed to determine which technique will work best for your project.


Quantitative research, unlike qualitative research, relies on a fixed questionnaire that should be structured to ensure it is administered the same way for each respondent to obtain a reliable measure of the market (filtering and randomisation excepted).

Although not hard to design, questionnaires require a few basic rules to be followed so that ambiguous results are avoided. Such as avoiding double meanings or leaving the respondent unable to answer.

A well designed questionnaire will be short, to the point, yet have a flow that the respondent (and interviewer for phone or face-to-face) can use to get through it quickly and accurately. Ideally a questionnaire should be designed with analysis and presentation in mind. Will I be able to use and explain the results? Have I covered off the key market metrics needed for analysis? Can I adequately segment and classify the different parts of the market?

Increasingly questionnaires are not just about measuring "x% of the population said...", but also involve modelling and forecasting behaviours from the answers given. Proper thought to the statistical output and modelling possibilities should be strongly considered when designing the questions, particularly if the questionnaire is to be used in any form of long-term tracking where changes are difficult and often costly to make.

As the questionnaire is central to the outcome, piloting and testing ensures that questions get to the data that needs to be investigated. The best piloting takes place face-to-face, but just running a survey for a few interviews online and allowing people to comment on the questions, then checking the data can help spot any unforeseen problems.

Analysis and reporting

Quality analysis is the unsung hero of quantitative research. It would be fair to say that most market research quantitative studies are under-analysed, usually because of pressures of time and a desire to get the results into a presentation, so creating a pile of charts, but not the intelligence or interpretation the business really needs - what does it mean?

Here under-analysed shouldn't be taken to mean cutting and re-cutting the data into millions of tiny subgroups looking for miniscule gems of information (which is very time intensive and usually not very productive). Under-analysed means cross-referencing one measure against another within the same survey (and with external data too). For instance, looking at average sales size for your company against your competitors, or finding out what proportion of your customers are active in a market (see market metrics).

The better the analysis, the shorter the questionnaire and the presentation will be. The reason being that good quality analysis means being able to focus on the important over the merely interesting. But to understand what is important, the analyst needs to have a good view of what the business is about and so what the audience need to hear.

Presenting quantitative information is also a challenge. It is by its nature numerical and not particularly visual (fancy graphics may dress up and mask this fact). The challenge for the researcher is to bring out and illustrate the story, not merely present the list of answers to the questionnaire.

If the research was mainly to collect information (such as a U&A) as a database, one question is whether you should ever present all the information to everyone all at the same go, or just need to communicate what you have and dip in and out. In particular, you might look at the tools that you have to access the data - for instance we often provide bespoke drill-down tools that help people explore the data in more detail or from a specific perspective in the future.

For help and advice on carrying out quantitative research projects on-line or off-line contact info@dobney.com