Market research is, in very general terms, a means for providers of goods and services to keep themselves in touch with the needs and wants of those who buy and use the goods and services.
It involves the systematic gathering, recording and analysing of information relating to the transfer and sale of goods and services from producer to consumer, together with systematic problem analysis, model building and fact finding for the purposes of improved decision making and control in marketing goods and services.
A huge range of companies and organisations carry out market research, answering questions (among many others) like:
- Why have the sales of my breakfast cereal decreased over the last few months?
- If I launch this new pasta sauce, will anyone buy it?
- If we build our new swimming pool here, will people be able to get to it easily enough, and will they actually use it?
- How much do people understand about our charity and how can we help them to understand more?
- I’ve got to change one of the ingredients in a drink I make, will my customers notice, and if so will it affect whether they buy the product?
A statistician working in market research can expect to use a multitude of different statistical techniques in order to solve the numerous challenges that are given by clients and researchers.
The proportion of time a market research statistician spends actually doing statistics depends on the company and the type of work it does, but may be anything up to about 80%. For this reason, the career is obviously very satisfying for people who want to continue really using statistics in a commercial environment.
As a market research statistician, you will be heavily involved with the research staff who run the individual projects. You will spend a lot of time working in effect as a consultant for these researchers. You will be involved in writing proposals describing how the market research will be carried out. These proposals will cover a number of areas of which the most important from the statistical point of view will be the overall research methodology and the calculation of sample sizes and related power for relevant tests. You will have to advise about the design of the investigations; for example, there might be complex rotation plans required if products are being tested or a number of different ideas are being considered in the same piece of research.
Once the data are collected and carefully checked, the statistical analysis itself can begin. The analysis may involve anything from the most simple tests to complex multivariate analyses or modelling. Part of the challenge for the statistician is firstly to explain the analysis and results to the researcher, who may well have no mathematical/statistical background, to enable the researcher to convey these the results to the cliient. There is also the opportunity to interact directlywit clients at the study debrief, where it is crucial that the various methods are explained using clear non-technical language in order to ensure that they are confident in the conclusions drawn.