This is the collection of secondary data, which has previously been collected by others and is not designed specifically for the study in question, but is nevertheless relevant. Secondary data is far cheaper and quicker to gather than primary data, but it can be out-of-date by the time that it is researched. The main sources of secondary data are reference books, government publications and company reports.
The primary and the secondary research will provide the business with much data relating to its markets and its consumers. This data can then be used to describe the current situation in the marketplace, to try to predict what will happen in the future in the marketplace, and to explain the trends that have occurred.
The business may also use the market research data to segment the market. This involves breaking the market down into distinct groups of consumers who have similar characteristics, so as to offer each group a product which best meets their needs. The main ways of segmenting a market are:
By consumer characteristics: this involves investigating their attitudes, hobbies, interests, and lifestyles.
By demographics: their age, sex, income, type of house, and socio-economic group.
By location: the region of the country, urban -v- rural, etc.
Effective segmentation of the market can lead to new opportunities being identified (i.e. gaps in the market for a product), sales potential for products being realised and increased market share, revenue and profitability.
Quantitative research
Quantitative research involves carrying out market research by taking a sample of the population and asking them pre-set questions via a questionnaire (normally 200+ respondents) in order to discover the likely levels of demand at different price levels, estimated sales of a new product, and the 'typical' purchaser of the company's products. The data is numerical and can be analysed graphically and statistically. There are several types of sample that can be used to gather quantitative data:
Random sampling - this gives each member of the public an equal chance of being used in the sample. The respondents are often chosen by computer from a telephone directory of from the Electoral Register.
Quota sampling - this method involves the consumers being grouped into segments which share certain characteristics (e.g. age or gender). The interviewers are then told to choose a certain number of respondents from each segment. However, the numbers of people interviewed in each segment are not usually representative of the population as a whole.
Cluster sampling - this normally involves the consumers being grouped into geographical groups (or 'clusters') and then a random sample being carried out within each location.
Stratified sampling - the consumers are grouped into segments again (or 'strata') based upon some previous knowledge of how the population is divided up. The number of people chosen to be interviewed from each 'strata' is proportional to the population as a whole.
Qualitative research
Qualitative research attempts to gain an insight into the motivations that drive a consumer to behave in a particular way. It is usually conducted through group discussions (often called focus groups) in order to discover the rationale behind consumers' purchases. The group discussion is often chaired by a psychologist in a relaxed manner, which should encourage the consumers to discuss their shopping habits and pre-conceptions concerning certain products and brands.
This involves attempting to estimate future outcomes (e.g. the level of sales). Forecasting can be done in a number of ways:
Extrapolation - this involves identifying the trend that existed in past data and then continuing this into the future. This is often done by using a software package to establish a line of best fit for past data, and then simply extending this line into the future.
The Delphi Technique - this involves using a panel of business and forecast 'experts' who discuss and agree long-range forecasting for important issues and events.
Market research - this can be used to try and establish the purchasing intentions of consumers.
Time Series analysis - this also attempts to predict future levels from past data. There are 4 main components of time-series data : the trend, cyclical fluctuations (due to the economic cycles of recessions and booms), seasonal fluctuations and random fluctuations.
Clearly, trying to predict and forecast what will happen in the future is not easy and many variables will change in both the short-term and in the long-term which will affect the accuracy of forecasts. It is always advisable for businesses to use a variety of forecasting techniques to arrive at suitable and acceptable figures for the future (e.g. costs, revenues, sales levels, profits, etc).
