Market Research:

Just as National Statistics may be one of your first ports of call when you want economic data, so it is too with firms. The Family Expenditure and General Household Surveys, Social and Regional Trends, are all important sources of information for businesses as well as students. In addition, specialist research services are available through the same organisation or many others in the private sector.

But when the data that's required is unavailable through any of these secondary sources, firms often decide to collect the data themselves. Market Research organisations traditionally were brought in to gather information on firms' customers and their profiles, preferences and intentions.

But with the growth in power and sophistication of organisations' information technology resources, has come an increased ability for firms to gather their own information directly from the customer. Think of how supermarket retailers gather information from their stock control systems to highlight rapid changes in consumer demand; or how the same companies have collated huge quantities of information on their customer base, from the records associated with incentive or reward schemes.

The techniques of gathering and collating market intelligence are briefly outlined here, but there are more details available on sampling elsewhere in the TimeWeb materials.

Information is generated not by studying the whole population of consumers, but by observations drawn from a sample of the population. This sample needs to be representative of the overall population from which it is drawn.

• Producing a simple random sample
  
  A sample is said to be drawn at random when every member of the population stands an equal chance of being selected as a member of the sample. Identifying the whole population can be hard, especially when that population is all the customers of a firm, for example. In cases such a this, it is appropriate to use systematic sampling. This is where a firm attempts to gather information on every 5th customer, for instance.
  
  
• Stratified sampling
  
  When the members of a population cannot be said to share common characteristics, in other words when it is heterogeneous, a stratified sample is often taken. In this method the population is divided into sub-groups, perhaps based on location, demographics or size. Sample units are then selected at random from each of these sub-groups.

More on sampling techniques.