Using this approach market researchers observe and record how consumers behave. They can count the number of people or cars that pass a particular location in order to assess the best site for a new business. Researchers can also observe people in shops to see how many look at a new display or take a product from the shelves. A simple stock check can also be used to record sales over a period of time. Results can be distorted, however, if people are aware of being watched and behave differently because of this. Also, observation does not give researchers the opportunity to ask for explanations of behaviour - it only records what actually happens.
2 Test marketing
This can take place after a decision has been made to produce a limited quantity of a new product but before a full-scale, national launch is made. It involves promoting and selling the product in a limited geographical area and then recording consumer reactions and sales figures. To be reasonably accurate the region selected must reflect as closely as possible the social and consumer profiles of the rest of the country. In many countries with regional TV stations, one of these regions is chosen as TV advertising can then be focused on just this one area. Test marketing reduces the risks of a new product launch failing completely but the evidence isnot always completely accurate if the total population does not share the same characteristics and preferences as the region selected.
3 Consumer surveys
These involve directly asking consumers or potential consumers for their opinions and preferences. They can be used to obtain both qualitative and quantitative research. For example, here are two questions asked in a recent survey of shoppers:
'How many foreign holidays did you take last year?' 'What do you look for in a ideal foreign holiday?
The first question is designed to obtain quantitative data which can be presented graphically and analysed statistically. The second question is designed to find out the key qualitative features of a holiday that would influence consumer choice.
There are four important issues for market researchers to be aware of when conducting consumer surveys:
· Who to ask? Given that in most cases it is impossible or too expensive to survey all potential members of a target market (called the survey population) it is necessary to select a 'sample' from this population. The more closely this sample reflects the characteristics of the survey population, then the more accurate is the survey likely to be.
· What to ask? The construction of an unbiased and unambiguous questionnaire is essential if the survey is to obtain useful results.
· How to ask? Should the questionnaire be self completed and returned by post or filled in by an interviewer in a face-to-face session with the respondent? Could a telephone survey be conducted instead?
· How accurate is ii? Assessing the likely accuracy and validity of the results is a crucial element of market research surveys.
Who to ask? - Sample size and sampling methods Sample size Generally speaking, the larger the sample, the more confidence can be given to the final results. In surveying consumer reaction to a new advertising campaign for a major brand of chocolate, a sample often people is unlikely to be sufficient.
Try to work out why. The first ten people chosen might show a positive reaction to the new advertisement. Yet another ten might show a majority with negative reactions. A sample often is too small to be confident about the result, as chance variations could easily occur as a result of the limited number of respondents chosen. A sample of 100 or even 1,000 will produce results that will reflect much more accurately the total preferences of the whole survey population. There will be much less risk of pure chance distorting the results. Obviously, a sample of 1,000 is more useful than one of 100, especially if the questions were being focused on particular age or income groups. Once a sample of 100 is broken down into, say, ten different age groups, then there will only be ten respondents from each age band. We have already stated that this is likely to be too few.
What prevents all primary research being based on a sample size of 1,000? Cost and time are the two major constraints here - the cost of research increases greatly with the sample size, especially when a specialist firm of market research analysts is used. The results of a survey might be needed quickly to assist managers in making rapid decisions - a sample size of 1,000 will take much longer to organise.