I want to base on a review of market research expenditure worldwide. It is a notorious truism that the market research industry is among the least well researched of any. Estimates of expenditure are hard to come by and even when found are distressingly subject to errors arising from faulty definitions, misinterpretation or lack of comparability. Nevertheless, persistent souls over the last 10 years have worked away at improving this situation and I want to present to you data which, while I would not go to the scaffold for it, is, I think, sound data and as comparable as it can be made. It is concerned with expenditure on commercial market research. That is to say, it excludes, so far as possible, expenditure confined within academic institutes; or made by agencies, official or otherwise, not involving any commercial market or social research spend; or by companies on internal market research departmental costs. I want to show you data worldwide and for broad geographical areas - for which I have data from 1977 and 1979, and then in a greater detail for Europe, where, because of work done for ESOMAR Congresses, I can show you a longer period.
A major and well-known "tool of our trade" is the collection of information through observation. The major applications of observation as an information-collection method may be classified into the categories of the audit, coincidental recording devices, and a general classification, direct observation. In this study we evaluate data which are obtained by audits which are performed on both distributors and consumers. The evaluation is performed from the point of view of a user of research. When we compare audit data with actual data, we will find a data bias. In this study the sources of data bias which may occur when audit data are used are carefully investigated. Empirical data are used to pinpoint the effects of a nonresponse bias in audit data. An informal analysis shows that the non-response bias is correlated with the factor price.
The purpose of the study was to select the most appropriate scale for the measurement of Africans' satisfaction with their transport services . Four types of scale (verbal, numerical, visual and graphic) were selected. Two types of each of these scales were tested, making 8 scales in all. The sample consisted of 273 respondents, all of whom made a similar journey to work each day. Five aspects of the respondents' transport to work served as the validation criteria for the scales. An important criterion for selecting the most appropriate scale was suitability for use among illiterates.
The purpose of the study was to select the most appropriate scale for the measurement of Africans' satisfaction with their transport services. Four types of scale (verbal, numerical, visual and graphic) were selected. Two types of each of these scales were tested, making 8 scales in all. The sample consisted of 273 respondents, all of whom made a similar journey to work each day. Five aspects of the respondents' transport to work served as the validation criteria for the scales. An important criterion for selecting the most appropriate scale was suitability for use among illiterates.
The following contribution should make it clear even to researchers outside this specific field that media planning studies have tested many different approaches to see whether they led to a more highly discriminating description of the media and its users. Part A deals with the question to which degree media selection can depend on descriptive and psychological material about the consumer; Part B deals with the quality of the individual media and with special relationships to them; Part C treats the possibility of linking different investigations (marriage), in view of the plethora of data demanded by media analyses.
The following contribution should make it clear even to researchers outside this specific field that media planning studies have tested many different approaches to see whether they led to a more highly discriminating description of the media and its users. Part A deals with the question to which degree media selection can depend on descriptive and psychological material about the consumer; Part B deals with the quality of the individual media and with special relationships to them; Part C treats the possibility of linking different investigations (marriage), in view of the plethora of data demanded by media analyses.
Our contribution on the subject of interviewer variability is made in the spirit of the opening quotation made and endorsed by two leading authorities in the field: conscious that we can add only a small part to the fabric of knowledge, but conscious, too, that every contribution is of value which throws light on the crucial interface between interviewer and respondent. In fact, a review of the literature reveals interviewer variability to be a more shadowy, more elusive subject than one would expect, with the very existence of the phenomenon open to doubt.
Computer-interactive (electronic) interviewing shows promise of producing higher levels of respondent interest fewer errors, and data with breadth and richness unattainable through conventional means, all at comparable cost. The Xerox Corporation undertook a methodological study to test the feasibility of electronic procedures in Germany and the UK during the Summer of 1979. The questionnaire dealt with office equipment and made use of Trade-Off Analysis. All interviewing took place in the respondents' own places of work. In each country equivalent samples received either paper-and-pencil or computer-administered interviews. The results favour the electronic approach. The electronic method produced superior prediction of respondents' subsequent choices among new product concepts. Those respondents receiving the computer-interactive interview also rated the task as both easier and more interesting than those receiving the paper-and-pencil interview.
The paper discusses the reasons why Consumer Panels are among the research techniques considered to be likely to lead in terms of future growth in market research, notably in relation to specialist Panels, increased practical sophistication of data analysis, and in the use of share prediction techniques. The paper goes on to discuss the development of panel prediction techniques from the early experiences of brand share prediction models in the 1960's through to the evolution of simulated market panels designed to maximise the speed with which the repeat purchase potential of a brand is known, whilst minimising expenditure or risk by avoiding putting the product on the open market until its successful potential is confirmed or not. The paper discusses a number of case histories from Sharescale Tests and, in particular, the unique opportunity to relate repeat purchasing to price.
Over the last ten years major progress has been made in the area of multidimensional preference models and related measurement and analysis procedures. This has resulted in a rich collection of models and methods that can give very valuable insights into the factors that determine consumers' preference and choices. In this paper a review of this methodology is presented. First the theoretical roots of multidimensional preference models in economics and social psychology are briefly reviewed. Then a taxonomy of preference models is given. The main purpose of the paper is to put the various models and methods into a common reference frame, clarifying their possibilities and limitations and in this way to provide support to the researcher who has to choose his preference analysis strategy.
This paper discusses the probable development in the decade ahead of electronic scanners as a source of market research data, and their possibly fundamental implications for packaged goods marketing - and market research practice. Problems of sample size, and the nature of many markets set limits to the application of consumer panel techniques. Data collection for both consumer and retail audits as currently conducted will become increasingly expensive as labour costs rise. Electronic data capture, by contrast, will get cheaper. Scanner data is already being used in America to evaluate the effectiveness of advertising campaigns, promotional techniques and pricing strategies as well as the more obvious considerations of display, in-store location and shelf allocation. As penetration of the retail trade grows and becomes more uniform across the country, scanners will become the standard source of audit data for packaged goods there - and eventually in Europe too.
One of the oldest techniques to test the level of attention of advertisements, yet nevertheless still often in practice today, is the portfolio test: the test person goes through a folder usually containing 10 advertisements. Afterwards it is ascertained which products and brands are remembered. A series of tests, respectively re-tests, in 1979 with over 2000 interviews has shown that this method can be unreliable: the same advertisements from test to test can obtain different recall and therefore different awareness values - despite using the same test design, same sample size, same sample structure, same test method, same institute. The paper shows factors which may be responsible for this phenomena and gives recommendations how to improve portfolio test reliability.