Typological analysis is a very general method with a wide range of very different applications. It can be used for any study of the general form described in the introduction to this note. We have already applied it to marketing and town-planning studies Gallup polls and agronomy survey data and have invariably found that users of the method were particularly impressed with its simplicity of interpretation despite the multi-dimensional nature of the problems dealt with. We have also found the method suitable for such different purposes as the acquisition of comprehensive knowledge or as a means of decision. We consider, therefore, that this method has its own place among the various other general operational research methods.
In October 1968, the Chairman and Managing Director of Alimentare Emiliana S.p.A., Mr. Limentani, had considered the advisability of entering into a new sector of pre-packed foods, for which the prospects of development in Italy seemed to be promising.
The aim of this work is ''building out" the simple Markovian model in a more realistic direction assuming that the time between sequencial purchases is variable. To study the complete relationship between "the marketing input variables" and brand transition, the author proposed to drop completely the analytical solution and instead construct a semi-mathematical model based on the use of an electronical computer. The solution has the following characteristics: 1. Each cell in the transition matrix is a function of series of marketing input variables. The type of function is proposed to be trigonometrical series. As a result of the transition matrix has to be dynamic (varying with time). The author is also discussing the interviewing methods by means of which one can collect the data-material for.
The questions posed by delegates , both in the working groups and in the final session on Saturday morning, fell into five main groups: these are summarised below, with the answers given by the authors.
The model is designed to aid a decision maker who is considering the expansion of existing facilities or building a new plant for old or new products. The model conditionally expands the facilities if the ratio of manufacturing capacity to potential customer demand is at the point the decision maker believes new facilities are required.
In this presentation, we have tried to see the usual applications of operations research in the marketing field. This field is in quick evolution. Among the most promising uses, we can cite: - the decision theory applied to operations with technical and commercial risks which escape from probability laws - the integrated systems of control and predictive planning in advertising and marketing fields that will probably give a new statute to test markets - investigation of the exposure concept (advertising and commercial) and the notion of response curve of a market versus exposures (consumers - advertising, sales promotion) (retailers - agents visits) - the behaviour studies of households in the consumer panels by statistical observations which will permit to simulate the market attitudes in relation with the launching strategy of the products. Those scientific approaches will be developed if the marketing executives are trained to take the best of a better organisation.
Propensity to purchase is defined as the probability of the occurrence of an event. An attempt is made to show that this probability is determinable and to a great extent without any need to rely on verbal statements, with their shortcomings of memory and rating errors. The possibilities are based on social-psychological model concepts , whereby the behaviour of an individual in a structured society cannot be chance-conditioned but obeys specific laws which can be defined by means of a suitable empirical methodology.
I think this has been a most interesting Seminar because it has brought together people who are approaching this area of Operational Research in Marketing from a number of different directions, or with a number of different objectives.
Before reporting upon the discussions of the different contributions, I would like to make some general remarks which result from our discussion group Nr. 3 and which do not refer to the one or the other specific report.
In the present state of the art of econometrics, it is not possible to quantify or validate the relationships sketched at the macro level. Data on certain of the other relationships point in certain directions and so we have some knowledge of how one variable affects another if at times we cannot be very precise about the strength of the effect. In other cases where we have little or no data, we have subjective ideas: if the model reacts particularly sensitively to a hypothesised relationship, we have to feed in different values for the variable in question and take a view about which is most likely. Our present forecasting method is based on the past relationships between rises in consumer expenditure, expenditure on foreign travel and number of inclusive tours. Naturally we monitor continually to refine and if necessary adjust this technique.