The results from Table 5 might be translated as follows into the language of marketing strategy. Within an industry, large firms make use of the cost effectiveness of network television, while small firms are pushed to local newspapers. Firms with high BRANDS or BREADS seem to be following a strategy of pursuing multiple market segments, whether they be geographic or product based. The firms segmenting their markets avoid network television, with its common messages beamed nationwide, in favor of spot television or magazines where messages may be better tuned to the target segments. The spot TV advertising messages themselves as well as the brand names advertised can differ by geographic area, and magazines can be selected with the target audience in mind. Innovating firms use electronic media to stress the dominant theme of newness and choose network television to take advantage of economies in repeating this common theme nationwide. Thus the media have differing capabilities for information transmittal which show themselves in tandem with their cost differences.
In view of the limited degrees of freedom and the confinement of attention to a single industry, my results can be taken only as exploratory. These qualifications aside, however, it does appear that the level and composition of firm advertising varies systematically among firms within industries. This variation seems to be consistent with the hypothesis that the firm's marketing strategy variables, of which advertising is one, are simultaneously determined. In addition, the evidence on variations in media composition is consistent with differences among media in their effective threshold levels of outlay and in their flexibility in addressing different advertising themes.
The results support the complex view of the firm's optimal advertising level presented in earlier sections. Translated into an interindustry context, they suggest that controls for the full array of firms' decision variables will be necessary to permit isolating the influence of individual decision variables. This proposition pertains not only to advertising, but also to pricing behavior and the determination of product quality and rates of outlay for research and development as well; and thus the concept of simultaneous strategy determination has wider significance for research in industrial organization.
Finally, it is hoped that the use of methodologies focusing within an industry will be stimulated by this study. Such methodologies can often prove to be a powerful way of studying individual firm behavior and avoiding the difficulties of controlling for industry variables. The