Marketing Science , Models , Monopoly Models , and Why We Need Them Steven

Despite popular belief, the marketing discipline was an offshoot of economics (Bartels 1951, Sheth et al. 1988). Early researchers, often educated in economics, felt economics was preoccupied with variables such as prices and costs. They focused on other variables. They also investigated marketing as a productive activity (i.e., adding value) by adopting different sets of assumptions other than economics (Bartels 1976). Were profit variations across firms caused entirely by random exogenous factors beyond managerial control, marketing would have no ‘‘valueadded,’’ and the discipline of marketing would have diminished purpose. Mathematical marketing models appeared much later as a precise, logical and scientific way to ‘‘add value.’’ By 1970, researchers developed mathematical models for many purposes, including better forecasting, integration of data, and understanding of markets. Just a few of the many high-impact pre-1970 marketing models include those of Bass (1969a, 1969b), Bass and Parsons (1969), Cox (1967), Frank et al. (1965), Friedman (1958), Green and Rao (1970), Kuehn (1962), Little and Lodish (1969), Little et al. (1963), Morrison (1969), Telser (1960, 1962), Urban (1969), Vidale and Wolfe (1957), and Winters (1960). Today, mathematical models in Marketing Science cover many topics. Just a few examples (presented alphabetically) include: advance selling (Xie and Shugan 2001); creating customer satisfaction (Anderson and Sullivan 1993); direct marketing (e.g., Rossi et al. 1996); forming empirical generalizations (e.g., Mahajan et al. 1995); identifying first-mover advantages (e.g., Kalyanaram et al. 1995); implementing a wide range of marketing instruments (e.g., Padmanabhan and Rao 1993, Shaffer and Zhang 1995); modeling customer heterogeneity (Gonul and Srinivasan 1993); implementing the new empirical IO (Kadiyali et al. 2000); creating new product development (e.g., Griffin and Hauser 1993); understanding channels (e.g., Messinger and Narasimhan 1995); understanding dynamic brand choice (e.g., Erdem and Keane 1996); and understanding market evolution (e.g., Dekimpe and Hanssens 1995). Mathematics, as the language of science, allows interplay between empirical and theoretical research (Hauser 1985). Marketing Science publishes both empirical and theoretical mathematical models. Precise definitions of mathematical marketing models are controversial (Leeflang et al. 2000). We could narrowly define models as mathematical optimizations of marketing variables. However, Marketing Science adopts a more generic definition, which includes mathematical representations that answer important research questions in marketing. That definition allows publication of mathematical models from a multitude of disciplines, including Management Science, Statistics, Econometrics, Economics, Psychometrics, and Psychology. Note that the following discussion only considers models used by researchers. Models for use by * Steven M. Shugan is the Russell Berrie Foundation Eminent Scholar of Marketing.