An Entropy Model of Brand Purchase Behavior

Abstract

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.. American Marketing Association is collaborating with JSTOR to digitize, preserve and extend access to A probabilistic model of consumer purchase behavior for frequently purchased, low cost items is presented. The concept of maximum entropy is used to specify the model. The only empirical data required are market shares; all other brand selection statistics, such as repeat and switch rates, are derived quantities. AnEntropy Model o f Brand Purchase Behavior ENTROPY CONCEPTIn this article a probabilistic model of consumer purchase behavior for frequently purchased, low cost items is presented. The model is completely determined by specifying only the market shares; all other brand selection statistics, such as repeat rates and switch rates, are derived from the model. The model presented is a heterogeneous multinomial preference model. It is assumed that each consumer has a set of preferences for the brands in the market, and there is a distribution of the preferences over the population. The probability of a consumer purchasing a particular brand is numerically equal to her preference for the brand. Rather than specifying the joint distribution of preferences and fitting the parameters to empirical data, the concept of entropy is employed and the distribution is selected that maximizes the entropy of the system subject only to the empirical market share values. In this way the model developed differs fundamentally from the probabilistic market models previously published [1, 3, 7].In the physical sciences, where it was first introduced, entropy is considered a measure of the degree of disorder, uncertainty, or randomness of a probabilistic system. It is assumed that at equilibrium all probabilistic systems will be at the maximum entropy consistent with the constraints on the system. In the early work in statistical thermodynamics, the entropy equations are derived from the likelihood function for the system, and maximum entropy is completely equivalent to the maximum likelihood estimate. A second interpretation of entropy is by Jaynes [4] who writes:"Just as in applied statistics the crux of a problem is often the devising of some method of sampling that avoids bias, our problem is that of finding a probability assignment which avoids bias while agreeing with whatever information is given. The great advance provided by information theory lies in the discovery that there is a unique, unambiguous criterion for the 'amount of uncertainty' represented by a discrete probability distribution, which agrees with our intuitive notions that a broad distribution represents more uncertainty than does a sharply peaked one, and satisfies all other conditions which make it reasonable." This meas...

“Listening In” to Find and Explore New Combinations of Customer Needs

“…For applications in marketing of reward functions based on information theory, see Hauser (1978) and Herniter (1973). For applications in psychology, see Prelec (2001).…”

The role of nonequilibrium thermo-mechanical statistics in modern technologies and industrial processes: an overview

“…It can also be noticed that HavrdaCharvat ITE [196] (or structural α-entropy, recently renamed q-entropy) provides the same probability distribution than the one that follows from Renyi ITE (they differ in the definition of the Lagrange multipliers that the extremum-principle approach introduces, and it can be noticed that Renyi ITE avoids difficulties associated to the other). It can be mentioned that they have been used in a particular problem (with "hidden constraints") of marketing of different brands of toilet soup [236], producing the same result [216], as it should.…”

An Information Theory Approach to Measuring Industrial Diversification

“…Year 1960196119621963196419651966196719681969197119721973Diversity Index 5.24055.18985.19215.17865.18815.17205.15144.9840 5.11135.11995.0960 5.0810 5.0604 5.0678 Year 19741975197619771978197919801981198219831984198519861987 The disaggregation of entropy for the above two groups is carried out using equations (2) to (5), and the results are presented in Table III, columns (2) to (7). The aggregated employment diversification indices initially presented in Table II are shown again in column (8).…”

“…In a marketing context, entropy can represent the distribution of consumer preferences for various brands. For example, Herniter (1973) uses entropy as a measure of uncertainty or disorder in the stochastic system that represents the consumer's preferences for special brands.…”