Dynamic Analysis of Consumer Response to Marketing Strategies

TLDR: In this article, the authors developed a methodology for modeling consumer response that integrates previous research in stochastic brand selection, diffusion of innovation, test market analysis, and new product design.

Abstract: This paper develops a methodology for modeling consumer response that integrates previous research in stochastic brand selection, diffusion of innovation, test market analysis, and new product design. The methodology makes it practical to extend brand selection models to include diffusion phenomena such as awareness, trial, and information flow. Purchase timing and brand selection are interdependent and both phenomena depend jointly on managerial controls such as advertising, coupons, price-off promotion, product positioning, and consumer characteristics. Within this general structure, we provide practical estimation procedures a least squares approximation to the maximum likelihood estimates to determine the parameters which link managerial controls to consumer response. Closed form solutions are derived for cumulative awareness, cumulative trial, penetration, expected sales, and purchases due to promotion-all as a function of time. We also provide simplified expressions for equilibrium t → ∞ market share. Tradeoffs among complexity of the diffusion process, number of managerial variables, nonstationarity, complexity of purchase timing, consumer segmentation, and sample size are made explicit so that the marketing scientist can customize his analyses to the managerial problems that he faces. The effects of sample size, data interval frequency, and collinearity in the explanatory variables are investigated with simulations based on a five-state consumer response process which depends on 8-10 marketing variables. The paper closes with a brief description of the application and predictive test of a consumer response model based on the methodology.