A markovian model of consumer buying behavior and optimal advertising pulsing policy

This paper deals with the problem of scheduling optimal advertising policy for a very general class of consumer buying behavior models. To avoid analysis of a complex multitude of social-psychological and cultural-environmental factors affecting the consumer’s decision we construct a stochastic model. Because of the diminishing effect of even advertising policy, we consider advertising pulsing policy (APP) as a means to increase advertising effectiveness. By reducing the probabilistic evolution of sales and consumers’ attitude over time to their means, multivariate linear least-square regression is used to estimate the market parameters and validate the model. The prescribed strategy scheduling maximizes the discounted profit function, which includes uncertainty in sales over a finite campaign duration. The superiority of APP over an even advertising policy is illustrated numerically.