|
Abstract:
Most of the marketing mix strategies that researchers have so far proposed assume that the parameters of the underlying diffusion model are known. However, it is very difficult to accurately establish those parameters in advance. A marketing mix strategy based on a faulty a priori estimate of new product diffusion may lead to lost profit or financial failure. Recent technological developments, such as the internet and smart cards, are increasing the benefit and reducing the cost of sophisticated dynamic marketing mix strategies. It is therefore important to develop procedures that allow firms to revise their strategies dynamically based on updated market information. This paper propose a general adaptive control procedure, based on the Augmented Kalman Filter (AKF[C-D]), that dynamically estimates and updates the parameters of diffusion models and revises the time path of the control variables (e.g., price and advertising) accordingly. Using data for four durable products, we demonstrate the superiority of AKF[C-D] over the Nonlinear Least Square (NLS) in updating the parameters of diffusion models with control variables. Using durable good pricing based on the Generalized Bass Model as an example, we show that the proposed adaptive procedure can significantly improve a vendor's profitability. Our results also show that using an inappropriate estimation method (e.g., Ordinary Least Square) for diffusion parameter updating in an adaptive procedure may actually lead to a performance worse even than a non-adaptive procedure. Our formulation of the proposed procedure is quite general, and it can be applied to different marketing mix decisions based on different diffusion patterns.
|
