Estimating and Tracking Marketing Effectiveness Using Adaptive High Fidelity Models
30 Pages Posted: 22 Feb 2018
Date Written: July 13, 2010
Abstract
The paper focuses on a methodology for creating adaptive high fidelity market response models that enable the measurement and tracking of the effectiveness of a firm’s marketing investments over long time horizons. The methodology uses a log-linear model with latent classes and is robust to changes in the effectiveness of the marketing-mix instruments used by the firm as well as to changes in the competitive landscape attributable to the addition and deletion of stock-keeping units. It is well suited for use within marketing decision support systems. The performance of the proposed methodology was tested at two Fortune 500 firms, and the results indicate that (a) in the short run, the model performance is equivalent, if not better, than the extant market response models; (b) the performance of the proposed system does not deteriorate over time even in dynamic market conditions; and (c) the parameter estimates are efficient as well as stable; instability/inconsistency in the parameter estimates often leads to a loss in trust of decision support systems.
Keywords: Marketing Decision Support System, Marketing-Mix Effectiveness, Moving Horizon, Effectiveness Tracking, Latent Class Regression
JEL Classification: M11, M30
Suggested Citation: Suggested Citation