A pulsing model of advertising competition: A game theoretic approach, part B — Empirical application and findings

Abstract

In part A of this study, we have theoretically established the assertion that the subgames of alternating pulsing competition (APC) and matching pulsing competition (MPC) may not have saddle points at the corners if at least one of the firms has a convex response function (see Result 4 in Mesak and Calloway, 1995). In this second part, we give numerical support to the above result and illustrate the use of linear programming (LP) to solve pulsing games associated with such cases. All findings obtained thus far are summarized afterwards within a theoretical decision-making framework. We also report the results of an empirical investigation aimed at estimating the Lanchester model in a duopolistic setting. Finally, we shed light on the lessons learned from the entire study and highlight possible directions for future research.