Time-Varying Effectiveness of Competitive Advertising

Several recent studies report a seemingly counterintuitive result of positive cross advertising effects. We examine this phenomenon, and document its prevalence and persistence using time series data for several mature packaged goods categories. For this purpose we develop a hierarchical dynamic model to capture time-varying market response to both advertising and pricing. The model is estimated using a novel matrix specification of a hierarchical dynamic model, which parsimoniously handles multivariate demand specifications with multiple marketing mix variables, dynamically accounting for both endogeneity and heterogeneity. We use our model to study the time-varying nature of cross effects of advertising, and find that positive cross effects persist over time in many categories. We additionally examine how brand growth or decline occurs due to own and competing brands\u27 effects. Share of voice, novelty, and variety in the media mix are positively associated with these cross effects. We find no evidence that cross effects are associated with competitive intensity. This study advances understanding of those situations in which increased advertising is likely to be effective at driving sales. It also can help managers with temporal allocation of brand-building advertising budgets, with the objective of increasing brand sales volume and market share in a competitive advertising environment