Dynamic Analysis of a Competitive Marketing System
AbstractThe focus of this paper is on models that accommodate dynamic phenomena and include consumer-focused and competitor-centered approaches. The consumer focus is represented in demand functions while the competitor orientation is captured in reaction functions. Although the extant literature has tended to restrict the competitive reactions to marketing decision variables, we also allow for feedback effects which model reactions to consequences of actions. These consequences may show both in own brand and in other brands' performance variables. We develop a VARX model with all relevant dynamic and interactive effects. Such a model is especially useful for situations in which causality, feedback and dynamic phenomena matter. Our model simultaneously includes all relevant individual brands, and we include lags that are brand-and variable-specific. We use Impulse Response Analysis to estimate the gross and net effects of marketing actions. To show how variables causally depend on each other, we use a Forecast Error Variance Decomposition. We apply the modeling process first to market-level tuna data, consistent with typical time series applications. Given a nonlinear model we use geometric averaging so as to avoid one source of aggregation bias. We apply the same modeling process then separately for each store so that we can also determine the heterogeneity in effects between stores. The variables of interest include two price promotion variables, one with support (feature and/or display), the other without. We find, as expected, that the nonsupported price elasticities are usually closer to zero than the supported price elasticities. Also, cumulative effects tend to be larger than the immediate effects, due to a tendency for some continuation of price discounting. Specifically, it appears that supported price cuts are often followed by non-supported price cuts. The Forecast Error Variance Decomposition results show that the variance of a supported price variable is explained largely by its own shocks, while a non-supported price variable's variance depends primarily on the same brand's supported price variable. The results based on store data show more evidence of dynamic effects than the market data do. Nevertheless, the average effects show somewhat similar magnitudes between the market and store-data. Importantly, there is a large amount of store heterogeneity in the effects. Thus, it appears that managers should consider the benefits from store-specific management of promotions relative to the costs.
Download InfoIf you experience problems downloading a file, check if you have the proper application to view it first. In case of further problems read the IDEAS help page. Note that these files are not on the IDEAS site. Please be patient as the files may be large.
Bibliographic InfoPaper provided by Yale School of Management in its series Yale School of Management Working Papers with number ysm226.
Date of creation: 02 Oct 2001
Date of revision:
Find related papers by JEL classification:
- M31 - Business Administration and Business Economics; Marketing; Accounting - - Marketing and Advertising - - - Marketing
- C32 - Mathematical and Quantitative Methods - - Multiple or Simultaneous Equation Models; Multiple Variables - - - Time-Series Models; Dynamic Quantile Regressions; Dynamic Treatment Effect Models; Diffusion Processes
You can help add them by filling out this form.
CitEc Project, subscribe to its RSS feed for this item.
- Dekimpe, M.G. & Hanssens, D.M., 2003. "Persistence Modeling for Assessing Marketing Strategy Performance," ERIM Report Series Research in Management ERS-2003-088-MKT, Erasmus Research Institute of Management (ERIM), ERIM is the joint research institute of the Rotterdam School of Management, Erasmus University and the Erasmus School of Economics (ESE) at Erasmus University Rotterdam.
If references are entirely missing, you can add them using this form.