(A) modeling for post-promotion effects with panel data

다국어 초록 (Multilingual Abstract)

The objective of this paper is to examine a dynamic sales model for consumer packaged goods for describing how the effects of sales promotions are distributed over time. This paper specifically focuses on post-promotion effects of sales promotions in the marketplace in Korea where there exists somewhat heavily competitive in-store sales promotions compared to those in Europe and America. The panel data for this study contains observations on individuals (stores) where each observed at several points in time. The dataset consists of store-level sales volume and price data from 22 stores nationwide for 52 weeks with promotional events (such as in-store advertisements and free-gifts promotions) on 5 hair-care national brands. The dynamic model with lagged variables is developed to detect in-store promotions' immediate and future effects such as own- and cross-brands current effects and post-promotion (lagged) effects. It also examines the lag structure to see how they affect on sales for each brand. Lastly, it finds that increased promotions reduce purchase rate under non-promoted condition, which is baseline sales in future periods.