Channel Interactions in a Vertically Differentiated Channel Structure: The Case of Factory Outlet Stores

Abstract

This study develops and estimates a model of multi-channel customer purchase behavior in a vertically differentiated channel structure. In recent years many retailers have added exclusively sourced factory outlet stores into their channel mix to achieve market expansion and customer segmentation. However, there are two main concerns for retailers who offer a lower quality alternative channel under the same brand name: cannibalization and the impact of the adoption of the lower quality channel on total spending with the retailer. The current literature has two conflicting predictions regarding the impact of adoption of the outlet channel on total spending with the retailer. Multichannel customer behavior research suggests that customer interaction using more channels would be beneficial for the retailer and lead to increased spending. Brand extension research, on the other hand, suggests that adoption of the lower quality channel might run the risk of brand dilution and lead to decreased spending. Using an individual level customer transaction data set from a large specialty apparel retailer, we investigate the impact of adoption of the outlet channel. We jointly model customer purchase timing, channel choice and expenditure decisions. Our results show that on average, overall spending with the retailer increases after the adoption of the outlet channel due to increased purchase frequency. More surprisingly, in addition to incremental spending at the outlet channel, there is a positive spillover that results in increased spending at the retail store channel. Ours is the first empirical study investigating multichannel customer behavior in a vertically differentiated channel structure. Our results showcase the effectiveness of this channel structure in customer segmentation. We find that by offering the lower quality, lower price channel the retailer is able to significantly increase the purchase frequency and spending of the customers on the lower end of the pre-outlet adoption expenditure distribution with no cannibalization on the higher end.