Investigating Post Product Recall Sales Dynamics Using Functional Data Analysis

Abstract

Extant marketing literature suggests that negative effects of product recalls are not limited to the brand or the product category in which a product is recalled, but also spills over to performances of other brands and product categories. Given such wide effects of a recall event where some products benefit while others face losses; it is imperative that retailers and brand managers should have a strategy on how to limit their damages by strategically managing their inventory and shelf space allocations. In this study, we propose that such planning requires a more in-depth examination of how sales of various products and categories change (i.e. sales dynamics) just after a product recall. Using functional data analysis, an emerging statistical methodology for the analysis of curves or functions (Ramsay and Silverman 1997), we examine the sales dynamics of affected and related brands during the recall of IAMS cat food in June 2010. Using a total of 18,508 transaction data from 482 UPCs and 62 brands across five related product categories, we (a) identify different patterns of sales recovery of the affected brands; (b) investigate how the patterns vary across similar and dissimilar products of competitors, and other categories. This study extends the current literature on product recalls and identifies marketing functions that play a significant role in post recall sales. The paper concludes with a discussion of the managerial implications on how to develop a post recall strategy to reduce the negative effect of the event.