Increasing Sales by Managing Congestion in Self-Service Environments: Evidence from a Field Experiment

Managing congestion in a self-service environment such as fitting rooms in apparel retailers is vital as retailers increasingly rely on their customers to perform many tasks independently. In this paper, we examine the impact of congestion in fitting rooms on the store performance for a retailer. Using point-ofsale (POS), traffic, and labor data from a retailer, we demonstrate an inverted-U relationship between fitting room traffic and sales; this shows that managing congestion in fitting rooms is critical for store performance. In addition, we find that a co-production environment, where associates help customers to complete their activities, is more effective in driving sales compared to self-service setting. We delve into plausible mechanisms to explain the observed inverted-U shape relationship using different queueing models. Finding that traditional models with passive or limited strategic consumer behavior (i.e., balking and reneging) do not explain the inverted-U shape relationship, we propose two alternate queueing models that are consistent with our data. Finally, we use a field experiment to show that increasing fitting room labor by one person to relieve congestion increases sales per hour by 15.98%. Our solution was adopted in a retail chain with around 100 stores.