Dealing with ‘demon customers’ (Selden, L., & Colvin, G. (2003). Angel customers & demon customers: Discover which is which and turbo-charge your stock. Penguin.) consumes substantial resources, and leads to continual inefficiency and sub-optimal profits for firms. Undesirable behavior exhibited by such customers can manifest in two major types of behavior—unwarranted returns and bad debt. Academic research has largely focused on the effect of returns, detailing conditions under which return behavior can be profitable (Petersen and Kumar, Journal of Marketing 73:35–51, 2009). Less is known about the impact of bad debt, even though predicting bad debt is an important marketing activity for firms in the direct response industry. Firms in this industry market products to a substantial segment of customers with limited access to credit card debt. For such customers transaction specific ‘credit rating’ is not predetermined (as in credit cards), but is rather an integral part of the targeting process through the ‘bill me later’ or ‘buy now pay later’ (BNPL) payment mechanism. Thus, little is also known about the following research question: what is the combined impact of both return and bad debt behaviors on firm profitability? We study this important question with the help of a unique dataset from a co-operative database, and find that direct response firms may benefit from reconsidering their return policies by factoring in customer bad debt propensities and costs. Our model and results continue to be relevant in the post-pandemic era of the 2020s with the rise of BNPL platforms such as Klarna, Affirm.