Self-service technologies (SSTs), such as airport check-in kiosks, can provide customers faster, better, and less expensive services. Yet sometimes customers experience service failures with these technologies. This study investigates the process by which customers recover from SST failures using their own effort (i.e., customer recovery) and explores their decisions to stay with or switch from the SST. Drawing from expectancy and attribution theories, we develop a process model centered on customer-recovery expectancy and test the model by tracking actual failure responses. The results show that internal attribution, perceived control over the SST, and SST interactivity all positively influence customer-recovery expectancy. In turn, expectancy affects customers’ recovery effort and recovery strategies, depending on the availability of competitive information. Furthermore, greater recovery effort increases the likelihood of staying with an SST, whereas more recovery strategies increase the likelihood of switching. The theoretical and managerial implications of these findings include ways to design SSTs to enhance recovery expectancy, a key mechanism of the recovery process, and thus to encourage customers to persist with the technologies.
Millions of customers receive services through automated machine and computer interfaces known as self-service technologies (SSTs). These interactive interfaces, including Internet-based e-tailers, free-standing kiosks, and mobile service applications, empower customers to obtain services without direct employee assistance (Meuter et al., 2005 and Zhu et al., 2007). Because of the speed, convenience, and cost savings they promise, purchases through SSTs have become sizable. Annual sales through kiosks are projected to exceed $1.0 trillion by 2014 (IHL Group 2012), and SSTs are said to be one of the “10 ideas that are changing the world” (Time 2008).
Nevertheless, SSTs can and do fail at times because of technical or human error. For example, 25% of online shoppers experience problems with websites (Forbes 2008), and only 18% of the time does interactive voice technology, such as automated customer phone service, work effectively (The Economist 2004). Such failures can result in missed sales opportunities, customer dissatisfaction, and technology abandonment. For example, more than 2,000 kiosks installed by the U.S. Postal Service are not in use today because of malfunctions and design issues (Selfserviceworld.com 2010).
In light of the risks of malfunctioning SSTs, such as dissatisfaction and lost sales, it is important to understand how customers respond to failures (Bolton et al., 2007, Puccinelli et al., 2009 and Verhoef et al., 2009). Because service personnel typically are not available to address SST errors when they occur, firms must motivate customers to recover from service failures on their own (known as customer recovery) and to stay with the technologies (Holloway and Beatty, 2003 and Meuter et al., 2000). In other words, they must encourage customers to recover from the failure (fix the SST problem) and not switch from the interface (leave the problem). Therefore, this study aims to determine the process by which customers engage in recovery or switching behaviors in response to SST failures.
The services literature identifies three types of recovery from service failures: recovery by the firm, recovery by the customer, and joint recovery by the firm and the customer (Bendapudi and Leone, 2003, Dong et al., 2008, Lusch et al., 2007, Meuter and Bitner, 1998 and Roggeveen et al., 2011). Most empirical studies have focused on recovery by the firm (e.g., Bitner, 1990, Grewal et al., 2008, McCollough et al., 2000 and Smith et al., 1999), though some have considered joint recovery (e.g., Dong, Evans, and Zou 2008). However, customer recovery—in which customers are the sole or principal actors in recovery—is relatively neglected. Gaining a better understanding of customer recovery is crucial, given the expanding role of SSTs in the service landscape.
To address this gap, we conduct an empirical study on customer recovery. We investigate a potential mechanism, called “customer-recovery expectancy” (CRE), that motivates customers to engage in the recovery process. CRE refers to the degree to which customers estimate that they will be effective in resolving the problem through their own actions and inputs. This internally focused evaluation differs from customers’ perceptions or evaluations of recovery actions taken by service firms or employees, which until now have been the primary focus of extant literature.
We attempt to answer four questions about CRE. Because CRE might propel customers to take actions on their own to address a service failure, our first research question is, what factors strengthen CRE? We consider three possible antecedents: internal attribution, perceived control over SSTs, and SST interactivity. Second, how does CRE motivate customers to fix a service failure? Here, we examine two fixing behaviors as possible consequences of CRE: customer-recovery effort and customer-recovery strategy. The former emphasizes working harder and longer to solve the problem; the latter involves searching for more appropriate solutions. Third, are these behaviors contingent on SST design features? We explore the possible moderating role of one design feature—namely, the availability of competitive information in the SST interface. Fourth, what are the effects of CRE and recovery responses on switching from the SST? We focus specifically on the likelihood that a customer will abandon the SST and demand employee assistance.
Our research thus makes several contributions to the services literature. First, we help fill a significant research gap by delineating the process of customer, rather than firm or joint, service recovery in the relevant context of SSTs. Second, we apply expectancy theory to introduce CRE as a mechanism for spurring customer-recovery actions. No previous study has used expectancy theory to explain service recovery. Third, this article offers a methodological advance in service failure research by analyzing responses to computer-simulated failures in a general population. We do so to increase the external validity and generalizability of our study. Table 1 highlights the knowledge gaps by presenting sample studies from the services literature.
