The use of AI chatbots in frontline customer service is beneficial as it can provide quick service responses, cost-saving on human employees and accelerate customers’ decision-making process. However, implementing chatbots can be a double-edged sword. On the one hand, companies benefit from the use of chatbots. On the other hand, it may hurt customer experience as customers perceive chatbots are less trustworthy and show less social presence. Service failures today have become more unpredictable with the increasing complexity of social environments. Aligning with the trends of online customer service, customers are most likely to encounter a chatbot when seeking online customer service to solve service failures. With most of the previous literature investigating customers’ perceptions of chatbots and chatbot-related service failures, little research has focused on the area where chatbots as service recovery agents and how customers perceive the use of chatbots handling their service requests after service failures.

Negative brand events can significantly harm the brand, affecting consumer cognitions, emotions, and behavior, leading to negative brand evaluations, negative emotions, brand avoidance, and retaliation. Thus, it is critical for brands to design and implement strategies that promote consumer forgiveness to mitigate those negative consumer responses. Despite the growing literature on the determinant factors of consumer behavior in a brand failure context, there is still scarce evidence regarding the impact of brand type on consumer responses, including consumer forgiveness, following a negative brand event. Considering that consumers tend to react differently to a brand perceived as arrogant (making them feel inferior), as opposed to a brand they strongly identify with, this paper focuses on failures of equal vs. higher status brands and investigates their impact on consumer forgiveness. Following the above, we expect that differences in brand status can shape consumer behavior following a failure or transgression.

Although the majority of prior literature has suggested the key reasons for consumer complaining in service failures are to vent negative emotions (e.g. anger, dissatisfaction) and to seek redress (Blodgett, Hill, & Tax 1997; Nyer 1997), some research has also pointed out that customers will give constructive suggestions to firms through complaining (Groth, 2005; Liu & Mattila, 2015). In this regard, consumer complaining can be classified into two types, namely, positive complaint (i.e., with constructive suggestion) and negative complaint (i.e., without constructive suggestion). Understanding what situations would dissatisfied consumers choose to give constructive suggestions in service failures would be of utmost important to firms. We suggest that a firm’s brand image may affect consumers’ intention to choose what types of complaint in service failures. In general, consumers may perceive a brand as having a competence image (e.g., professional and efficient) or having a warmth image (e.g., friendly and approachable) (Kervyn, Fiske and Malone, 2012). Comparatively speaking, a warmth image is associated with friendship and caring, whereas a competence image is associated with expertise. Therefore, it is possible that consumers would have a higher empathy and intention to help a firm with a warmth (vs. competent) image, and be more likely to choose positive complaint when a failure happens to this firm. An experimental study confirmed this prediction. In addition, we found that although a firm’s competence (vs. warmth) image does not affect the likelihood to give constructive suggestions in service failures, it leads to a higher level of return intention. Implications and future research directions will be discussed.

The service failure phenomenon is a long-recognized problem in hospitality industry’s marketing, and has consequently attracted significant research attention (Chan, Wan, & Sin, 2007). Once service failures occur, customers usually assess the causes of the problem. Researchers have thus studied the impact of service failures on customer failure attribution and their behavioral outcomes toward the service provider. Usually, studies examine consumer psychological processes when only one service firm is involved. However, it is unclear whether customer failure assessments are the same when they have to assess more than one service at the same time. According to a review article by Cohen, Prayag, and Moital (2014), consumer behavior has been extensively examined in the field of tourism in many aspects (e.g., decision making, motivations, satisfaction, and loyalty); however, research assessing failure attribution within tourist satisfaction literature is still rare. Moreover, in service marketing studies, it is somewhat surprising that existing service research has overlooked the fact that customers may confront failure situations where there are two or more service providers involved (Weber & Sparks, 2010). Hence, this study makes two key contributions. First, it addresses scholars’ calls for more research assessing failure attribution within a tourist satisfaction context. Second, it contributes to our understanding of consumer behavior in tourism industry by studying customer perceptions of service failures within service networks, where at least two firms are involved in the incident. This study uses in-depth interviews for data collection. And interview results indicate that relational and network characteristics have a significant influence on how customers attribute service failures to different service providers.

A growing body of research has discovered that even the trivial attribute of similarity, that is, incidental similarity, will have significant favorable impact on initial social interactions (Burger et al. 2004; Guéguen, Pichot, and Le Dreff 2005; Jiang et al. 2010; Martin and Guéguen 2013). Incidental similarities are chance similarities between individuals, such as a shared first name or birthplace, which create a sense of association between two people (Burger et al. 2004). Prior research shows that people who perceive they share a birthday, first name, or similar fingerprints with a stranger are more likely to comply with the stranger’s request (Burger et al. 2004; Guéguen, et al. 2005) and are more willing to respond to the stranger’s questions on intimate topics (Martin and Guéguen 2013). Jiang et al. (2010) applied these findings in a service interaction context and found that incidental similarities between consumers and service providers will increase consumers’ liking for the services and purchase intentions. The key reason for these favorable impacts is that an incidental similarity creates a sense of connectedness between two strangers. This sense of connectedness creates a unit relationship between two strangers that is not shared by other people around them. Therefore an incidental similarity generates a fleeting sense of liking and it has been linked to positive affect (Burger et al. 2004) and interpersonal attraction (Insko and Wilson 1977). Although existing research suggests that incidental similarities lead to favorable reactions to the similar other, we propose that the effects of incidental similarities are not invariably favorable. Incidental similarities can elicit unfavorable effects, and can make an otherwise disinterested observer become involved in an exchange between a stranger and a company that (s)he merely witnessed. For example, a service failure involves the service provider and the suffering customer. From the perspective of an individual observing the failure, the nature of the effect of incidental similarities would depend on whether one feels a sense of association with the provider or the customer. Imagine a situation in which someone observes a customer being told that a table he had reserved is actually not available. If the observer notices that customer’s surname happens to be the same as his own, he may be disposed to view the situation from the customer’s perspective and blame the provider for the failure. However, if the observer notices from the provider’s name tag that they happen to have the same surname, he may feel more similar to the provider and attributing him less responsibility for the failure.