When service robots make mistakes: how does customers’ mood regulation affect their continuance intention to adopt?
1. Phenomenon under investigation
The rapid development and implementation of RAISA (service robots, artificial intelligence (AI) and service automation) have changed how services being delivered and experienced (Prentice et al., 2020). Thus, academics and practitioners propose that service robots become an integral part of human life (Tung and Law, 2017), and human labour will be replaced (Tuomi, Tussyadiah, and Stienmetz, 2021). However, customers’ adoption level of service robots is still quite low. Scholars attempt to examine factors that affect customer adoption of service robots, so that they can enhance their adoption. Factors related to robot characteristics, customer characteristics, and robot-customer interaction are found crucial in determining their adoption (Mende et al., 2019). Specifically, factors such as ease of use (Turja et al., 2020), gender (Seo, 2022), and anthropomorphism level (Chen et al., 2022) can affect customers’ service robot adoption to various degrees. However, this technology is still in its infancy and thus, unsuccessful outcomes of service robots, such as glitches, unexpected negative events and robots not living up to promises must be considered as they will significantly influence customers’ assessments of the service provider (Holloway and Beatty, 2003; Dabholkar and Spaid, 2012). A few scholars look into issues related to robot failure and how it affects customers’ adoption. For instance, Belanche et al. (2020) found that respondents made stronger attributions of responsibility for the service performance toward human employees than toward robot employees, particularly after a service failure. Moreover, Yang et al. (2022) examined the effectiveness of humour in robot failures suggesting that its effectiveness varies based on failure severity level. However, extant work fails to explore whether customer adoption of service robots will be affected after service failures and how customers’ affective states and individual traits affect their intentions to continue to adopt service robots.
2. Potential contributions and research questions
Factors such as a person’s affective state and mood (as well as their emotional reactions to them) largely impact their decisions to accept new technologies (Djamasbi et al., 2010). In particular, their continuous adoption of specific technology is also determined by their perceptions, such as trust (Tussyadiah et al., 2020). In reality, the affective state and “how people feel” when using technology are decisive factors, as emotional systems help define our rational decisions in conjunction with rational thought (Hanoch, 2002; Muramatsu and Hanoch, 2005). Recent studies have investigated the factors motivating customers to use these robots in service interactions (e.g., Lu et al., 2019; Liu et al., 2022). However, most of work lacks to take consumers’ affective states and individual characteristics into consideration when examining factors on their adoption intention after a service failure. Given the importance, this study aims to better understand how customers’ mood regulation influences their continued adoption of service robots after encountering unsuccessful service with a frontline robot. This research serves as a valuable contribution to the field of RAISA because it advances our understanding of the impacts that an individual’s mood has on their personality traits, which, in turn, influences their decisions to adopt new technology.
3. Theoretical Foundations
Customers’ mood regulation and adoption
Mood causes a differential impact on behaviour and advise that this has a complex relationship with people’s personality traits (Karimi and Liu, 2020). As non-specific affective states, moods can have a powerful impact on both behaviour and cognition, previous studies advising that mood can not only impact behaviour to a significant degree but also the cognition (Lischetzke and Eid, 2003; Das and Fennis, 2008). Biss et al. (2010) highlight that these enduring affective states may have positive or negative valence.
The process that people use to manage their affective states is known as mood regulation (Koole, 2010). According to Forgas (1995), the Affect Infusion Model (AIM), puts forward two different mechanisms that describe how mood impacts decisions and judgements, affect-as-information and affect-priming. In affect-as-information mechanisms, the affective state is used as a shortcut to infer evaluations and inform decisions. Therefore, consumer behaviour and choices are triggered by the behaviour that results from the influence of mood (Geen, 1995). Since consumers have different information processing behaviours (Karimi et al., 2018; 2020), their mood can affect their adoption decisions in various ways.
The mediating role of affective states
User perceptions of technology characteristics are impacted by affective states (Darban and Polites, 2016). Considering the acceptance and adoption of technologies, previous studies (e.g., Hoong et al., 2017; Verkijika, 2020) identified that people’s emotions and feelings are the most pivotal factor, outweighing other factors like the perception of risks or benefits (Chuah, 2021). Specifically, when people are in a positive mood, this affects their level of acceptance or support (Karimi and Liu, 2020). Meanwhile, Jobin et al. (2019) advise that research has not only found that people’s evaluation of information is guided by affective reactions but also that this consequently influences their acceptance of technology.
The moderating role of service failure
Service failure will become unavoidable as service robot technology becomes more commonplace; suggesting that customer satisfaction will be adversely affected by service robot failures such as preparing the wrong meal, providing incorrect directions or over-charging customers (Yam et al., 2021). Customers have certain expectations and when service falls below this standard, they will react accordingly (Hoffman and Bateson, 2010). Thus, when customers become dissatisfied through such service failures, they frequently respond with anger (Sliter et al., 2010; Wilson and Holmvall, 2013).
However, Schwarz and Clore (1983) propose that when people experience a positive mood, their affective state acts to inform their behaviour-related judgements; although satisficers may be affected by the informational impact, their pragmatism and judgement mean that they use their positive mood as information and decide to continue to use service-bots. Thus, customers who are experiencing a positive mood will be more forgiving of the failures of service robots (Yam et al., 2021). Building upon the above-mentioned conclusions and reasoning, we therefore propose the following hypotheses:
Hypothesis 1. In a positive (vs. negative) mood regulation, customers are more willing to adopt service robots after a service failure.
Hypothesis 2. The effect proposed is mediated by customers’ processed affective states.
Hypothesis 3. Robot service failure moderates the indirect effect of mood regulation on intention to adopt.
This paper will conduct one correlational research (Study 1) and two experiments (Study 2A, Study 2B) to evaluate the hypotheses. Specifically, Study 1 investigates the correlations between the mood regulations that clients experience and their likelihood of continuing to use service robots. Study 2A uses an experimental method to investigate how different types of mood regulations affect participants’ affective responses, while Study 2B investigates how experimentally manipulated mood regulations facilitate customers’ intention to continue using service robots through the mediator (affective states) and the moderator (service failures). This study plans to conduct a lab experiment in a Chinese university and also an online experiment to hire samples from MTurk to take part this study.
5. Expected findings and conclusion
In conclusion, service recovery is one of the most critical initiatives required to address faults in the service delivery process and turn service failures into positive service outcomes (Chen and Tussyadiah, 2021). This study strives to add new knowledge to the pertinent literature in this field in several ways. This research will shed lights on how mood control can encourage customers’ continued intention to embrace service robots; the expected findings will experimentally evaluate and record the mediating function of customers’ affective states. Moreover, this study attempts to contribute to the literature via experiments to investigate the relationship between service failure and mood regulation. Expected findings will provide important insights into consumer behaviour and suggest that affective states plays a critical role in determining consumers’ preferences and behaviours when they encounter service failures.