2.1 Salespeople as a source of social discomfort during service encounters

Customers' sense of social comfort1 in the service context is defined as a psychological state “wherein the customer's anxiety concerning a service has been eased, and [wherein] he/she enjoys peace of mind and is calm and worry-free concerning service encounters with service providers” (Spake et al., 2003, p. 321). Social comfort has been consistently recognized as an important relational factor that shapes customer–brand relationship by influencing customers' affective and behavioral responses toward the brand, including commitment, satisfaction, trust (Gaur et al., 2009), word-of-mouth (Lloyd & Luk, 2011), and brand loyalty (Butcher et al., 2001). Unlike physical and physiological aspects of comfort, social comfort is a psychological outcome of social interactions between customers and salesperson. In a traditional service environment that requires human contact, a salesperson, as a direct communication bridge between brands and customers, can influence customers' affective responses through his/her attitude, behavior, and even appearance. For instance, the physical attractiveness of a salesperson activates self-presentation concerns among customers and eventually reduces patronage intentions (Wan & Wyer, 2015). Moreover, a salesperson's competence and relaxing manner increase the social comfort that customers feel during the service experience (Lloyd & Luk, 2011), whereas physical proximity to a salesperson is considered invasive and decreases social comfort in the expressive consumption context (Otterbring et al., 2021).

In sum, previous findings underscore the crucial role of salespeople as a source of social dis/comfort during service encounters. In the traditional service environment, customers had no choice but to endure social discomfort in their relationships with frontline employees. Now, however, with the help of digital innovation, customers are provided with technology that complements or even replaces human labor in various service sectors (Huang & Rust, 2018). Despite this change in service environments, the existing literature mainly examines the antecedents and outcomes of social comfort in traditional settings. Prior research in the self-service technology setting has focused on cognitive antecedents of technology adoption (e.g., Technology Acceptance Model; Davis, 1989; Marangunić & Granić, 2015). Although some studies have recognized the affective (Kulviwat et al., 2007) and social (Nasco et al., 2008) aspects of technology adoption, relatively less attention has been paid to the role of social comfort in the process. To be specific, prior self-service literature has examined social influence as an outcome of the subjective norm (Demoulin & Djelassi, 2016; Nasco et al., 2008; Park et al., 2019; Venkatesh et al., 2012) or social anxiety related to one's fear for the negative evaluation of fellow consumers (Dabholkar & Bagozzi, 2002). Only a few studies have directly addressed the impact of interaction-induced social discomfort in technology-mediated service encounters. Two qualitative studies revealed that customers use self-service technology to avoid social discomfort stemming from interactions with a “moronic salesperson” (Meuter et al., 2000), especially among customers with high dispositional social anxiety (Delacroix & Guillard, 2016). This view is echoed by Giebelhausen et al. (2014), who empirically explored the role of technology as a means to relieve social discomfort arising from a salesperson's negative behaviors that are unacceptable in terms of courtesy, responsiveness, and knowledge or from violations of the social exchange norms that guide customer–employee relationships (e.g., the absence of a warm welcome).

Taken together, past research has explored the impact of social discomfort in populations with specific individual traits, in situations where salesperson misbehaviors occur, or in the presence of fellow consumers who trigger technology-related social anxiety (see Table 1). The present study attempts to extend previous research on social influence by examining social discomfort as a sociocultural factor that reflects the change in consumer behavior in the digital age, and by demonstrating that typical in-role salesperson behaviors can also result in social discomfort among customers.

Although prior studies have mainly examined the impact of salesperson misbehaviors (Giebelhausen et al., 2014; Meuter et al., 2000), customers may also feel social discomfort from interacting with a salesperson just doing his/her job (e.g., help and persuade customers to promote sale) when the social nature of service encounters becomes salient. Indeed, service encounters go beyond mere transactions or economic exchanges in the sense that customers perceive a salesperson as a social actor and a target of impression management (Wan & Wyer, 2015). Unlike in economic exchanges where people make decisions based on monetary costs and benefits, in a social exchange of service encounter, customers seek relational rewards of respect and social approval even during a brief encounter with salespeople (i.e., social exchange theory; Blau, 1964; Bradley et al., 2010). Driven by these social needs to make good impressions of themselves, customers feel personal obligations to return the favor of a salesperson by making a purchase. However, in many cases, customers cannot achieve this goal and delay purchase decisions for various reasons: they may have uncertainty about the product, dislike its high price, or want to try out alternative options. In apprehension of such cases when they may suffer from this sense of indebtedness, customers feel social discomfort from interactions with salespeople. Indeed, though limited to those with high dispositional social anxiety, Delacroix and Guillard (2016) revealed that customers feel social discomfort from the situation they have to say no to a salesperson who has faithfully attended to their needs. The impact of social discomfort triggered by in-role behaviors of a salesperson becomes even more important in this digital age where customers can use self-service technology as a means to avoid uncomfortable human contact. With the growing use of self-service technology under digital transformation, customers feel greater social discomfort due to fewer opportunities to make human contact and less need for salespeople as brand informants.

2.2 Relational cost of interpersonal contact: Social discomfort

Under the new social distancing rules imposed during the COVID-19 pandemic, contact-free technology has become more important in service interactions. Technology has replaced human contact everywhere from the factory floor to the retail store. However, the coronavirus has merely accelerated a change in shopping habits that had been transforming the service environment ever since the retail self-checkout system was introduced in the early 1990s. Self-service technology, as a technology interface that allows for service delivery without face-to-face interactions between service providers and customers (Meuter et al., 2000), has gained popularity among service providers that pursue cost efficiency (Autor & Salomons, 2018). In this technology-infused service environment, modern consumers routinely interact with automated kiosks and tablets powered by artificial intelligence technology. More interestingly, customers now use self-service technology as a means “to avoid interacting with a moronic salesperson” (Meuter et al., 2000, p.55). Indeed, this trend of customer avoidance of human contact and of preference for technology was captured by researchers and practitioners as an issue reflecting the irreversible change in consumer behavior in the digital age (Hendriksen, 2018; Torres, 2019).

Digital innovations have provided alternative options for consumers who suffer from social discomfort and would rather interact with an automated machine than a salesperson. With less chance to socialize and interact with other people in the offline world and the diminishing role of salespeople as a source of brand information, modern consumers tend to avoid human contact and use self-service technology instead. Specifically, with the help of the Internet and high-tech devices such as smartphones, which allow for ubiquitous communication unbounded by time and space, people can communicate with each other and fulfill their social needs in their own space, without having to endure the social discomfort prevalent in the outside world (Hendriksen, 2018). As a result, digital consumers tend to cling to a few people with whom they feel comfortable and avoid having new relationships with strangers. With this tendency of digital cocooning (Kobayashi & Boase, 2014) or network privatism (S. W. Campbell, 2015), consumers have become more vulnerable to social discomfort arising from interactions with salespeople.

Digitalization has also changed consumers' decision-making process by adding a step of online information search. Traditionally, consumers acquired most of their product information on-site, and frontline employees played an important role as brand informants. Now, consumers are provided with a pool of information through the digital infrastructure of online search engines and social network services. Indeed, it has become a global trend to share personal consumption experiences on social media on a real-time basis, and these shared online reviews and ratings are considered important and credible sources of information (Nielsen, 2015). With the increased accessibility of both firm-provided and customer-created information, customers are in less need of recommendations or explanations from salespeople and are more likely to consider attention from salespeople to be invasive and bothersome.

In summary, with fewer opportunities to make contact with people outside of their comfort zone and less need for salespeople as brand informants, consumers in the digital age are easily intimidated by human interactions and instead turn to self-service technology, which offers a respite from the social discomfort in service interactions. Building on these arguments, we propose that customers would feel greater social discomfort when interacting with a frontline employee than when facing a kiosk in an unmanned store.2 This increased sense of social discomfort would eventually result in less favorable brand evaluation, which predicts long-term customer–brand relationships by measuring customers' affective responses toward the brand (i.e., brand liking, trust, desirability) and future purchase intentions. In other words, the type of service providers (human vs. self-service technology) would influence the quality of the service interaction and eventually impact customers' willingness to continue their relationships with the brands.

H1: Interpersonal contact (vs. self-service technology) decreases the social comfort that customers feel during the service encounter and subsequently overall brand evaluation.

2.3 Relational benefit of interpersonal contact: Perceived empathy

Despite the relational cost of social discomfort, human contact is still an inevitable part of the service experience in that customers want a fellow social being who can empathize with their needs and emotions. A sense of empathy, the ability to understand and respond appropriately to the affective states of others (Hoffman, 2001), is another important relational outcome of social interactions, which individuals learn and develop from the collective knowledge shared in their social relationships (i.e., theories of social construction; Gerth & Mills, 1953; Ruiz-Junco, 2017). Empathy in service encounters is defined as a service provider's ability to sense and react to a customer's thoughts, feelings, and experiences during a service encounter (Castleberry & Shepherd, 1993). Based on this genuine concern for and personal attention to customer welfare, an empathic frontline employee can better understand customer needs and therefore provide customized solutions (Giacobbe et al., 2006). In a social exchange of service encounter, these empathic concerns from salespeople result in gratitude and pleasure, which customers perceive as the relational rewards that motivate them to maintain long-term relationships with the brand/company that is represented by its salesperson (i.e., affect theory of social exchange; Lawler, 2001). Indeed, salesperson empathy has been acknowledged as playing an important role in relationship marketing by increasing customer satisfaction and brand loyalty (Bahadur et al., 2018; Wieseke et al., 2012).

Although it is theoretically possible for artificial intelligence technology to learn social intelligence, most of the commercial self-service machines are yet to reach this stage of development and mechanically repeat standardized messages rather than actually interacting with customers as social beings. Indeed, self-service machines, unlike human beings, have low social presence (Yoganathan et al., 2021) and are less likely to be perceived as a social actor that can empathize with consumers. Even for tech-savvy consumers with high technology readiness, a human touch is still an important part of the service delivery process, not only because it provides proper solutions for their problems but also because it fulfills the need for feelings of care and personal connection (Makarem et al., 2009). Indeed, recent conceptual papers have consistently recognized the empathic communication skills of human employees as a competitive advantage over artificial intelligence technology (Huang et al., 2019; Larivière et al., 2017; Pedersen, 2021; Solnet et al., 2019). In this study, we provide empirical evidence for this relational cost of using self-service technology, which is no less important than revealing its relational benefit (H1). Therefore, we propose that a customer's perception of a service provider's ability to empathize (i.e., the sense of being cared for and connected to the service provider, which we will refer to as “perceived empathy,” hereafter) increases when customers interact with a frontline employee (vs. self-service technology). This increased sense of perceived empathy eventually results in the more favorable brand evaluation.

H2: Interpersonal contact (vs. self-service technology) increases perceived empathy and subsequently overall brand evaluation.

2.4 The moderating effect of perceived persuasive intent

According to our conceptualization, interpersonal contact with a salesperson activates ambivalent affective responses from customers. While customers feel socially uncomfortable around salespeople, they still need human attention to fulfill their needs for caring and personal connection. Despite the emotional difficulties involved, we still need human contact in the service environment to compensate for the weakness of contact-free technology. Should this be the case, the objective of this study is to identify the moderating factor that can minimize the cost and maximize the benefit of interpersonal contact during service encounters. Specifically, we examine how a customer's access to a salesperson's ulterior persuasive intent impacts the quality of the relationship between the customer and salesperson.

In the traditional service environment, a salesperson plays an important role in persuading customers to make a purchase. For this reason, a salesperson is by definition one who sells, often characterized by his/her strong motivation to sell products. Indeed, in an attempt to influence customers to make a purchase decision, salespeople use various persuasion tactics such as making flattering remarks (Chan & Sengupta, 2010), providing information about limited-time offers (Balachander & Stock, 2009), and being attentive to customers by maintaining physical proximity (Otterbring et al., 2021). After being exposed to such persuasive tactics and becoming aware of the salesperson's ulterior persuasive intent, customers develop persuasion knowledge as a coping strategy to achieve their own consumption goals and avoid being swayed by the salesperson's attempts at persuasion (i.e., persuasion knowledge model; Friestad & Wright, 1994).

This process of using persuasion knowledge to identify and cope with the persuasive intent of a salesperson is emotionally demanding for customers who wish to avoid disappointing salespeople, just as salespeople wish to avoid having their offers rejected by customers (i.e., sales call anxiety; Verbeke & Bagozzi, 2000). To be specific, when exposed to the persuasive intent of a salesperson, customers are reminded of the social exchange characteristics of their relationships with the salesperson. In this social relationship governed by the rule of reciprocity, customers are often pressured by the sense of obligation to return the favor by making a purchase (i.e., social exchange theory; Clark & Mils, 1993). As a result, they feel greater social discomfort from interacting with a salesperson because they fear that they may have to disappoint the salesperson. However, when customers are less exposed to the persuasive intent of a salesperson, the relational cost of interpersonal contact will be reduced or even disappear, such that customers will feel comfortable interacting with a salesperson. Indeed, when the source of social discomfort (i.e., the salesperson's implicit display of persuasive intent) is removed, customers have no reason to avoid interacting with salespeople, who tend to be more reliable and understanding than automated kiosks.

H3: Customer perception of a service provider's persuasive intent moderates the impact of interpersonal contact on overall brand evaluation via social comfort. H3a: When a service provider's ulterior persuasive intent is more accessible to customers, interpersonal contact (vs. self-service technology) decreases social comfort and subsequently overall brand evaluation. H3b: When a service provider's ulterior persuasive intent is less accessible, interpersonal contact (vs. self-service technology) increases social comfort and subsequently overall brand evaluation.

Furthermore, when the persuasive intent underlying the salesperson's behaviors is apparent, customers become suspicious of the salesperson's manipulative intent (i.e., persuasion knowledge model; M. C. Campbell & Kirmani, 2000), and realize that the salesperson is driven by the motivation to fulfill his/her own self-interests rather than by genuine concern for the welfare of customers. Hence, we predict that when the persuasive intent of a service provider is more (vs. less) accessible, the relational benefit of interpersonal contact on perceived empathy will diminish.

H4: Perceived persuasive intent moderates the impact of interpersonal contact on overall brand evaluation via perceived empathy. H4a: When a service provider's ulterior persuasive intent is more accessible, interpersonal contact (vs. self-service technology) decreases perceived empathy and subsequently overall brand evaluation. H4b: When a service provider's ulterior persuasive intent is less accessible, interpersonal contact (vs. self-service technology) increases perceived empathy and subsequently overall brand evaluation.

Figure 1 displays the conceptual framework and the hypothesized relationships. In this study, three experimental studies examine the relational cost and benefit of interacting with self-service technology during service encounters. Using a scenario-based experiment, Study 1a reveals the impact of interpersonal contact on overall brand evaluation via social comfort (H1) and perceived empathy (H2). In Study 1b, we eliminate an alternative explanation by experimentally controlling for the impact of customer perception of service providers' persuasive intent. Study 2 expands the findings by exploring the moderating effect of perceived persuasive intent (H3 and H4).

3.1 Method

3.2 Results

3.2.2 The mediating role of social comfort and perceived empathy

A parallel mediation analysis was conducted using PROCESS Model 4 (Hayes, 2018) to explore whether social comfort (H1) and perceived empathy (H2) mediate the impact of interpersonal contact on overall brand evaluation. To enhance the internal validity of the data, we controlled for the effect of dispositional social anxiety (Delacroix & Guillard, 2016), technology readiness (Lin et al., 2007), and age (Simon & Usunier, 2007), which are known to influence self-service technology adoption.3 Using 5000 bootstrapping samples, the procedures showed a significant indirect effect of social comfort () and perceived empathy (; low = 0, high = 1). Specifically, a high level of interpersonal contact during the service encounters led to greater social discomfort (, t = −9.74, p < 0.001), which in turn decreased overall brand evaluation (, t = 7.93, p < 0.001), consistent with H1. Further, participants in the high interpersonal contact condition felt greater perceived empathy toward the company (, t = 6.69, p < 0.001), which in turn enhanced overall brand evaluation (, t = 4.23, p < 0.001; see Figure 2; for standardized coefficients, see Appendix SD), consistent with H2.

3.3 Discussion

The results of Study 1a reveal the emotional consequences of a technology-infused service environment, with a focus on social comfort and perceived empathy. On the one hand, using self-service technology offers customers a respite from the social discomfort that they feel in traditional service settings. On the other hand, when customers have direct contact with a frontline employee, interpersonal contact increases perceived empathy, such that customers feel that the company cares about and empathizes with their needs. These results highlight both the dark and bright sides of human contact during service encounters, supporting H1 and H2.

Nevertheless, one may suggest an alternative explanation for the observed effect. In Study 1a, we posited a situation in which participants in the high interpersonal contact condition were highly exposed to a salesclerk's persuasive intent to make sales. It is difficult to tell whether the observed effect is due to differences in the accessibility of ulterior persuasive intent or in the level of human interaction. To eliminate the alternative explanation, we turn to Study 1b, which experimentally controls for the impact of customer perception of service providers' persuasive intent.

4.1 Method

4.2 Results

4.2.2 The mediating roles of social comfort and perceived empathy

The results of a bootstrapping analysis (PROCESS Model 4; 5000 resample; Hayes, 2018) with technology readiness and age as covariates showed a significant indirect effect of social comfort () and perceived empathy (; low = 0, high = 1). Specifically, a high level of interpersonal contact during the service encounters decreased social comfort (, t = −5.88, p < 0.001) and in turn decreased overall brand evaluation ( = 0.33, t = 5.17, p < 0.001), consistent with H1. The results also supported H2, showing that participants in the high interpersonal contact condition thought more highly of the company's ability to empathize with customer needs (, t = 4.86, p < 0.001), and in turn evaluated the brand more favorably (, t = 4.78, p < 0.001; see Figure 3; for standardized coefficients, see Appendix SD).

4.3 Discussion

Study 1b replicates the findings from Study 1a while ruling out the alternative explanation. The results show that customers feel greater social discomfort when a salesclerk is eager to persuade them to buy a product, whereas persuasive messages from a tablet PC are considered relatively less burdensome. These results suggest that the same persuasive message triggers different emotional responses from customers, depending on the type of service agent that delivers the message. Furthermore, we reconfirmed the finding that interpersonal contact has a positive impact on customers' perceptions of the company's empathic capabilities. When customers have interactions with frontline employee, they feel that their needs will be met with greater care and empathy in their relationships with the service provider.

5.1 Method

5.2 Results

5.2.2 The mediating roles of social comfort and perceived empathy

The results of a mediation analysis with technology readiness and age as covariates confirmed a significant indirect effect of social comfort () and perceived empathy () in the relationship between interpersonal contact (low = 0, high = 1) and overall brand evaluation. Specifically, a high level of interpersonal contact during the service encounters led to greater social discomfort among customers (, t = −1.98, p < 0.05), and in turn had a detrimental impact on their evaluations of the brand (, t = 11.43, p < 0.001), supporting H1. Further, participants in the high interpersonal contact condition perceived greater empathy from the service provider (, t = 12.24, p < 0.001), and in turn made more favorable evaluations of the brand (, t = 7.90, p < 0.001; see Figure 4), consistent with H2.

5.2.3 The moderating effect of accessibility of persuasive intent

We conducted a mediated moderation analysis (Hayes, 2018; PROCESS Model 8; 5000 resamples) with covariates of technology readiness and age. The results showed a significant interaction effect of interpersonal contact (low = 0, high = 1) × persuasive intent (low = 0, high = 1) on overall brand evaluation via social comfort () and perceived empathy (), supporting H3 and H4. As illustrated in Figure 5, interpersonal contact and persuasive intent had a significant interaction effect on perceived social comfort () and empathy (), which in turn significantly influenced customers' evaluations of the brand, respectively.

To look more deeply into the moderation effect, we present a graph plotting significant interactions using the estimates obtained from the simple slope analyses. Figure 6 shows that when the service provider's persuasive intent was highly accessible, interpersonal contact decreased the social comfort felt during the service encounter. However, the pattern of this result was reversed in the less accessible condition: participants felt more comfortable when they had interactions with a frontline employee (vs. a self-service machine). These results suggest that there is a chance to overcome the downside of the traditional service environment by changing the behaviors of salespeople.

In addition to reconfirming the finding that interpersonal contact improves customer evaluations of the company's level of care and empathy, we discovered the fact that sales-pressing persuasive gestures from a salesclerk reduce the beneficial effect of human contact on perceived empathy (see Figure 7).

5.3 Discussion

Study 2 not only replicates the findings from the previous studies but also shows that accessibility to persuasive intent moderates the impact of interpersonal contact on brand evaluation. Specifically, Study 2 provides further evidence of the relational benefit and cost of technology-involved service encounters: that is, using self-service technology increases social comfort and decreases perceived empathy, compared to human interactions. More importantly, Study 2 provides insights that can help in managing service personnel. The study demonstrates the moderating effect of perceived persuasive intent and shows that greater empathy on the part of service personnel can compensate for the weaknesses of self-service technology. The results also reveal that the beneficial impact of interpersonal contact on perceived empathy declines as the persuasive intent of a service provider becomes accessible. More interestingly, the results show that the detrimental impact of human interactions on social comfort disappears when salespeople cease their attempts to persuade customers to make purchases. Indeed, customers feel even more comfortable interacting with a frontline employee than using self-service technology when the persuasive intent is not accessible, suggesting that customers feel uncomfortable interacting with a salesperson because of the social nature of service encounters. A salesperson's display of persuasive intent highlights the social exchange characteristics of the service interaction and increases a sense of obligation to make a purchase. Consequently, customers feel uncomfortable when they have to leave the store empty-handed, to the disappointment of a salesperson. When the salesperson does not use persuasive tactics, however, service interactions become less stressful because the main source of social discomfort disappears.

6.1 Theoretical implications

Our research not only reveals that social comfort is an underlying mechanism behind customer preference for self-service technology over human contact but also recognizes that self-service technology suffers from an inability to empathize with customer needs. In addition to exploring the relational benefit and cost of contact-free technology, our research identifies the moderating effect of a service provider's persuasive intent. The findings of our study have several theoretical implications.

First, our research contributes to the self-service technology literature by exploring the role of social discomfort as a relational cost of human interactions that motivate customers to prefer self-service technology over human contact. Based on the social exchange theory, we identified the relational benefit of using self-service technology to spare customers the trouble of interacting with a salesperson, which eventually improves overall brand evaluation in a situation where typical salesperson behaviors might increase social discomfort.

Second, our research illuminates the relational cost of using self-service technology by demonstrating the indirect effect of perceived empathy. That is, customers think highly of a service provider's ability to empathize with their needs, which in turn results in more favorable brand evaluations. This finding contributes to the literature by extending the scope of prior research which has addressed this issue at a conceptual level or in the traditional setting. In this way, the present research reveals that perceived empathy plays an important role even in technology-infused service environments as a competitive strength of human contact that complements the weakness of self-service technology.

Third, this study sheds additional light on customers' use of persuasion knowledge by demonstrating the moderating effect of the perceived persuasive intent of salespeople. Interpersonal contact decreases social comfort when the persuasive intent is highly accessible. When the persuasive intent is less accessible, however, the relational cost of human interactions disappears, and customers feel socially comfortable interacting with salespeople. While prior research mainly focused on the cognitive aspect of the use of persuasion knowledge in one-time encounter, we offer a new perspective on the role of self-service technology as a buffer against the activation of persuasion knowledge and its negative emotional consequence (i.e., social discomfort), which eventually impacts customer-brand relationships in the long run.

6.2 Practical implications

The findings also provide several implications for marketing and human resources managers. Specifically, the results reveal the underlying mechanism behind the growing customer desire for zero contact by demonstrating that customers feel social discomfort even from in-role behaviors of salespeople. Frontline employees need to understand that some customers dislike aggressive attempts to provide brand information but appreciate caring and empathic personal connections with brand representatives.

Considering that human contact is still an inevitable part of the service environment, our finding has an important implication for managers of frontline employees. Customers are found to feel comfortable interacting with salespeople when the persuasive intent is less accessible, suggesting that there is a chance to overcome the downside of the traditional service environment. We suggest that brand managers work closely with their human resources and operations groups to hire employees with high emotional intelligence and to advise these employees against the negative emotional consequences of aggressive persuasion tactics.

Finally, Study 2 reveals that a customer's affective response toward a persuasive message differs depending on the type of the sales agent. Whereas exposure to a salesperson's persuasive intent decreases social comfort among customers and eventually diminishes the persuasive effect of the message, a machine-delivered persuasive message is perceived as neutral information and has no significant impact on social discomfort and overall brand evaluation. This finding suggests that managers can reduce the negative consequences of a persuasive message by delivering it through self-service technology rather than a salesperson. Future research can probe this difference in customers' perceptions of persuasive intent among more advanced types of artificial intelligence technology such as chatbots and service robots, which are considered effective marketing tools that can complement human labor during the pandemic and afterward.

6.3 Limitations and future research

We also recognize several limitations that provide room for future research. First, future research may extend our findings by examining the relational cost and benefit of interpersonal contact in other service settings and among customers with individual differences (e.g., high or low prior knowledge). For instance, it will be worthwhile to explore the moderating effect of service complexity or product type (e.g., luxury vs. nonluxury5) and figure out the particular service areas where customers would prefer self-service technology to avoid social discomfort from human contact.

Second, whereas our research examined the salesperson behaviors and attitudes that trigger social discomfort among customers, further research could investigate the impact of other easily manageable salesperson characteristics (e.g., the brand-aligned physical appearance of a salesperson).

Third, the experiments in the current research were conducted with participants from Western countries. Future research can extend our findings by exploring the moderating effect of cultural differences. Such differences may influence the extent to which interpersonal contact activates social discomfort, such that customers from individualistic (vs. collectivistic) cultures are less likely to feel socially uncomfortable around salespeople with self-concepts that are independent from their group identities (Kitayama & Markus, 1999).

Fourth, although this study has focused on the negative impact of digital technology that increases social discomfort in weak-tie relationships, it would also be interesting to explore the positive aspect of network privatism that technology-mediated communication can enhance social bonds in strong-tie relationships. For instance, future research can extend our findings by examining the impact of relationship norms (communal vs. exchange) on social discomfort among digital consumers.