Online brand communities (OBCs) represent a strategic asset for firms to enhance their consumers’ service experience. However, market research highlights that interactions in OBCs have become increasingly hostile, with over 50% of users in the UK and the US experiencing different forms of online aggression on a regular basis (Ditch, 2018; PEW, 2021). Specifically, recent reports show that this phenomenon has further intensified since the start of the COVID-19 pandemic, as more consumers visit and interact in OBCs (BBC, 2021). At the same time, service providers have struggled to adjust to the new service regulations and demands, leading to an increase in service failures and, inevitably, dissatisfied and complaining consumers (Ozuem et al., 2021). Recent research further suggests that service providers regularly fail to respond to these complaints (Wang et al., 2021), and that such failures can lead to an increase in subsequent hostile interactions among consumers, hereafter referred to as consumer-to-consumer (C2C) conflicts (Bacile, 2020; Dineva et al., 2017).
In this paper, we offer first insights into the sources of these conflicts, and how service providers respond to them, with a particular focus on COVID-19-related incidents. While service recovery research has examined B2C service recovery interactions (Gelbrich and Roschk, 2011), such as the provider offering an apology or compensation for a service failure, less is known about the hostile C2C interactions that occur afterwards (Bacile et al., 2018), and particularly how these may be managed by the provider (Dineva et al., 2021). To address this lack of knowledge, we observed 13 OBCs on Facebook that are hosted by service firms, purposefully sampling C2C conflicts related to the COVID-19 pandemic. Using double coding and a hybrid approach to thematic analysis (Fereday and Muir-Cochrane, 2006), five C2C conflict sources and five response strategies emerged from the data. First, we identified “brand attack”, “brand dissatisfaction”, “brand scepticism”, “brand contention” and “brand defence” as distinct sources of C2C conflict that lie on a continuum with individualistic orientation on the one extreme and collectivistic on the other. Second, we uncovered several response strategies – “non-engaging”, “automatic”, “bolstering”, “direct/indirect asserting”, and “factual/empathetic/apologetic informing” utilized by service providers, which we differentiated based on their passive versus active approach and authoritative versus cooperative orientation.
Our findings therefore contribute to the service failure literature by addressing the general lack of knowledge on the sources of post-failure C2C conflicts, offering a first step to conceptualizing a relatively novel phenomenon. We further contribute to the service recovery research by offering real-life evidence on how firms respond to these conflicts, an area that has so far been limited to fictional scenarios (Bacile et al., 2018), which address service failures but not the C2C conflict itself. For service practitioners, we are first to offer an insight into different response options that social media service teams can utilize to handle C2C conflicts, and specifically those relating to the current and urgent COVID-19 challenge.

Stereotypes and discrimination in the marketplace have recently received more attention in marketing discussions. In service literature, perceptions of discrimination have been investigated as occurring at different stages on the consumer journey: pre-purchase, including the treatment received from the salesperson; purchase; and post-purchase, encompassing service recovery (Baker et al 2008). Despite the importance of the last stage of the consumer journey in the loyalty loop, studies investigating discrimination focus mainly on racial issues and do not consider the compensation offered to socially dissimilar consumers.
We suggest that in a comparison of the compensation received after a service failure, when the target consumer is from the upper class and receives more compensation than someone they compare themselves with socially, or when the target consumer is from a lower social class and receives less compensation, then perceptions of discrimination will arise. This perceived discrimination could lead to intentions to spread NeWOM, but probably only by those who did not receive the benefit of compensation.
We performed two experimental studies involving service scenarios (restaurant and hotel) with manipulation of financial compensation after a service failure (10% vs.40% vs. 70% discount), followed by a social class comparison (upper vs. lower vs. similar) between two consumers.
Study 1 used Mturk respondents (n=230, 56%male, Mage= 36, s.d.=11.81). Those participants who received lower compensation than a target consumer from an upper social class perceived discrimination (F (1,229) = 16.72, p= .001, Mup= 5.46, Mcontr= 3.36, Mlow= 3.79), as did participants who received greater compensation than someone from a lower class (Mup= 3.93, Mcontr= 3.33, Mlow= 5.21). Moderated mediation analysis (Process Model 7) confirmed that the indirect effect of compensation on NeWOM following discrimination was moderated by social class (index: CI -.39 to -.14) when the target individual was perceived to be from a lower class (b= .20, CI .09 to .34) and when he/she was perceived to be from an upper social class (b= -.21, CI -.36 to -.10). Hence, and surprisingly, when discrimination is perceived to exist, not only would individuals who were discriminated against intend spreading NeWOM, but so would those who perceived discrimination against someone else.
Study 2 used data collected on Prolific (n=120, 58% male, Mage= 30, s.d.= 10.14). Perceptions of discrimination were similar to those in Study 1. Interestingly, participants who received higher compensation than someone from the lower class reported greater intentions to adopt NeWOM (M= 2.71, SD= 2.03) than respondents who received lower compensation than someone from an upper class (M= 2.03, SD= 1.20). As with Study 1, the results suggest that the indirect effect of compensation on NeWOM following discrimination was moderated by social class (moderated mediation index: CI -.38 to -.03).
This paper contributes to the literature about service recovery by exploring the idea that offering differential treatment to dissimilar customers after a service failure may result in negative reactions, not only from the customer who was prejudiced, but also from customers who benefited from the situation.

Relevance of the phenomenon being studied

The rise of new media allows customers to share their negative experiences with a certain brand or with a large audience. Marketers have different options to address the complaints and also satisfy the mass audience on social media. In practice, many companies use a humorous response strategy to address complaints. We want to shed further light on this topic and show that the effect of a humorous response is very different for products and services. Moreover, although the observers might be entertained by a humorous response, other negative aspects, as being perceived as less reliable might be possible.
Research questions

(1) Is a humorous response strategy to online complaints appropriate for services (vs. products)?
(2) How does a humorous response to an online complaint affects different brand personality dimensions?

Potential contributions to the field

We contribute to the field as we show that service organization should be cautious in applying humorous response strategies on social media. We show that a humorous response does not affect all brand personality dimensions in the same way and provide an explanation based on the halo effect.

Theoretical foundations

Scholars suggest that the positive effect of humor can be seen as a halo effect. Grounded in social psychology, the halo effect describes the phenomenon of “the influence of a global evaluation on evaluations of individual attributes of a person” In the same way as for individuals, we propose that humor does not only trigger positive associations with brands, but can also be associated with negative traits.

Methodology

We performed two scenario based online experiments during the time from spring 2019 to January 2020. In study 1 we tested the basic assumption that a humorous response (vs. neutral response) affects brand personality dimensions differently. The goal of study 2 was to see if a humorous answer has different effects depending on the context (service vs. product).

Findings and discussion

We show that a humorous response forms brand personality different depending on the context. In a service context (vs. product) a humorous response elicits fewer positive outcomes. Our results are in line with past research on humor use in advertisements and a social context. In advertisement research it could be showed that humorous ads are less effective when consumers are more involved. Since, services are characterized with higher risk and involvement, humor might be less beneficial as well. Moreover, in social context humorous individuals are not only associated with positive attributes, but also seen as more arrogant and less reliable.

Conclusion

We show that despite positive outcomes humor also brings undesirable outcomes such as the brand might be perceived as less conscientious. Past research focused on the potential negative effects of humor on purchase intention and memory but we are, to the best of our knowledge, the first who showed that brand personality can suffer under certain conditions as well. Moreover, we show that service (vs. product) organization must be cautious and might be apply different response strategies.

With 4.2 billion active social media users, social media platforms have become an important place for brands to interact with and support (potential) customers. On Twitter alone, over 400 million brand-related tweets are published annually, with brands seeing a 15 percent volume increase of inquiries and mentions during the first wave of the Covid19 pandemic. In this context, brands are investing significantly in social media customer care (SMCC hereafter) to address consumers’ specific questions, compliments, and complaints on social media.

Many brands, however, struggle to deal with online customer messages. The lack of best practices and guidelines for effective SMCC is disturbing given its impact on focal and observing customers. While SMCC research is rapidly developing in a variety of disciplines, including service, marketing, hospitality and tourism, management information systems and communication sciences, individual papers cover specific and isolated aspects of SMCC and lack definitional and conceptual precision. This situation, in turn, limits our understanding of the subject and fails to put forth an integrated view of SMCC that would serve as a strong theoretical foundation. In response, this article seeks to contribute to the literature in three ways.

First, to overcome definitional ambiguities, we look at the conceptual foundations of SMCC, summarize and evaluate the various ways SMCC has been defined, and propose a new and comprehensive definition of SMCC, entailing the act of engaging in social media interactions initiated by individual customers to support them and address their messages to develop customer and brand equity. We also discuss how SMCC connects to and differs from related concepts such as electronic word-of-mouth, service recovery, crisis communication and social media marketing to further establish conceptual clarity.

Second, building on our SMCC definition, we consolidate prior findings from 140 articles selected through a structured literature review process. This results in a conceptual framework anchored in communication and relationship marketing theory, and organized around (1) experience (i.e., cognitive, affective, social), (2) design (i.e., SMCC content and framing), (3) operational (i.e., SMCC source, channel, and timing), and (4) contextual (i.e., customer message and brand characteristics) elements driving brand-relevant outcomes. In connection, we distill a set of concrete managerial take-aways, reflecting the do’s and don’ts of SMCC.

Third, we connect our synthesis of existing SMCC academic knowledge with insights derived from practitioner reports and a set of 13 expert interviews to identify core themes—SMCC source status (manager vs employee), initiation (proactive vs reactive), nature (human vs AI), transparency (public vs private), and customer message legitimacy (fair vs unfair)—that reflect prior research inconsistencies and unique academic-practice gaps where novel research is particularly needed. The identified inconsistencies and gaps together have the potential to guide SMCC practice and serve as input for the further expansion of research on SMCC by pointing to promising research directions.