Humor and webcare. Supervisor: Dr. C.C. Liebrecht Second Reader: Dr. J. Bartels. Simone Schaffelaars ANR January, 2018

Transcription

1 ` Humor and webcare The effect of humor in webcare responses to negative electronic word-of-mouth on the attitude towards the webcare response for familiar and unfamiliar brands Master Thesis Communication and Information Sciences Specialization Business Communication and Digital Media Faculty of Humanities Tilburg University, Tilburg Supervisor: Dr. C.C. Liebrecht Second Reader: Dr. J. Bartels Simone Schaffelaars ANR January, 2018

2 Abstract Due to the rise of web 2.0, consumers and brands interact more frequently via social media platforms. Many research has investigated which tone of voice companies should apply in those online interactions, and has shown that a conversational human voice (CHV) can be beneficial for organizations. Despite the fact that humor is a part of the CHV, and companies are increasingly adopting humor to cope with electronic word of mouth (ewom), little research exists about the use of humor specifically in the context of webcare. In this paper, the use of humor in webcare responses towards negative ewom (NeWOM) messages was investigated with the use of a 3 (humor: humorless vs. affiliative humor vs. self-defeating humor) x 2 (brand familiarity: familiar vs. unfamiliar) between-subjects design. The results of an online questionnaire, in which 166 respondents participated, suggest that humor in webcare responses does not lead to a more favorable attitude towards the webcare response compared to humorless responses. This is applicable to both familiar as well as unfamiliar brands, thus brand familiarity does not play a role. However, the use of humor does positively influence the perceived CHV, and a higher CHV leads to a more favorable attitude towards a webcare response. Yet, the use of humor also has a negative influence on appropriateness, which in turn influences attitude towards the webcare response negatively. Future research is needed to investigate in which context humor would be appropriate, and is thus suitable for a webcare response. Keywords: Webcare, Humor, NeWOM, CHV, Appropriateness, Brand familiarity 2

3 Table of contents Introduction... 4 Theoretical framework... 6 NeWOM and webcare... 6 Conversational human voice... 8 Appropriateness Humor and types of humor Brand familiarity Hypotheses and conceptual model Method Design Participants Materials Pre-test Measures Procedure Results Demographic variables Hypothesis testing Differences in NeWOM messages Conclusion and discussion Limitations and explanations for findings Future research References Appendix A Pre-test materials Appendix B Pre-test questionnaire Appendix C Measures Appendix D Final questionnaire Appendix E Final materials Appendix F Pre-test results Appendix G Reliability

4 Introduction More than ever, consumers are able to voice their opinions about a brand (either positive or negative) in the form of electronic word-of-mouth, also known as ewom (Gruen, Osmonbekov, & Czaplewski, 2006). Also, questions from consumers to organizations appear more often online. As an example, a student from Tilburg University posted a question on the University s Facebook page: Can we print again tomorrow, since some of us like to hand in their thesis?! ;). Many companies have employees who engage in online interactions with (complaining) consumers, by actively searching the web to address certain consumer feedback (e.g., questions, concerns, and complaints), also known as webcare (van Noort, Willemsen, & Antheunis, 2013). This is also the case for Tilburg University. The question from the student was followed by a webcare response from the University as illustrated in Figure 1. Figure 1. Humorous webcare response (Tilburg University, n.d.). In online webcare conversations with consumers, companies tend to communicate with a more conversational human voice (CHV) instead of their corporate voice (Searls & Weinberger, 2000; Kelleher & Miller, 2006). One of the ways to apply a more CHV is to use humor (Kelleher & Miller, 2006), just like Tilburg University did in the example above. By applying humor, the company s voice becomes more humanized, which makes consumers think they are communicating with a person rather than a company (Kwon & Sung, 2011). 4

5 Research has shown that the use of a CHV leads to a positive brand attitude, a positive reputation, and positively influences the credibility and trust of a company (Kerkhof, Beukenboom, Utz, & de Waard, 2010; Yang, Kang, & Johnson, 2010). However, little research exists about the use of humor specifically. Previous research on humor in responses to negative online reviews has shown that humorous responses were followed by more positive reviews from consumers. Besides that, compared to non-humorous responses, humorous responses exhibited more positive effects on consumers brand attitude (Zhu, Zhang, & Gong, 2017). However, little research has been conducted for humor in a webcare context specifically, while humor has been broadly investigated in other fields such as customer service. For example, previous research by Mathies, Chiew, and Kleinaltenkamp (2016) has shown that the use of humor by customer service employees may improve customer experience. However, Söderlund, Oikarinen, and Heikka (2017) state that jokes in customer service encounters reduce customer satisfaction. Those mixed results indicate that that the right adoption of humor might be a real challenge for companies. Especially since humor is not appreciated to the same extent by all individuals (Xiao, Cauberghe, & Hudders, 2017). Moreover, humor can be interpreted differently than intended, which might lead to miscommunication (Hay, 2001). There are different factors that are likely to affect consumers their response to humor (Fugate, Gotlieb, & Bolton, 2000). Especially, specific brand characteristics such as brand familiarity might play an important role in the appreciation of, and attitude towards humorous responses. Previous research investigated the role of brand familiarity when communicating with a CHV (Gretry, Horváth, Belei, & van Riel, 2017). They specifically examined informal language as part of the CHV. Results showed that a more informal communication style is perceived more appropriate for familiar brands compared to unfamiliar brands. This is in line with the Role Theory that posits that successful interactions depend on whether relationship 5

6 partners behave appropriately according to their specific role in a relationship (Sarbin & Allen, 1968; Schewe, 1973; Solomon, Surprenant, Czepiel, & Gutman, 1985). Therefore, an informal style is more appreciated when someone is familiar with a brand. Taking the Role Theory into account, it is plausible that the use of humor will be found more appropriate and is thus more appreciated for familiar brands as well. To contribute to more understanding about how consumers perceive humorous webcare responses and their attitude towards those responses, this research aims to investigate the effect of humor in webcare responses towards NeWOM on the attitude towards the webcare response. By doing so, this research contributes to more insight into the use of humor, specifically in the field of webcare where insight about humor is lacking. This research aims to answer the following research question: What is the effect of humor in webcare responses towards NeWOM on the attitude towards the webcare response and what is the role of brand familiarity in this relationship? The results of this research can positively contribute to the webcare challenges that many companies face daily. In particular, webcare practitioners will gain more insight in the overall effect of humor in webcare responses, but also for specific humor styles. With those insights, companies can decide what strategy to apply in their webcare responses. Theoretical framework NeWOM and webcare Due to the rise of social networking sites, the number of online consumer-brand interactions is exploding (Forrester, 2014). The amount of online conversations is not only rising, but the presence of social media has also changed the way organizations and consumers interact with one another (Dijkmans, Kerkhof, Buyukcan-Tetik, & Beukeboom, 2015). While in the past communication between consumers and companies was mostly oneway interaction in which the organization controlled the content, it is now mostly the 6

7 consumer who is in charge (Hinchcliffe, 2006). Social media enables consumers to interact with the company in such a way that two-way-interaction takes place, and consumers are increasingly taking an active role in online brand and product conversations by sharing their experiences. This results in a major growth of ewom, in which the consumers control the generated content (Hennig-Thureau et al., 2010). To maintain and reinforce a good reputation, various companies have employees who communicate online with complaining consumers, and thus practice webcare. Webcare aims to serve three main goals: customer care, marketing, and public relations (van Noort, Willemsen, Kerkhof, & Verhoeven, 2014). With regard to customer care, the goal of webcare is to signal customer problems and solve it to the satisfaction of the customer and meet his expectations. Webcare also has a marketing function, because the complaints or suggestions people post online can be used to improve certain products and or services. Lastly, public relations is one of the main goals of webcare, and can be used as a means of reputation management but also relationship management. In practice, this means that companies can respond to NeWOM before something might escalate into a large crisis. In this way, a company s reputation is managed which is the most important goal of webcare according to van Noort et al. (2013). Previous research suggests that a response to a NeWOM message is more appreciated compared to the absence of a response (van Noort & Willemsen, 2012). As a result, various scholars have addressed the question what response strategy to apply in webcare (Huibers & Verhoeven, 2014; Grégoire, Salle, & Tripp, 2015). Furthermore, other factors in webcare responses have been investigated, such as a reactive and proactive approach (van Noort & Willemsen, 2012), the channel and platform (Ghosh, 2017; Schamari & Schaefers, 2015), timing and strength of the response (Ghosh, 2017), and the nature of the ewom (Joireman, Grégoire, Devezer, & Tripp, 2013). Moreover, a considerable amount of literature on webcare 7

8 has focused on the tone of voice, and especially the conversational human voice (Kelleher, 2009; Kelleher & Miller, 2006; Kerkhof, Beugels, Utz, & Beukeboom, 2011; Schamari & Schaefers, 2015; Park & Lee, 2012; Kwon & Sung, 2011; Gretry et al., 2017, van Noort et al., 2014). In line with previous research, this study also explores the role of CHV in webcare responses. Conversational human voice The communication style companies adopt online is known to be an important characteristic in online communications (Searls & Weinberger, 2000). With regard to webcare, this means that besides a suitable response strategy, the tone of voice in this strategy is also of main importance. Drawing upon previous research by van Noort and Willemsen (2012), two main communication styles can be distinguished: a corporate tone of voice and a conversational human voice (CHV). Kelleher (2009) describes a conversational human voice as an engaging and natural style of organizational communication as perceived by an organization s publics based on interactions between individuals in the organization and individuals in publics (p.177), whereas communication via a corporate voice is perceived as dominant, profit-driven and persuasive (Levine, Locke, Searls, & Weinberger, 2000). However, the definition of the CHV is problematic since a CHV consists of numerous different elements, and many different terms are used to refer to a CHV, such as human voice, anthropomorphism, personalization, personal tone of voice, and so on. For companies to demonstrate a level of CHV, they welcome a conversational communication, which means that their communication suggests that they are open to dialog (Kelleher, 2009). However, the use of humor, admitting mistakes, treating other as human, and providing links to competitors also contribute to a perceived CHV (Kelleher, 2009; Kelleher & Miller, 2006). The difficulty to agree upon one definition is also noticeable in previous research on CHV. Whereas some scholars investigate a CHV with use of visuals (Park & Lee, 2012), others investigate other 8

9 factors that contribute to the human voice such as message personalization (Schamari & Schaefers, 2015; Crijns, Cauberghe, Hudders, & Claeys, 2017) and informal language (Gretry et al., 2017). The CHV has been a main topic of interest in the field of online corporate communication, and results suggest that a CHV is effective in online communication, due to several positive effects. Firstly, it has been shown that a CHV can lead to a higher level of trust, satisfaction, and commitment (Kelleher, 2009; Beldad, de Jong, & Steehouder, 2010; Sweetser & Metzgar, 2007; Yang et al., 2010). Secondly, a human communication style enhances consumers evaluations of the firm adopting the CHV (Kelleher, 2009; Kelleher & Miller, 2006; van Noort & Willemsen, 2012). Thirdly, a CHV enhances the perceived effectiveness of the communication (Weinberg & Pehlivan, 2011). However, there are also adverse side effects on the human voice. Previous research by Koot (2013) has shown that messages that are both personalized and contain informal language, negatively influence brand perception. Also, Demmers, van Dolen, and Weltevreden (2014) linked highly personalized messages to negative effects since those messages lead to an invasion of privacy. The CHV has mostly been investigated as an overall concept, while it consists of a range of different elements that can contribute to a more humanized tone of voice. Based on previous research, it can be argued that one of those elements is the use of humor (Kelleher, 2009; Kelleher & Miller, 2006). Therefore, it is expected that the perceived CHV will be rated higher for humorous webcare responses compared to humorless responses. In turn, it is likely that the relationship between a humorous webcare response and the consumers attitude towards that response is mediated via the perceived CHV. However, little research has focused on the actual effects of humor as a part of the CHV. Especially in a webcare context, research about humor is lacking. To date, no study has investigated whether humor indeed leads to a higher perception of the CHV. This study will fill this gap, by investigating whether 9

10 humor leads to a higher perceived CHV. Furthermore, this study will examine whether the perceived CHV mediates the relationship between a humorous response and the consumers attitude towards that response. Appropriateness There are more factors that are likely to play a mediating role in peoples responses to humor (Fugate et al., 2000). One of those factors is the appreciation of humor, which according to Weinberger and Gulas (1992) depends on perceived appropriateness. Perceived appropriateness can be explained by the expectation-disconfirmation theory, which states that when performances or outcomes meets or exceeds peoples expectations, the evaluation will be positive and people will be satisfied (Burgoon, 1993). In general, consumers expectations create a certain frame of references to which they compare other behavior and outcomes. When communication meets consumers expectations, the communication is perceived as appropriate. However, when communication does not meet those expectations, people will experience negative emotions and stress (Xia, 2013). Previous research investigated the mediating role of appropriateness in responses of companies towards NeWOM (Xia, 2013). The results suggest that the positive effect of a certain response strategy towards NeWOM is mediated by perceived appropriateness. This can be attributed to the fact that, based on social norms for interpersonal communication, all consumers have certain expectations on the way a company should react towards NeWOM (Xia, 2013). Perceived appropriateness also mediates the effect of informal language in webcare on brand trust (Gretry et al., 2017). Their results could be explained by the Role Theory, which posits that successful interactions depend on appropriate behavior from communication partners depending on their relationship (Sarbin & Allen, 1968; Schewe, 1973; Solomon et al., 1985). In practice, this means that when communication partners are unfamiliar with one another, they stand farther apart. Therefore, it is less appropriate to 10

11 approach each other more informal (Little, 1965; Willis, 1966). This is in line with daily faceto-face social interactions, in which people address strangers more formally and with more distance than their friends (Gretry et al., 2017). Perceived appreciation and appropriateness have also been investigated for humor specifically, although it was not in a webcare context. Previous research examined humor in service advertising and asked participants to indicate the level of appropriateness of humor used for different service companies (Fugate et al., 2000). Results suggest that participants deemed it inappropriate to use humor when advertising professional services such as a physician, education, and veterinarian. Humor turned out to be appropriate for services such as bars and restaurants. Also telephone organizations scored above the median with regard to humor appropriateness. Thus, it can be concluded that the nature of the service of the company makes humor less or even more appropriate, and that perceived appropriateness functions as a mediator between humorous advertising and consumers responses towards this advertisement (Fugate et al., 2000). Those results are in line with previous research that suggests that appreciation of humor in advertising might affect consumers attitude and response to humorous advertising (Weinberger & Gulas, 1992). Drawing upon previous research about the mediating role of appropriateness in research on humor (Fugate et al., 2000; Weinberger & Gulas. 1992), companies responses towards NeWOM (Xia, 2013), and informal language (Gretry et al., 2017), it is likely that the effect of humor on the attitude towards the webcare response is mediated via perceived appropriateness as well. Humor and types of humor Scholars have defined the construct humor differently in their previous research. Crawford (1994) states that humor consists of verbal and nonverbal communication elements, which produce a positive cognitive or affective response from listeners. Reyes, Rosso, and Buscaldi (2012) defined humor as the presence of amusing effects, such as laughter or well- 11

12 being sensations (p. 2). Booth-Butterfield and Booth-Butterfield describe humor as a deliberately use of verbal and non-verbal messages that elicit laughter or other forms of behavior related to joy or surprise (1991, p. 206). When looking at all three definitions they consist of a main similarity, the fact that humor leads the receiver to feel positive emotions such as laughter. Thus, humor aims to positively influence the receivers mood and can be defined as amusing communication that produces positive emotions and cognitions in the individuals or group (Romero & Cruthirds, 2006). The existence of those positive emotions can be attributed to the Relief Theory, which states that positive emotions arrive as a result of humor because humor and laughter reduce the feeling of tension and stress (Berlyne, 1972). The positive mood and feelings that arise from humor can subsequently lead to a positive attitude towards an organization and/or message (Tom & Eves, 1999). Therefore, humor might lead to relationship enhancement between consumers and organizations. Especially since humor can encourage the growth of friendship between communication partners (Civikly, 1986). Humor that aims to enhance the relationship with others increases other s feelings of well-being. It reduces conflicts, strengthens ties between individuals, and increases one s attractiveness to the other (Martin et al., 2003). Relationship enhancement is in line with the webcare goals PR, marketing, and customer care as mentioned by van Noort, et al. (2014). All three goals emphasize the importance of positively influencing consumers evaluations of the organization. Especially for PR, relationship management, by managing and restoring the relationship with stakeholders (Coombs, 2002), turns out to be of great importance. Thus, the use of humor and the enhanced relationship that might arise from it can positively contribute to an organizations webcare goals. Although the different definitions of humor consist of several similarities, defining humor is difficult since it depends on the receiver how humor will be interpreted (Meyer, 2000; Hay, 2001). As shown in previous research, humor is not appreciated to the same extent 12

13 by all individuals (Xia et al., 2017), and humor from the perception of the sender might not go along with the perception of the receiver. According to Hay (2001), there are four implicatures associated with full support of humor: recognition of humor, understanding humor, appreciating humor, and agreeing with any message associated with it. Therefore, humor might be threatening when the receiver does not understand the joke (Zajdman, 1995). Moreover, when people interpret humor differently than intended, humor is likely to lead to miscommunication (Hay, 2001). Humor has been a main topic of interest in the field of marketing research, which has shown positive effects for humor use. It has been shown that humor enhances the attitude towards an advertisement (Weinberger & Gulas, 1992; Speck, 1987; Eisend, 2009). Besides that, it has been shown that humor in advertising attracts attention, enhances liking, and positively influences consumers buying decision (Koneska, Teofilovska, & Dimitrieska, 2017). Drawing upon the mentioned researches in the marketing field, and due to the fact that humor leads to a positive attitude towards a message (Tom & Eves, 1999), it is plausible that the use of humor in webcare responses will lead to a positive attitude towards the webcare response as well. Especially because the social media context is a suitable medium to be more informal and use humor (Kelleher, 2009; Kim, Zhang, & Zhang, 2016). There are different models to classify humor. Humor could, for example, be classified into different types such as sarcasm, wordplay, and exaggeration (Taecharungroj & Nueangjamnong, 2015; Catanescu & Tom, 2001). Classifying humor into specific types might, however, be too simplistic since there are more characteristics that play a role in the classification of humor. Martin et al. (2003), classified humor based on certain characteristics, which led to four main humor styles: self-enhancing, affiliative, aggressive, and self-defeating humor. As shown in Table 1, Martin et al. (2003) made two distinctions with regard to valence and relationship. Concerning valence, humor can either be relatively benign and 13

14 benevolent (positive), or potentially detrimental or injurious (negative). Regarding the relationship, humor can be used to enhance the self or to enhance the relationship with others. Table 1 Four dimensions of humor as defined by Martin et al. (2003) Valence Relationship Enhance yourself Enhance relations with others Positive Self-enhancing humor Affiliative humor Negative Aggressive humor Self-defeating humor Martin et al. (2003) distinguish two different humor styles that aim to enhance the relationship with others, affiliative, and self-defeating humor. Those two styles are interesting to investigate in a webcare context, since they can contribute to the webcare goals as defined by van Noort et al. (2014). Affiliative humor entails saying funny things, amuse others, facilitate relationships, and to reduce interpersonal tensions (Lefcourt, 2001). This can be applied to a webcare context, by directing the humor towards the consumer rather than to the company itself. In this way, the company aims to improve their relationship with the consumer. The webcare response of Tilburg University as presented in Figure 1 is such an example. Self-defeating humor is directed to the company itself. It attempts to amuse others by doing or saying funny things at one s own expense as a means of ingratiating oneself (Martin et al., 2003). Thus, making fun of your own company s failures may be an example of self-defeating humor in webcare responses. Even though the humor styles differ with regard to valence and direction (directed to another and directed to self), both humor styles are used for the same purpose and are thus interesting to compare. According to Martin et al. (2003), both affiliative and self-defeating humor aim to enhance the relationship with others. However, different humor styles might have different effects (Martin et al., 2003). Vigsø (2013) states that the use of self-defeating humor may be 14

15 effective because it arouses the consumers attention and expresses the organization s willingness to rectify the problem in a humorous way. According to Martin et al. (2003), only affiliative humor is expected to be related to positive emotions and relationship satisfaction while self-defeating humor is expected to negatively relate to relationship satisfaction and is expected to relate to negative emotions. Taking those expectations of Martin et al. (2003) into account, it is plausible that the attitude towards a webcare response will be more favorable when affiliative humor is used compared to self-defeating humor, since self-defeating is a negative form of humor that is expected to cause negative emotions. Therefore, this study will investigate whether the positive affiliative humor leads to a more favorable attitude towards the webcare response compared to the negative self-defeating humor style. Brand familiarity The effects of humor are not inconclusive, and more factors might contribute to the successful use of humor. One factor that could be of influence in the relationship between humor and the attitude towards the webcare response is brand familiarity. This refers to the extent of a consumer s experiences with a product (Alba & Hutchinson, 1987). In other words, it represents the consumer s degree of acquaintance with a certain brand. Previous research investigated whether brand familiarity moderated the effect of an informal humanized communication style on brand trust (Gretry et al., 2017). Results suggested that the use of informal language is more appropriate and thus more appreciated for familiar brands in comparison to unfamiliar brands. Those results might be explained through relationship strength (Xia, 2013). Consumers relationship with the brand is found to moderate consumers reactions to companies responses by influencing the standards of tolerance. Specifically, a strong brand relationship, which appears more often when being familiar with a brand, will increase those standards (Santos & Boote, 2003). Also in the field of marketing research, it has been shown that brand familiarity affects 15

16 consumers judgments and evaluations of brands. Specifically, research has shown that consumers respond differently when exposed to an advertisement of a familiar brand compared to an unfamiliar brand (Campbell & Keller, 2003). Besides that, it has been shown that consumers react differently to pronoun variations used in a brand advertisement of familiar and unfamiliar brands (Sela, Wheeler, & Sarial-Abi, 2012). In both cases, the familiar brand was favored compared to the unfamiliar brand regarding brand attitude. Drawing upon previous research, it is stated that familiarity with a brand, positively influences a consumer s confidence with a brand, as well as brand attitude and brand trust (Laroche, Kim, & Zhou, 1996; Campbell & Keller, 2003; Sela et al., 2012; Grety et al., 2017). Brand familiarity has also been studied in the context of ewom specifically. Sundaram and Webster (1999) investigated whether the impact of positive and negative ewom on brand evaluations, such as purchase intentions and brand attitudes, was moderated by the role of brand familiarity. With regard to NeWOM, consumers evaluations became less favorable for both the familiar as well as the unfamiliar brands. However, NeWOM had a more detrimental effect on unfamiliar brands than familiar brands. Also for Positive ewom (PeWOM), both the unfamiliar as well as the familiar brands benefited from PeWOM, but the gain was more exposed for the unfamiliar brand. Therefore, the results suggest that brand evaluations for unfamiliar brands are more susceptible to change as a result of ewom, compared to brand evaluations of familiar brands (Sundaram & Webster, 1999). Previous research has shown interesting effects of brand familiarity on consumers judgments and evaluations of brands. However, the use of humor has not yet been extensively investigated with regard to brand familiarity, while the use of humor might not work similarly for familiar and unfamiliar brands. Especially, since previous research by Sundaram and Webster (1999) has shown that unfamiliar brands are more affected by ewom than familiar brands. Thus, the way familiar and unfamiliar brands respond towards ewom messages 16

17 might also be perceived differently. Moreover, research has shown that the use of humor is more successful for existing and famous products, rather than new relatively unknown products (Koneska et al., 2017). It might be plausible that this also applies to existing and famous brands compared to relatively unknown brands. Taking the Role Theory into account, and drawing upon the previous reasoning, it is plausible that the use of humor as a part of a CHV will be more appreciated for familiar brands compared to unfamiliar brands. The moderating role of brand familiarity might also be applicable to the two different humor styles. Although self-defeating humor is a negative humor style, which is expected to lead to a less favorable attitude than affiliative humor (Martin et al., 2003), it might be that the effect of self-defeating humor is different for familiar and unfamiliar brands. The use of selfdefeating humor might lead to a more favorable attitude towards the webcare response for familiar brands than for unfamiliar brands, due to increased standards of tolerance for familiar brands (Santos & Boote, 2003). Hypotheses and conceptual model This research aims to investigate the effect of humor in webcare responses towards NeWOM on consumers attitude towards a webcare response. Based on existing insight about the operation of humor, different factors that are likely to play a role in this relationship are included, such as different humor styles, appropriateness, CHV, and brand familiarity. Drawing upon the theoretical framework, the following hypotheses are formulated: H1: A webcare response with the use of humor will lead to a more favorable attitude towards the webcare response compared to a webcare response without humor. H1a: The effect of humor on the attitude towards the webcare response is different for familiar and unfamiliar brands. Especially, the use of humor will lead to a more favorable attitude towards the webcare response for familiar brands, compared to unfamiliar brands. 17

18 H2: The attitude towards the webcare response will be more favorable for webcare responses with the positive affiliative humor, compared to those with the negative self-defeating humor. H2a: The effect of specific humor styles is different for familiar and unfamiliar brands. Especially, the use of self-defeating humor will lead to a more favorable attitude towards the webcare response for familiar brands compared to unfamiliar brands. H3: The effect of humor on attitude towards the webcare response is mediated via the perceived CHV. H4: The effect of humor on attitude towards the webcare response is mediated via perceived appropriateness. Figure 2. Conceptual model of the study Method Design In order to examine the effect of humor on the attitude towards a webcare response, and whether this effect differs with regard to brand familiarity, a 3 (humor: no humor vs. affiliative humor vs. self-defeating humor) x 2 (brand familiarity: familiar vs. unfamiliar) between-subjects design was conducted with perceived appropriateness and perceived CHV 18

19 as mediators. Thus, this experiment consisted of 6 conditions, of which each participant was exposed to one. Participants In total, 203 respondents participated in the study. For the specific manipulations, those participants had to meet two requirements. Firstly, it was important that participants in the humorless conditions reported they had not seen humor, while participants in the affiliative and self-defeating conditions should have noticed the use of humor. Secondly, it was desired that participants were familiar with the familiar brands while the opposite should have been the case for the unfamiliar brands. In total, 166 participants met those requirements and were taken into account for further analysis. This final sample consisted of 50 males and 116 females, with a mean age of years (SD = 11.41), who were rather equally divided among the six conditions (number of participants per conditions varied from 25 to 30). Overall, most participants were high educated HBO Bachelors (37,3%), followed by WO masters (30.1%), and WO bachelors (7,8%). Out of the 166 participants, 148 reported being familiar with webcare (89,8 %). However, webcare engagement was less common among the participants. Almost half of the participants (47%), reported to have never engaged in a webcare conversation. Analyses showed that gender (X² (5) = 5.14, p =.399), age F(5,165) =.38, p =.862, and level of education (X² (35) = 41.04, p =.223) of the participants were equal among all 6 conditions. Thus, the conditions could be compared in a reliable manner. Materials The questionnaire consisted of 3 fictive webcare conversations between consumers and three different telecom organizations (either familiar or unfamiliar). Organizations in the telecom industry were chosen since they perform well on webcare with regard to content and timing of the response, and the use of a CHV in their responses (Gouduberconnected, 2017). 19

20 Of those 3 conversations, two were experimental and one conversation was a filler which was always showed secondly to the participants and aimed to distract respondents from the true task. The two experimental webcare conversations both consisted of a complaint from a consumer that contained a certain level of dissatisfaction. The complaints, about an unjustified invoice and poor telecom reception, were fictive but based on actual complaints posted on Facebook pages of different telecom organizations. All the webcare conversations were adjusted to a Facebook layout in such a way that the consumer posted the complaint on the Facebook wall of the specific telecom organization. Facebook has been chosen for this study because it is the biggest platform for both creating as well as consuming electronic word-of-mouth (Kaplan & Haenlein, 2010; Wise, Albahash, & Park, 2010). By adjusting the conversations to a Facebook format, the webcare conversations looked more realistic. To rule out any confounding effects of gender, the names and profile pictures of the consumers posting the ewom were gender neutral, meaning they could either be male or female. Depending on the condition, the webcare responses were either from a familiar brand or an unfamiliar brand. The familiar brands used in this study were Vodafone, T-Mobile (filler), and Tele-2. Those specific brands have been chosen since they belong to the biggest telecom providers in the Netherlands, and participants in the pre-test reported that the use of humor would be suitable for those telecom providers. All webcare responses from the familiar brands were made with the use of the official brands Facebook name and Facebook profile picture. The unfamiliar brands, Telecoms, Call4ALL (filler), and CallPhone were fictive. By choosing non-existing brands, it was very unlikely that participants would know the unfamiliar brands. Also the profile pictures were fictive and made using online logo creators. Depending on the condition, the complaint was either followed by a humorless response, an affiliative joke or a self-defeating joke. In all three responses, the company first stated that they were sorry to hear the complaint and asked the consumer to send their 20

21 personal information in a private message. In the affiliative response, the organization also made a joke that aimed to enhance the relationship with the consumer without damaging the consumer or the organization itself. Those jokes were about the consumers situation, and positive since they did not harm any party involved. In the self-defeating condition, the joke that was made also aimed to enhance the relationship with the consumer, but at the expense of the company. Those jokes were about the organization, and negative since they harmed the company. The two humor styles thus only differed with regard to the topic of the joke (either about the consumer or about the organization), and valence (either positive or negative). To ensure that the affiliative and self-defeating humor styles could be compared as accurately as possible, and to rule out any other explanations, wordplay was used for both humor styles. A certain word from the consumer s complaint was used and reframed into another meaning. Pre-test To ensure a successful humor manipulation, a pre-test was conducted. Seven different webcare situations were tested for all three condition (no humor, affiliative, and selfdefeating), which makes 21 webcare conversations in total which can be found in Appendix A. All the tested webcare situations were conversations between consumers and telecom organizations. Twenty respondents participated of whom 4 males and 16 females, with a mean age of years (SD = 13.35). Out of the 20 participants, 17 reported to have seen such webcare conversations in their daily lives. Overall, most participants were high educated WO masters (45%), followed by HBO masters (10%), and HBO bachelors (25%). Based on the 21 tested conversations, the final materials were chosen. The more elaborated results from the pre-test can be found in Appendix F. To ensure that the jokes were correctly operationalized, the pre-test measured different aspects of the jokes. In total, participants were asked to answer 8 items per conversation, all on a 7-point scale (1 = totally disagree, 7 = totally agree), which can be found in Appendix B. 21

22 Firstly it was important that the use of humor was present in both humor conditions and absent in the humorless condition, and that the jokes were preferably perceived as funny. To ascertain that the conversations in the humorless condition were not perceived as humorous, a repeated measures ANOVA was performed to investigate whether humor use and funniness differed among the three conditions. For all seven types of conversations, the humorless condition and humor conditions differed significantly on humor use (mean differences ranged from 3.40 to 4.00, p <.001) and funniness (mean differences ranged from 2.55 to 2.95, p <.001). Moreover, no significant differences were found between affiliative and self-defeating humor. Therefore, the humor vs. no humor manipulation was successful. Further analyses were used to assess whether the manipulation between the affiliative and self-defeating jokes was also successful. Jokes from both humor styles should aim to enhance the relationship with the consumer as defined by Martin et al. (2003), which was measured with two items: The organization makes this joke to give the other a positive feeling, and Via this joke, the organization wants to bond with the consumer. Drawing upon the classification of Martin et al. (2003), the affiliative and self-defeating jokes should not differ significantly with regard to relationship enhancement, and a high score was preferred. However, one significant difference was found between the self-defeating jokes and affiliative jokes in the fifth conversation with regard to relationship enhancement F(1,19) = 4.48, p =.048. Thus, the humor style manipulation was not successful for this conversation. The topic of, and victim in the joke were important characteristics to distinguish the affiliative from the self-defeating jokes. The topic of the jokes was measured using two questions: The functioning of the organization is the topic of this joke, and The consumers situations is the topic of this joke. The victim in the jokes was measured using the following question: The organization disadvantaged itself with this joke. In the pre-test, a low score represented an affiliative joke, and a high score represented a self-defeating joke. According 22

23 to the humor classification of Martin et al. (2003), a significant difference was desired when looking at the topic of the joke and the victim in the joke. However, no significant differences were found for the topic between the self-defeating and affiliative jokes from conversation number 3 F(1,19) = 2.70 p =.117, and conversation 4 F(1,19) = 2.32 p =.144. Besides that, no significant differences were found for the victim between the affiliative and self-defeating jokes in conversation 2 F(1,19) = 3.15, p =.092. Up till this point, there remain three possible successful webcare conversations who could all characteristics for affiliative and selfdefeating humor (conversation number 1, 6, and 7). Lastly, neither the affiliative nor the self-defeating jokes should harm the consumer, since they could then not be attributed to affiliative or self-defeating humor. This was measured using the following question: The consumer is being harmed in this joke. A low score was preferred on this item, and the two humor styles should not differ significantly on this item. However, for the first conversation, a significant difference was found, in such a way that the affiliative joke harmed the consumer more than the self-defeating joke F(1,19) = 9.23, p =.007. To conclude, as shown in Table 2, the affiliative vs. self-defeating humor manipulation was only successful for conversation 6 and 7, since they were the only two who only differed significantly with respect to the topic of and the victim in the joke. Furthermore, they did not differ on all other items. 23

24 Table 2 Pre-test results for the 14 humorous webcare conversations ( 7 situations x 2 humor styles) Topic Victim Harming consumer Amusing Affiliative Self-defeating p value Affiliative Selfdefeating 1 M = 3.58 M = 4.78 p =.004 M = 4.10 M = 3.00 M = 4.35 M = 5.15 p =.014 M = 4.65 M = 4.40 p value p =.007 p = n.s. 2 M = 3.40 M = 4.25 M = 4.80 M = 5.15 p =.002 p = n.s. M = 3.55 M = 4.55 M = 2.85 M = 4.65 p = n.s. p = n.s. 3 M = 4.15 M = 4.05 M = 4.95 M = 5.50 p = n.s. p =.001 M = 3.10 M = 4.85 M = 2.65 M = 5.20 p = n.s. p = n.s. 4 M = 4.18 M = 4.30 M = 4.80 M = 5.30 p = n.s. p =.005 M = 3.25 M = 4.60 M = 2.70 M = 4.65 p = n.s. p = n.s. 5 M = 4.10 M = 4.40 M = 5.30 M = 5.85 p =.004 p =.001 M = 3.30 M = 4.45 M = 2.80 M = 5.20 p = n.s. p = M = 3.73 M = 3.75 M = 5.10 M = 5.40 p <.001 p <.001 M = 3.25 M = 4.75 M = 2.90 M = 5.10 p = n.s. p = n.s. 7 M = 3.65 M = 4.30 M = 5.00 M = 5.50 p =.001 p =.003 M = 3.80 M = 4.80 M = 2.90 M = 5.05 p = n.s. p = n.s. Note. Items were measured on a 7-point scale. N.s. = not significant. Therefore, conversation number 6 (about poor telecom reception) and 7 (about an unjustified invoice) were used as materials in the final experiment. An example of conversation 7 is shown in Figure 3. The materials used in this experiment were in Dutch since humor and especially wordplay are closely related to language, and translation of humor might lead to misinterpretation. The final materials can be found in Appendix E. 24

25 Figure 3. Humorless (unfamiliar), affiliative (familiar), and self-defeating (familiar) webcare responses of conversation 7 about an unjustified invoice. Measures To measure CHV, appropriateness, and attitude, different scales were used which were partly based on previous literature. All items were measured on a 7-point likert scale (1=totally disagree; 7=totally agree), and can be found in Appendix C. 25

26 To measure the perceived CHV, participants were asked to answer questions based on previous research from Kelleher and Miller (2006). They identified 11 items that measure a CHV. Participants were asked to answers five of them, such as Communicates with a sense of humor, and Tries to make communication enjoyable. Those five items have been chosen since they are most closely related to the use of humor compared to other items in the scale from Kelleher and Miller (2006) such as Provides link to competitors and Would admit a mistake. The overall reliability of the scale was good, α =.80 (M = 5.09, SD = 0.80). Perceived appropriateness was measured with four items ( Meets my expectations, Correspond to how I expect it to communicate, Is appropriate, and Is suitable for this situation ). The first three items were adapted from a previous study by Gretry et al. (2017). The fourth item was added to make sure there were two items regarding expectations and two items regarding appropriateness. The overall reliability of the scale was good, α =.89 (M = 4.78, SD =1.14). To measure the attitude towards the webcare response, participants were asked to answer statements such as My attitude towards the webcare response is (1) good, (2) enjoyable, (3) unpleasant (r), and (4) positive. Those items were based on previous research from Gardner (1985), Mitchell (1986), and Coyle and Thorson (2001). The overall reliability of the scale was good, α =.86 (M = 4.87, SD = 0.97). Since all scales had a high reliability, the items per scale were merged into one mean variable and used for further analyses. Reliability scores per webcare conversation can be found in Appendix G. Procedure To gather enough participants, a network sampling approach was used. The link to the online questionnaire was distributed via social media platforms such as Facebook and Twitter, and send to the respondents directly via , WhatsApp, and Facebook Messenger. Firstly, 26

27 participants were shown an introduction in which the length, topic, and procedure were explained to them. They were also informed about the possibility to quit the survey or finish it at another time, and their anonymity was guaranteed. Subsequently, three webcare conversations were shown. One of those was a filler, which contained a question posted by a consumer followed by a neutral, humorless response. The two experimental NeWOM conversations were randomized. After each conversation, participants were asked to answer multiple questions regarding CHV, appropriateness, and their attitude towards the webcare response. To ensure a successful humor and familiarity manipulation, two questions ( Did one or more conversations that you saw contain humor?, Which of the brands from the survey did you know on forehand? ) were asked as a final check after participants were exposed to the webcare conversations. Participants knowledge of webcare and level of engagement in webcare were questioned as well. Finally, some demographic questions were asked, such as age, gender, and level of education. At the end of the survey, the participants were debriefed about the fact that the conversations were fictive. This also accounts for the brands in the unfamiliar condition. The final questionnaire took approximately 10 minutes and can be found in Appendix D. Results Demographic variables Prior to the main analysis, multiple analyses were performed to investigate whether the demographical variables had any influence on the dependent variable. Three independent t-tests were performed with gender as independent variable, and attitude, CHV, and appropriateness as dependent variables. Because the assumption of normality was not met for gender (z-score Kurtosis = -3.33), attitude (z-score Skweness = -2.95), CHV (z-score Skweness = -2.12), and appropriateness (z-score Skweness = -2.28), the 95% Confidence Intervals will be reported. On average, females scored higher on attitude towards the webcare 27

28 response (M = 5.01, SD = 0.94), than males (M = 4.56, SD = 0.96), t(164) = 2.78, p =.006, BCa 95% CI [-.763, -.130]. Females also reported a higher score with regard to appropriateness of the webcare response (M = 4.91, SD = 1.13), than males (M = 4.46, SD = 1.10), t(164) = 2.37 p =.019, BCa 95% CI [-.825, -.076]. Lastly, CHV scores were higher for females (M = 5.27, SD = 0.77), compared to males (M = 4.99, SD = 0.85), t(164) = 2.08, p =.039, BCa 95% CI [-.544, -.014]. Besides that, a correlation test revealed a significant correlation between age and attitude towards the webcare response r = -.24, p =.002, BCa 95% CI [-.400, -.050], age and CHV r = -.29, p <.001, BCa 95% CI [-.483, -.094], and age and appropriateness r = -.18, p =.019, BCa 95% CI [-.349, -.005]. Level of education only correlated significantly with CHV r =.20, p =.010, BCa 95% CI [.003,.347]. Since the demographic variables age and gender had an effect on all three dependent variables, both variables were used as covariates in future analyses. The main analyses were first performed without covariates. Subsequently they were performed again, this time including the covariates age and gender. However, no differences were found between the analyses with and without covariates. Therefore, those covariates will not be discussed any further. Hypothesis testing In order to test the first hypothesis that a webcare response with the use of humor will lead to a more favorable attitude towards the webcare response compared to a humorless response, an independent sample t-test was performed. On average, the participants in the no humor condition had a more positive attitude towards the webcare response (N = 54, M = 5.04, SD = 0.88), than those in the humor conditions (N = 112, M = 4.80, SD = 1.00). This difference, 0.24, was not significant t(164) = 1.49, p =.138, BCa 95% CI [.238,.159]. Therefore, no support was found for the first hypothesis. To investigate the role of brand familiarity in this relationship, a moderation analysis 28

29 was performed using PROCESS model number 1 by Preacher and Hayes (Hayes, 2013) with brand familiarity as moderator, humor (humor vs. no humor) as independent variable, and attitude as dependent variable. The overall model was not significant, R 2 =.03, F(3, 162) = 2.24, p =.086, indicating that the effect of humor on attitude towards the webcare response is not different for familiar and unfamiliar brands. As can be seen in Table 3, no main effects nor an interaction effect was found. This led to the rejection of hypothesis 1a. Table 3 Linear Model of Predictors Constant 4.88 [4.73, 5.03] V1 Brand familiarity (centred) [-0.34, 0.26] V2 Humor (centred) 0.25 [-0.05, 0.55] V1 X V2 Brand familiarity * humor [-1.17, 0.03] b [BCa 95% CI] SE t p < An independent sample t-test was also performed to test the second hypothesis that the attitude towards the webcare response would be more favorable for webcare responses with the positive affiliative humor, compared to those with the negative self-defeating humor. On average, the participants in the affiliative condition (N = 54) had a more positive attitude towards the webcare response (M = 4.83, SD = 0.97), than those in the self-defeating humor condition (N = 58, M = 4.77, SD = 1.04). This difference,.06, was however not significant t(110) =.32, p =.747, BCa 95% CI [.238,.159]. Therefore, no support was found for the second hypothesis, which means that the attitude towards the webcare response was not more favorable for responses with affiliative humor, compared to responses with self-defeating humor. 29

30 Additionally, the moderating role of brand familiarity was investigated using PROCESS model number 1 by Preacher and Hayes (Hayes, 2013). Brand familiarity was entered as moderator, humor (affiliative vs. self-defeating) as independent variable, and attitude towards the webcare response as dependent variable. The overall model was not significant, R 2 =.00, F(3, 112) = 0.30, p =.826, indicating that the effect of the two different humor styles on attitude towards the webcare response is not different for familiar and unfamiliar brands. This led to the rejection of hypothesis 2a. To examine whether there is a link between humor and attitude that can be explained by CHV (H3), and appropriateness (H4), two mediation analyses were performed using PROCESS model number 4 by Preacher and Hayes (Hayes, 2013). In the first model, CHV was entered as a mediator. Results show that humor was indeed positively related to CHV (b = 0.64, p <.001), and CHV was in turn positively significantly related to attitude towards the webcare response (b = 0.84, p <.001). There was no significant total effect (b = -0.24, SE = 0.16, p =.137), indicating that the use of humor did not lead to a higher attitude towards the webcare response. This effect became significant when adding the mediator CHV to the model, the direct effect was significant (b = -0.78, SE = 0.13, p <.001). Given the results it can be concluded that CHV explains the link between humor and attitude towards the webcare response, because there was a significant total indirect effect (b = 0.54, SE = 0.12, 95% BCa CI [0.33, 0.79]). Thus, support was found for the third hypothesis. In the second model, appropriateness was entered as a mediator. Results show that humor was negatively related to appropriateness (b = -0.77, p <.001), and appropriateness was in turn significantly positively related to attitude towards the webcare response (b = 0.76, p <.001). There was no significant total effect (b = -0.24, SE = 0.16, p =.138), however there was a significant direct effect (b = 0.34, SE = 0.09, p =.002). Since there was a significant total indirect effect (b = -0.58, SE = 0.14, 95% BCa CI [-0.84, -0.31]), it can be concluded 30

31 that appropriateness explains the link between humor and attitude towards the webcare response. Thus, support was found for the fourth hypothesis. Differences in NeWOM messages Per condition, participants saw two experimental webcare conversations, with two different wordplays. Those jokes only differed with regard to the wordplay that was made, but were equal with regard to humor style (affiliative or self-defeating) and sender (familiar or unfamiliar brand). Based on the analyses so far, it remains unclear whether the responses to the two different NeWOM messages led to different results, which might impact the generalizability of the results. To investigate whether the responses to the two different NeWOM messages (unjustified invoice vs. poor telecom reception) in the same condition led to different scores for CHV, attitude, and appropriateness, a repeated measures ANOVA was performed with type of NeWOM message as within subject factor. In all four humor conditions, the ANOVA revealed a significant main effect of the type of NeWOM message on CHV (mean differences ranged from 0.55 to 0.65, p ranged from <.001 to.009), appropriateness (mean differences ranged from 0.78 to 1.36, p ranged from <.001 to.015), and attitude (mean differences ranged from 0.93 to 1.33, p ranged from <.001 to.002). As can be seen in Table 4, the responses to the NeWOM message about the poor telecom reception scored significantly higher in all four humor conditions on CHV, appropriateness, and attitude compared to the NeWOM message about the unjustified invoice. Moreover, in the humorless conditions (condition number 1, and 4), no significant differences were found. Thus, the humorous responses to the NeWOM message about poor telecom reception performed better on CHV, appropriateness, and attitude than the humorous responses to the NeWOM message about the unjustified invoice. Additional analyses were performed to investigate whether there would be a main effect of humor on the attitude towards the 31

32 webcare response when only including the NeWOM situation about the poor telecom reception. However, those analyses still did not reveal a significant main effect for humor (humor vs. no humor), t(164) = 1.48, p =.142, and humor styles (affiliative vs. selfdefeating), t(110) = 0.29, p =.769 on attitude towards the webcare response. Table 4 Scores for CHV, appropriateness and attitude for the two different NeWOM messages per condition CHV Appropriateness Attitude Reception Invoice 1 M = 4.57 M = 4.80 p value Reception Invoice p = n.s. M = 5.03 M = 5.24 p value Reception Invoice p = n.s. M = 5.04 M = 4.63 p value p = n.s. 2 M = 5.71 M = 5.14 p =.009 M = 5.18 M = 4.18 p =.015 M = 5.53 M = 4.35 p = M = 5.50 M = 4.94 p =.004 M = 4.90 M = 4.12 p =.006 M = 5.28 M = 4.35 p < M = 4.76 M = 4.88 p = n.s. M = 5.37 M = 5.57 p = n.s. M = 5.24 M = 5.27 p = n.s. 5 M = 5.72 M = 5.17 p <.001 M = 5.20 M = 3.84 p <.001 M = 5.32 M = 4.15 p < M = 5.81 M = 5.16 p <.001 M = 4.99 M = 3.86 p =.002 M = 5.39 M = 4.06 p <.001 Note. Condition 1 and 4 are humorless conditions. Significant differences thus only appeared in the humor conditions (2, 3, 5, and 6). All items were measured on a 7-point scale. N.s. = not significant. Conclusion and discussion In an attempt to contribute to more understanding about humor in webcare, an experimental study was performed among 166 participants. The aim of this study was to understand the relationship between humor and humor styles, and the attitude towards the 32

33 webcare response. Therefore, this study gave answer to the following research question: What is the effect of humor in webcare responses towards NeWOM on the attitude towards the webcare response and what is the role of brand familiarity in this relationship? The results showed that a webcare response with the use of humor does not lead to a more favorable attitude towards the webcare response compared to a humorless response. Therefore, contrary to expectations, no support was found for the first hypothesis, which stated that webcare responses with the use of humor would lead to a more favorable attitude towards the webcare response compared to humorless responses. Those results are not in line with previous research that showed that humor can lead to a positive attitude towards messages (Tom & Eves, 1999). Results in the marketing field suggest that humor enhances the attitude towards an advertisement (Weinberger & Gulas, 1992; Speck, 1987; Eisend, 2009), and enhances liking (Koneska et al., 2017). However, based on the current results, those beneficial effects for humor on attitude are not applicable to a webcare context as well. Analysis has shown that this non-appearing relation between humor and attitude can be explained by appropriateness, which led to support of the fourth hypothesis. Humor is significantly related to appropriateness, in such a way that responses with humor were rated as less appropriate compared to humorless responses. Subsequently, appropriateness was related to attitude, which means that a higher score on appropriateness, led to a more favorable attitude towards the webcare response. Therefore, it can be concluded that appropriateness has a mediating role in the relationship between humor and attitude. This seems to be consistent with previous research that has also shown the mediating role of appropriateness in the relation between a certain response strategy towards online criticism and a positive outcome (Xia, 2013), and between informal language and brand trust (Gretry et al., 2017). Also CHV has a mediating role in this relationship between humor and attitude towards the webcare response, which led to support of hypothesis 3. Results have shown that 33

34 the use of humor positively relates to the perceived CHV, which means that the perceived CHV was higher in webcare responses with humor compared to humorless response. Those results further support the idea that the use of humor contributes to a CHV (Kelleher, 2009; Kelleher & Miller, 2006). The results showed that this perceived CHV positively relates to attitude towards the webcare response. This positive relation confirms the proven positive effects of a CHV in previous research such as enhanced consumer evaluation (Kelleher, 2009; Kelleher & Miller, 2006; van Noort & Willemsen, 2012), and higher ratings of trust, satisfaction, and commitment (Kelleher, 2009; Beldad et al., 2010; Sweetser & Metzgar, 2007; Yang et al., 2010). Besides that fact that the use of humor in general does not lead to a more favorable attitude, also the different humor styles show no significant relation with the attitude towards the webcare response. Therefore, no support was found for the second hypothesis that the affiliative humor style would result in a more favorable attitude toward the webcare response compared to the self-defeating humor style. Despite the fact that different humor styles can have different effects (Martin et al., 2003), this study did not find support for differences between affiliative and self-defeating humor. These results differ from the humor classification of Martin et al. (2003), who expect affiliative humor to be related to positive emotions, and self-defeating to negative emotions. Answering the question which role brand familiarity plays in the relationship between humor and attitude towards the webcare response, results showed no significant moderating effect of brand familiarity in this relationship, which led to the rejection of hypothesis 1a. The results are not in line with the Role Theory that posits that the adoption of a more informal communication style is more suitable for familiar brands compared to unfamiliar brands (Sarbin & Allen, 1968; Schewe, 1973; Solomon et al., 1985). Previous research by Gretry et al. (2017) found that the use of informal language as part of a CHV is more appreciated for 34

35 familiar brands compared to unfamiliar brands. In contrast to those earlier findings, however, people in the current experiment did not appreciate an informal, humorous reaction more when it was made by a familiar brand, compared to an unfamiliar brand. The results also did not show a moderating role for brand familiarity on the effect of humor styles on attitude towards the webcare response, which leads to rejection of hypothesis 2a. Thus, self-defeating humorous responses from familiar brands did not lead to a more favorable attitude towards the webcare response compared to self-defeating responses from unfamiliar brands. Contrary to expectations, the increased standards of tolerance for familiar brands (Santos & Boote, 2003), did not play a role in the appreciation of self-defeating humor. To conclude, in response to the research question, the results of this research have shown that the use of humor does not lead to a more favorable attitude toward the webcare response, and that brand familiarity does not play a role in this relationship. However, the use of humor does lead to a higher perceived CHV, which in turn leads to a more favorable attitude towards the webcare response. Not only CHV mediates this relationship, also the level of appropriateness has a mediating role in such a way that humor is perceived as less appropriate, which in turn leads to a less favorable attitude towards the response. No differences were found between the affiliative and self-defeating humor styles. Both humor styles were perceived equally, both for familiar as well as for unfamiliar brands. Limitations and explanations for findings The mediating role of appropriateness that was found in this study might be attributed to the fact that this research investigated the use of humor in a NeWOM context. Perhaps, people find a humorous response less appropriate when they express a certain level of dissatisfaction, because a humorous response makes a mockery of seriousness (Lockyer & Pickering, 2001), while people want their complaint to be taken seriously. In line with this 35

36 reasoning, it is plausible that the use of humor will be rated as more appropriate in a neutral or positive ewom message, since people are not dissatisfied in those situations, and thus more open to the use of humor. Type of ewom might also explain the fact that brand familiarity was no mediator in the effect between humor and attitude towards the webcare response. Previous research found that the use of informal language as part of the CHV was more suitable for familiar brands compared to unfamiliar brands (Gretry et al., 2017). However, Gretry et al. (2017) investigated the use of informal language and the role of brand familiarity in the context of PeWOM. It might be plausible that the use of humor in a NeWOM context is not appreciated anyway, and that the sender of this message (either familiar or unfamiliar) might not matter. Another possible explanation for the undetected moderating effect of brand familiarity might be that consumers in this experiment were familiar with the brands, but did not necessarily hold a positive attitude against them. When looking at the study by Gretry et al. (2017), their familiar brand used in the study was, as they state themselves, a popular hotel brand. They also emphasized this popularity towards their participants via the online questionnaire they used. This might have led to the results that it was more appreciated for this popular hotel to use informal language compared to the unfamiliar hotel. However, it is questionable whether this moderating role can be attributed to brand familiarity, or perhaps brand attitude or even brand popularity. In the current study, the brands were chosen based on their level of familiarity and not on their level of popularity. Also brand perception was not taking into account in the current research. However, it is plausible that the attitude towards a brand is already more favorable on forehand for certain brands, which in turn might lead to a higher level of appreciation when this brand applies humor in its webcare response. The different approach with the study by Gretry et al. (2017) regarding brand familiarity, might explain the fact that no moderating role was found for brand familiarity in the current study. 36

37 This research did not find evidence that a webcare response with affiliative humor would results in a more favorable attitude towards the webcare response compared to a response with self-defeating humor. This might be explained by the fact that those two humor styles have the same goal. Although affiliative humor is directed to others and positive, while self-defeating humor is directed to the self and negative, they both aim to enhance the relationship with others (Martin et al., 2003). Results from the pre-test suggest that participants noticed that both the affiliative, as well as the self-defeating jokes were made to enhance the relationship with the consumer. Thus, it is plausible that people can recognize the motivation of the organization to use humor. This might have led to the fact that the attitude that people hold against the humorous webcare responses in the current study did not differ significantly between the two humor styles. Differences between humor styles might, however, have occurred when a humor style with the purpose to enhance the self was included, such as self-enhancing humor (Martin et al., 2003). When receiving a response with a joke that is made to enhance the organization itself, this might lead to a less favorable attitude compared to a joke that has the purpose to enhance the relationship with one another. As a final limitation, the materials used in the survey met all the criteria to be attributed to affiliative or self-defeating humor, but the pre-test revealed that the wordplay jokes were not necessarily rated as funny. However, an important aspect in the definition of humor is that humor leads the receiver to feel positive emotions such as laughter (Crawford, 1994; Reyes et al., 2012; Booth-Butterfield & Booth-Butterfield, 1991; Romero & Cruthirds, 2006). Regrettably, it cannot be stated with certainty that the jokes used in this experiment elicited such positive feelings, since the materials used in this study were not necessarily rated as funny. Future research The current research provides new insights for the use of humor in the field of 37

38 webcare. It has been shown that there is no overall main effect for humor use on the attitude towards the webcare response. However, results have shown that the use of humor indeed leads to a higher CHV, and that CHV is a mediator in the relation between humor and attitude towards the webcare response. This mediating role also accounts for perceived appropriateness. The current research thus contributes to more understanding of the operation of humor in webcare, and especially of humor as a part of the CHV. However, based on the current results, there are some suggestions for future research. The fact that humorous responses are perceived as less appropriate compared to humorless responses might suggest that a company should not apply humor in webcare responses. However, humor also leads to a higher CHV, which in turn leads to a more favorable attitude. Therefore, it is interesting to investigate in which context humor would be appropriate. Although the current results have shown that humor is not appropriate in a response to NeWOM, it remains unclear in which context the use of humor would be appropriate. For future research, it might be interesting to look at other types of ewom, such as PeWOM and neutral ewom messages. The fact that humor ridicules the level of seriousness (Lockyer & Pickering, 2001), might be of less importance in a neutral, and especially PeWOM context. When taking other types of ewom into account, research might reveal a difference in appropriateness for different ewom context. Such an insight might help organizations to decide in which situation it is more appropriate to give a humorous response. Based on the current research, it remains unclear what explanation can be attributed to the fact that no main effect was found for humor use on attitude towards the webcare response. Based on the results, it can be stated that there are no differences between the affiliative and self-defeating humor styles, although literature research suggested that affiliative humor would be evaluated more positive compared to self-defeating humor (Martin et al., 2003). However, it is questionable whether the specific humor type used in this 38

39 experiment, wordplay, may be of influence on those results. According to Catanescu and Tom (2001), different humor types have different effects. The results of this study are limited since only wordplay was used in order to examine the effect of affiliative and self-defeating humor. There are more humor types that are applicable for both affiliative as well as self-defeating humor. For example, sarcasm, which is a nuanced form of language in which individuals state the opposite of what is implied (Rajadesingan, Zafarani, & Liu, 2015). This humor type can be seen as a form of verbal aggression (Katz, Blasko, & Kazmerski, 2004). Therefore, sarcasm in a webcare response might elicit stronger reactions compared to wordplay, which might lead to more outspoken attitudes towards the webcare response. For future research, it is recommended to include sarcasm as humor type. In this way, it can be stated with more certainty that there is no difference between affiliative and self-defeating humor, or whether this perhaps depends on the specific type of joke that is used. By doing so, the insights about humor, and especially affiliative and self-defeating humor, will be more generalizable. The use of sarcastic humor in future research might also contribute to more insight about the role of brand familiarity. Results from the current research has shown no moderating role of brand familiarity on the relation between humor and attitude towards the webcare response. However, previous research has shown that people who are familiar with one another are found to be better at recognizing sarcasm compared to people who are unfamiliar with one another (Cheang & Pell, 2011). This could be of importance since the recognition of humor is a key feature in the successful use of humor (Hay, 2001). Thus, the role of brand familiarity might be more outspoken for sarcastic humor compared to wordplay. When conducting similar research with sarcastic humor, this research might reveal a difference in the appreciation of humor for familiar and unfamiliar brands. In order to gain more insight into the operation of humor in webcare specifically, this research investigated humor in a NeWOM context. As shown in previous research, and again 39

40 in this study, the effects of humor are not inconclusive. Many factors influence the effect of humor. This research contributed to more understanding of those factors. Results have shown that the relationship between humor and the attitude towards the webcare response is influenced by CHV since humor leads to a higher level of perceived CHV. This is of added value since, to date, no study investigated whether humor indeed influenced the perceived CHV. Although results showed a negative relationship between humor and appropriateness, again those insights are of added value in humor research in webcare, since it contributes to more understanding of the underlying mechanisms in the operation of humor. Even though this study managed to identify two important factors in the operation of humor, humor and its effects are still difficult to capture. Mainly because humor is not appreciated to the same extend by all individuals (Xia, 2013), which makes a successful adoption of humor almost a personal manner. Also, humor recognition, understanding, and agreeing with the message associated with it, are factors that are needed for full support of humor and are likely to be personal (Hay, 2001). So, planning to apply humor in webcare? Well, take care.. 40

41 References Alba, J. W., & Hutchinson, J. W. (1987). Dimensions of consumer expertise. Journal of Consumer Research, 13(4), Beldad, A., de Jong, M., & Steehouder, M. (2010). How shall I trust the faceless and the intangible? A literature review on the antecedents of online trust. Computers in Human Behavior, 26(5), doi: /j.chb Berlyne, D. E. (1972). Humor and its kin. The Psychology References of Humor, 1, Booth Butterfield, S., & Booth Butterfield, M. (1991). Individual differences in the communication of humorous messages. Southern Journal of Communication, 56(3), doi: / Burgoon, J.K. (1993). Interpersonal expectations, expectancy violations, and emotional communication. Journal of Language and Social Psychology, 12(1-2), doi: / X Campbell, M. C., & Keller, K. L. (2003). Brand familiarity and advertising repetition effects. Journal of Consumer Research, 30, doi: / Catanescu, C., & Tom, G. (2001). Types of humor in television and magazine advertising. Review of Business, 22(1), Cheang H. S., & Pell, M. D. (2011). Recognizing sarcasm without language: A cross- linguistic study of english and cantonese. Pragmatics & Cognition, 19(2), doi /pc cheissn Civikly, J. M. (1986). Humor and the enjoyment of college teaching. New Directions for Teaching and Learning, (26), doi: /tl

42 Coombs, W.T. (2002) Assessing online issue threats: Issue contagions and their effect on issue prioritisation, Journal of Public Affairs, 2(4), doi: Coyle, J. R., & Thorson, E. (2001). The effects of progressive levels of interactivity and vividness in web marketing sites. Journal of Advertising, 30(3), Crawford, C. B. (1994). Theory and implications regarding the utilization of strategic humor by leaders. The Journal of Leadership Studies, 1(4), Crijns, H., Cauberghe, V., Hudders, L., & Claeys, A-S. (2017). How to deal with online consumer comments during a crisis? The impact of personalized organizational responses on organizational reputation. Computers in Human Behavior, 75, doi: /j.chb Demmers, J., van Dolen, W. M., Weltevreden, J. W.J. (2014). Bedankt voor het compliment! : het effect van bedrijfsreacties op positieve online word-of-mouth. Ontwikkelingen in het marktonderzoek, Dijkmans, C., Kerkhof, P., Buyukcan-Tetik, A., & Beukeboom, C. (2015). Online conversation and corporate reputation: a two-wave longitudinal study on the effects of exposure to the social media activities of a highly interactive company. Journal of Computer-Mediated Communication, 20(6), doi: /jcc Eisend, M. (2009). A meta-analysis of humor in advertising. Journal of the Academy of Marketing Science, 37(2), doi: /s y Forrester. (2014). Predictions 2015: Social media grows up. Retrieved from: Predictions+2015+Social+Media+Grows+Up/fulltext/-/E- res

43 Fugate, D. L., Gotlieb, J. B., & Bolton, D. (2000). Humorous services advertising: what are the roles of sex, appreciation of humor, and appropriateness of humor?. Journal of Professional Services Marketing, 21(1), doi: /J090v21n01_02 Gardner, R. C. (1985). Social psychology and second language learning: The role of attitudes and motivation. Arnold Ghosh, T. (2017). Managing negative reviews: the persuasive role of webcare characteristics. Journal of Internet Commerce, 16(2), doi: / Gouduberconnected. (2017). Digital dialogues pulse #4. [Infographic] Retrieved from: Grégoire, Y., Salle, A., & Tripp, T.M. (2015). Managing social media crises with your customers: the good, the bad, and the ugly. Business Horizons, 58(2), doi: /j.bushor Gretry, A., Horváth, C., Belei, N., & van Riel, A. C. (2017). Don't pretend to be my friend! When an informal brand communication style backfires on social media. Journal of Business Research, 74, doi: /j.jbusres Gruen, T.W., Osmonbekov, T., and Czaplewski, A.J. (2006). ewom: the impact of customer-to-customer online know-how exchange on customer value and loyalty. Journal of Business Research, 59(4), doi: /j.jbusres Hay, J. (2001). The pragmatics of humor support. Humor, 14(1), doi: /humr Hennig-Thurau, T., Malthouse, E. C., Friege, C., Gensler, S., Lobschat, L., Rangaswamy, A., & Skiera, B. (2010). The impact of new media on customer relationships. Journal of Service Research, 13(3), doi: /

44 Hinchcliffe, D. (2006). All we got was web 1.0, when Tim Berners-Lee actually gave us web 2.0. [Blog post]. Retrieved from: Huibers, J., & Verhoeven, J. (2014). Webcare als online reputatiemanagement. Tijdschrift voor Communicatiewetenschap, 42(2), 165. Joireman, J., Grégoire, Y., Devezer, B., & Tripp, T. M. (2013). When do customers offer firms a second chance following a double deviation? The impact of inferred firm motives on customer revenge and reconciliation. Journal of Retailing, 89(3), doi: /j.jretai Kaplan, A. M., & Haenlein, M. (2010). Users of the world, unite! The challenges and opportunities of Social Media. Business Horizons, 53(1), doi: /j.bushor Katz, A. N., Blasko, D. G., & Kazmerski, V. A. (2004). Saying what you don't mean: Social influences on sarcastic language processing. Current Directions in Psychological Science, 13(5), doi: /j x Kelleher, T. (2009). Conversational voice, communicated commitment, and public relations outcomes in interactive online communication. Journal of Communication, 59(1), doi: /j x Kelleher, T., & Miller, B. M. (2006). Organizational blogs and the human voice: Relational strategies and relational outcomes. Journal of Computer Mediated Communication, 11(2), doi: /j x 44

45 Kerkhof, P. Beugels, D., Utz, S., & Beukeboom, C. J. (2011). Crisis PR in social media. An experimental study of the effects of organizational crisis responses on Facebook. Paper presented at the 61st Annual ICA Conference, Boston (USA), Kerkhof, P.. Beukeboom, C. J., Utz, S., & de Waard, J. (2010). Het vermenselijken van een bedrijf: Effecten van persoonlijke vs. onpersoonlijke bedrijfsreacties op online klachten van consumenten. [Humanizing an organization: effects of personal and impersonal reactions on negative consumer reviews]. Paper presented at the 9th Etmaal van de Communicatiewetenschap, Ghent, Belgium. Kim, S., Zhang, X. A., & Zhang, B. W. (2016). Self-mocking crisis strategy on social media: Focusing on Alibaba chairman Jack Ma in China. Public Relations Review, 42(5), doi: /j.pubrev Koneska, L., Teofilovska, J., & Dimitrieska, S. (2017). Humor in Advertising. European Journal of Economics and Business Studies, 8(1), Koot, F. (2013). Organizations with a Human Voice: Antecedents and Consequences of a Conversational Human Voice in the Webcare Responses of Organizations. (Master s thesis, UVA Amsterdam). Retrieved from: Kwon, E. S., & Sung, Y. (2011). Follow Me! Global Marketers Twitter Use. Journal of Interactive Advertising, 12(1), doi: / Laroche, M., Kim, C., & Zhou, L. (1996). Brand familiarity and confidence as determinants of purchase intention: An empirical test in a multiple brand context. Journal of Business Research, 37(2), doi: / (96) Lefcourt, H. M. (2001). Humor: The psychology of living buoyantly. New York: Kluwer Academic Publishers. 45

46 Levine, R., Locke, C., Searls, D., & Weinberger, D. (2000). The cluetrain manifesto: The end of business as usual. New York times book review, 105(13), Little, K. B. (1965). Personal space. Journal of Experimental Social Psychology, 1, doi: / (65) Lockyer, S., & Pickering, M. (2001). Dear shit-shovellers: Humour, censure and the discourse of complaint. Discourse & Society, 12(5), doi: / Martin, R. A., Puhlik-Doris, P., Larsen, G., Gray, J., & Weir, K. (2003). Individual differences in uses of humor and their relation to psychological well-being: Development of the Humor Styles Questionnaire. Journal of Research in Personality, 37(1), doi: /S (02) Mathies, C., Chiew, T. M., & Kleinaltenkamp, M. (2016). The antecedents and consequences of humour for service: a review and directions for research. Journal of Service Theory and Practice, 26(2), doi: /JSTP Meyer, J. C. (2000). Humor as a double edged sword: Four functions of humor in communication. Communication Theory, 10(3), Mitchell, A. A. (1986). The effect of verbal and visual components of advertisements on brand attitudes and attitude toward the advertisement. Journal of Consumer Research, 13(1), Park. H. & Lee, H. (2012). Show us you are real: the effect of human-versus-organizational presence on online relationship building through social networking sites. Cyberpsychology behavior and social networking, 16(4), doi: /cyber

47 Rajadesingan, A., Zafarani, R., & Liu, H. (2015). Sarcasm detection on twitter: A behavioral modeling approach. In Proceedings of the Eighth ACM International Conference on Web Search and Data Mining (pp ). ACM. Reyes, A., Rosso, P., & Buscaldi, D. (2012). From humor recognition to irony detection: The figurative language of social media. Data & Knowledge Engineering, 74, doi: /j.datak Romero, E. J., & Cruthirds, K. W. (2006). The use of humor in the workplace. The Academy of Management Perspectives, 20(2), doi: /AMP Santos, J., & Boote, J. (2003). A theoretical exploration and model of consumer expectations, post purchase affective states and affective behaviour. Journal of Consumer Behaviour, 3(2), Sarbin, T. R., & Allen, V. L. (1968). Role theory. In G. Lindzey, & E. Arsonson (Eds.), The Handbook of Social Psychology (pp ). Reading, MA: Addison Wesley. Schamari, J. & Schaefers, T. (2015). Leaving the home turf: how brands can use webcare on consumer-generated platforms to increase positive consumer engagement. Journal of Interactive Marketing, 30, doi: /j.intmar Schewe, C. D. (1973). Selected social psychological models for analysing buyers. Journal of Marketing, 37, doi: / Searls, D. & Weinberger, D. (2000). Markets are conversations. In R. Levine, C. Locke, D. Searls, & D. Weinberger (Eds.), The cluetrain manifesto: The end of business as usual,

48 Sela, A., Wheeler, S. C., & Sarial-Abi, G. (2012). We are not the same as you and I: Causal effects of minor language variations on consumers' attitudes toward brands. Journal of Consumer Research, 39(3), doi: / Söderlund, M., Oikarinen, E. L., & Heikka, E. L. (2017). Jokes in the store and its effects on customer satisfaction. The International Review of Retail, Distribution and Consumer Research, 27(3), doi: / Solomon, M. R., Surprenant, C.m Czepiel, J. A., & Gutman, E. G. (1985). A role theory perspective on dyadic interactions: The service encounter. Journal of Marketing, 49(1), doi: / Speck, P. S. (1987). On humor and humor in advertising (Doctoral dissertation, Texas Tech University). Sundaram, D. S., & Webster, C. (1999). The role of brand familiarity on the impact of word- of-mouth communication on brand evaluations. Advances in Consumer Research, 26, Sweetser, K. D., & Metzgar, E. (2007). Communicating during crisis: Use of blogs as a relationship management tool. Public Relations Review, 33(3), doi: /j.pubrev Taecharungroj, V., & Nueangjamnong, P. (2015). Humour 2.0: styles and types of humour and virality of memes on facebook. Journal of Creative Communications, 10(3), doi: / Tilburg University. (n.d.). Webcare conversation [Printscreen]. Retrieved from 48

49 Tom, G., & Eves, A. (1999). The use of rhetorical devices in advertising. Journal of Advertising Research, 39, Van Noort, G., & Willemsen, L. M. (2012). Online damage control: The effects of proactive versus reactive webcare interventions in consumer-generated and brand-generated platforms. Journal of Interactive Marketing, 26(3), doi: /j.intmar Van Noort, G., Willemsen, L.M., & Antheunis, M.L. (2013). Negative Word of Mouth on Social Network Sites: How to Change Online Brand Opinion Dynamics with Effective Webcare, paper presented on the European Marketing Academy Conference (EMAC), Istanbul, Turkey Van Noort, G., Willemsen, L. M., Kerkhof, P., & Verhoeven, J. W. (2014). Webcare as an integrative tool for customer care, reputation management, and online marketing: a literature review. In Integrated Communications in the Postmodern Era (pp ). Palgrave Macmillan UK. Vigsø, O. (2013). Ironic Crisis Communication?. Nordicom Review, 34(2), doi: /nor Weinberger, M. G., & Gulas, C. S. (1992). The impact of humor in advertising: A review. Journal of Advertising, 21(4), doi: / Weinberg, B. D., & Pehlivan, E. (2011). Social spending: Managing the social media mix. Business Horizons, 54(3), doi: /j.bushor Willis, F. N. (1966). Initial speaking distance as a function of the speakers' relationship. Psychonomic Science, 5(6), doi: /bf

50 Wise, K., Alhabash, S., & Park, H. (2010). Emotional responses during social information seeking on Facebook. Cyberpsychology, Behavior, and Social Networking, 13(5), doi: /cyber Xia, L. (2013). Effects of companies' responses to consumer criticism in social media. International Journal of Electronic Commerce, 17(4), doi: /JEC Xiao, Y., Cauberghe, V., & Hudders, L. (2017) Humour as a double edged sword in response to crises versus rumours: The effectiveness of humorously framed crisis response messages on social media. Journal of Contingencies and Crisis Management, doi: / Yang, S. U., Kang, M., & Johnson, P. (2010). Effects of narratives, openness to dialogic communication, and credibility on engagement in crisis communication through organizational blogs. Communication Research, 37(4), doi: / Zajdman, A. (1995). Humorous face-threatening acts: Humor as strategy. Journal of Pragmatics, 23(3), doi: / (94)00038-G Zhu, H., Zhang, Y., & Gong, X. (2017). Does company s humor resolve consumer complaining: The match of humor types and relationship norms. Acta Psychologica Sinica, 49(4), doi: /SP.J

51 Appendix A Pre-test materials Network complaint waarom heb ik al dagen geen bereik terwijl jullie toch zo n goed netwerk claimen te hebben? Kan niet eens een belletje doen? No humor Affiliative humor Self-defeating humor wat wat wat vervelend vervelend om te horen. vervelend om te horen. Het om te horen dat ons netwerk net Kun je je gegevens in een moet inderdaad maar net n(i)et werkt ;). En dat we onze claim privébericht naar ons werken bij je ;). Kun je je dus niet kunnen waarmaken. Kun je sturen? Dan gaan we op gegevens in een privébericht je gegevens in een privébericht naar zoek naar een oplossing. naar ons sturen? Dan gaan we ons sturen? Dan gaan we op zoek op zoek naar een oplossing. naar een oplossing. Payment complaint onze onze onze excuses. ik krijg hier net een excuses Kun je je excuses. Goed dat je je Wat vervelend dat we kennelijk niet hoge rekening binnen, maar ben gegevens in een privébericht krachten hebt gebundeld om in staat zijn om klantgegevens en helemaal niet over mijn bundel gegaan! naar ons sturen? Dan gaan ons te contacten ;). Kun je je rekeningen op een juiste manier te Ik ben niet van plan dit te gaan betalen.. we op zoek naar een gegevens in een privébericht bundelen ;). Kun je je gegevens in oplossing. naar ons sturen? Dan gaan wij een privébericht sturen? Dan gaan op zoek naar een oplossing. we op zoek naar een oplossing voor je. 51

52 Wrong information complaint wat wat wat vervelend om ik krijg net een SMS vervelend om te horen dat er vervelend om te horen dat er te horen dat er onduidelijkheid is van jullie met de informatie dat ik al onduidelijkheid is over je onduidelijkheid is over je over je mobiele tegoed. Er heeft hier door mijn internet heen ben voor deze mobiele tegoed. Kun je je mobiele tegoed. Kun je je intern nog iemand een lesje rekenen maand. Maar in de mobiele app staat dat gegevens in een privébericht gegevens in een privébericht tegoed ;). Kun je je gegevens in een ik pas 600 MB heb verbruikt. En ik heb naar ons sturen? Dan gaan naar ons sturen, want je hebt privébericht naar ons sturen? Dan een bundel van 3000 MB. Dan heb ik we het voor je oplossen. nog een oplossing tegoed ;). gaan we het voor je oplossen. toch nog 2400 MB over en klopt het toch weer niet wat jullie nu weer tegen me zeggen?! Bad consumer service complaint dat is ik probeer jullie dat is niet hoe het hoort. dat is niet hoe het hoort. Goed niet hoe het hoort. Goed dat je ons vandaag al de hele dag telefonisch te Kun je je gegevens in een dat je een bericht stuurt, daar een bericht stuurt, daar kun je bereiken, maar ik hang nu wederom 40 privébericht naar ons kun je heel wat mee bereiken momenteel kennelijk meer mee minuten in de wacht. Ik wil nu graag op sturen? Dan neem ik zo ;). Kun je je gegevens in een bereiken dan wanneer je ons korte termijn iemand spreken! Wat zijn gelijk persoonlijk privébericht naar ons sturen? telefonisch benadert ;). Kun je je jullie slecht te bereiken zeg.. telefonisch contact met je Dan neem ik zo gelijk gegevens in een privébericht sturen? op. persoonlijk telefonisch contact Dan neem ik zo gelijk persoonlijk met je op. telefonisch contact met je op. 52

53 Not working website complaint we we begrijpen wat een (t)error zeg Klopt het dat jullie website eruit ligt begrijpen dat dit vervelend dat het vervelend is dat je in ;). Ik denk dat ik vast Ik wil graag een nieuw is. Kun je je gegevens in deze (t)error situatie zit ;). Kun ga aanvragen, want als onze klanten abonnement afsluiten, maar dat gaat een privébericht naar ons je je gegevens in een geen abonnementen meer kunnen jammer genoeg niet. Het enige dat is te sturen? Dan gaan we op privébericht naar ons sturen? afsluiten gaat er toch echt iets goed zien krijg is error zoek naar een oplossing. Dan gaan we op zoek naar een mis hier. Kun je je gegevens in een oplossing voor je. privébericht sturen? Dan gaan we het voor je uitzoeken. Poor telecom reception dat is dat is niet dat is niet de ik heb een probleem niet de bedoeling. Kun je de bedoeling. Goed dat je een bedoeling. We zijn hier als met bellen. Elke keer als ik gebeld word je gegevens in een lijntje naar ons uitgooit door telecomorganisatie duidelijk over de door iemand en ik neem op hoor ik privébericht sturen? Dan een berichtje te sturen ;). Kun lijn gegaan ;). Kun je je gegevens helemaal niks meer aan de andere kant gaan we het voor je je je gegevens in een in een privébericht sturen? Dan gaan van de lijn.. Ik hoop dat jullie dit oplossen. privébericht sturen? Dan gaan we het voor je oplossen. kunnen oplossen, want ik betaal we het voor je oplossen. natuurlijk niet aan jullie als telecomprovider terwijl ik niet eens wat hoor aan de andere kant van de lijn! 53

54 Unjustified invoice complaint we we we begrijpen Jullie kunnen naar je geld fluiten deze begrijpen dat dit vervelend begrijpen dat dit een flu(i)t dat dit vervelend is. Het afsluiten van Ik krijg hier een is. Kun je je gegevens in situatie is voor je ;). Kun je je de juiste abonnementen blijkt voor rekening binnen omdat ik internet heb een privébericht naar ons gegevens in een privébericht ons zo te zien geen fluitje van een gebruikt tijdens mijn vakantie, maar sturen? Dan gaan we het sturen? Dan gaan we het voor cent ;). Kun je je gegevens in een daar had ik gewoon een abonnement voor je uitzoeken. je uitzoeken. privébericht sturen? Dan gaan we het voor afgesloten voordat ik ging. Ik ga voor je uitzoeken. niet voor jullie fouten betalen! 54

55 Beste participant, Appendix B Pre-test questionnaire Hartelijk dank voor het deelnemen aan deze enquête. Dit onderzoek wordt gehouden als onderdeel van mijn afstudeeronderzoek voor de master Bedrijfscommunicatie en Digitale Media. Het onderzoek gaat over webcare, in andere woorden, online gesprekken tussen consumenten en bedrijven op sociale media. In mijn onderzoek staat het gebruik van humor in webcare centraal. Om voorafgaand aan mijn hoofdonderzoek te testen of de grappen in dit onderzoek geslaagd zijn, heb ik uw hulp nodig. U krijgt zo dadelijk verschillende webcare gesprekken te zien van verschillende telecom organisaties. Hierin uit de consument een klacht naar de organisatie, gevolgd een reactie vanuit het bedrijf. U krijgt meerdere malen dezelfde klacht te zien, gevolgd door verschillende reacties. Na het zien van de webcare gesprekken volgen er verschillende stellingen. Om uw antwoorden op deze stellingen niet te beïnvloeden zijn de namen van de klant en het bedrijf zwart gemarkeerd. De enquête neemt ongeveer 10 minuten van uw tijd in beslag. Als u vragen heeft kunt u contact opnemen via Nogmaals hartelijk dank! Simone Schaffelaars <Webcare conversatie wordt getoond, zie Appendix A> Het bedrijf gebruikt humor in de reactie Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens Ik vind de reactie van het bedrijf grappig Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens Het functioneren van de organisatie is het onderwerp van deze grap Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens 55

56 De situatie van de consument is het onderwerp van deze grap Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens Met deze grap benadeeld de organisatie zichzelf Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens De consument wordt benadeeld door deze grap Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens Het bedrijf maakt deze grap om een ander een goed gevoel te geven Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens Met deze reactie wil de organisatie een band aangaan met de consument Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens Heeft u een betere suggestie voor deze grap? Optioneel.. Rank vraag Order de onderstaande webcare-gesprekken door de grappigste reactie bovenaan te zetten (1) en de minst grappige reactie onderaan (7) Affiliative condition Rank vraag Order de onderstaande webcare-gesprekken door de grappigste reactie bovenaan te zetten (1) en de minst grappige reactie onderaan (7) Self-defeating condition Geslacht o o Man Vrouw Leeftijd In cijfers 56

57 Hoogst genoten opleidingsniveau Indien u momenteel studeert, kies het opleidingsniveau dat u momenteel volgt o o o o o o o o o VMBO HAVO VWO HBO Bachelor HBO Master WO Bachelor WO Master PhD Anders Ziet u wel eens dit soort online conversaties tussen consumenten en bedrijven? o o Ja Nee Als u denkt aan humor in webcare reacties van telecom organisaties, bij welke telecom organisatie(s) vindt u het gebruik van humor passen? Meerdere antwoorden mogelijk.. Heeft u grappen gezien in de webcare reacties waarvan u denkt dat deze online niet zullen voorkomen?.. Heeft u nog opmerkingen/suggesties aan de hand van deze vragenlijst? Optioneel.. 57

58 Appendix C Measures Perceived Conversational Human Voice (1= zeer mee oneens; 7= zeer mee eens) 1. Het bedrijf nodigt uit tot conversatie 2. Het bedrijf gebruikt humor 3. Het bedrijf maakt communicatie aangenaam 4. Het bedrijf staat open voor een gesprek met de consument 5. Het bedrijf communiceert met een menselijke toon Perceived appropriateness (1= zeer mee oneens; 7= zeer mee eens) 1. Het bedrijf voldoet met deze reactie aan mijn verwachtingen 2. De reactie van het bedrijf is gepast 3. Het bedrijf reageert op een manier zoals ik van een bedrijf verwacht 4. De reactie van het bedrijf is geschikt voor deze situatie Attitude towards the webcare response (1= zeer mee oneens; 7= zeer mee eens) Mijn houding ten opzichte van de webcare-reactie is: 1. Goed 2. Plezierig 4. Onaangenaam 5. Positief 58

59 Beste participant, Appendix D Final questionnaire Hartelijk dank voor het deelnemen aan deze enquête. Dit onderzoek wordt gehouden als onderdeel van mijn afstudeeronderzoek voor de master Bedrijfscommunicatie en Digitale Media aan Tilburg University. Het onderzoek gaat over webcare, dat zijn online gesprekken tussen consumenten en bedrijven op sociale media zoals Facebook. (Familair) U krijgt zo dadelijk 3 korte webcare-gesprekken te zien tussen consumenten en de Nederlandse telecombedrijven Vodafone, T-Mobile en Tele-2. Na het zien van de webcaregesprekken volgen er verschillende stellingen om te beantwoorden. Er zijn geen goede of foute antwoorden op deze stellingen mogelijk, ik ben enkel benieuwd naar uw mening. (Unfamiliar) U krijgt zo dadelijk 3 korte webcare-gesprekken te zien tussen consumenten en de (jonge) Nederlandse telecombedrijven Call4All, Telecoms en CallPhone. Na het zien van de webcaregesprekken volgen er verschillende stellingen om te beantwoorden. Er zijn geen goede of foute antwoorden op deze stellingen mogelijk, ik ben enkel benieuwd naar uw mening. De enquête neemt 5 minuten van uw tijd in beslag. Deelname aan deze enquête is vrijwillig en uw gegevens zijn geheel anoniem. U kunt te allen tijde stoppen met deze vragenlijst, of de enquête op een later moment afronden. Als u vragen heeft kunt u contact opnemen via Nogmaals hartelijk dank voor uw deelname! Simone Schaffelaars <Per conditie worden er 3 verschillende webcare conversaties getoond (zie Appendix E), gevolgd door de volgende vragen> Geef aan in hoeverre u het eens bent met de volgende stellingen. Het bedrijf nodigt de consument uit tot conversatie Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens 59

60 Het bedrijf gebruikt humor Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens Het bedrijf maakt communicatie aangenaam Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens Het bedrijf staat open voor een gesprek met de consument Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens Het bedrijf communiceert met een menselijke toon Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens Geef aan in hoeverre u het eens bent met de volgende stellingen. Het bedrijf voldoet met deze reactie aan mijn verwachtingen Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens De reactie van het bedrijf is gepast Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens Het bedrijf reageert op een manier zoals ik van een bedrijf verwacht Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens 60

61 De reactie van het bedrijf is geschikt voor deze situatie Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens Mijn houding ten opzicht van de webcare-reactie is: Goed Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens Plezierig Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens Onaangenaam Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens Positief Zeer mee Mee Enigszins Neutraal Enigszins Mee Zeer mee oneens oneens mee oneens mee eens eens eens Tot slot vraag ik u een aantal persoonlijke gegevens in te vullen. 1. Geef aan welke organisaties u voorafgaand aan dit onderzoek al kende? (Familiar condition) Meerdere antwoorden mogelijk o Tele-2 o T-Mobile o Vodafone 61

62 1. Geef aan welke organisaties u voorafgaand aan dit onderzoek al kende? (Unfamiliar condition) Meerdere antwoorden mogelijk o Call4ALL o Telecoms o CallPhone o Geen van bovenstaande 2. Van welke telecomprovider bent u zelf klant? o Vodafone o T-Mobile o Telecoms o Tele-2 o KPN o Call4All o Ben o Hollands Nieuwe o CallPhone o Anders 3. Kwam er in één of meerdere conversaties die u heeft gezien humor voor? o Ja o Nee 4. Was u bekend met online conversaties tussen consumenten en bedrijven ( webcare ) voorafgaand aan deze enquête? o Ja o Nee 62

63 5. Heeft u zelf wel eens deelgenomen aan webcare-gesprekken? o Nooit o Zelden o Soms o Regelmatig o Vaak 6. Wat is uw geslacht? o Man o Vrouw 7. Wat is uw leeftijd? In cijfers.. jaar 8. Wat is uw hoogst genoten opleidingsniveau? Indien u momenteel studeert, kies het opleidingsniveau dat u momenteel volgt o VMBO o HAVO o VWO o HBO Bachelor o HBO Master o WO Bachelor o WO Master o PhD o Anders 63

64 9. Van welke sociale media maakt u gebruik? Er zijn meerdere antwoorden mogelijk o Facebook o Twitter o Snapschat o Instagram o Ik gebruik geen sociale media o Een ander medium dan bovenstaand (Familair) U bent aan het einde gekomen van deze enquête, dank voor uw tijd! Ik wil enkel nog benadrukken dat de webcare-conversaties die u zojuist heeft gezien fictief zijn. De desbetreffende telecom-organisaties hebben deze conversaties nooit daadwerkelijk gevoerd. Ze zijn enkel gemaakt voor dit experiment. Vergeet niet om op de pijltjes te klikken om uw antwoorden te versturen. (Unfamiliar) U bent aan het einde gekomen van deze enquête, dank voor uw tijd! Ik wil enkel nog benadrukken dat de webcare-conversaties die u zojuist heeft gezien fictief zijn, net als de telecom-organisaties gebruikt in deze enquete. De desbetreffende telecomorganisaties zijn fictief en hebben deze conversaties nooit daadwerkelijk gevoerd. Ze zijn enkel gemaakt voor dit experiment. Vergeet niet om op de pijltjes te klikken om uw antwoorden te versturen. 64

65 Condition 1 (Familiar, no humor) Appendix E Final materials 65

66 Condition 2 (Familiar, affiliative humor) 66

67 Condition 3 (Familiar, self-defeating humor) 67

68 Condition 4 (Unfamiliar, no humor) 68

69 Condition 5 (Unfamiliar, affiliative humor) 69

70 Condition 6 (Unfamiliar, self-defeating humor) 70

71 Filler familiar condition (used in condition 1-3) Filler unfamiliar condition (used in condition 4-6) 71