Brand transgressions, characterized by service failure, are a frequent theme for marketing scholars. Their impact on satisfaction, trust and brand loyalty is of high interest. However, in assessing the influence of those events on third-party consumers, the literature is still lacking. The purpose of this paper is to explore how social distance explains the reactions of close and distant third-party consumers toward other consumers during a brand transgression event. Anger is analyzed as a driver of this process.
Two experiments were conducted. Both studies presented a 3 (social distance: victim vs close third party vs distant third party) by 2 (severity: low vs high) between-subjects design. Respondents were asked to read a transgression scenario in a mobile phone service (study 1) and in a restaurant (study 2) and then completed scales that measured their affective reactions and evaluations of the relationship – satisfaction, trust, and loyalty intention – with the transgressing brand.
The results showed that transgression severity intensifies the effect of the brand transgression on consumer’s anger. Victims and close third parties demonstrated higher levels of anger compared to distant third-party consumers. In the case of severe transgressions, an experience of anger contagion between victims and close third-party consumers was responsible for the negative effect on the relationship evaluation of the transgressing brand compared to distant third-party consumers.
This study extends previous research about how social distance influences consumer-brand relationships and demonstrates the mediating role played by affective anger contagion.
Mantovani, D., Korelo, J. and Ibarra, J. (2018), "Effects of brand transgressions on third-party consumers", Marketing Intelligence & Planning, Vol. 36 No. 3, pp. 306-317. https://doi.org/10.1108/MIP-11-2017-0276Download as .RIS
Emerald Publishing Limited
Copyright © 2018, Emerald Publishing Limited
Transgressions are an intrinsic and inevitable perceived violation of rules that guide relationships (Aaker et al., 2004). The focus of attention on transgressions lays on the dyadic relationship between brands consumers, and a transgression takes place when the brand causes damage to the consumer who undergoes an experience of loss and distress. Despite of a good number of studies on the topic (Harrison-Walker, 2012; Kim and Ulgado, 2012; Sinha and Lu, 2016), more research is needed to address the issue when another consumer that have no prior problem with the brand is aware of brand’s transgression events toward third-party consumers (e.g. a friend of the mistreated consumer). In that case, it is assumed that the relationship between consumers and brands may become triadic (e.g. brand-consumer-third party consumer). If the brand produces bad outcomes to a mistreated consumer, the effect could extend to other consumers, who in turn could respond to the brand, depending on how socially close or distant they are from the victimized consumer. In this regard, this research aims to explore the aftermath of brand’s transgression over other consumers who are not directly the target of brand’s wrongdoing.
These types of incidents are more frequent than marketers conceive. This paper proposes that the socially closer the third-party consumers are from the victim of the brand transgression, the more similar their ratings about the brand will be in the face of a severe transgression (e.g. decrease of brand loyalty). This is due to the similar affective state that the brand transgression produces in both consumers. We focus on anger as a driver mechanism affecting brand’s assessment. Consumers that attribute the responsibility of mistakes to brands will experience anger, and socially close consumers will present similar pattern accounted by the contagion of anger. Contrary, when third-party consumers are socially distant from the victim, the event of transgression will not affect their rationalization. Therefore, they will not experience anger because the event is processed in an abstract level when compared to those that are close to the victim. This research builds on the existing literature in two ways. First, this study contributes to the body of knowledge on consumer-brand relationships, identifying the assessment of the relationship made by third-party consumers when they are aware of a brand transgressing on other consumers. Past research has demonstrated that others might influence our behavior and preferences (Linke, 2012; Mattila et al., 2014), but this research provides initial evidence of this influence in the consumer-brand relationship field. Based on previous studies on marketing relationship, consumer-brand relationship is often measured through satisfaction, trust, and loyalty (Fornell et al., 1996; Beerli et al., 2004; Sirdeshmukh et al., 2002). Therefore, these variables are the focus of this study for relationship evaluation.
Second, the study explores the role of a specific emotion – consumer anger – in service transgression and its impact on third-party consumers, a topic which has been largely overlooked in the literature. The cost of consumers’ angry reactions is higher than managers could expect (Funches, 2016). Besides the complaint websites (e.g. Complaints.com), which give voice to consumers to post details about the services and product failures, it is also common to see customers posting on their own social networks complaints about a transgressing company. Therefore, their close friends will also be aware of the brand failure. Past research has typically examined the affective reaction of the consumer that suffered the transgression (Harrison-Walker, 2012) and its potentially damaging effects (Grégorie et al., 2009; Funches, 2011). However, third-party consumers also experience anger when they are aware of others being treated unfairly (Mattila et al., 2014). Our research sheds light on the conditions under which this phenomenon occurs. Based on construal level theory (Trope et al., 2007; Trope and Liberman, 2010), this study argues that only close third party, but not the distant consumers, will be influenced by a brand transgression because of the emotional contagion between them (Berger, 2013). Therefore, they will both react negatively and, as a consequence, will present a similar decline in the post-event evaluations of the brand.
This research also provides important practical implications for companies, which should motivate brand managers to establish plans to identify and track brand failures according to the severity level.
Transgressions in consumer-brand relationships
Transgressions are considered harmful for both brands and consumers because they bring undesirable consequences, such as a negative impact on the brand’s image (Aaker et al., 2004; Trump, 2014) and threats to the continuity of the relationship (La and Choi, 2012). Past research shows that transgressions impact specific aspects of consumer-brand relationship evaluation, such as satisfaction (Fornell et al., 1996; Cambra-Fierro and Melero-Polo, 2017), trust (Sirdeshmukh et al., 2002), and loyalty (Beerli et al., 2004). For instance, La and Choi (2012) demonstrated that service failure has a high impact on trust and in loyalty intention. Kim and Ulgado (2012) also showed the negative impact of transgression on satisfaction. This negative impact is influenced by factors such as transgression controllability, brand relationship strength (Sinha and Lu, 2016), and recovery performance (Grégorie et al., 2009; Strizhakova et al., 2012; Kim and Ulgado, 2012). Research on the attribution-affect-behavior consequence has showed that when consumers blame the transgressing brand, they get angry (Funches, 2011; Strizhakova et al., 2012; Du et al., 2014), demonstrate greater dissatisfaction, intention to complain, negative word of mouth, and their willingness to repurchase is reduced (Harrison-Walker, 2012). These causal inferences guide consumers’ post-transgression behavior (Sinha and Lu, 2016).
Another important characteristic of transgressions is their variation in terms of severity. Transgression severity refers to a customer’s perceived intensity as it relates to a service problem. The more intense the service failure is, the greater will be the customer’s perception of loss (Tsarenko and Tojib, 2015). Research has demonstrated a strong correlation between the severity of brand service failure and consumer dissatisfaction and negative responses toward the company or brand (Kim and Ulgado, 2012; Strizhakova et al., 2012; Tsarenko and Tojib, 2015). The most common and explored perspective when analyzing the effects of brand transgression is the dyadic relationship between the brand and the customer (Aaker et al., 2004). Beyond this relationship, this study suggests that transgressions also affect third-party consumers (Casidy and Shin, 2015). The differences and similarities between the negative reactions of victims and third-party consumers are explained by social distance (Trope et al., 2007; Trope and Liberman, 2010) and the emotional contagion between these groups (Berger, 2013; Hasford et al., 2015).
Previous research has demonstrated that unfair treatment experienced by others triggers negative reactions in third-party consumers (e.g. anger), even when they have experienced fair treatment from the transgressing brand (Mattila et al., 2014; Casidy and Shin, 2015). This research suggests that the negative reaction of third-party customers depends on the social distance between them and the customer who was mistreated by the brand. Third-party consumers that are socially close to the victimized customer tend to interpret the event with greater similarity to them, whereas, consumers that are socially distant do not interpret the brand transgression as so harmful. This effect is explained by the psychological distance theory, or more specifically, the dimension of social distance that proposes how people construe their mental representations about an event (Trope et al., 2007; Trope and Liberman, 2010). The more psychologically distant an event, the more abstract it will be represented in the mind. Contrary, the more psychologically close the event, the more concrete it will be construed in peoples’ minds (Trope et al., 2007; Park and Morton, 2015).
If the brand transgression is severe, the intensity of individuals’ emotional affiliation with other people can amplify the social influence (Berger, 2013). Research also demonstrates that consumers’ negative reactions may rely on the intensity of the transgression (Tsarenko and Tojib, 2015), such that the higher the perceived severity of the transgression, the higher the levels of negative emotions. When consumers observe an unfair treatment caused by a severe transgression, they will also react with negative emotions. In this case, the social distance will influence other consumers’ affective reactions, such that close third-party consumers will experience anger similar to the victim, and as a consequence, they will present similar negative perceptions of the relationship with the brand (Linke, 2012). On the other hand, distant third-party consumers will take the perspective of the transgression as an isolated circumstance, not triggering anger, which by consequence will lead them to a milder evaluation of the relationship with the brand after a transgression. Therefore, the following hypothesis is proposed:
In face of severe transgressions, victims and close third-party consumers will demonstrate higher levels of anger compared to distant third-party consumers.
When forming evaluations of satisfaction, trust and loyalty about a brand, consumers use experienced emotions as a source of information (Funches, 2016; Hasford et al., 2015). Past research (Funches, 2011; Harrison-Walker, 2012; Strizhakova et al., 2012) shows that anger is responsible for the impact of service quality on relationship evaluation, often measured in terms of satisfaction, trust, and loyalty intention. Once emotional contagion is also triggered by the capacity of empathy, or by imagining another person’s feelings, the perspectives of close others are strongly considered, which enhances the emotional transfer (Parkinson, 2011). Because of affective sharing and affective similarity attraction, individuals’ emotions can promptly transfer to other group members (Berger, 2013; Du et al., 2014; Hasford et al., 2015). For example, Du et al. (2014) found that the impact of group emotional contagion is more intense in a familiar group than in an unfamiliar group. The authors argued that familiar groups (e.g. friends and family members) are more likely to have similar opinions and behavior because they are more susceptible to emotional contagion with each other than with distant others. As a consequence, victims and close third-party consumers will share similar affective reactions, and their evaluations about satisfaction, trust and loyalty about in face of a severe transgression will be similarly mediated by the feeling of anger. However, distant third-party consumers assess an event more abstractly and feel lower levels of anger, presenting a more positive assessment of the brand relationship after a transgression.
This research proposes that anger will negatively mediate the impact of the interaction between social distance and severity on relationship evaluation. When the transgression is severe, the victim’s anger, driven by the mechanism of affective sharing and contagion (Berger, 2013; Linke, 2012; Du et al., 2014), will transfer to close third-party consumers, and as such, they will provide the same negative evaluation of satisfaction, trust and loyalty, unlike distant third-party consumers. This effect might be contingent on the transgression severity. Events that are not severe in terms of harming the consumer have a weaker effect on the relationship, since the consumer experiences less anger in that situation. As a result, this study suggests the following hypothesis:
In face of severe transgressions, an experience of anger contagion between victims and close third-party consumers will result in a stronger negative effect on the relationship evaluation – satisfaction, trust, and loyalty – of the brand compared to distant third-party consumers.
To test the two hypotheses of the research, a scenario-based design in the domain of mobile phone services was selected, using a fictitious brand, named COMCEL. This sector is characterized by high levels of failures toward consumers and high rates of complaints, resulting in one of the lowest rates of satisfaction on the ACSI index (mobile phone services ranked 38th among 43 industry sectors in the ACSI, 2016 index).
The sample was composed of 176 undergraduate students (Mage=21.9 years, 55.4 percent women). The experiment presented a 3 (social distance: victim vs close third party vs distant third party) by 2 (severity: low vs high) between-subjects design. Social distance manipulation procedure was based on the concepts of Trope et al. (2007) and Trope and Liberman (2010). A recall task was used to induce the close third-party condition, in which participants were asked to think of a real and close person in their lives and to explain the reasons the person is important to them (e.g. what is the relationship between you and the person?; In general, what do you feel when you are in his/her presence?; What would be the main similarity between you and this person?).
In the distant third-party condition manipulation was adapted to use a recognition mechanism in which participants were asked to read the scenario about a known but socially distant person, a student who took the same course as you during a different semester. Participants in the victim condition were not subjected to the social distance manipulation and were directly asked to read and imagine themselves in the brand transgressing situation.
After the social distance manipulation, we addressed the transgression of the service in two levels (low vs high). We asked participants to imagine themselves using a service from a cell phone company COMCELL, and that everything about the service was going ok until that moment. But one day the company made a mistake: printed wrong name (low-severity condition); $100 was charged on the cell phone bill for an average of $30 of regular consumption and also, the name of the victim was included in companies’ list of consumers in debt (high-severity condition). The severity was adapted for each condition of social distance, where the target of the mistake was: the participant (victim condition); a close person (close third-party condition); or a distant person (distant third-party condition).
These manipulations were selected based on a separated pre-test (n=30, 64 percent male, Mage=24.2 years) to identify the severity associated with the transgression situations. The results indicated no difference in the perceptions of scenario reality (Mlow=6.03, SD=1.09; Mhigh=5.92; SD=1.36); t(28)=−1.01, p=0.515), where 1=it is not a possible real situation and 7=it is a possible real situation. Most important, participants assigned to the high-severity condition reported higher means of severity perception (Mhigh=6.18, SD=1.01) compared to the low-severity condition (Mlow=2.76; SD=1.33; t(28)=7.92, p<0.001).
After reading the scenario, participants answered questions involving the measurements of the experiment. Anger was operationalized with one question about how much anger the participant would feel in the situation described in the scenario. Satisfaction was measured with the scale adapted from Fornell et al. (1996) containing three items: ideal conformity, global satisfaction, and expectations with the brand, all collapsed to a satisfaction index (α=0.88). Trust was measured following the proposal of Sirdeshmukh et al. (2002), with three items collapsed in an index: the brand’s competence, integrity, and concern with customers (α=0.90). Loyalty was measured using the scale adapted from Beerli et al. (2004), which also included three items collapsed in an index: intentions to solve the problem first with the brand, keep using the brand services, and recommend the brand to others (α=0.92). All items varied from 1=“completely disagree” to 7=“completely agree.”
The manipulation check of social distance and severity were performed with t-tests for independent samples. Only the differences between two groups were compared (close vs distant third-party consumers), since there was no social distance for the victim group. The results of social distance showed the manipulation worked. There was a significant difference between the measures of close third (Mclose third=6.03, SD=1.09) and distant third (Mdistant third=3.34, SD=1.45; t(114)=11.34, p<0.01). The manipulation of the transgression severity worked as well. There was a significant difference for participants who responded in the low-severity condition (Mlow=3.52, SD=1.62) compared to the high severity (Mhigh=6.08; SD=1.18; t(174)=−11.93, p<0.01).
We performed a 3 (social distance)×2 (severity) two-way ANOVA on satisfaction, trust, and loyalty. We adopted an ANOVA rather than a MANOVA because the focus on this research is to analyze each dependent variable isolated from the other two. That is a more adequate approach to the mediated model that our hypotheses predict. Despite this fact, the results from MANOVA presented no changes in the adjusted model. The results from ANOVA showed a marginal main effect of social distance on satisfaction (F(2, 170)=2.43, p=0.091, ηp2=0.028) and trust (F(2, 170)=2.65, p=0.074, ηp2=0.030) and a significant main effect on loyalty (F(2, 170)=3.43, p=0.035, ηp2=0.039). Also, a significant main effect of transgression severity on satisfaction was found (F(1, 170)=56.89, p<0.01, ηp2=0.251), trust (F(1, 170)=39.66, p<0.01, ηp2=0.189), and loyalty (F(1, 170)=77.18, p<0.01, ηp2=0.312). Additionally, there was a significant interaction effect on satisfaction (F(2, 170)=3.29, p=0.040, ηp2=0.037), and loyalty (F(2, 170)=5.65, p=0.004, ηp2=0.062), but no interaction effect for trust (F(2, 170)=1.85, p=0.161). These results are in accordance with the rationalization that only through anger (our predicted mediator) the manipulated conditions can affect the dependent variables.
To test H1, a 3 (social distance)×2 (severity) ANOVA on anger was performed. Results show a marginal main effect of social distance (F(2, 170)=2.39, p=0.095, ηp2=0.027), a significant main effect of transgression severity (F(1, 170)=43.46, p<0.01, ηp2=0.204), and a marginal interaction effect (F(2, 170)=2.35, p=0.098, ηp2=0.027). Pairwise comparisons showed that the marginal interaction effect is due to differences concentrated within social distance groups in the high-severity transgression condition. In the low-severity condition, there was no difference between groups (Mvictim=2.73, SD=1.80; Mclose third=3.24, SD=2.01; Mdistant third=2.87, SD=2.33; F(2, 170)=0.52, p=0.598). On the contrary, in the high-severity condition there was a difference between the distant other (Mdistant third=4.07, SD=2.11) when compared with the victim (Mvictim=5.53, SD=1.98) and the close other (Mclose third=5.17, SD=1.64; F(2, 170)=4.09, p=0.018, ηp2=0.046). Moreover, each social distance group individually showed a significant difference when comparing the low and high conditions of anger: higher anger for the victim (F(1, 170)=29.74, p<0.01, ηp2=0.149), the close third party (F(1, 170)=13.83, p<0.01, ηp2=0.075), and the distant third party (F(1, 170)=5.24, p=0.023, ηp2=0.030) for the high transgression severity. These results show initial support for the mediation of anger.
To confirm that anger mediates the effect of the interaction between social distance and severity on our dependent measures (H2 prediction), an interaction term of social distance and severity of transgression was created and performed a mediation analysis (model 4) using the procedure recommended by Hayes (2013). We used an interaction term and a simpler model of mediation instead of model 7 – Hayes (2013), because we are interested in the variance explained within the condition of high severity compared with the low severity in a model with three conditions in the independent variable coded as follow. To test the victim when compared with the third party, social distance was coded with two contrast variables (1, −1, 0 – for close third party and 1, 0, −1 – for distant third party). A 95 percent conﬁdence interval (CI) of the parameter estimates was obtained by running the resampling 10,000 times (Figure 1).
The results in Figure 1 confirm that anger mediates the effect of the interaction from social distance and transgression severity on the three dependent variables (satisfaction, trust, and loyalty). More importantly, this process is only significant when comparing the victim and distant third party (Figure 1(a)), which was confirmed by the indirect effect on satisfaction (Coef.=−0.28; CI −0.563-−0.083), trust (Coef.=−0.24; CI −0.488-−0.070), and loyalty (Coef.=−0.31; CI −0.615-−0.094). As expected, the model that compared the victim and close third party (Figure 1(b)) showed no significant indirect effect on satisfaction (Coef.=0.08; CI −0.111-0.263), trust (Coef.=0.07; CI −0.092-0.237), and loyalty (Coef.=0.09; CI −0.123-0.299). These results support anger as our hypothesized mechanism and explain why the victim and the close third party present lower satisfaction, trust, and loyalty when compared with the distant third-party consumers.
The results of experiment 1 provided evidences that anger triggered the decrease of satisfaction, trust, and loyalty with the brand after the transgression. Particularly, the victim and the close third-party consumer shared more anger when the transgression was severe, compared with the distant third-party consumer. When transgression severity was low, anger was not triggered, and there were no differences among the groups on the relationship with the brand. This outcome is important because it indicates in which circumstances third-party consumers could be contaminated by an event affecting other consumers.
More important, these results are contingent to anger contagion (Mattila et al., 2014; Casidy and Shin, 2015). As predicted the effect of the interaction of social distance and transgression severity on brand relationship evaluation occurs through the same affective reactions of the victim and the close third-party consumers. In fact, sharing the same emotions will provide similar negative evaluation of the transgressing brand (Berger, 2013; Du et al., 2014). The interaction of social distance and transgression severity was not significant on trust evaluation and there was a marginal main effect of social distance on satisfaction and trust. Those results were predicted theoretically where the interaction would present the effect in the dependent variables only through the mediator. Although these results support the hypotheses, it is an open question whether these findings would replicate in a different scenario. To address this goal, the scenario manipulation in the experiment 2 was changed.
Experiment 2 replicates the results of the previews study and uses another domain of brand transgression.
The sample consisted of 194 college students (Mage=22.3 years old; 52.6 percent men). The experiment used a similar design to the previous experiment. Social distance was manipulated using a similar procedure form experiment 1. To manipulate severity of transgression, the scenarios were adapted from Kim and Ulgado (2012) in which participants were asked to think about service at a restaurant. This is a service where consumers can easily change the service provider, compared to the mobile phone services. If there is a failure, the relationship could be more influenced, even when a low-severity failure occurs. Therefore, this service context is useful to investigate the consistency of the previous findings in study 1. Also, this scenario is more suitable and relevant to our sample reality (college students).
In the low-severity condition, they were asked to think about a situation where the waiter just knocked over water and wet the clients’ coat. In the high condition, the waiter took too much to bring the ordered food and charged $30 more for the service. These manipulations were selected based on a separated pre-test (n=30, 50 percent female, Mage=22.7 years) to identify the severity associated with the transgression situations. The results indicated no difference in the perceptions of scenario reality (Mlow =5.73, SD=1.51; Mhigh =5.80; SD=1.42); t(28)=−0.39, p=0.54), where 1=it is not a possible real situation and 7=it is a possible real situation. Most important, participants assigned to the high severity condition reported higher means of severity perception (Mhigh=5.64, SD=1.23) compared to the low-severity condition (Mlow=2.62; SD=1.52; t(28)=5.99, p<0.001).
After reading the scenario, participants rated satisfaction (α=0.88), trust (α=0.90), and loyalty (α=0.91). Anger was measured with a single item, similarly to experiment 1.
Social distance presented a significant difference between close third (Mclose third=5.69, SD=1.33) and distant third (Mdistant third=3.26, SD=1.13; t(122)=10.94, p<0.01). The transgression severity presented a significant difference between participants who responded to the low condition (Mlow=1.88, SD=1.09) compared with the high condition (Mhigh=4.54, SD=1.52; t(188)=−13.94, p<0.01).
Similarly to experiment 1, a 3 (social distance)×2 (severity) ANOVA on satisfaction, trust, and loyalty was performed. This time, the results showed a marginal main effect of social distance on satisfaction (F(2, 184)=2.59, p=0.078, ηp2=0.027) and no significant main effect for trust (F(2, 184)=0.90, p=0.408) and loyalty (F(2, 184)=1.70, p=0.186). Also, there were no significant interaction effects on satisfaction (F(2, 184)=2.33, p=0.100), trust (F(2, 184)=0.62, p=0.541), and loyalty (F(2, 184)=1.25, p=0.288).The results only showed the significant main effect of severity on satisfaction (F(1, 184)=87.10, p<0.01, ηp2=0.321), trust (F(1, 184)=31.73, p<0.01, ηp2=0.147), and loyalty (F(1, 184)=76.83, p<0.01, ηp2=0.295). These patterns are similar to our predicted causal path in H1 and H2.
As expected, the 3 (social distance)×2 (severity) ANOVA on anger showed a significant main effect for social distance (F(2, 184)=7.10, p<0.01, ηp2=0.072) and severity (F(1, 184)=28.12, p<0.01, ηp2=0.133) and a significant interaction effect (F(2, 184)=8.47, p<0.01, ηp2=0.084).
Pairwise comparisons showed that the differences between groups are concentrated in the high-severity transgression condition. In the low-severity condition, there was no difference between social distance groups (Mvictim=1.85, SD=1.20; Mclose third=1.94, SD=1.22; Mdistant third=1.94, SD=1.20; F(2, 184)=0.04, p=0.962). In the high-severity condition there was a higher anger for victim (Mvictim=4.15, SD=1.94) when compared with third close (Mclose third=3.03, SD=1.92) and third distant (Mclose third=2.04, SD=1.37; F(2, 184)=14.95, p<0.01, ηp2=0.140). When low- and high-severity differences are analyzed individually within each social distance group, the victim presented higher anger in the high- vs low-severity condition (F(1, 184)=38.32, p<0.01, ηp2=0.172). Participants in the close third-party condition also presented higher anger in the high- vs low-severity condition (F(1, 184)=8.26, p<0.01, ηp2=0.043). That difference did not happened for distant third-party participants (F(1, 184)=0.06, p=0.804).
These results suggest anger as a mediator with respect to the high-severity transgression, specifically for participants who were psychologically closer to the event of transgression. Next, the role of anger accounting for the effect on dependent variables was confirmed. This analysis followed a similar procedure to the previous examination. Figure 2 shows the results.
The results in Figure 2(a) show that the mediation of anger occurred only when the victim with the distant third-party consumer were compared. More interesting, this is only valid when the transgression severity is high, verified by the interaction with social distance. There was an indirect effect (Coef.=−0.39; CI −0.679-−0.184) for satisfaction, (Coef.=−0.30; CI −0.532-−0.145) for trust, and (Coef.=−0.43; CI −0.788-−0.190) for loyalty. The model that compared the victim and close third-party consumers (Figure 2(b)) showed no significant indirect effect on satisfaction (Coef.=−0.01; CI −0.221-0.195), trust (Coef.=−0.01; CI −0.179-0.142), and loyalty (Coef.=−0.01; CI −0.246-0.213). As in experiment 1, these results support anger as a mediator that justifies the decrease in the value of satisfaction, trust, and loyalty for the victim when compared to the distant third party. These results support H2.
The results of the two studies confirmed that when transgressions are severe, anger mediates the impact of social distance on satisfaction, trust, and loyalty toward the brand. However, the interaction of social distance and transgression severity on satisfaction, trust and loyalty was not significant, the result showed anger contagion might be essential for the impact of this interaction on brand relationship evaluation to be significant. It is necessary that close third-party consumers and victims of transgressions actually feel the same affective reactions about the transgressing brand for the effect to emerge. Research on emotional contagion (Berger, 2013; Du et al., 2014) suggests that emotions are more likely to be shared among similar others. This work shows that when the victim’s anger transfers to close third-party consumers, they will provide the same negative evaluation of the brand relationship.
General discussion and implications
Although previous studies have investigated brand transgressions, they have mainly focused on the interactions between the company and individual customers or a group of customers. This study provided a perspective of the third-party consumers, investigating how third-party consumers, close and distant, react to a transgression of a brand toward other consumers.
This research advances the understanding of the potential consequences of severe brand transgressions by integrating social distance theory and affective responses. Little research has been done that goes beyond the dyadic relationship between consumers and brands (see Mattila et al., 2014 for an exception). This study suggests that viewing brand transgressions from different perspectives plays a critical role in the marketing brand relationship theory and is useful for future studies in the area of company and brand transgressions and recovery. Past research using emotional contagion theory has focused on a group of consumers (Du et al., 2014), but has not considered other consumers that are not a victim of the brand’s wrongdoing. This study demonstrates that close third parties that are potential or actual customers will also be affected by the brand’s wrongdoing. Thus, the possible damage caused by a transgression can become much bigger than marketers and brands expect.
The social distance between consumers also contributes theoretically to consumer-brand relationship literature when dealing with negative events (Aaker et al., 2004; Grégorie et al., 2009; Trump, 2014). Construal level theory (Trope et al., 2007; Trope and Liberman, 2010) suggests that one’s behavior is more likely to be influenced by situational factors when at a close rather than a distant psychological distance. Specifically, this study shows evidence that the construal level makes consumers more susceptible not only to behavioral influences from others, but to behaviors performed by brands in relation to close and distant others.
From a managerial perspective, companies are constantly monitoring customer complaints, but may be missing opportunities to respond to customers in a way that close third-party consumers could be aware of this resolution. Therefore, online complaints could be an opportunity to improve the relationship not only with the mistreated customer, but also with the others close to these customers. Most managers keep their focus on customers’ complaints and how to resolve services failures. We recommend managers to keep attention on the perceived severity of the transgression because the implications for emotional contagion for close others are more pronounced. The emotional contagion may impact relationships that would not be at risk at first sight. This work indicates that brand transgressions consequences should be viewed as a means by which customers engage to each other, rather than only as a way they interact with the company.
Limitations and future directions
Even though a special effort was taken to construct realistic scenarios, a limitation of this research is the use of scenario-based experiments with transgression manipulation. It is possible that the effects related to this information were stronger than they would be in everyday life, especially for third-party consumers. For this reason, replicating this study under even more naturalistic conditions than done herein would provide a more accurate test of transgressions’ effects on the consumer-brand relationship. Nevertheless, this research shows evidence that the more severe the transgression, the more intense the impact on consumers’ evaluations. Studies 1 and 2 suggest that this effect is consistent across different consumption situations.
Another limitation is the convenience sampling of undergraduate students used for both studies. This group may not represent the behavior of other groups in the society. According to Peterson (2001), it would be important to investigate the predicted variables relationships with a different and more representative sample to add higher validity to the findings. Despite that drawback, the author argue that student samples are very suitable to address internal validity, specifically because they are more homogeneous and in some cases present effect sizes comparable to real consumer samples. One major concern running, for instance, a field experiment would be manipulating the service failure without penalties to the service provider and to the sample. Also, the social distance manipulation could become hard to conduct and ineffective in that situation.
Kim and Ulgado (2012) warned that some transgressions, such as a wedding photographer who loses the film or digital images of an event, are considered so severe that they are impossible to recover from. Future studies could address this issue and investigate if distant third-party consumers would also be affected by such severe transgressions. Also, future researchers could further investigate when distant consumers are influenced. This research has discussed the situation in which distant consumers are influenced by unknown consumers who complain about a service provider when they are about to consume or buy the same service (e.g. book a hotel). But in this situation, they see themselves as possible future victims, and that is why they are influenced. The underlying mechanism is probably different, since there is no emotional contagion. Other possible emotions may be taking place in the distant third party affective reactions, such as fear of future losses or regret, for example.
Aaker, J., Fournier, S. and Brasel, S.A. (2004), “When good brands do bad”, Journal of Consumer Research, Vol. 31 No. 1, pp. 1-16.
ACSI (2016), “American customer satisfaction index report”, available at: http://marketing.theacsi.org/acton/attachment/5132/f-0048/1/-/-/-/-/ACSI%20Telecommunications%20Report%202016.pdf (accessed September 9, 2016).
Beerli, A., Martín, J.D. and Quintana, A. (2004), “A model of customer loyalty in the retail banking market”, European Journal of Marketing, Vol. 38 Nos 1/2, pp. 253-275.
Berger, J. (2013), Contagious: Why Things Catch On, Simon & Schuster, New York, NY.
Cambra-Fierro, J. and Melero-Polo, I. (2017), “Complaint-handling as antecedent of customer engagement: do consumer characteristics matter?”, Marketing Intelligence & Planning, Vol. 35 No. 2, pp. 277-295.
Casidy, R. and Shin, H. (2015), “The effects of harm directions and service recovery strategies on customer forgiveness and negative word-of-mouth intentions”, Journal of Retailing and Consumer Services, Vol. 27 No. 1, pp. 103-112.
Du, J., Fan, X. and Feng, T. (2014), “Group emotional contagion and complaint intentions in group service failure: the role of group size and group familiarity”, Journal of Service Research, Vol. 17 No. 3, pp. 326-338.
Fornell, C., Johnson, M.D., Anderson, E.W., Cha, J. and Bryant, B.E. (1996), “The American customer satisfaction index: nature, purpose, and findings”, Journal of Marketing, Vol. 60 No. 4, pp. 7-18.
Funches, V. (2011), “The consumer anger phenomena: causes and consequences”, Journal of Services Marketing, Vol. 25 No. 6, pp. 420-428.
Funches, V. (2016), “A model of consumer anger”, Journal of Marketing Management, Vol. 4 No. 1, pp. 31-42.
Grégorie, Y., Tripp, T.M. and Legoux, R. (2009), “When customer love turns into lasting hate: the effects of relationship strength and time on customer revenge and avoidance”, Journal of Marketing, Vol. 73 No. 6, pp. 18-32.
Harrison-Walker, J.L. (2012), “The role of cause and affect in service failure”, Journal of Services Marketing, Vol. 26 No. 2, pp. 115-123.
Hasford, J., Hardesty, D.M. and Kidwell, B. (2015), “More than a feeling: emotional contagion effects in persuasive communication”, Journal of Marketing Research, Vol. 52 No. 6, pp. 836-847.
Hayes, A.F. (2013), Introduction to Mediation, Moderation, and Conditional Process Analysis, The Guilford Press, New York, NY.
Kim, N. and Ulgado, F.M. (2012), “The effect of on-the-spot versus delayed compensation: the moderating role of failure severity”, Journal of Services Marketing, Vol. 26 No. 3, pp. 158-167.
La, S. and Choi, B. (2012), “The role of customer affection and trust in loyalty rebuilding after service failure and recovery”, The Service Industries Journal, Vol. 32 No. 1, pp. 105-125.
Linke, L.H. (2012), “Social closeness and decision making: moral, attributive and emotional reactions to third party transgressions”, Current Psychology, Vol. 31 No. 3, pp. 291-312.
Mattila, A., Hanks, L. and Wang, C. (2014), “Others service experiences: emotions, perceived justice, and behavior”, European Journal of Marketing, Vol. 48 Nos 3/4, pp. 552-571.
Park, S. and Morton, C.R. (2015), “The role of regulatory focus, social distance, and involvement in anti-high-risk drinking advertising: a construal-level theory perspective”, Journal of Advertising, Vol. 44 No. 4, pp. 338-348.
Parkinson, B. (2011), “Interpersonal emotion transfer: contagion and social appraisal”, Social and Personality Psychology Compass, Vol. 5 No. 7, pp. 428-439.
Peterson, R.A. (2001), “On the use of college students in social science research: insights from a second-order meta-analysis”, Journal of Consumer Research, Vol. 28 No. 3, pp. 450-461.
Sinha, J. and Lu, F. (2016), “‘I’ value justice, but ‘we’ value relationships: self-construal effects on post-transgression consumer forgiveness”, Journal of Consumer Psychology, Vol. 26 No. 2, pp. 265-274.
Sirdeshmukh, D., Singh, J. and Sabol, B. (2002), “Customer trust, value, and loyalty in relational exchanges”, Journal of Marketing, Vol. 66 No. 1, pp. 15-37.
Strizhakova, Y., Tsarenko, Y. and Ruth, J.A. (2012), “I’m mad and I can’t get that service failure off my mind: coping and rumination as mediators of anger effects on customer intentions”, Journal of Service Research, Vol. 15 No. 4, pp. 414-429.
Trope, Y. and Liberman, N. (2010), “Construal-level theory of psychological distance”, Psychological Review, Vol. 117 No. 2, pp. 440-463.
Trope, Y., Liberman, N. and Wakslak, C. (2007), “Construal levels and psychological distance: effects on representations, prediction, evaluation and behavior”, Journal of Consumer Psychology, Vol. 17 No. 2, pp. 83-95.
Trump, R.K. (2014), “Connected consumers’ responses to negative brand actions: the roles of transgression self-relevance and domain”, Journal of Business Research, Vol. 67 No. 9, pp. 1824-1830.
Tsarenko, Y. and Tojib, D. (2015), “Consumers’ forgiveness after brand transgression: the effect of the firm’s corporate social responsibility and response”, Journal of Marketing Management, Vol. 31 Nos 17/18, pp. 1851-1877.
About the authors
Danielle Mantovani is an Adjunct Professor of Marketing at the Universidade Federal do Paraná, College of Business Administration, Curitiba, Brazil. Her research interests are concentrated in the area of consumer decision making.
José Carlos Korelo is an Adjunct Professor of Marketing at the Universidade Federal do Paraná, College of Business Administration, Curitiba, Brazil. His research interests lies in the area of consumer decision making.
Jenny Ibarra is a Master of Science at the Universidade Federal do Paraná, College of Business Administration, Curitiba, Brazil. Her research interests lie in the area of consumer behavior.