DON’T EXCLUDE ME: BEING EXCLUDED IN A BRAND COMMUNITY OWNED BY OTHER USERS LEADS CONSUMERS TO AVOID THE BRAND

People are social animals and need to interact in communities to feel included. However, sometimes they face exclusion situations in many interactions. Nevertheless, little is known about how this issue affects consumers’ purchases after being excluded from a brand community. In this study, We performed three laboratory experiments to demonstrate the proposed effects of social exclusion on consumer choice and sequential mediation. In this study, we demonstrated that after being excluded from a brand community, consumers perceive their relationship poorer compared with included consumers (study 1), exhibit lower perceptions of an ideal relationship with the brand (study 2), and are more prone to purchase a product in a rival brand (study 3). Furthermore, the sequential mediation of ideal perceptions of brand relationship and quality drives consumer intention to purchase a product from a rival brand. These findings contribute to the social exclusion theory and brand community literature by demonstrating how consumers felt after being excluded from a brand community and the psychological mechanisms underlying this effect.


Resumo
As pessoas são animais sociais e precisam interagir em comunidades para se sentirem incluídas. No entanto, às vezes eles enfrentam situações de exclusão em muitas interações. No entanto, pouco se sabe sobre como essa questão afeta as compras dos consumidores após serem excluídos de uma comunidade de marca. Neste estudo, realizamos três experimentos de laboratório para demonstrar os efeitos propostos da exclusão social na escolha do consumidor e na mediação sequencial. Neste estudo, demonstramos que, após serem excluídos de uma comunidade de marca, os consumidores percebem seu relacionamento pior em comparação com os consumidores incluídos (estudo 1), exibem percepções mais baixas de um Furthermore, the sequential mediation of ideal perceptions of brand relationship and quality drives consumer intention to purchase a product from a rival brand. These findings contribute to the social exclusion theory and brand community literature by demonstrating how consumers felt after being excluded from a brand community and the psychological mechanisms underlying this effect. Keywords: Social Exclusion; Brand Community; Brand Relationship; Belongingness; Ostracism. relacionamento ideal com a marca (estudo 2) e são mais propensos a comprar um produto de uma marca rival (estudo 3). Além disso, a mediação sequencial de percepções ideais de relacionamento com a marca e qualidade impulsiona a intenção do consumidor de comprar um produto de uma marca rival. Esses achados contribuem para a teoria da exclusão social e a literatura da comunidade de marca ao demonstrar como os consumidores se sentiram após serem excluídos de uma comunidade de marca e os mecanismos psicológicos subjacentes a esse efeito.

INTRODUCTION
The human being is a social animal and needs to interact with others in the community to feel included (DeBono & Muraven, 2014). In this way, a typical attitude when facing exclusion is to search for manners to solve this issue (Whitehead, 2003). Generally, people search for strategies to reaffiliate with others or for objects representing a social group to fill that necessity of belongingness (Brown, Sacco & Medlin, 2019;Chen, Wan & Levy, 2017). Furthermore, social exclusion leads people to different behavior and consumption to reduce the discrepancy between ideal and current social relationships (Mandel et al., 2017).
Based on this perspective (Baumeister & Tice, 1990;Baumeister, Twenge & Nuss, 2002a), the present research examines how social exclusion in a brand community influences people's choices related to the brand. As a ubiquitous force changing people's behaviors, attitudes, and psychological and cultural aspects (Baumeister et al., 2002;Verkuyten & Martinovic, 2017), ostracism permeates many aspects of consumer's lives, including their consumption behavior (Mead, Baumeister, Stillman, Rawn, & Vohs, 2011;Ward & Dahl, 2014). As social rejection and ostracism can become more frequent in some communities (e.g., brand communities), it is critical to understand its effects on consumption (Mead et al., 2011).
This research will focus on excluding other users in a brand community. For instance, imagine a consumer with a strong brand relationship and needing help with his cellphone. However, he searches for a brand community, and nobody helps him. Naturally, this issue will lead the consumer to feel excluded in that community (Sinha & Lu, 2019). Thus, how will this consumer feel after the community's exclusion, even with its strong relationship with the brand? This research addressed this gap and revealed some aspects of the consumer-brand relationship after social exclusion in the brand community and how this episode of ostracism can affect his subsequent choices.
Prior research in marketing indicates that socially excluded people in a brand community tend to increase their desire for the brand because of the perception of exclusiveness (Morgan, Ward & Dahl, 2014). Also, social exclusion affects people's preference for product density (Su, Wen & Jiang, 2019), anthropomorphized products (Chen, Wan & Levy, 2017) affect people's cognitive capabilities (Baumeister, Twenge & Nuss, 2002), and, lead people to search for affiliations (Mead et al., 2011). Hence, there are theoretical reasons to expect social exclusion to affect consumer relationships and choices negatively.
In this way, we employed social exclusion theory (Baumeister & Tice, 1990;Baumeister, Twenge, & Nuss, 2002;Twenge, Catanese & Baumeister, 2002) and brand relationship theory (Aggarwal & Larrick, 2012;Alvarez & Fournier, 2016) to investigate those issues. We argue that brand relationship quality and ideal perception of this relationship vary among socially excluded consumers compared with included ones. Specifically, social exclusion decreases relationship quality and increases the discrepancy between the ideal and current relationship.
Previous studies demonstrated that social exclusion in a brand community could increase people's desire for the brand (Ward & Dahl, 2014). However, we propose that the effects of exclusion in a community are carried over to the brand, and socially excluded consumers will avoid it. Furthermore, they will opt for a rival brand to buy a product due to their decreased brand relationship quality. Therefore, we performed three studies using social exclusion in a fictitious brand community to test these effects.
Our findings contribute to social exclusion literature by demonstrating that the exclusion performed by a brand community leads consumers to choose rival brands. Further, we demonstrated that the social exclusion made by a third party directly affects the consumer-brand relationship. Expressly, when other users exclude consumers, they decrease their Ideal perceptions of the brand, affecting their relationship quality. Moreover, to our knowledge, this research is the first endeavor exploring the serial mediation of perceptions of the ideal relationship and relationship quality in social exclusion literature. Finally, we contribute to brand relationship literature by demonstrating that brand relationship quality has another antecedent: the ideal relationship's perception. Implications for managers and directions for future research are also discussed.

THEORETICAL BACKGROUND AND HYPOTHESES DEVELOPMENT Social Exclusion
Social exclusion refers to being ostracized by other people, a community, or a brand (Beard et al., 2022;Williams & Nida, 2022). It is a multifaceted variable, reflecting individual differences in their emotions (Poon & Teng, 2017), cognitive responses (Baumeister, Twenge & Nuss, 2002), the necessity for affiliation (Lutz & Schneider, 2021), self-regulation (Baumeister et al., 2005), and social resources (Sacco et al., 2012;Dorsch, Törnblom & Kazemi, 2017). Prior research suggests that the effects of social exclusion are distinct and lead people to different behaviors (Baumeister & Tice, 1990). Thus, the present research focuses on social exclusion made by other users in a brand community and how it affects subsequent purchase behaviors in a related context. Moreover, previous studies reported a broad range of how social exclusion feelings can be triggered. For instance, the person can be excluded by others in a tossing ball game (Walasek, Juanchich & Sirota, 2019) by recalling an exclusionary event (Bernstein et al., 2010) by seeing exclusion scenarios (Sommer & Baumeister, 2002) or by being ignored by others in social interactions (Sinha & Lu, 2019). This research makes people think they were ignored to induce social exclusion feelings. Specifically, we made them think they needed some help in the community, and other users ignored them.
Moreover, recent research suggests that social exclusion is complex and affects people's behavior and relationships (Critcher & Zayas, 2014). Specifically, those researches explore the effect of social exclusion directly made by the brands and do not explore how some elements that connect with the brand also affect the consumer-brand relationship (e.g., brand community made by other users). To address these questions, we adopted an experimental approach to operationalize social exclusion performed by a brand community and test the different possibilities of consumer choice.

Social Exclusion and Brand Relationship
The brand relationship covers many elements that benefit both parties by increasing consumer satisfaction and brand loyalty (Atulkar, 2020;Ebrahim, 2020). In addition, the kinds of relationships are strong or weak depending on consumers' proximity to a brand, making them very similar to relationships between humans (Aggarwal, 2004;Aggarwal & Larrick, 2012).
For instance, Aggarwal (2004) demonstrated two distinct relationships consumers could have with brands. The first one is a more functional relationship that focuses on self-interested benefits and the costs of their relationship with the brand. In the second type, the consumers have a hedonic motivation. In this case, they develop an emotional connection and expect mutual support from the brand and the community.
Note that one kind of relationship does not replace the other. However, it can vary according to the circumstances (Kumar & Kaushik, 2020;Zhang et al., 2020). Moreover, we intend to investigate the effects of social exclusion in a brand community on the consumer-brand relationship. In this case, the relationship we are talking about is communal. Due to the characteristics of brand communities (e.g., giving support to the member), the communal relationship better represents our approach.
Furthermore, the relationship quality is high when people interact deeply with the brand and the community (Wisker, 2020;Ibrahim & Aljarah, 2021). While loyalty programs can directly manage the brand relationship, interactions with communities indirectly impact its quality (Kumar & Kumar, 2020;Rodrigues, Brandão & Rodrigues, 2020). Thus, the communities can be considered a distinct form of creating or destroying consumers' relationships with the brand.
Furthermore, ostracism within brand communities weakens and destroys consumers' relationships. In this case, the relationship quality is not owned just by the brand but also by independent communities that the company does not manage (Rodrigues, Brandão & Rodrigues, 2020). From a practical perspective, an important reason to examine this issue is that organizations often do not control the acts of other users, and it can lead to switching brands rather than increasing the desire for products (Wang & Ding, 2017;Ward & Dahl, 2014).

Social Exclusion and Consumer Choice
A product choice is an act of some deliberative process, and it is triggered by different reasons, such as consumer self-regulation (Coleman, Royne & Pounders, 2020; Vosgerau, Scopelliti & Huh, 2020), reducing the self-discrepancy (Mandel et al., 2017), compensatory consumption (Mourey, Olson & Yoon, 2017), demonstrate social status (Song et al., 2016) and, regulate bad experiences (Schiessl, Korelo & Dias, 2022). In this vein, choices can help consumers cope with ostracism by avoiding the trigger of exclusion and restoring belongingness (Chen et al., 2017;Mead et al., 2011). However, previous research demonstrated that they do it in different forms. For instance, consumers search for affiliation products (Thomas & Saenger, 2020) or prefer anthropomorphized items (Chen et al., 2017).
Socially excluded people are often perceived to have more intentions to purchase products from the same brand due to reaffiliation (Ward & Dahl, 2014). Furthermore, we argue that socially excluded consumers can act in two distinct ways, purchase a product in the same or the rival brand. Previous studies demonstrated that consumers choose the same brand when the brand is vital. It is also contingent when the exclusion is made by communities held by the brand (Liu et al., 2022).
Conversely, social exclusion leads people to avoid those who caused the ostracism by prompting them to switch their choice to a new decision (Hess & Pickett, 2010). Thus, we argue that excluded consumers perceive less possibility of engaging in that community again. At the same time, the community affects their relationship with the brand and weakens their attitudes toward the company (Yaakobi & Williams, 2016;Liu et al., 2022). In this sense, when people are excluded from a brand community, it can affect their choices of products made by the same brand. Formally.

H1:
When consumers face a social exclusion situation (vs. inclusion), they tend to avoid the brand related to the community that made the exclusion and purchase the product from a rival brand.

Serial Mediation of Ideal Relationship Perception and Brand Relationship Quality
Previous studies demonstrated that consumer-brand relations have similarities with humanhuman relationships (Aggarwal, 2004). This paper uses two aspects regarding the relationship between brands and consumers. The first one is the brand relationship quality. We define Brand Relationship Quality as the intensity and depth that people are social, emotional, and behaviorally connected with the brand (Xie, Kwok & Wu, 2019;Zhang et al., 2020). Thus, when the brand's connection is high and more profound, consumers perceive high quality in their relationship.
The second construct we explored in this research is the perception of the Ideal Relationship with the brand. When a person relates with someone, it is natural that he creates ideal expectations. Generally, these expectations are Ideal assumptions people have regarding their relationship. Thus the Perceptions of Ideal relationships are projections made by partners regarding Ideal warmthtrustworthiness, vitality-attractiveness, and status-resources they expect in their relationships (Fletcher et al., 1999;Overall, Fletcher & Simpson, 2006;Charlot, Balzarini & Campbell, 2020).
However, note that there is a crucial difference between both constructs. In the ideal relationship perceptions, people project the ideal warmth-trustworthiness, vitality-attractiveness, and status resources expected by their partners. In this case, they are not evaluating their relationship in the present. Instead, they focus on how the relationship should be. Conversely, in relationship quality, consumers focus on the current relationship with the brand and see how the relationship actually is.
Furthermore, we posit that excluded people tend to perceive less relationship quality and imagine poor ideal relationships after being excluded. The main reason for the existence of a brand community is the integration and the communal relationship between members (Muniz & O'Guinn, 2001;Zhang et al., 2018). Nevertheless, when there is some issue in that community (e.g., social exclusion), natural thinking is that the relationship's quality is not as good as expected.
In a communal relationship, people help each other and interact with the other members by increasing the quality of their relationships (Aggarwal, 2004). However, a communal interaction does not guarantee the relationship's good for the partners. We argue that in this case, when someone is seeking help in a brand community and does not have any answer, it can lead them to perceive a more inferior relationship with others, even in a community where the relationship should be communal (Aggarwal, 2004;Kumar & Kumar, 2020;Zhang et al., 2020).
As in traditional relationships, consumers create ideal standards in brand relationships to maintain this relationship over the years (Kumar & Kaushik, 2020;Zheng, Zhang & Song, 2020). For instance, imagine a romantic relationship where a couple is together for one week. In the beginning, both people think about their relationship's future, having kids, purchasing a home, and how their partner could be. Generally, people create a general standard for their partner and expect him to act according to this expectation (Overall, Fletcher & Simpson, 2006). We argue that this same effect happens in brand-consumers relationships.
When consumers build their relationships with the brand, they also create ideal standards and expect that the brand follows them to produce and maintain good relationships (Kumar & Kaushik, 2020;Wisker, 2020). For example, in the brand community, consumers expect a good relationship with other users, and the brand is responsible for maintaining those social connections with all users. However, in an exclusion situation, the expectations created by consumers are lost, and they perceive a gap between their current and Ideal perceptions of the brand (Overall, Fletcher & Simpson, 2006).
The communal relationship is based on mutual support by creating brand and community expectations regarding this endorsement. Therefore, a brand community exclusion generates a gap between the Ideal relationship and what consumers effectively receive from the brand. Specifically, the difference between ideal expectations and experience leads consumers to perceive poor relationship quality (Song, Tao & Wen, 2021;Chang & Kim, 2022).
The mutual support consumers expect in a communal relationship creates a necessity of being accepted by others. People with good relationships and pleasant interactions in those communities expect it to fill this necessity (Moynihan, Igou & van Tilburg, 2017). In this case, the sense of belongingness induces consumers to idealize inclusion in that community. Conversely, the threat of social exclusion makes the relationship far from the Ideal one and reduces perceived relationship quality.
Based on these theoretical aspects, we argue that the perceived differences between the ideal relationship are an antecedent of relationship quality. Thus, when consumer faces a social exclusion situation, they tend to perceive a lack in their relationship with the brand, which leads them to perceive less quality in their relationship with the brand and explains why people choose a rival brand to purchase a product. After this theoretical background, we present our last research hypothesis.
H2: Ideal perception of relationship and brand relationship quality sequentially mediate the effect of social exclusion on people's intentions to buy a rival brand.
In figure 1, we demonstrate our complete research model. Source: Prepared by the author (2022).

OVERVIEW OF STUDIES
In study 1, we performed a single-factor manipulation with two conditions (social exclusion vs. social inclusion) and tested the effect of social exclusion on brand relationship quality. In study 2, we replicated the direct effect of social exclusion on brand relationship quality and demonstrated the mediation effect of consumers' perception of difference in ideal vs. current relationships. Finally, in study 3, we tested the direct effect of social exclusion on consumer choice and the sequential mediation of ideal perception and brand relationship quality (H2).

Study 1: Measured the Effect of Social Exclusion on Brand Relationship Quality
Study 1 provides initial evidence for the effect of social exclusion on brand relationship quality. Thus, we expect social exclusion in a brand community (vs. inclusion) lead consumers to perceive less (vs. more) quality in their relationship.

Method
Participants and Design. Eighty-four Brazilian Facebook users participated in our study (78.6% female, Mage = 28.23; SD = 10.40). All participants competed for a U$ 10.00 voucher after finishing the research. In addition, we advertise the research in the entire Brazilian territory for three days to reach more participants. People were randomly assigned to two conditions (Social exclusion vs. Social Inclusion).
Procedure. We informed participants that this study was to investigate some consumption habits. We asked them to imagine they had a notebook made by a fictitious brand with which they had a long history and a good relationship. Further, we instructed them to imagine they needed help with the image editing software, and they searched for a brand community to solve this issue. However, we told participants that nobody in the community answered the question, and they excluded him. It is a similar procedure used by Sinha and Lu (2019) to induce feelings of social exclusion. Nevertheless, many people answered the question and made him feel included as included consumer. Afterward, we requested them to write a phrase describing how they felt excluded or included in that community. After writing the phrase, they answer the brand relationship quality scale adapted from (Algesheimer, Dholakia & Herrmann, 2005). with five items. This brand is part of my life, without it I feel that something is missing; I really like this brand; I would recommend this brand strongly to my friends; I trust in this brand; This brand makes me feel comfortable. Responses were rated along scales ranging from 1 (strongly disagree) to 7 (strongly agree). Further, the items were averaged to create a brand relationship quality index (α= 0.89). At last, they answered the manipulation checks, control, and demographic questions.
Manipulation check. Based on a 7-point scale anchored at "1 strongly disagree to 7 strongly agree" participants rated their feelings of social inclusion in a single question, "I felt included" and social exclusion, "I felt excluded". As expected, in the inclusion situation people rated more feelings os social inclusion (Msocial_inclusion = 5.73 vs Msocial_exclusion = 3.66 p = 0.000). Conversely, in the exclusion condition, people rate more feelings of social exclusion (Msocial_exclusion = 4.48 vs Msocial_inclusion = 1.53 p = 0.000).
Brand relationship quality. As proposed, people in the excluded situation tend to perceive their brand relationship quality as poorer than included people due to their feelings against the brand community. There is a significant difference between socially excluded (vs included) condition on brand relationship quality (Msocial_exclusion 5.60, SD = 1.60; Msocial_inclusion 4.84, SD = 1.28; p = 0.019).
Discussion. Study 1 revealed that social exclusion in a brand community conducts consumers to perceive less quality in their relationship with the brand. However, we have not explored the psychological mechanism that underlies this perception. To address this gap, we argued that the disparity between current and ideal relationships could explain why social exclusion diminishes relationship quality. Study 2 solves this issue by measuring this variable and testing the mediation effect of ideal relationship perception.

Study 2: Mediation of Ideal Relationship Perception
In study 2, we explored the direct effect of social exclusion on brand relationship quality. We aim to replicate the direct effect and test the difference between the current and ideal perception of the brand relationship as a mechanism that mediates this effect.

Method
Participants and Design. we recruited one hundred thirty-six living in Brazil (89% female, Mage = 53.05; SD = 14.22) from Facebook. They participated in the study for a chance to win a U$ 10.00 voucher after concluding the research. We advertise the research in the entire Brazilian territory for one week to reach more participants. They were randomly assigned to two conditions (social inclusion vs. social exclusion).
Procedure. After manipulating exclusion (similar to study 1), participants answered the brand relationship quality scale (α= 0.916) and the ideal relationship scale. To measure Ideal perceptions of their relationship, we first asked participants to imagine an Ideal relationship they would like to have with any brand. Next, we instructed them to rate the differences they perceived between the ideal and the current relationship with the brand community after the exclusion of the following items, "This brand supports me in difficult times"; "This brand is sensitive to understanding my problems"; "This brand likes its consumers"; "This brand is outgoing"; "This brand is very good with its customers"; "This brand does a good job" (view Overall, Fletcher & Simpson, 2006 to more details). Finally, the items were averaged to create an Index of Ideal perceptions of relationships (α= 0.963). After this, they answer manipulation checks, control, and demographic questions.

Results
Manipulation check. Based on a 7-point scale anchored at "1 strongly disagree to 7 strongly agree" participants rated their feelings of social inclusion in a single question "I felt included" and social exclusion, "I felt excluded". As expected, in the inclusion situation people rated more feelings os social inclusion (Msocial_inclusion = 5.74 vs Msocial_exclusion = 4.11 p = 0.000). Conversely, in the exclusion condition, people rate more feelings of social exclusion (Msocial_exclusion = 4.80 vs Msocial_inclusion = 2.09 p = 0.000).
Brand relationship quality. As expected, there is a significant difference between social excluded (vs included) people in brand relationship quality (Msocial_exclusion = 4.28; SD = 1.78 and Msocial_inclusion = 5.26; SD = 1.64; p = 0.001). This result replicates the effect shown in study 1.
Ideal relationship perception. We employed model 4 from (Hayes, 2018)to test the mediation with 10.000 bootstrapped samples. We performed an analysis with social exclusion as the independent variable, ideal relationship perception as a mediator, and brand relationship quality as the dependent variable. The total effect of this model was significant (Effect = -0.972; se = 0.294; t = -3.297; df = 136; p = 0.001). The direct effect of social exclusion on brand relationship quality was no significant (Effect = -0.136 se = 0.227; t = -0.599; df = 134; p = 0.549). The Indirect effect of social exclusion on brand relationship quality via ideal relationship perception was significant (Effect = -0.836 CI 95% [-1.267, -0.2442]).
Discussion. Study 2 provided additional evidence for our prediction that social exclusion leads consumers to perceive poor relationship quality. Furthermore, this study also tested a psychological mechanism underlying the reduction in this effect. In this way, when the gap between the current and ideal brand relationship increases, less relationship quality consumers perceive. These results complement previous findings in this research and bring more pieces of evidence to develop study 3.

Study 3: Serial Mediation
In Study 3, we aim to replicate the findings of studies 1 and 2. Further, we intend to demonstrate the sequential mediation in our model pictured in figure 1. At last, we will measure a consumption variable to understand the choice after being excluded. After the exclusion situation, we will analyze if consumers choose a product made by the same or a rival brand.

Method
Participants and Design. Similar to study 1 and 2, we recruited ninety Facebook users in Brazil (88.9% females, Mage = 39.15; SD = 7.97). Participants competed for a U$ 10.00 voucher after finishing the research. In addition, we advertise the research in the entire Brazilian territory for one week to reach more participants. Participants were randomly assigned to one of both states (social exclusion vs Social Inclusion).
Procedure. The manipulation in this study was similar to the study 1 and 2. After people described their feeling about social exclusion, we asked them to imagine they needed a new laptop, and we showed two products with very similar characteristics and asked them to choose just one option. The choices were coded as 1 choice of the same brand and the rival brand as 0. After that, they answered the brand relationship quality scale (α = 0.964), perception of an ideal relationship with the brand (Similar to study 2) (α = 0.960), manipulation check, control, and demographic questions.

Results
Manipulation check. Based on a 7-point scale anchored at "1 strongly disagree to 7 strongly agree" participants rated their feelings of social inclusion in a single question "I felt included" and social exclusion, "I felt excluded". As expected, in the inclusion situation people rated more feelings os social inclusion (Msocial_inclusion = 5.75 vs Msocial_exclusion = 2.86 p = 0.000). Conversely, in the exclusion condition, people rate more feelings of social exclusion (Msocial_exclusion = 5.42 vs Msocial_inclusion = 3.83 p = 0.000).
Choice of the same or rival brand. To analyze the direct effect of social exclusion on consumer purchase intention in the same or the rival brand, we used logistic regression and put the independent variable social exclusion coded as 1 and social inclusion coded as 0. Moreover, we inserted the binary choice option in the dependent variable: 1 person who chose the same brand and 0 for those who chose the rival brand. The logistic regression results revealed a significant effect of social exclusion on consumers intentions to choose the rival brand (x²(1) = -1.21; p = 0.014, R² Nagelkerke 0.099) with a significant odds ratio (OR = 0.296; CI = 0.113; 0.778). Thus, participants seemed more inclined to purchase a rival brand after being excluded.
Serial mediation. Before running the regression modeling, we performed an exploratory factor analysis with the Items of brand relationship quality and perceptions of ideal relationship to verify if the scales' items represent the same construct. As expected, the factor analysis loaded the items in two dimensions by indicating that they measure different aspects of the relationship with the brand.
Furthermore, we used model 6 from (Hayes, 2018) with 10000 bootstrap samples to test the entire model. In the independent variable, we insert social exclusion coded as 1 and social inclusion coded as 0. In the dependent variable, we put the choice in the same brand coded as 1 and the choice in the rival brand coded as 0. Furthermore, we inserted the ideal relationship perception and brand relationship quality for serial mediation. After running the process the direct effect of social exclusion on brand choice was not significant (Effect = -0.755 se = 0.542; z = -1.392; df = 87; p = 0.163). The Indirect Effect of serial mediation was significant (Effect = -0.362; CI 95% [-1.057, -0.006]).
Discussion. Study 3 provided evidence to predict that social exclusion leads people to increase their preference to buy a product from a rival brand. We also demonstrated a serial mediation of ideal perception of brand relationship and brand relationship quality. After being excluded, people tend to buy more the rival brands because the ideal perception of brand relationships decreases, reducing brand relationship quality. The reduction in these two variables leads people to prefer the rival brand and avoid purchasing the same brand.

GENERAL DISCUSSION
Prior research suggests that when consumers are excluded, they prefer products from the same brand because exclusion increases their desire (Wang & Ding, 2017;Ward & Dahl, 2014). Building on social exclusion theory (Baumeister & Tice, 1990;DeWall et al., 2008), the present research complemented those findings and systematically investigated how social exclusion influences consumer relationships with brands and their choices.
Across three studies, we show that when people are excluded from a brand community, they negatively perceive their relationship with the brand, leading them to buy a rival product in a subsequent choice. The findings also revealed that the difference in the ideal relationship is an antecedent of brand relationship quality in the mediation process. These findings contribute to the literature on brand relationship quality and social exclusion, which also have important practical implications.

Theoretical Contributions
The present research contributes novel insights into social exclusion and consumer behavior literature. First, we demonstrated that social exclusion in a brand community leads consumers to choose a product from a rival brand. These findings complemented previous research by showing that social exclusion leads people to switch their choice after this threat (Su et al., 2017). However, previous literature explored exclusion performed by brands. We contribute to this literature stream by demonstrating that the community created by other users also leads the consumer to avoid the brand and purchase a rival one.
In this case, the community is not owned by the brand but affects how consumers perceive their relationship with it. Moreover, after being excluded by other users in a brand community, consumers have not seen an opportunity to restore their previous relationship and choose a rival brand. These findings contribute to brand relationship literature and demonstrate that the third party affects the relationship quality between the brand and consumers. Furthermore, this evidence demonstrated that social support is essential to maintain consumers' use of the brand.
Furthermore, previous literature demonstrated that people engage in a brand community because they need to feel included by other people. Moreover, after entering the community, they expect to have a communal relationship with the other members (Aggarwal, 2004;Muniz & O'Guinn, 2001). In the communal relationship, people expect others' support and help with issues regarding the brand. However, we demonstrated in this research that sometimes the community can threaten the communal relationship among members when they exclude or ignore someone in the community. Thus, social exclusion threatens people's sense of belonging, affecting brand relationships.
Moreover, we showed that this interruption in communal relationships within a brand community is severe and directly affects people's relationship with the brand by choosing the rival. We argue that this effect happens because the social exclusion threat affects their sense of belongingness, reducing their relationship quality and their perceptions of the ideal relationship with the brand. These findings complement previous literature by demonstrating other aspects of the brand that affect consumers' choices and relationships. Specifically, the brand community also has a crucial role in guiding people's relationship with the brand and their decisions under social exclusion.
We complement previous literature demonstrating the serial mediation effect of ideal brand relationship and brand relationship quality. For example, previous studies demonstrated that being rejected in a community could increase people's desire for the brand (Wang & Ding, 2017). However, we demonstrated that rejection could negatively affect people's relationships with the company. Furthermore, we show that it happens because people perceive a lack of an ideal and current relationship with the brand. A brand that does not manage its relationship with the consumer through the brand community affects consumers' standards defined earlier (Overall, Fletcher & Simpson, 2006). Notably, the findings demonstrated that a relationship with the brand is very similar to relationships between humans, and the acts of other users can influence people's behaviors toward the brands.
This research is the first to demonstrate this serial mediation in the social exclusion literature. We demonstrated new mechanisms to explain social exclusion's effect on consumer behaviors regarding the brand. Furthermore, we complement previous literature by demonstrating that brand relationship quality has one antecedent. Specifically, consumers' relationship quality depends on how the relationship is close to their ideal notion of a good relationship with the brand. This effect happens in social exclusion because entering a brand community creates ideal expectations regarding their relationship with other members. Thus, the relationship is far from ideal when excluded and precedes poor relationship quality.
In adopting a social exclusion from the community perspective, the present research differs from past research examining the effects of other users on consumer brand choice. For example, drawing the same theory, Wang and Ding (2017) demonstrated that when strong brands exclude consumers, they desire to buy products from the same brand. Similarly, exclusion by luxury brands increases consumers' desire to acquire products that make the exclusion (Ward & Dahl, 2014). However, we look through other lenses, showing how consumers behave after being excluded by other users in a brand community. This different perspective complements previous research and shows that consumers avoid behavior when the exclusion is represented in other users' figures. These findings also have implications for marketers.
At last, the results found in this paper support the idea that consumers will have the same outcome even though they know they can search for the same information on other sites or communities. As we argued earlier, the consumer relationship is essential to maintaining well-being and happiness (Kumar & Kumar, 2020;Song, Tao & Wen, 2021). In this case, when the community excludes any individual, they lose their social support, leading them to change the brand.

Practical Implications
Previous studies demonstrated that inducing consumers to feel excluded can increase their intentions toward the brand. It happens because brands create a sensation of scarcity and uniqueness, which leads people to desire solid and luxury brands (Wang & Ding, 2017;Ward & Dahl, 2014). However, our core finding on socially excluded consumers having a greater propensity for choosing a rival brand complement those findings and has important practical implications for marketing strategies.
In many cases, consumers can be ostracized in communities created by brands or other users. In this case, the exclusion is not made directly by the brand. However, they need to monitor the interaction between users in the communities to integrate the members better and avoid episodes of exclusion. Furthermore, an exclusion made by other users can negatively affect people's intentions to buy the same brand. Therefore, if the brand intends to sell to all consumers, it must manage its users' social interaction. For instance, the brand can designate some people to moderate online brand communities. As a result, those people can maintain the interactions in the group working well (Frith, 2014).
At last, we demonstrated that social exclusion in brand communities decreases consumer-brand relationship quality. In particular, we revealed in a serial mediation that social exclusion increases the gap between current and ideal relationship expectations with the brand, reducing relationship quality. With this in mind, we argue that brands can use common strategies to improve relationship quality to reestablish consumer relationships and avoid buying from a rival brand. For instance, the brand can provide hedonic experiences to restore the consumer's sense of ideal relationships, increasing the perceived relationship quality (Huber et al., 2010;Kuehnl, Jozic & Homburg, 2019).

Future Research
Notwithstanding our findings on the social exclusion effect on consumers' choices, several areas merit further research. First, our exclusion stimuli consisted of utilitarian goods (e.g., laptops). People generally do not have a solid emotional connection with this kind of goods (Krishen, Berezan & Raab, 2019). Thus, future research could test our framework using a different range of products. For example, hedonic products (e.g., a trip) could produce different results after social exclusion. Which, in turn, could have distinct underlying mechanisms.
Second, we tested consumers' binary conditions after excluding our dependent variable. In this case, we aim to understand their intention to switch to another brand after the ostracism episode. Maybe, due to the consumer's emotional aspects triggered by the exclusion situation, some consumers could adopt different choices. For instance, some consumers could defer buying the product because of emotional impairments. This possibility merits further investigation, especially for consumers with more robust emotional responses toward the brand and the community (Kumar & Nayak, 2018;Martins & Patrício, 2018).
Finally, focusing on social exclusion and brand relationships, we do not discount the possibility of brand personality and the different effects that strong brands could have on the model. For instance, different brand personalities have a different impact on consumer evaluations of the brand. In particular, exciting personalities make consumers more tolerant of some issues. Conversely, sincere personalities lead a consumer to tolerate this misconduct (Fournier & Alvarez, 2012;MacInnis & Folkes, 2017). Therefore, future research could explore the effect of social exclusion in different brand communities, for instance, communities of an exciting brand personality. Maybe, people in this community feel less of the effect of ostracism.
In the same way, exclusion performed in a weak brand community can affect more consumers who intend to buy the same brand. In contrast, strong brands can influence consumers' decisions due to rival weaker brands (Wang & Ding, 2017). These two possibilities also merit investigation. At last, all studies were performed on Facebook. This social media is a common place to have brand communities (Gummerus et al., 2012;Munnukka, Karjaluoto & Tikkanen, 2015), which can be a limitation of our research because people see many brand communities in this social media. Future studies can investigate our model in new environments and explore other forms of brand communities.

Research ethic statement
This paper has not been published elsewhere in any form. It has not been submitted simultaneously to any other publication outlet.

Author contribution statement
I confirm sole responsibility for the following: study conception and design, data collection, analysis and interpretation of results, and manuscript preparation/revision.

Declaration of conflicting interests
I declare no potential conflicts of interest regarding this paper.

Funding
This work was supported by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior -CAPES.