JAPR_1-1_183.pdf377.9KB

INTRODUCTION

Today's marketers are eager to utilize social media as a viable marketing communication channel. Social media is defined as “a group of Internet-based applications that build on the ideological and technological foundations of Web 2.0, and that allow the creation and exchange of User Generated Content” (Kaplan & Haenlein, 2010, p. 61). Social media platforms are changing the way brands connect to their consumers, providing a shift from the traditional one-way communication to an expanded dialogue between a company and its consumers (Qualman, 2009). 

In 2011, advertisers spent more than $2 billion on social media advertising, and predictions indicate that they may be spending up to $5.59 billion by 2014 (eMarketer, 2012). More importantly, in addition to advertising spending, 84% of Fortune Global 100 companies are using at least one social media platform – namely Twitter, Facebook, YouTube, or a corporate blog – to interact with their consumers (Burson-Marsteller, 2011). Despite the increase and popularity of social media, marketers are still unclear about how to effectively integrate social media into their advertising strategy (Kuhn & Burns, 2008; Verna, 2007). Existing studies in this area typically investigate from the viewpoint of consumers, focusing on user-generated contents (e.g., Cheong & Morrison, 2008; Daugherty, Eastin, & Bright 2008) and electronic word-of mouth (e.g., Lee & Youn, 2009; Sohn, 2009). At present, only a few studies have looked at the utilization of social media from a brand’s perspective (e.g., Ahuja & Medury, 2010; Cho & Huh, 2010) 

 These days, consumers are able to watch what a brand does, listen to what a brand says, and expect to make a meaningful connection with a brand via social media. Before the emergence of social media, the disclosure of personal and intimate information, such as behind-the- scenes content, “making of” videos, employee stories, and inside stories that were not directly related to a firm's commercial activities, was difficult to distribute via traditional media channels. With the availability of various social media platforms, marketers can now easily produce and deliver personal and intimate stories about the brands. Thus, creating effective and persuasive content on behalf of a brand to attract consumers via social media becomes an important task for today's marketers.

 The purpose of this study is to investigate the influence of a brand's disclosure of perceived backstage information through social media. More specifically, this research is aimed at answering the following three research questions:

 RQ1 : How does the degree of brand information disclosure (i.e., perceived backstage visibility) via social media influence consumer responses toward a brand?

 RQ2 : How does the scarcity of brand information disclosure via social media influence consumer responses toward a brand?

 RQ3 : Would the consumer's general advertising skepticism moderate the impact of disclosure of brand’s information disclosure?

 Motivated by the integral role of self-disclosure in interpersonal relationships, the present study is based on four key premises. First, a consumer and a brand can form a relationship, which resembles interpersonal relationships. Second, given the first premise, disclosing information about a brand can be understood as a type of brand information disclosure. Third, literature on self-disclosure in a human relationship context can help us understand consumer-brand relationships. Fourth, when a brand discloses itself to consumers in terms of depth (i.e., front- and perceived back-stage) in an exclusive manner (i.e., scarcity of information), consumers are likely to use them accordingly in evaluating the brand.

 This research hopes to contribute to the theoretical understanding of self-disclosure in the consumer-brand context. More specifically, the present research is an exploratory attempt at investigating the impact of disclosing the perceived backstage of a brand as a marketing communication strategy. The results of the current study advance our understanding of how consumers interpret different types of messages provided by marketers via social media and use them in evaluating brands. In addition, the findings of this research contribute to our understanding of a brand’s perceived backstage information and suggest practical guidelines for its strategic use in social media

THEORETICAL BACKGROUND

Consumer-Brand Relationship

The consumer-brand relationship research suggests that the relationship between consumers and brands is similar to that of a human relationship (Fournier, 1998). In her inspiring exploratory study, Fournier (1998) argued that brands have an active relationship with consumers, providing a framework for characterizing and understanding consumer-brand relationships. According to this consumer-brand relationship paradigm, consumers often form relationships with brands in much the same way as they do with other human beings in a social context (Aggarwal, 2004). Based on this approach, scholars in marketing and consumer behavior have suggested that consumer-brand relationships share qualities similar to those of human relationships, as consumers consider brands viable relationship partners. 

Self-Disclosure

 Given that consumer-brand relationship shares similar characteristics with interpersonal relationships, previous findings on interpersonal relationships can provide further insights into investigating the relationship between a consumer and a brand. In the interpersonal relationship context, self-disclosure is a key construct. It has been defined as any information about oneself that a person verbally communicates to another person (Cozby, 1973; Wheeless, 1976). Derelega and Grzelak (1979) define self-disclosure as “any information exchange that refers to the self, including personal states, dispositions, events in the past, and plans for the future” (p. 152). In every sort of interpersonal relationship, the exchange of self-disclosure plays an important role (Rubin, 1973). Given that brands and consumers form a relationship that is similar to an interpersonal relationship, a brand’s information disclosure can be understood from the perspective of self-disclosure in interpersonal relationships.

 In addition, previous scholars suggest that self-disclosure is a multi-dimensional construct since not every piece of self-disclosing information is identical in terms of its intimacy (Altman & Taylor, 1973; Journald, 1971; Pearce & Sharp, 1973; Wheeless, 1976). The depth of information, which refers to the level of topic intimacy in self-disclosing information, especially has been suggested as a critical parameter of intimate self-disclosure (Altman & Taylor, 1973). For example, information and facts about oneself might be more or less personal, such as “my hometown” compared to “details about my sex life.”

 More importantly, relational outcomes of self-disclosure vary depending upon the degree of self-disclosure. Previous studies have reported that information receivers would show greater intimacy (e.g., Altman & Taylor, 1973; Reis & Shaver, 1988), liking (Chaikin & Delega, 1974; Chelune, 1977), and trust (e.g., Pederson & Higbee, 1969; Wheeless, 1976) toward the discloser who presents more personal and intimate information.

 Thus, in an interpersonal context, it is suggested that the degree of self-disclosing communication and its relational consequences vary depending upon the degree of intimacy. Likewise, when a brand discloses information about itself to consumers, the degree of self-revealing information can vary in terms of the depth (i.e., intimacy). Consequently, it is highly conceivable that the relational outcomes of brand information disclosure can be different depending upon the degree of self-disclosing information about a brand.

Dramaturgical Theory

 Then, what qualifies as more or less personal and intimate information of a brand? This study proposes the “perceived backstage” of a brand as the more personal and intimate aspects of the brand. Based on Goffman's (1959) Dramaturgical Theory, marketing scholars (Grayson, 1998; Gummeson, 1990; Lovelock, 1991) have identified frontstage, backstage, and perceived backstage. Frontstage is defined in terms of “performances and regions perceived by the customer that are believed to have been created or altered for a customer audience,” whereas backstage refers to the firm's “performances and regions not perceived by the consumer” (Grayson, 1998, p.130). In addition to this conceptualization, Grayson (1998) proposed a third region called perceived backstage. Perceived backstage characterizes the firm’s “performances and region perceived by the customer and believed not to have been created or altered for customer audiences” (Grayson, 1998, p. 130). Consumers are exposed to this perceived backstage when marketers attempt to give the impression of revealing the backstage of the firm to consumers. For example, a clean and decorated reception area of a local car repair shop is considered a frontstage, while the repair area is considered a backstage of the shop. When this car repair shop has its ad agency create a “behind-the-scenes” promotional video presenting the image of a clean and organized repair area, which is shown on a display in the reception space, customers can access the perceived backstage by seeing this “behind-the-scenes” video (Grayson, 1998).

 This idea of a perceived backstage explains the brand’s usage of social media, where marketers easily upload and distribute their backstage information, such as behind-the-scenes content, employee stories, or inside stories, about a brand to their consumers. In other words, social media becomes a viable venue where consumers can access brands’ perceived backstage. For example, when Disney fans visit its corporate blog they can find behind-the-scenes information about the Disneyland themepark, such as video clips revealing how the Mickey Mouse-looking giant pumpkins were carved, sanded, finished, painted, and installed with cranes for the Halloween celebration. Another example of disclosure of a brand's backstage activities can be easily found on YouTube, where consumers can watch numerous “makings of” commercials. For instance, T-mobile successfully generates thousands of consumer responses on the popular video sharing site by disclosing the making of its “T-Mobile Dance” television commercial. Thus, social media provides backstage passes to those consumers who want to become part of “the scene” of their favorite brands (Lutters & Ackerman, 2003).

 The privilege of having these backstage passes to a brand is perceived to be even more valuable when only a small number of people are allowed the same access. According to Brock’s commodity theory (1968), resources acquire value to the extent that they are perceived to be scarce or unavailable to others. Based on this assertion, previous self-disclosure studies have reported the scarcity effect of self-disclosure (e.g., Derelega & Grzelak, 1979; Petty & Mirels, 1981). That is, being selected as the recipient of a scarce message leads recipients to have a more positive evaluation of the discloser (e.g., Archer & Cook, 1986). In general, people evaluate the discloser more favorably if they appear to be selective (Kleinke, 1979). Thus, in the consumer-brand relationship context, it is highly conceivable that the relational outcomes of brand information disclosure will be even greater when the disclosing information is only available to selected consumers, such as members of an online brand community, rather than when the disclosing information is available to a public audience.

 Based on aforementioned discussion, the following hypotheses are proposed:

 H1 : The greater the degree of brand information disclosure, the stronger consumers feel (a) intimacy, (b) liking, and (c) trust toward the brand.

 H2 : The higher the scarcity of brand information disclosure, the stronger consumers feel (a) intimacy, (b) liking, and (c) trust toward the brand.

Skepticism toward the Brand Information Disclosure

People try to interpret a speaker's various reasons and goals for sharing intimate information (Derlega & Berg, 1987). Judgments and evaluations of self-disclosers vary according to individual characteristics of persons who make the evaluations (Kleinke, 1979). As a moderating factor between brand information disclosure and outcomes, this study proposes the individual’s ad skepticism. 

 Obermiller and Spangenberg (1998) define skepticism toward advertising as an individual’s tendency toward disbelief of advertised claims. This disbelief is basic to the marketplace and varies within consumers, based on a variety of social and economic factors. Advertising skepticism is suggested to be a key attribute in most advertising as consumers are not likely to respond to advertising in the desired manner if they do not believe what it says (Beltramini, 1994). Moreover one’s acceptance of an advertised claim is related to the amount of persuasive influence the advertisement has on the viewer (Obermiller & Spangenberg, 1998). Similarly, the expected relational outcomes of brand information disclosure may not be observed if consumers do not believe the self-disclosing information.

 Thus, it is highly possible that consumers with high skepticism toward the brand information disclosure are less likely to have relational outcomes toward the brand. Therefore the following research hypotheses are proposed:

 H3: The effect of the degree and the scarcity of brand information disclosure on (a) intimacy, (b) liking, and (c) trust toward the brand will vary according to the consumers’ advertising skepticism level.

METHODOLOGY

Given the objective of this study, a between-subjects experiment is conducted to examine the above-presented hypotheses. An experimental design is employed to test hypotheses, because it affords the possibility of manipulating and controlling the most important factors for the researcher (Kirk, 1968). 

Development of Stimulus Material

 Based on previous academic work (e.g., Chang & Chieng, 2006; Fournier & Yao, 1997) and the results of the pretest, coffee products were chosen for the main experiment. To eliminate any confounds based on prior brand familiarity and brand preference, a fictitious brand name, “Mike's Coffee,” was created. The behind-the-scenes information on coffee products from Mike’s Coffee was presented in a single blog posting with brief headline, a relevant picture, and body text. A blog was selected for this study out of various social media platforms because (1) the corporate blog is the optimized vehicle for providing a large amount of information, compared to Twitte's 140-character or Facebook's 1000-character limit per post; and (2) among various available social media platforms, the corporate blog has received the most attention as a tool for providing various brand-related content to consumers in a personal diary format (Ahuja & Medury, 2010; Cho & Huh, 2010).

Independent Variables and Manipulations

▪Degree of information disclosure

 Degree of information disclosure was conceptualized as the extent to which social media content provides perceived backstage information about a brand. Degree of brand information disclosure was manipulated at two levels, with high degree and low degree conditions. The low degree condition was operationalized as a blog posting containing frontstage information that provided the very basic information a brand usually makes available to consumers. The high degree condition was operationalized as a blog posting containing frontstage information as well as behind-the-scenes information about the frontstage information, which thus provided more than the basic information that a brand usually makes available to consumers. Thus, the blog postings in the high degree condition included deeper and broader information than the blog postings in the low degree condition.

▪Scarcity of information disclosure

 Scarcity of brand information disclosure in the current study was conceptualized as the extent to which social media content was provided exclusively. For the purpose of this study, scarcity of brand information disclosure was manipulated at two levels, high and low. The high scarcity condition was operationalized as a blog posting indicating that the information is exclusively available for consumers who are invited to and join the membership-based blog. The low scarcity condition was operationalized as a blog posting indicating that the information is available elsewhere as well. Scarcity was thus manipulated by indicating the private or public nature of the information in the blog posting.

▪Consumer skepticism

Consumer skepticism in the current study was conceptualized and measured as the individual's general tendency toward disbelief regarding advertised claims. Subject skepticism as a consumer was assessed as the sum of the nine items that comprise Obermiller and Spangenberg's (1998) advertising skepticism scale, using a 5-point scale ranging from strongly agree to strongly disagree, where the higher the score, the higher the skepticism. Then, consumer skepticism scores were divided into two groups (i.e., low vs. high skepticism) on the basis of a median split. 

Study Design

▪Experimental Design

 A between-subjects 2×2×2 experimental design was employed to test the hypotheses. The three factors were the degree of information disclosure (high vs. low), the scarcity of information disclosure (high vs. low), and the consumers’ general advertising skepticism (high vs. low). The degree and scarcity of brand information disclosure were manipulated variables, whereas skepticism was a measured variable that yielded low versus high advertising skepticism groups on the basis of a median split.

▪Participants

 A total of 256 participants were recruited from an online consumer panel consisting of Web users. As panel members, subjects agreed to participate in this online experiment for various rewards. Among 400 consumer panel members who submitted their data, 256 were usable and included in the final sample after eliminating 56 respondents who submitted incomplete surveys and 88 respondents who completed the survey faster than the researcher's estimated minimum response time (i.e., less than 10 minutes). Among the participants, 35.2% (n=90) were male and 64.8% (n=166) were female. The subjects came from the general population, which contributed to the external validity of the study results, although they were a bit younger than the general population. The participants' ages ranged from 19 to 36 with an average of 27. The majority of the participants were Caucasian (73.4%), followed by African-American (13.7%), Hispanic (5.1%), and Asian (5.1%). More than half of the participants had at least some post-secondary education. In addition, the majority of the subjects reported that they spend more than one hour using social media on the average weekday.

▪Procedure

 Participants were informed that the purpose of the study was to understand general consumer-related issues in the marketplace. The whole experiment was carried out online. Since the experiment was conducted online, signed informed consent was not obtained. Instead, participants’ voluntary act of clicking on the ‘next’ button and filling out the questionnaire was considered to constitute informed consent.

 Prior to being assigned to one of four experimental conditions, subjects were exposed to the consumer-brand relationship scenario that described their relationship with a fictitious coffee brand (Mike’s Coffee). Since relationships are formed over long periods of time, it might seem difficult to manipulate the consumer-brand relationship over the course of an experiment. However, previous studies indicate that even without an actual long-term relationship having been formed by participants, the effects of consumer-brand relationships nevertheless can be observed, which suggests that a relationship can in fact be primed in experimental studies (Aggarwal, 2004). Thus, participants were presented with descriptions to prime a consumer-brand relationship and were asked to imagine themselves in the situation described. After reading the description, the participants were asked several questions designed to measure the effectiveness of the relationship priming.

 Participants were then randomly assigned to one of four experimental conditions (i.e., Degree x Scarcity) and then exposed to a blog posting that contained information about iced coffee from Mike's Coffee's. After exposure to the stimuli, subjects were asked to answer questionnaires including the consumer skepticism measure, the dependent measures, and demographic questions.

Dependent Variables

▪Liking toward a brand

 The likeability scale was employed to measure consumers' liking toward the brand. The scale is composed of four single-word descriptors with a 7-point response format. The scale was originally developed to assess a person's opinion of an object with an emphasis on attraction. Moon (2000) used them in order to measure intimate information exchange between consumers and websites.

▪Intimacy toward a brand

 Price and Arnould's (1999) perceived friendliness scale was employed to measure the extent to which a person considers another person to be likeable and pleasant to be around. After careful review of the existing instruments, this friendliness scale was adopted instead of a scale of intimacy in interpersonal relationship, because the focus of the original scale was on the perceived friendliness of a service provider by a client. The six-item, five-point scale measures the extent to which a consumer perceives the brand as friendly. The items were modified in order to fit the consumer-brand relationship context.

▪Trust toward a brand

 A scale of trustworthiness toward the retailer (DeWulf, Odekerken- Schroder, & Iacobucci, 2001) was adopted to measure the level of trust toward a brand. Three Likert-type statements with a five-point response format were used to assess consumers' belief in a brand’s reliability and integrity.

DATA ANALYSIS AND RESULTS

Reliability Analysis

 Internal consistency of major constructs used in the study was examined, and the results are shown in Table 1. As an independent variable, advertising skepticism was measured. As dependent variables, intimacy toward a brand, liking toward a brand, and trust toward a brand were measured. As shown in Table 1, advertising skepticism and all dependent measures were reliable across conditions.

Table 1: Reliability Analysis of Measurement Items

Manipulation Check

 To ensure that participants appreciated the degree manipulation, subjects were asked to indicate (1) degree of disclosure and (2) scarcity of disclosure by answering a series of manipulation check questions. Subjects who incorrectly identified their condition were eliminated and excluded from analysis.

 An additional manipulation check was conducted to ascertain whether degree of disclosure had been successfully manipulated in terms of perceived amount of behind-the-scenes information. In order to check subjects' perceived amount of behind-the-scenes information, they were asked to indicate their level of agreement with the following statements: (1) “The blog posting gives out behind-the-scenes information about Mike's Coffee.” (2) “The blog posting provides consumers with the inside scoop of Mike's Coffee.” These instruments were developed to explore subjects' perceived amount of behind-the-scenes information. Ratings were anchored from 1 (strongly disagree) to 5 (strongly agree). A one-way ANOVA result showed that the degree manipulation was effective in terms of perceived amount of behind-the-scenes information. Participants in the high degree condition reported a greater perceived amount of behind-the-scenes information than did participants in the low degree condition [F (1, 255) = 81.859, p<.001, M_{high_degree} = 4.51, M_{low_degree} = 3.48]. Similarly, participants in the high degree condition indicated a greater amount of inside scoop on Mike's Coffee than did participants in the low degree condition [F (1, 255) = 30.509, p<.001,  M_{high_degree 4.36,} M_{low_degree} = 3.73].

 For the consumer skepticism variable, high and low groups were created by splitting the sample into two groups (high and low) based on a median split. The mean score of the low skepticism group (M = 4.05) was significantly higher than that of the high skepticism group (M = 2.46), F (1, 255) = 518.72, p<.001. The lower the score, the higher the degree of skepticism. In addition, an ANOVA revealed no differences in skepticism across the four degree-scarcity conditions.

Hypothesis Testing

 First, the zero-order correlations among the dependent variables were examined. All dependent variables were significantly positively correlated with one another at the .01 level.

A 2×2×2 multivariate Analysis of Variance (MANOVA) was performed with the manipulated independent variables Degree (Low, High) and Scarcity (Low, High), the measured independent variable Skepticism (Low, High), and the seven dependent variables. As shown in Table 2, a significant main effect of Degree (Wilks' Λ = .93, p <.05) and Skepticism (Wilks' Λ = .84, p < .001) was obtained, as were trends for two-way interactions for Degree x Scarcity (Wilks' Λ = .95, p = .094) and Scarcity × Skepticism (Wilks' Λ = .95, p = .082). 

Table 2: Multivariate Analysis of Variance Results

 Hypotheses 1a-3a predicted that the degree and the scarcity of brand information disclosure will each increase consumers’ feelings of intimacy toward a brand, and that these effects will be moderated by consumers' skepticism level. To test these hypotheses, several univariate ANOVAs were performed on intimacy toward a brand, including tests for Degree, Scarcity, Skepticism, and the two- and three-way interactions among these variables. As suggested by the ANOVA results in Table 3, the main effect for Degree was significant, F(1,248) = 9.18, p = .003, as was the Skepticism main effect, F(1,248) = 13.43, p < .001. Consistent with predictions, participants in the high degree of disclosure condition reported greater intimacy toward the brand (M = 4.48) than did participants in the low disclosure condition (M = 4.24). Also participants in the low skepticism condition reported greater intimacy toward the brand (M = 4.50) than did participants in the high skepticism condition (M = 4.22).

Table 3: Univariate Analysis of Variance Results

 In addition, the Degree × Scarcity interaction was significant, F (1,248) = 5.10, p < .05. Participants in the high degree/high scarcity condition reported greater intimacy toward the brand (M = 4.55) than did participants in the high degree/low scarcity condition (M = 4.41), low degree/high scarcity condition (M = 4.13), and low degree/low scarcity condition (M = 4.35). Upon closer examination of the interaction effects, when a brand provided its perceived backstage information exclusively, subjects felt the strongest intimacy toward the brand. A notable finding is the potential backlash of the scarcity of disclosure in the low degree disclosure condition. That is, when the information contains only frontstage information of the brand, making the information exclusive was not effective for generating consumer intimacy toward the brand. The main effect of Scarcity and the Degree × Skepticism and Scarcity × Skepticism interactions were not significant. Table 4 displays the means for all dependent variables by each of the eight conditions.

Table 4: Cell Means and Sample Sizes

 Hypotheses 1b-3b predicted that degree and scarcity of brand information disclosure will increase consumers' feelings of liking toward a brand and that these effects will be moderated by consumers' skepticism level. To test these hypotheses, univariate ANOVAs were performed on Liking toward a brand for Degree, Scarcity, Skepticism, and their two- and three-way interactions. As seen in Table 3, the main effect for Degree was significant, F(1,248) = 8.77, p = .003, as was the Skepticism main effect, F(1,248) = 29.67, p < .001. Consistent with predictions, participants in the high degree condition reported greater liking toward Mike's Coffee (M = 6.38) than did participants in the low disclosure condition (M = 6.06). In addition, participants in the low skepticism condition reported greater liking (M = 6.52) than did participants in the high skepticism condition (M = 5.93). High scarcity participants were not significantly more likely to feel liking (M = 6.22) than were low scarcity participants (M = 6.23).

 Furthermore, a significant Degree × Scarcity (F(1,248) = 7.67, p = .006) interaction was detected. Participants in the high degree/high scarcity condition reported greater liking toward the brand (M = 6.53) than did participants in the high degree/low scarcity condition (M = 6.24), low degree/high scarcity condition (M = 5.91), and low degree/ low scarcity condition (M = 6.22). That is, when a brand provided its perceived backstage information exclusively, subjects felt the strongest liking toward the brand. However, as observed in intimacy toward a brand, when the information reveals frontstage information about the brand only, exclusive feature information was not effective in generating greater liking toward the brand.

Another notable finding was the Degree x Skepticism interaction, F(1,248) = 5.21, p = .023. Participants in the low degree/high skepticism condition reported less liking toward the brand (M = 5.64) than did participants in the high degree/high skepticism condition (M = 6.21), high degree/low skepticism condition (M = 6.55), and low degree/low skepticism condition (M = 6.48). This finding indicates that highly skeptical consumers tend to feel significantly stronger liking toward the brand when they perceived the brand’s perceived backstage comparing to when they perceived frontstage-only information. 

 Hypotheses 1c-3c predicted that degree and scarcity of brand information disclosure will increase consumers' feelings of trust toward the brand, and that these effects will be moderated by consumers' skepticism level. To test these hypotheses, univariate ANOVAs were performed on Trust for Degree, Scarcity, Skepticism, and their two- and three-way interactions. As seen in Table 3, the main effect for Degree was significant, F(1,248) = 6.64, p < .01. Participants in the high degree condition reported greater trust (M = 4.42) than did participants in the low degree condition (M = 4.20). The main effect for Skepticism was also significant, F(1,248) = 22.90, p < .001. Participants in the low skepticism condition reported greater trust (M = 4.51) than did participants in the high skepticism condition (M = 4.11). High scarcity participants were not significantly more likely to feel trust (M = 4.35) than were low scarcity participants (M = 4.27).

 Another notable finding was the Degree × Scarcity interaction, F(1,248) = 4.24, p < .05. Participants in the high degree/high scarcity condition reported greater trust toward the brand (M = 4.46) than did participants in the high degree/low scarcity condition (M = 4.37), low degree/high scarcity condition (M = 4.08), and low degree/low scarcity condition (M = 4.33). That is, when a brand provided its perceived backstage information exclusively, subjects felt the strongest trust toward the brand. However, as observed in intimacy toward a brand and liking toward a brand, when the information reveals frontstage information about the brand only, exclusive nature of information disclosure was not effective in generating greater trust toward the brand. The Degree x Skepticism and Scarcity x Skepticism interactions were not significant.

DISCUSSION

 This study investigated the influence of brand information disclosure through social media on consumer evaluation of a brand. Specifically, this study explored how the degree and scarcity of information disclosure influenced consumers’ intimacy, liking, and trust toward a brand. Furthermore, the study investigated the moderating role of consumers' advertising skepticism on their brand evaluations.

 First, in line with the predictions, the results showed that subjects exhibited enhanced relational outcomes, such as greater intimacy, liking, and trust toward the brand when there was greater brand information disclosure. The results revealed significant main effects for degree of information disclosure on intimacy toward the brand (H1a), liking toward the brand (H2a), and trust toward the brand (H3a). Consumers in the high degree of disclosure condition felt greater intimacy, liking, and trust toward the brand than did consumers in the low disclosure condition. The results may be interpreted as follows: When consumers perceived a higher degree of information disclosure by recognizing the perceived backstage story of the brand in addition to its frontstage counterpart, they were more likely to evaluate the brand in more intimate, favorable, and trustable terms than when they perceived a lower degree of information disclosure by merely seeing the frontstage of the brand.

 Second, the study predicted that subjects in the high scarcity of information disclosure condition would feel more intimacy (H1b), liking (H2b), and trust (H3b) toward the brand than in the low scarcity condition. The results showed that these hypotheses are supported only in particular situations. Significant interaction effects of degree and scarcity were detected across proposed relational outcomes. In particular, consumers in the high degree/high scarcity condition reported greater intimacy, liking, and trust toward the brand than did consumers in other conditions. It was noteworthy that consumers exhibited the lowest intimacy, liking, and trust in the low degree/high scarcity condition. Thus, the results may be interpreted as follows: When a brand offered a higher degree of information disclosure in an exclusive manner, consumers felt the strongest liking toward the brand. In contrast, when a brand disclosed a lower degree of information in an exclusive way, consumers felt the least liking toward the brand. This finding suggests that there may be a potential backlash of the scarcity of disclosure in situations where the information is perceived as the very basic information that a business usually makes available to consumers. That is, when the information contains a relatively lower degree of disclosure, such as frontstage information, making the information exclusive is not effective for generating positive consumer responses.

 One possible reason for failure to detect a main effect of scarcity relates to the characteristics of the social media used. It is conceivable that the subjects comprehended social media as more open platforms than other media environments. Since social media or Web 2.0 are a collection of new internet applications that emphasize user participation, connectivity, user-generation, information sharing, and collaboration (Henderson & Bowley, 2010), subjects may have perceived scarcity involving exclusive information sharing as unremarkable for a social media environment. Nonetheless, the results of the study suggest that scarcity of information can be an effective communication strategy to build a strong connection with consumers when the brand discloses particularly unique and rare information about the brand.

 Third, a significant interactive effect between degree and skepticism was detected for consumers’ liking toward the brand. Subjects in the high degree/low skepticism condition showed the strongest liking toward the brand, followed by those in the low degree/low skepticism condition, high degree/high skepticism condition, and low degree/high skepticism condition. This finding can be interpreted as follows: The relatively highly skeptical consumers tended to feel significantly stronger liking toward the brand when they perceived higher disclosure of the brand compared to when they perceived lower disclosure. On the other hand, the relatively less skeptical consumers generally liked the brand more than the highly skeptical consumers. In addition, while these less skeptical consumers evaluated the high degree of information disclosure more positively than the low degree of disclosure, the difference in their liking toward the brand in the low degree disclosure and high degree disclosure conditions was smaller than for the highly skeptical consumers. Thus, the highly skeptical consumers demonstrated a more sensitive response toward the degree of information disclosure than the less skeptical consumers.

 Another interesting and unexpected finding in this study was the significant main effect of consumer skepticism across various relational outcomes such as intimacy, liking, and trust. Consumers who had more disbelief toward advertising claims in general were less likely to feel intimacy, liking, and trust toward the brand than consumers who had less disbelief toward advertising claims.

 Overall, the findings from the current study illustrate that degree of brand information disclosure is a significant influencer on consumers' brand evaluations in a social media environment. In addition, findings highlighted the influential role of the scarcity of information disclosure, depending upon the degree of information disclosure. Moreover, the findings evidenced how consumers' general advertising skepticism can play a significant role when consumers are exposed to information from the brand via social media.

Theoretical Contributions

 Based on the findings from this study, several theoretical contributions are evident. The foremost theoretical contribution of the current research is that of adapting theories of self-disclosure to a highly relevant consumer-brand relationship context. As the consumer-brand relationship paradigm suggests, the relationship between consumers and brands is similar to that of human relationships (Fournier, 1998). Accordingly, consumers often form relationships with brands in much the same way as they do with other human beings in a social context (Aggarwal, 2004). The findings of the current investigation extend previous consumer-brand relationship research by suggesting the existence of relationships between consumers and brands in newly emergent social media environments. More importantly, by adopting self-disclosure theories from interpersonal contexts, this study proposed a significant role of brand information disclosure in enhancing consumer evaluation and building a stronger consumer-brand relationship.

 In doing so, one of the biggest contributions of the current research is that it sheds light on the degree of information disclosure in the consumer-brand relationship context. As previous studies suggested, relational outcomes of self-disclosure vary depending upon the degree of self-disclosure (Altman & Taylor, 1973; Journald, 1971; Pearce & Sharp, 1973; Wheeless, 1976). In this theoretical light, this study proposed consideration of the perceived backstage of a brand as the more personal and intimate aspects of the brand. By adopting the frontstage/perceived backstage conceptualization from the highly relevant service marketing context (Grayson, 1998), the current study attempted to elucidate different degrees of brand information disclosure. The results of this study echo the findings of previous self-disclosure research in the interpersonal relationship field that information receivers would show greater intimacy (e.g., Altman & Taylor, 1973; Reis & Shaver, 1988), liking (e.g., Chaikin & Delega, 1974; Chelune, 1977), and trust (e.g., Pederson & Higbee, 1969; Wheeless, 1976) toward a discloser who presents more personal and intimate information. That is, when consumers perceive a higher degree of information disclosure by recognizing the perceived backstage of the brand in addition to its frontstage counterpart, they are more likely to evaluate the brand in more intimate, favorable, and trustworthy terms than when they perceive a lower degree of information disclosure from merely seeing the frontstage of the brand.

 Another noteworthy theoretical contribution is that the current study highlights the impact of information scarcity in the consumer-brand relationship context. Previous self-disclosure studies reported a scarcity effect of self-disclosure (Derelega & Grezelak, 1979; Petty & Mirels, 1981), suggesting that being selected as the recipient of a scarce message leads to having a more positive evaluation of the discloser (e.g., Archer & Cook, 1986). The findings of this study support previous similar arguments by showing that the relational outcomes of brand information disclosure were maximized when the disclosing information was only available to selected consumers rather than when the disclosing information was available to a public audience. Furthermore, the current investigation extended our understanding of scarcity in self-disclosure by demonstrating that scarcity of information disclosure is effective only when the disclosing information is perceived as highly personal and intimate.

 Lastly, the results suggest a caveat to previous findings on consumer skepticism toward advertising. As Obermiller and Spangenberg (1998) evidenced, the notion that more skeptical consumers evaluate advertising claims more negatively than less skeptical consumers was also observed in this study. More skeptical consumers are less likely to feel intimacy, liking, and trust toward a brand than less skeptical consumers are toward the same information disclosure from a brand.

 In sum, the findings of this study not only substantiate previous findings on the impact of self-disclosure, but also extend the literature by demonstrating the effectiveness of degree and scarcity of self- disclosure, as well as consumer skepticism in the context of consumer -brand relationships in a social media environment.

Managerial Implications

 Overall, this research is one of the first attempts to examine the use of behind-the-scenes contents on social media sites to understand how advertisers can use this new platform more effectively. By contributing to the limited research available, this research allows advertisers to examine and understand the implications and effectiveness of information disclosure on social media.

 This research offers several practical implications for social media marketers who struggle to attract prospects and retain fans. In particular, as evidenced by the findings of this study, social media content could enhance consumer-brand relationship quality in a more positive direction by providing perceived backstage brand information. In particular, findings from the current study suggest that the brand needs to be more transparent and authentic in a social media environment to be more intimate, likable, and trustable. To improve the effects of marketing communication efforts via social media, the most important recommendation this research provides is to actively disclose behind- the-scenes information about a brand. Consumers feel closer to the brand and like it more when the brand discloses information about itself and its business activities.

 Secondly, this study highlights the effectiveness of scarcity of information disclosure when the information contains exceptional content. Given that the characteristics of social media are comparatively open to the public and easy to access, making information access limited is not recommendable in today's “open” media landscape. However, in particular, when the information contains more than basic content that discloses an in-depth look into a brand, making this information scarce could be an effective strategy. Therefore, marketers should use the exclusivity of information for such goals.

 Lastly, as evidenced in this study, consumer skepticism is a strong predictor that may influence marketing communication campaigns via social media. Even though social media are generally not perceived as advertising, skeptical consumers may maintain their skepticism toward any information that may be generated by a brand.

Limitations and Future Research

 As with most discovery-based investigations, several limitations are associated with this study. First, this study examined the impact of perceived backstage information using a single product category and a single social media platform (i.e., a corporate blog). Further research with a larger set of brands across different product/service categories is needed to expand the generalizability of the findings. In addition, future research should explore the effects of brand information disclosure on the Internet beyond corporate blog platforms.

 Second, as does most experimental research, this study uses only fictitious brands as stimuli. Therefore, this study must accept reduced external validity to ensure internal validity. In particular, the consumer- brand relationship took place in a hypothetical context using a fictitious brand. To evaluate the predictions in the study more fully, future research should be conducted in real-world contexts.

 Another limitation involved the use of an online experiment. Participants in this online experiment were simply seeking rewards from the online panel company. Therefore, there are methodological concerns about the internal and external validity of the results. However, this study excluded all participants who did not take the online experiment within the designated time period and subjects who did not perceive the stimuli as intended. In addition, the study provided a unique opportunity to study an understudied area of the discipline. Therefore, the researcher believes that the potential benefits of the study outweigh validity concerns. Further research should explore the disclosure of perceived backstage information about the brand by employing different methods, such as survey or content analysis.

 Lastly, due to the exploratory nature of this study, the findings are somewhat limited in their ability to help us fully understand how the disclosure of brand information can systemically influence various consumer behaviors. Based on the current findings, future research should propose and test a more comprehensive theoretical model to explain consumers’ responses toward information disclosure from brands.

 Further experiments should be conducted to compare consumers who have a shallow brand relationship to those who have a strong brand relationship. It also may be worthwhile to investigate the effects of social media content as an advertising vehicle. Because social media contents are under the control of marketers, the characteristics of social media content and advertising are similar except the budget for buying media spots. Thus, exploring the effect of social media content on brand attitude and purchase intention would be another meaningful step for future research.