INTRODUCTION

Much has been written about how customers perceive brands and about positive outcomes in terms of brand loyalty.1, 2, 3, 4 However, an area that has been almost totally neglected by marketing researchers is brand identification. For a number of years, identification based on social identity theory has been examined in the sociological and psychological disciplines and lately in organization behaviour and human resource management.5, 6, 7, 8, 9 There are a number of reasons why brand identification should receive more attention from marketing managers and researchers. For example, brand identification may lead to important positive behavioural outcomes such as brand loyalty.10 According to social identity theory, people articulate their sense of self and typically go beyond their personal identity to develop a social identity.8, 9 In the marketing context, consumers do so by identifying and associating themselves with brands that reflect and reinforce their self-identities.11 In addition, a consumer may regard the brand as an extension of the self.12 This is of great importance because it is vital for marketing managers and researchers to understand the dynamics of self, the symbolic meaning of goods and the role played by brands.13 Another reason for this research study is that despite the recent increased interest in the concept of brand personality,14, 15, 16 there is also a lack of research that investigates the effect of brand personality congruence on consumer attitudes and behaviours17 and brand identification and brand loyalty are such outcomes. It is of interest and importance to look at this issue of customer identification with the brand using social identity theory, and the outcomes of the study will be theoretically and managerially relevant.

The study closest to the current study was conducted by Kim et al,18 but their study did not examine brand personality congruence and reputation as antecedents of brand identification. In addition, there are a number of problems associated with their research study. The fit measures for their structural model (for example, GFI=0.82, AGFI=0.76, CFI=0.88) are well below even minimum levels,19 indicating ample room for improvement. Furthermore, they examined brand personality using Aaker's scale,15 but failed to integrate it into their final structural model. Instead, they used different and untested items to assess some of the fundamental dimensions. Furthermore, they used a small convenience sample of 130 university students, which falls short of normal recommendations concerning the sample size for structural equation modelling.19 The use of a convenience student sample could be seen as problematic because it has been pointed out that using a student sample does decrease generalizability of the findings,20 and that the usage of a convenience sample does render any study as exploratory in nature.21 In addition, some items are problematic. For example, ‘I will continue to use this brand because I am satisfied and acquainted with the brand’. This item was used to measure ‘brand loyalty’ but it addresses satisfaction as well. Nevertheless, this was an important study in a somewhat neglected research area.

Consequently, it is the objective of this study to propose a model and assess the effects of reputation and brand personality congruence on brand identification and brand loyalty. In the following section, social identity theory and brand identification are explored in detail. Then the effects of reputation and brand personality congruence on brand identification are discussed. This leads to the development of a model that includes brand loyalty as an outcome of brand identification. The model and a rival model are tested with a sample of car owners in Germany. This will be followed by a discussion of the most important findings, as well as theoretical and managerial implications of our study.

CONCEPTUAL BACKGROUND

Social identity theory and brand identification

The concept of brand identification is built on social identity theory, which has been employed widely in other disciplines.5, 6, 9 Identification based on social identity theory is in essence a perception of oneness with a group of persons.9 It has been emphasized that identification fulfils the need for social identity and self-definition.10 Using the identification concept in a brand-customer context can be justified in terms of the social identity theory itself, where identification with an organization can also happen without a need to interact or have formal ties with an organization.5 This perspective is supported by Scott and Lane,22 who propose that people can seek identification even when they are not formal organizational members: ‘To the extent that the group category is psychologically accepted as part of the self, an individual is said to be identified with the group’ (p. 46). Consequently, it can be said that identification is a perceptual construct where the individual is not compelled to expend effort towards group goals.9 Previous research studies in marketing have been conducted mainly in the non-profit sector.23, 24, 25

Social identification takes place when an individual identifies with a certain group. Consumers who identify with a certain group will define themselves in relation to that group and in distinction from members of a rival social group.7 Social identity theory is useful for understanding consumer behaviour because a consumer need not overtly act together with other consumers (nor even know them), but nevertheless sees themselves as part of a social group. We analyse this for consumers who are owners of a corporate brand. We discuss an important aspect of marketing that is just coming to the fore in marketing research: brand identification.

Development of hypotheses

In the following, we begin to discuss the concepts that emerge from the literature. In particular, we propose that reputation and brand personality congruence will influence brand identification, which in turn affects brand loyalty.

Reputation

Bhattacharya et al 23 suggested that the reputation of an organization often serves as an indicator of organizational success, because well-regarded organizations are assumed to be successful. If the reputation is perceived as successful and well known by consumers, this may also enhance their pride in identifying with a brand that has a good reputation.10, 26 A number of different stakeholders may evaluate the reputation of a brand (for example, industry experts, stockholders, employees). As the focus of this study is on consumers, we focus on the extent to which consumers hold the brand in high regard. Reputation is sometimes assessed based on personal experiences. However, it could similarly be derived from word-of-mouth information or reports in the media.27 Bergami and Bagozzi28 point out that when a person associates himself with a well-regarded brand, then positive identification is generated. This leads people to associate themselves with well-regarded brands to increase their self-esteem. On the basis of previous studies,10, 11, 23, 28 we advance the following hypothesis:

Hypothesis 1 (H1):

  • Reputation is related positively to brand identification.

Brand personality congruence

The idea of brand personality developed out of the notion of brand image in the early 1950s.29 Accordingly, brand personality stems from one stream of research that emphasized human qualities of brands inspiring personality-led branding and the creation of brand personality as a concept. Hence, generally, brand personality is viewed as brand image component.30 It is noteworthy that, because of the strong link to marketing communications, brand personality became synonymous with all the brand's non-functional benefits.31 These were then communicated, mainly by utilizing user imagery (the portrayal of a specific brand user who matches the brand's perceived personality). This links the consumer's personality to the brand's personality. Brand personality is usually defined as ‘the set of human characteristics associated with a brand’ (p. 347).15 This manner of attributing human features to inanimate objects has been found relevant: a number of studies have shown that a consumer's attitude is influenced by matching the perceived product user image with their own self-concept.32, 33 A brand personality scale enables researchers to quantitatively relate brand personality to self-congruence.12, 32, 33 This can then be related to symbolic consumption benefits. Previous research has shown how self-concept is employed as a cognitive referent in assessing symbolic cues.33 The personality of a brand can indeed be seen as a set of product-symbolic cues. The greater the match between the brand personality and the consumer's self-concept, the more likely that the consumer has a favourable attitude towards the brand. This process is typically called ‘self-congruity’. Sirgy32 identifies four aspects of self-concept in explaining consumer behaviour. These are the ideal self-image, the actual self-image, the social self-image and the ideal social self-image. Our study focuses on ideal self-congruity because it has been found that ideal self-congruity is most appropriate for brand preference decisions.4, 32, 33 Ideal self-congruity relates to fit between how consumers would like to see themselves in relation to brand personality. The ideal self affects behaviour via the need for self-esteem. Consumers have ideal images of themselves, and realizing these images, through buying certain brands to boost their self-esteem. In this way, buying a brand that has a personality consistent with their ideal self aids them to feel good about themselves. For example, a consumer has an ideal self-image of becoming upmarket, and owning an aspirational car brand may help the consumer to fulfill this ideal self-image of becoming upmarket.30, 33 Therefore, it is hypothesized that:

Hypothesis 2 (H2):

  • Brand personality congruence is related positively to brand identification.

Brand loyalty

Customers can be promoters of a company's brands and services, and display loyalty behaviours.34 Organizational identification research indicates that when customers identify with a company, they have a tendency to purchase more and recommend the company's products more.10 In a study among brand communities, it was found that customers identifying with a brand community also tend to be supportive and make positive recommendations about the brand.25 These arguments find backing from previous research where university students were found to enact supportive behaviours towards the organization they identify with and provide positive word-of-mouth.24 A number of past studies10, 11, 28 have emphasized that identification may have favourable impacts on loyalty behaviours. Accordingly, it is hypothesized that:

Hypothesis 3 (H3):

  • Brand identification is related positively to brand loyalty.

We have presented our hypotheses in a model, depicted in Figure 1.

Figure 1
figure 1

Model of brand identification.

Full size image

METHODOLOGY

Research setting

Car ownership was the area of investigation. The choice of product class was based on a number of considerations. For example, cars provide ample opportunities for brand identification, as Hamm and Cundiff35 demonstrated that cars are rated among the top products that are most descriptive of the self. In addition, other studies25, 36 have investigated car ownership, and our findings will therefore contribute to developing understanding of this important sector of the global economy. We have obtained the cooperation of a major car manufacturer in Germany to collect empirical data via self-administered postal questionnaires. However, for confidentiality reasons the name of the company has been withheld. In order to certify further that the questionnaire was clear and meaningful to respondents, it was pre-tested at an authorized dealer of the brand.37 The sample consisted of 30 subjects whose characteristics resembled those from the target population in terms of car ownership, gender and age. Respondents had very few difficulties in completing the questionnaire. Hence, only minimal modifications were necessary to devise the final questionnaire and no second test was required. Using the company's national customer database as a sampling frame, primary data for this study were collected from a random quota sample of 5000 brand owners. The sample was drawn via a standardized selection process that minimizes systematic sampling bias. Out of the 5000 sent questionnaires, 1170 usable responses were received, yielding a response rate of 23.4 per cent. We investigated the likelihood of non-response bias using the extrapolation method.38 On the basis of our analyses, we can deduce that non-response bias is not a problem in our study.

Measures

The study uses multi-item scales to measure the constructs in our model. All items in the questionnaire were measured on a 5-point Likert-type scale anchored from ‘strongly disagree’ (1) to ‘strongly agree’ (5). All items were translated using the back-translation method.39 Some of the measures were available in the literature, although most were adapted to suit this particular context.

Previous studies27, 40 that measured reputation mostly used an index or parts of it, based on Fortune magazine's ‘Annual survey of America's Most Admired Corporations’ index. However, this index has received a lot of criticism.41 On the basis of the criticism, it seems appropriate to employ different measures. We, therefore, decided to use two items from Mael and Ashforth's42 measure because it has shown very encouraging results.23, 26 In addition, the items used are very similar to Selnes’2 reputation measure. Despite its popularity, it was decided not to use Aaker's15 brand personality scale. There are a number of reasons for this. The number of traits used (42 in total) makes it inappropriate for the current study because it would lead to an excessively long questionnaire, and consequently may lead to respondent fatigue. Second, research has shown that Aaker's brand personality scale is peculiarly sensitive to the cultural context in which it is used, and it has never in fact been completely replicated either inside or outside the United States.16, 18, 30, 41 Therefore, we used the brand personality scale of Hieronimus43 as it is the first validated brand personality measurement instrument in a German-speaking context; it also obtained strong support from Waller et al 14 The scale consists of two dimensions: security and temperament. The following items were used for security: successful, down-to-earth, honest and original. For temperament we used: spirited, imaginative, daring and passionate.

In order to assess congruence, we chose the traditional method, which is used in almost all studies30, 32 because it allows marketers to ascertain which dimensions are responsible for self-congruity. The dimensions that are most congruent with the self-concept can then be used in marketing communication campaigns. In this study, a two-step procedure was followed.32 First, respondents evaluated the brand with respect to the set of brand personality characteristics. Subsequently, the ideal self-concepts of the respondents were assessed with the same characteristics. After that, congruity is calculated by mathematically computing a discrepancy score for each characteristic, and then summing across all characteristics of the two dimensions. The following mathematical index was used to construct the two indexes:

where P i is rating of brand personality on characteristic i and S i is rating of ideal self-concept on characteristic i.

Brand identification was assessed with three items based on Mael and Ashforth's42 identification scale. A two-item scale was developed that measures behavioural loyalty, and the items are similar to the promoting scale developed by Arnett et al,24 the customer extra role behaviours scale developed by Ahearne et al 10 and the loyalty scale of Selnes.2

RESULTS

Measurement model

Structural equation modelling with LISREL 8.8 and the maximum likelihood method44 was employed to analyse the data. We evaluated the measurement model and then the structural model following the guidelines suggested by Anderson and Gerbing.45 The model is assessed with the goodness of fit index (GFI), adjusted goodness of fit index (AGFI), comparative fit index (CFI), standardized root mean residual (SRMR), root mean square error of approximation (RMSEA) and Akaike's46 information criterion (AIC), which addresses the issue of parsimony. The measurement model provided an excellent fit to the data: (χ 2 (21)=74.75, GFI=0.99, AGFI=0.97, CFI=0.99, SRMR=0.02, RMSEA=0.04 and AIC=122.75). All values of GFI, AGFI, CFI, SRMR and RMSEA meet the standards suggested by Hu and Bentler:47 0.95 for GFI, AGFI and CFI, 0.08 for SRMR, and 0.06 for RMSEA. There are no guidelines for the AIC, but lower values indicate better fit.47 All loadings being significant support convergent validity48 are reported in Table 1.

Table 1 Confirmatory factor analysis
Full size table

The internal consistency of the constructs was evaluated with Cronbach's α, composite reliability (CR) and average variance extracted (AVE), and the values are reported in Table 1. All Cronbach's α values are above the level set by Nunnally49 of 0.60. The estimates of CR should be above 0.60 and for AVE above 0.50.50 All values are above the stipulated criteria, and consequently are indicative of good internal consistency. In addition, we found support for discriminant validity by following the guidelines of Fornell and Larcker.51

Structural model estimation

The fit of the structural model was good: (χ 2 (23)=178.00, GFI=0.97, AGFI=0.94, CFI=0.97, SRMR=0.04, RMSEA=0.07 and AIC=222.00), and is reported in Table 2. In light of the theoretical foundation and the values of the fit indexes, no re-specifications were made to the model. In the following, we will test our hypotheses and consequently report the standardized path coefficients. As shown in Table 2, all our hypotheses are empirically supported. The effect of reputation on brand identification is significant (standardized path coefficient=0.52). This means that the higher the reputation of the brand, the more the customer will identify with the brand. Brand personality congruence has also a significant effect on brand identification (standardized path coefficient=0.24). Yet our findings indicate clearly that reputation has a stronger influence on brand identification. Brand identification, in turn, has a positive effect on brand loyalty (standardized path coefficient=0.60). This implies: the higher the identification of customers with a brand, the more loyal they are to a brand.

Table 2 Structural parameter estimates and fit indexes
Full size table

Rival model

It is not enough to evaluate the quality of a hypothesized model, and researchers should always compare it with rival models.52 In our preceding discussion, we presented theoretical support for positioning brand identification as a mediating variable in our hypothesized model. A rival model would hypothesize additional paths from reputation and brand personality congruence to brand loyalty. There is support in the literature for the rival model. According to Selnes,2 there is a direct influence of reputation on brand loyalty. In addition, a few authors3, 4 have shown that there are direct effects of self-congruity on brand loyalty, even though Magin et al 30 did not find empirical support for it. This leads us to add the two paths to brand loyalty. We have presented the rival model in Figure 2.

Figure 2
figure 2

Rival model of brand identification.

Full size image

Our theoretical model will be compared with the rival model. The results from the analysis for the rival model are shown in Table 2. We compare the hypothesized model with the rival model on the following criteria: (1) overall fit of the model, as measured by the GFI, AGFI, CFI, SRMR, RMSEA and AIC (2) number of the model's significant structural paths, and (3) ability to explain the variance in the outcomes of interest, as measured by squared multiple correlations (SMCs) of the outcome variables.

With regard to overall model fit, the rival model achieved an excellent fit, and outperforms the theoretical model on all fit indexes. All structural paths of the rival model are significant, indicating their value empirically. In addition, the SMCs in the rival model have changed. The SMC for brand identification is slightly lower (0.48 versus 0.45). However, the SMC of brand loyalty has increased considerably (0.36 versus 0.46). This implies that the explanatory power of reputation, brand personality congruence and brand identification as antecedents of brand loyalty are stronger than the explanatory power of brand identification alone. These results indicate that the rival model fits the data better than the hypothesized model.

DISCUSSION AND MANAGERIAL IMPLICATIONS

It appears that the time has come for marketers to benefit from this important feature of customer behaviour: brand identification. For some time, marketing academics10, 53, 54, 55, 56, 57 have pointed out that the concept of identification has great potential to explain and predict consumer behaviour. Our study has made a significant contribution to the marketing discipline in empirically demonstrating that brand identification can be a pertinent concept even in cases where there is no formal membership. For too long the importance of brand identification has not been properly addressed by the marketing literature. Only recently the marketing discipline and its top journals have taken up identification of customers with organizations and brands.10, 11, 24, 25, 54, 55, 56 Our study contributes to this important development. Indeed, it may be useful to build upon our findings in future by testing our model across different countries and industries to verify the generalizability of our results.

A number of important managerial implications will be discussed in the remaining part of this article. Managers should not rely only on customer satisfaction to evaluate their relationships with customers. Companies can develop strong relationships with customers via brand identification. It has been shown that in such cases customers are willing to pay higher prices, and this positively influences the financial performance of companies.55

It is vital that managers assess the level of brand identification in regular customer surveys, and the measurement instrument for this has been provided in our article. Managers are advised to work on developing their customers’ brand identification. A main focus of this could be to increase the salience of the brand. This can be achieved by highlighting positive comparisons between the in-group (the focal brand) and out-groups (other brands) in their marketing communications strategy.55 Furthermore, brand communities58 could prove useful to increase the level of brand identification because these communities would enhance the feeling of being part of a group.25

The results of this study suggest that when consumers perceive a brand as reputable, they report a higher level of brand identification and brand loyalty. Consequently, managers who enhance their products’ and services’ reputations could benefit from stronger brand identification and brand loyalty. Concentrating on external marketing communications might positively influence the brand's reputation. This should intensify the level of identification that consumers feel with a particular brand by making them feel like the brand's proud owners. In order for consumers to maintain their self-enhancement and satisfy their self-esteem, it is important for them to identify with well-regarded brands.11 Even so, reputation is not influenced only by a company's direct communications with different stakeholder groups. Certain information sources, such as newspaper articles, are clearly outside a company's sphere of influence. For this very reason, these sources may be better trusted and therefore be more influential in determining the perceived reputation of a brand. As independent external information sources sometimes have an even bigger influence on reputation than company-controlled communication efforts,26 it is clear that efforts should be made to strengthen reputation via marketing communication channels. Clearly, a positive reputation allows a company not only to keep its current customers, but also to attract new ones. In addition, a positive reputation can assist as a value signal where there is information overload or complexity. As most brands tend to have similar functional features, it is important for managers to focus more on emotional issues in order to clearly differentiate their brand from competitors. This is precisely where investment in developing brand personality to match target customers has value as a strategy in order to encourage brand identification. Previous studies11, 18, 23 have shown that consumers’ identity-based thinking can have a great influence on their resistance to change, and this in turn may deepen brand loyalty. Self-congruence theory postulates that consumers prefer those brands carrying symbolic value that enable them to portray or enhance their actual or desired self-image in order to derive emotional (that is, non-functional) benefits. Thus, this theory explains how brand personality drives brand identification and brand loyalty through self-congruity. It has been pointed out that consumers who experience a high level of self-congruity have a propensity to favourably process the utilitarian attributes of a product.33 This means that consumers who perceive a high match with their ideal self-image are likely to establish a positive attitude towards the brand. This is known as ‘first impressions’ in social psychology. This first impression has an effect on further information processing.32 If the first impression is good, then it is very likely that additional information processing will be positive as well. Consequently, it is very important to focus on a high level of self-congruity with the target market in any advertising messages because this would also influence reception of messages about the functional qualities of a brand. If there is a major gap between brand personality and the ideal self, then advertising messages directed at consumers could be a major cause for irritation.59 Consequently, managers are encouraged to assess the personality of their brands and to evaluate the ideal self of their consumers and their target market. This will help to design advertising campaigns and select communication channels that reinforce the match between the personality of the brand and the consumer's ideal self.