A brand's image is a key component of brand equity, or brand value (Aaker, 1991). Keller (1993) defines brand image as different types of brand associations as well as the favourability, strength and uniqueness of these associations. Brand associations are developed based on product‐related and non product‐related attributes. Brand personality, user imagery, and usage imagery are the three main non product‐related attributes in Keller's, 1998 model of brand equity. Biel (1992) claims that user imagery is a key form of brand image while Shimp (2010) states that usage imagery has an important contribution to a brand's image.

Attitude toward a brand, or Brand Attitude, is another key component for valuing a brand's equity. Mitchell and Olson (1981) define Brand Attitude as an individual's overall evaluation of a brand. This means that attitude toward a brand mainly depends on a consumer's own perceptions regarding a brand, and are argued to be a reliable predictor of consumers' behaviour toward brands (Shimp, 2010). Most researchers (e.g. Aaker, 1996; Faircloth et al., 2001; Helgeson and Supphellen, 2004) regard these two as distinctive concepts where brand image helps develop potential buyers' overall evaluation or attitude toward a brand.

The Brand Loyalty construct signifies the degree of attachment a customer has for a brand and it is closely linked to use experience. Dick and Basu (1994) and Neal and Strauss (2008) both propose that Brand Loyalty have attitudinal and behavioural dimensions. The attitudinal dimension describes a consumer's overall satisfaction while the behavioural dimension represents the tendency of a consumer to purchase a particular brand repeatedly over time.

Some previous studies (e.g. Starr and Rubinson, 1978) find a positive relationship between attitude and loyalty toward a brand. A similar relationship appears to exist between a brand's image and loyalty toward the brand (e.g., Mustafa, 1999). Researchers such as Keller (1993) view loyalty toward a brand as an outcome of effectively managing knowledge about a brand's images and attitude toward the brand. Therefore, both a brand's image and the buyers' attitude toward a brand contribute to loyalty toward the brand (Faircloth et al., 2001).

It has been long suspected that consumers use products and brands to form and sustain their image (Veblen, 1899). The brand's image can provide “self‐expressive” benefits where consumers can use visually consumed brands to express their own image (Aaker, 1996). A brand's ability to express one's self‐image can be explained by self‐congruity theory (Sirgy et al., 1997). A theory derived from psychology, self‐congruity theory proposes that consumers like to compare themselves with a brand and see if the brand matches their concept of themselves (Sirgy, 1986).

Past research (e.g. Sirgy, 1986) has suggested that consumers have a better liking (attitude towards) for brands that are perceived to have strong favourable human characteristics that are congruent with his or her self‐concept. Various studies report that Brand Personality Congruity has a positive effect in attitude toward the brand (e.g. De Chernatony and Riley, 1998; Harris and Fleming, 2005). Kim et al. (2001) also finds that Brand Personality Congruity has a positive effect in loyalty toward the brand. We propose that:

H1. Brand Personality Congruity has a significant and positive effect in Brand Attitude.

H2. Brand Personality Congruity has a significant and positive effect in Brand Loyalty.

Brand User Imagery Congruity refers to the relationship between a buyer's self image and the perceived image of the user (Sirgy et al., 1997). High self‐congruity would mean that a consumer perceives the user of a particular brand closely matches his or her image. Consumers are more likely to purchase a particular brand if they perceive it to be consistent with their self images (Liu et al., 2008).

Sirgy et al. (1997) have shown evidence of using this construct to predict consumers' attitude and purchase intentions toward brands. Lim and O'cass (2001) find consumers have more favourable preferences and purchase intentions toward a brand that has a user image that is congruent with the perceived image of the brand's target market. Liu et al. (2008) find that Brand User Imagery Congruity has a significant and positive effect in consumers' attitude toward an automotive brand whilst Brand Personality Congruity does not. Both Aaker (1996) and Keller (1998) assert that they need not be congruent. For example, Levi's brand personality (tough, durable and simple) is based on the firm's heritage of providing clothes for miners in the 1849 gold rush in California, while its user imagery now tends to be urban, hip and unisex (Aaker, 1996). User imagery is said to be developed based on consumers' perception of the typical user of a brand whereas brand personality is developed based on interpreting the brand as having a human personality (Parker, 2005). Helgeson and Supphellen (2004)'s study has empirically supported that Brand Personality Congruity and Brand User Imagery Congruity may be two distinctive constructs. We hypothesize that:

H3. Brand User Imagery Congruity has a significant and positive effect in Brand Attitude.

H4. Brand User Imagery Congruity has a significant and positive effect in Brand Loyalty.

Biel (1992) defines brand usage imagery as the perceived stereotype of the situations in which a brand or product is generally used. For example, formal wear can be easily associated with a particular function, which is important and serious. Like user imagery, usage imagery is typically formed either through direct contact with the brand or through indirect means like word‐of‐mouth communications (Keller, 1998).

Brand Usage Imagery Congruity refers to the relationship between a consumer's expectations of the typical situations that a brand would be used by him/her, and the general perceptions of the brand's usage situations. Even though there has been substantial discussions on this area (e.g. Biel, 1992; Sotiropoulos, 2003), limited research has been conducted. The studies done to date (e.g. Liu et al., 2008) tend to focus on the two other congruity concepts of personality and user imagery.

Research has found that consumers tend to rely on external cues if they do not have sufficient self‐schemas that are congruent with the brand's image (Aaker, 1999). However, consumers more easily visualize brand usage imagery of when and how the brand should be used. Therefore, an association needs to exist between usage imagery and the potential buyers' self‐congruity. Sotiropoulos (2003) finds that liking the brand is expected to be higher when a respondent is presented with some usage imagery rather than with none under conditions of high self‐congruity. Therefore, we propose that:

H5. Brand Usage Imagery Congruity has a significant and positive effect in Brand Attitude.

H6. Brand Usage Imagery Congruity has a significant and positive effect in Brand Loyalty.

We also hypothesize a direct relationship between Brand Attitude and Brand Loyalty. Figure 1 illustrates the relationships examined in this study.

H7. Brand Attitude has a significant and positive effect in Brand Loyalty.

There have been a number of studies on luxury brands, the relationship with premium pricing, and the consumption behaviour of prestige‐seeking people (e.g. Vigneron and Johnson, 1999; Yeoman and McMahon‐Beattie, 2006). The 2008 global financial crisis has slowed luxury brand consumption in most developed economies. However, some emerging economies such as China and Russia are still experiencing continuous growth in luxury consumption. For example, China still maintained a three percent growth in luxury consumption after the global financial crisis, and it has now become the third biggest luxury market (Chinadaily, 2008).

A luxury brand is defined as going beyond functionality and emphasizes the status and image of an individual (Nueno and Quelch, 1998). As compared to necessity products, luxury products (usually goods) often place more emphasis on image rather than objective physical attributes. Important reasons for owning luxury products are to:

• ·

  show success and social status;

• ·

  reward own self; and

• ·

  provide me with confidence (KPMG, 2006).

However, what one person perceives as luxury may be ordinary to another. To minimize confusion, Vigneron and Johnson (1999) suggest a brand should consist of five unique values in order to be considered as luxury; conspicuous value, unique value, social value, hedonic value and quality value. Most luxury fashion brands provide all these five values that help develop a brands' overall image (Berry, 1994). Therefore, this study focuses on luxury fashion brands to better understand the relationships between imagery, self‐congruity, attitude and loyalty toward a brand.

This study uses a sample of university students in Australia. There are several reasons for the choice beyond convenience. University students represent an important present and future market segment that are “most likely to buy luxury goods and services” (Hauck and Stanforth, 2007, p. 179). The chosen university for this study is one of the most exclusive and expensive universities in one of Australia's most expensive cities in 2009‐2011. Due to the difficulty of getting admitted for enrolment and the associated education costs, the students in this sample tend to come from families with above average economic and social status. A major purpose of this study is to understand the young affluent's attitudes toward luxury brands because their early perceptions tend to have effects in their long run behaviours (Unity Marketing, 2007). Finally, university students are relatively homogenous in terms of their age, intelligence and income so this similarity can reduce the potential effects of these potential covariates in the results.

After conducting interviews of experts, this study chose luxury brand watches and sunglasses as two product categories examined. O'Cass and Frost (2002) used these two categories in a study to examine the effects of brand associations in the status and conspicuous consumption of luxury brands. A pre‐test among 50 undergraduate students found that Calvin Klein (CK) and Channel were perceived as well recognised luxury brands for watches and for sunglasses.

The study adopts Aaker's (1997) brand personality (BP) scale. The scale has five dimensions and a total of 42 traits. Numerous studies have suggested precautions in adopting the BP scale for different cultural contexts (e.g. Aaker et al., 2001; Sung and Tinkham, 2005). The Australian culture, however, is viewed to be not distinctively different from the US and UK cultures based on the key cultural dimensions (e.g. Hofstede, 2001; Liu et al., 2009). Thus, Aaker's BP scale was adopted for this study. As some traits are product categories specific, ten of the 42 traits are removed based on expert opinions. The second pre‐test was conducted to test the BP scale, and further four traits were removed from the measure due to their low factor loadings of less than 0.3 using factor analysis. The questions about self‐personality use the same response scale as brand personality. The absolute differences between brand personality and self‐personality represent the respondent's Brand Personality Congruity.

The measurement for Brand User Imagery Congruity is adopted from Sirgy et al. (1997), and includes four statements that are each rated on a five point Likert scale. The five item measurement of Brand Usage Imagery Congruity was developed based on a method similar to the one Sirgy et al. (1997) use to develop the scale for Brand User Imagery Congruity. The measurement for Brand Attitude is adopted from MacKenzie and Lutz (1989) and includes three semantic differential responses. The six‐item measurement for Brand Loyalty is adopted from Netemeyer et al. (2004) and Zeithaml et al. (1996). A third pre‐test was conducted to ensure the clarity of each item to respondents and found no problems with the respondents' interpretation or ability to respond to the questions.

This study uses self‐administered surveys completed during a lecture. A total of 391 students participated in the study. After removing the questionnaires that have less than 50 per cent of the questions answered, 264 valid questionnaires remain for further analyses. International students' responses were not used for this study because the Aaker scale used has been shown to be culturally specific. Also, many of the international students that come to the university sampled come from Asian countries where counterfeit luxury goods were the norm (O'Cass and Frost, 2002).

A mixture of 51 per cent males and 48 per cent females offer a good balance of gender in the sample. Around 58 per cent of the respondents are between 16 and 20 years old with the rest between 20 and 30 years old. Twenty three percent (23 per cent) of the respondents report spending an average of 100 Australian dollars per month on non‐necessities. Thirty seven percent (37 per cent) report spending between 100 to 200 dollars, while 23 per cent report between 200 to 300 dollars spent on non‐necessities. The remaining 17 per cent report an average monthly expenditure of more than 300 dollars on non‐essentials.

Confirmatory factor analyses using the alpha factoring extraction method with varimax rotation are conducted to test the constructs. Three out of the four items provide a single significant factor for each brand for Brand User Imagery Congruity. Another factor analysis finds a single factor for Brand Usage Imagery Congruity, Brand Attitude and Brand Loyalty. For both CK and Channel luxury brands, a single factor is an effect for four of the five Aaker Brand Personality Congruity dimensions (see Table I). The standardized reliability coefficients for all of the measures are larger than 0.8, indicating a good level of inter‐item reliability.

Structural Equation Modeling (SEM) using the maximum‐likelihood estimation method is used to test the hypotheses (see Figure 2 for the full model). Composite scores are used to represent each construct. A separate SEM model is tested for CK and Chanel brands. The model fit indices show that each model provides a good fit to the data (CK: X²/df=1.575, p=0.078, RMR=0.060, RMSEA=0.055, CFI=0.973, GFI=0.961, TLI=0.930; Chanel: X²/df=1.512, p=0.098, RMR=0.063, RMSEA=0.052, CFI=0.987, GFI=0.970, TLI=0.967).

Tables II and III present the effects of all independent variables in Brand Attitude and Brand Loyalty for the CK and Chanel brands. Brand Personality Congruity is not a significant effect in the Brand Attitude for either brand. Therefore, hypothesis one is not supported. The results show that Brand Personality Congruity is a significant effect in Brand Loyalty for CK but not a significant effect in Chanel Brand Loyalty. Thus, hypothesis two is only partially supported. Brand User Imagery Congruity is a positive effect in Brand Attitude for both CK and Chanel brands. Hypothesis three is supported. Brand User Imagery Congruity is a significant effect in Brand Loyalty for the Chanel brand, but is not a significant effect in Brand Loyalty for the CK brand. Thus, hypothesis four is only partially supported.

The construct Brand Usage Imagery Congruity is a significant effect in Brand Attitude for both CK and Chanel brands. Thus, hypothesis five is supported. Brand Usage Imagery Congruity is a significant effect in Brand Loyalty for the CK brand, but is not a significant effect for the Chanel brand's loyalty. Therefore, H6 is partially supported. Finally, the results suggest that Brand Attitude is a significant effect in the Brand Loyalty for both CK and Chanel brands. Therefore, hypothesis seven is supported.

An examination of variances reveals that Brand Personality Congruity, Brand User Image Congruity and Brand Usage Imagery Congruity are significantly associated. For the CK brand, Brand Personality Congruity and Brand User Imagery Congruity have the highest correlation, followed by the association of Brand Usage Imagery Congruity and Brand User Imagery Congruity. The lowest association is Brand Usage Imagery Congruity with Brand Personality Congruity.

For the Chanel brand, Brand Usage Imagery Congruity and Brand User Imagery Congruity have the highest correlation, followed by the association of Brand Usage Imagery Congruity with Brand Personality Congruity. Brand Personality Congruity and Brand User Imagery Congruity have the lowest significant correlation for the CK brand. Finally, none of the covariates (such as age, gender, and reported expenditures on non‐necessities) is a significant effect in the findings.

Brand User Imagery Congruity and Brand Usage Imagery Congruity are found to be stronger predictors of attitude and loyalty toward brands than Brand Personality Congruity. The construct of Brand Personality Congruity is not a significant effect in the attitude toward either brand. For Brand Loyalty, Brand User Imagery Congruity is a significant effect for the Chanel brand while Brand Usage Imagery Congruity is a significant effect for the CK brand. Although the Brand Personality Congruity construct has a significant effect in loyalty toward the CK brand, its significant path value (0.099) is slightly lower than the Brand Usage Imagery Congruity path value (0.128).

These findings support the claims of Sirgy et al. (1997) that the similarity between a consumer's own self‐image and the typical user's image can play a significant role in consumers' attitude or purchasing intentions. The findings are also consistent with Liu et al.'s (2008) results where Brand User Imagery Congruity is a significant effect in attitudes toward automobile brands. Although the literature regarding usage imagery congruity is limited, the findings are consistent with Sotiropoulos's (2003) report that usage imagery has a significant effect in both attitude toward and purchase intentions for a brand (Sirgy et al., 2008).

The finding that a potential buyer's Brand Personality Congruity is not an effect in their attitude toward the luxury brands tested is consistent with Liu et al.'s experience (2008) with automobile brands. The inconsistent and relatively weak (in this study) effect of brand personality supports other research (e.g. Kressmann et al., 2006). However, the effect of Brand Personality Congruity in the loyalty of a brand may be due to the product categories tested.

Buyer imagery plays an important role in their information processing and their attitude formation (Mazzocco and Brock, 2006). Brand personality, brand user imagery, and brand usage imagery are brand concepts heavily based on imagery (Shimp, 2010). These three brand self‐congruity concepts are potential factors of a brand's image but may have different effects in decision‐making. For example, user imagery and usage imagery focus on the typical user or the typical usage. The imagery processing associated with user and usage would appear to be more easily judged than imagining the brand as a human being (Parker, 2005). More concrete imagery tends to develop stronger attitudes than abstract imagery (MacInnis and Price, 1987; Keller and McGill, 1994).

There are several practical implications of these findings. As brands are becoming less and less different in terms of product attributes, how to develop or enhance brand images via non‐product attributes is becoming increasingly important for marketing any brand (O'Cass and Frost, 2002). The findings that Brand User and Brand Usage Imagery could enhance a consumer's positive attitude and loyalty towards a brand suggests that visual coverage of these elements is not only important, but needs to be consistent with the imagery desired by the potential market. Therefore, these two self‐congruity concepts may be key factors in maintaining a long‐term relationship between the consumer and the brand. This leaves brand personality as a possibly less useful area of perceptual management.

Another important implication lies in the product category effects for user and usage imagery. As discussed earlier, user imagery appears to play a significant role for developing Brand Loyalty in the category of sunglasses whilst usage imagery is significant in the category of watches. Because most luxury brands (like Chanel and CK) market in multiple product categories, attention should be paid to the core perceptions of the brand when designing communication strategies for different categories. Ideally, user and usage imagery should be applied in marketing communications in order to develop a unique position in the market. However, this ideal solution may not be viable in many circumstances. For example, it is often difficult to portray both user and usage imagery in a 30‐second commercial. Under this circumstance, brand managers should carefully consider the differentiating effects of user and usage imagery in different product categories and select the best alternative.

There are a number of possible limitations that need to be noted beyond the caution about a limited number of product categories being studied. The sample was only one of many targets for luxury goods, and only two luxury brands were tested. Students enrolled in other universities or in other regions of Australia may have different responses. The scales for self‐concept have been shown to be sensitive to use in different cultures and the perception of these luxury brands.

This study was conducted in a university setting. The respondents' views may be different if the data was collected in another format or venue. Although the average amount spent on non‐essential products was not a significant covariate, their frequency of purchasing luxury brands was not obtained or accounted for in the results. Future studies should explore the effects of purchase experience.

The self‐congruity concept applied in this study refers to the actual self. Sirgy (1986) claimed that the effect of actual self is often more important than other types of self, such as ideal self, social self, or ideal social self. However, due to the strong symbolic values associated with luxury brands, the effect of ideal or social self may be more important for luxury brands than that for non‐luxury brands. Hence, future studies may include other types of self to broaden our understanding of the impact of self‐congruity in luxury consumption.

Another interesting future pursuit would be a cross‐cultural study. People in a collectivist society are often more influenced by others than those in an individualistic society (Hofstede, 2001; Choi et al., 2005). This difference implies that user imagery may play a more important role than usage imagery or brand personality in a collectivist society, than in an individualistic society. The current study utilizes a sample from Australia, a population that is regarded as an individualist culture. A cross‐cultural study including both individualistic and collectivist cultures would be able to test whether culture has an impact in the relationships tested in this study. As most luxury brands such as Chanel and CK target consumers globally (Wiedmann et al., 2007), a cross‐cultural study would be able to provide important insights to luxury brands on the issue of localization vs standardization.

The objective of this study was to examine the combined effects of the three brand self‐congruity concepts of potential buyers in their reported attitude and loyalty toward luxury fashion brands. The findings show that user and usage imagery are both more of an effect than their perceptions of the brand's personality. These images appear to be more readily formed than perceptions of other attributes and could be applied to non‐luxury brands and categories to create a competitive advantage over other competing brands.

About the authors

Section:

Dr Fang Liu is an Associate Professor of Marketing in the School of Business, University of Western Australia. Dr Liu's research interests centre on brand strategies, marketing communications, and cross‐cultural consumer behaviour. She has published dozens of peer‐reviewed articles, book chapters, and conference proceedings in these areas. Prior to her academic career, Dr Liu has extensive working experience in the area of international trade.

Dr Jianyao Li (PhD in Marketing, The University of Western Australia) is Assistant Professor in Marketing and Hospitality Management at the Business School, Sun Yat‐sen University in China. His main research areas include higher education marketing, international marketing, and service marketing. He has published a number of referred conference papers and his work has also appeared in international journals such as International Marketing Review, and Asia Pacific Journal of Marketing and Logistics. Jianyao Li is the corresponding author and can be contacted at: jianyao@mail.sysu.edu.cn

Dr Dick Mizerski is a Winthrop Professor and has the University Chair in Marketing at the University of Western Australia. He has special expertise in consumer decision‐making, advertising and brand management. He has worked as a government regulator of marketing in the USA, and has over 35 years of marketing consulting experience with some of the world's largest marketers of luxury products.

Huangting Soh has obtained an honors degree in Marketing at the University of Western Australia. She is currently working at a market consulting firm in Singapore.