text new page (beta)
English (pdf)
Article in xml format
Article references
Automatic translation
Send this article by e-mail
Cited by SciELO 
Similars in
SciELO 
Permalink
INTRODUCTION
The concept of 'brand ambassador' plays a vital role in the fashion industry, although the literature shows at least three distinct interpretations. The most common usage is that of a celebrity endorser formally designated by the fashion brand (Barron, 2019; Caan & Lee, 2023; Wigley, 2015). The term is also applied to a social media fashion influencer who takes the initiative or is invited by a company to create online content endorsing their brand (Pedroni, 2022; Sanmiguel et al., 2018; Smith et al., 2018).
A third and arguably older interpretation is the "employee brand ambassador." This usage is less frequent in the fashion literature, even though the major brands are keenly aware of front- line employees and sales associates' importance in connecting with customers (Hui & Yee, 2015; Ton et al., 2010).
In recent years, the literature on internal branding has dramatically advanced the study of employee brand ambassadors. Numerous studies, particularly from the hospitality industry, have systematically examined the organizational practices that shape employee attitudes and behavior in the service of the brand (King, 2010; Terglav et al., 2016; Chung & Byrom, 2021; Xiong, 2023).
Hospitality organizations are well aware of the importance of creating a differentiated (i.e., branded) customer experience in order to achieve a competitive position and the critical role employees play in that endeavor (Hurrell & Scholarios, 2014; Garavan et al., 2022; King & Murillo, 2022; Qureshi et al., 2022).
Brands in the fashion industry share this awareness, especially those operating branded retail organizations. They invest substantial resources in recruiting and training personnel to develop employee brand ambassadors meaningfully connecting with their customers in retail settings (Ton et al., 2010). Indeed, "fashion retailers are not just in the business of selling clothes; they are selling a lifestyle" (Cutcher & Achtel, 2017, p. 2), an endeavor that requires employees who are knowledgeable and committed to the fashion brand they represent.
A literature review found no articles about fashion organizations' efforts to develop their employees into competent brand ambassadors using internal branding, which suggests a relevant gap. Given the critical role employees play in representing the fashion brand to customers in retail settings, this study examines the internal branding processes of a sportswear fashion firm with a global brand, an extensive retail network, and well-established onboarding practices and training materials. Accordingly, survey data were collected from employees at the corporate office and the retail network in Mexico, a significant consumer market for this brand. Results show that the company's brand training efforts significantly influence both retail and corporate employees and that they display greater brand understanding, identification, and commitment, which in turn contribute to brand-aligned behavior. The study thus finds evidence of internal branding processes and a clear intent to turn employees into competent brand ambassadors.
This study starts with a literature review of both fields addressed in the research: on the one side, the vision of the fashion and retail industry about the role of employees as brand ambassadors, and on the other side, an evolution of the concept of internal branding. Both fields allow us to introduce the hypotheses. Then, the methodology is explained, and the primary data and models are presented. Finally, the discussion introduces a brand ambassador model for the fashion industry and targets new possible directions for researchers in the area.
THEORETICAL BACKGROUND
Employee brand ambassadors in the fashion industry
For fashion organizations, the human factor is one of the pillars of branding, given that front- line employees act as the interface between the brand's internal and external environments, thereby strongly influencing customers' perceptions (Brodie et al., 2009; Chung et al., 2020). Exemplary service organizations have historically agreed that the brand's values should be communicated through employees' interactions with customers (e.g., Dekker, 2014; Wirtz & Zeithaml, 2018). Therefore, depending on the employees' internalization of the brand's values, their behaviors will reinforce or undermine the brand's advertised values during customer interactions.
Herein lies the marketing rationale for recognizing and leveraging the role of employee brand ambassadors (Gelb & Rangarajan, 2014; Schmidt & Baumgarth, 2018), for it is they who must bring the brand alive towards customers and deliver on the promises made through external marketing communications (Brodie et al., 2009). Furthermore, organizations usually consider all their employees, not just front-line, as brand ambassadors (Schmidt & Baumgarth, 2018), following the logic that corporate staff must lead by example. In addition, companies should have recovery strategies for unsatisfied clients with service breakdown prevention actions and face-to-face training to guarantee high-quality service (Woodside & Mir-Bernal, 2020).
For fashion organizations, contact between customers and company employees usually occurs in branded retail stores. Most fashion brands operate at least some branded stores, as these are "where they can express brand identity and reinforce a brand's power" (Arrigo, 2018, p. 121). These stores leverage the role of company employees as fashion brand ambassadors, relying on the significant impact their behavior can have on customers' shopping experience and brand perception (Gammoh et al., 2014; Kim & Kim, 2012; Ton et al., 2010). Interpersonal interaction between the salesperson and the customer influences customers' attitudes and loyalty toward the brand (Brexendorf et al., 2010).
Salespersons' identification with the brand positively influences their brand-supporting behaviors and sales efforts (Badrinarayanan & Laverie, 2011). Any misalignment between the company and the salespersons' perceptions of the brand has potential adverse effects not only on salesforce satisfaction, commitment, and performance but also on how the brand is represented in front of customers (Anisimova & Mavondo, 2010; Gammoh et al., 2014). Aligning salespeople's views of the brand to what is being communicated to external stakeholders is essential to deliver a consistent and strong brand message in the marketplace (Gammoh et al., 2014). The persuasion mechanism of influence in young people has been reviewed with the emerging term of influencer and its difference from the concept of influence and the effects on young new consumers (Sanmiguel & Sádaba, 2018).
Major fashion brands are aware of the critical role played by their employees in portraying the brand values (Pettinger, 2004; Storemark & Hoffmann, 2012; Sum & Hui, 2009) and historically have invested substantial resources to turn them into effective brand ambassadors (Ton et al., 2010). However, empirical research on these internal branding processes within fashion firms must be more extensive. This study draws from the internal branding research stream, which has systematically examined these processes in service industries.
Internal branding
Internal branding is "the activities undertaken by an organization to ensure that the brand promise, reflecting the espoused brand values that set customers' expectations, is enacted and delivered by employees" (Punjaisri & Wilson, 2011, p. 1523). This research area, also known as internal brand management, studies the organizational processes that firms have enacted to ensure their employees have the knowledge and attitudes they need to conduct themselves as brand ambassadors (King et al., 2023).
Because hospitality firms strive to create differentiated customer experiences through comprehensive training (Hurrell & Scholarios, 2014; Garavan et al., 2022), internal branding research has focused historically on the hospitality industry (Punjaisri & Wilson, 2011; Terglav et al., 2016; Zhang & Xu, 2021; Qureshi et al., 2022). However, the fundamental constructs have been studied in other service industries, including airlines (Vatankhah & Darvishi, 2018; Murillo, 2022), financial services (Garas et al., 2018; Taku et al., 2022), universities (Clark et al., 2020; Murillo & Atristain-Suárez, 2023), healthcare (Huang & Lai, 2018), ride-sharing platforms (Murillo & Terán-Bustamente, 2020); retail (Murillo, 2020; Porricelli et al., 2014), and public sector organizations (Leijerholt et al., 2022).
Over the years, internal branding researchers have systematically identified the various practices service organizations implement to achieve brand-aligned employee behavior and validated multi-item scales to measure the associated constructs (Piehler et al., 2016; Xiong et al., 2013). These practices include brand-oriented recruitment (King & So, 2015; Murillo & King, 2019), brand-oriented training (Huang & Lai, 2018; King & So, 2015; Murillo & King, 2019), and brand-oriented leadership (Terglav et al., 2016; Ayrom & Tumer, 2021; Xiong, 2023).
Among these, brand-oriented training is the most frequent organizational practice aimed at shaping employees' brand behaviors. In that sense, it can be the core practice in the internal branding toolkit. Indeed, some studies do not measure brand training separately but only a global internal branding construct with a prevalence of brand training items (Huang & Lai, 2018; Punjaisri & Wilson, 2011). In studies that explicitly measure brand training, the construct appears under different labels, such as brand training (Murillo & King, 2019), brand-oriented training (King & So, 2015), brand knowledge dissemination (King, 2010), and brand-centered training (Buil et al., 2016).
To deliver a differentiated customer experience, employees must thoroughly understand the fashion brand. It includes factual knowledge of the brand identity, heritage, values, and the promises made to customers through external advertising (Brodie et al., 2009). In addition, the employee must learn the specific behaviors the brand translates into during customer service encounters (Fleming & Witters, 2012).
Employee brand understanding is, therefore, a prerequisite of brand promise delivery among fashion employees (Piehler et al., 2016; Xiong et al., 2013). Defined here as the cognitive representation of the brand within employees' minds (Baumgarth & Schmidt, 2010), brand understanding is the proximal outcome organizations seek through brand training, particularly among new hires (King & So, 2015; Murillo & King, 2019).
Studies of brand understanding have shown that enhancing employees' cognitive grasp of the brand brings about several positive impacts, including greater brand identification (Piehler et al., 2016), brand commitment (Xiong et al., 2013; Piehler, 2018; Ruzzier et al., 2021; Taku et al., 2022), and brand performance (Xiong et al., 2013; Piehler et al., 2016; Piehler, 2018; Zhang & Xu, 2021).
These studies confirm that brand training is the proximal driver of brand understanding (King & So, 2015; Murillo & King, 2019; Murillo, 2022). Furthermore, recent studies taking a more fine-grained longitudinal view have shown that even for simple jobs, such as restaurant servers, a thorough grasp of the desired customer experience (i.e., the brand promise) and their role in delivering it can take months to be fully internalized, thus emphasizing the importance of a sustained training effort (Murillo & King, 2019). Based on these previous studies, the first hypothesis is formulated thus:
H1 Brand training has a positive impact on brand understanding.
Another proximal outcome of brand training is brand identification, defined here as "the degree to which a person defines himself-or herself by the same attributes that he or she believes define a brand" (Hughes & Ahearne, 2010, p. 84). In contrast to other conceptualizations of brand identification (Piehler et al., 2016), it is viewed here as a primarily cognitive attitude resulting from increased knowledge about the fashion brand, which constitutes the core content of brand training programs.
This brand identification view aligns with the cognitive dimension of organizational identification in Bergami and Bagozzi's (2000) cognitive, evaluative, and emotional social identification model. In the context of the fashion employee, brand identification involves the integration of perceived brand identity into self-identity (Hughes & Ahearne, 2010). Since brand identification is a cognitive response to knowledge about the brand, it is enhanced by brand training and increased understanding of the brand (Piehler et al., 2016). Therefore, the following hypotheses are advanced.
H2 Brand training has a positive impact on brand identification.
H3 Brand understanding has a positive impact on brand identification.
A core employee attitude identified since early internal branding studies is brand commitment, defined as "an employee's psychological attachment or feeling of belonging to the brand" (Xiong et al., 2013, p. 349). In contrast to brand understanding and identification, brand commitment has been conceptualized as an affect-based attitude that has brand understanding and brand identification as proximal antecedents (Punjaisri & Wilson, 2011; Piehler et al., 2016; Terglav et al., 2016; Murillo & Terán-Bustamante, 2020; Ruzzier et al., 2021; Zhang & Xu, 2021). Accordingly, it is proposed:
H4 Brand understanding has a positive impact on brand commitment.
H5 Brand identification has a positive impact on brand commitment.
The outcome sought by internal branding programs is to elicit brand-aligned behavior from employees, which most studies refer to as employee brand performance. Ideally, this behavior should stem from both the mind and the heart. In the final analysis, internal branding "deals with implementing the brand cognitively, affectively and behaviorally at the employee level" (Piehler, 2018, p. 217). It is easy to see how this agenda would be critical to a fashion brand and how cognitively and affectively engaged employees provide the best interface between the brand and external customers.
Multiple studies have found substantial and significant impacts of brand understanding on brand performance (Murillo & King, 2019; Murillo & Terán-Bustamante, 2020; Piehler, 2018; Xiong et al., 2013; Zhang & Xu, 2021). Brand identification has also shown significant direct effects on performance (Helm et al., 2016; Punjaisri & Wilson, 2011; Taku et al., 2022), although the relationship failed to reach significance in other studies (Piehler et al., 2016). Lastly, the positive relationship between brand commitment and brand performance has been confirmed on numerous occasions from the earliest internal branding studies (Xiong et al., 2013; Du Preez et al., 2017; King, 2010; Piehler et al., 2016; Qureshi et al., 2022; Zhang & Xu, 2021). Based on previous research, the following hypotheses are proposed.
H6 Brand understanding has a positive impact on brand performance.
H7 Brand identification has a positive impact on brand performance.
H8 Brand commitment has a positive impact on brand performance.
Based on the above discussion, the proposed research model is displayed in Figure 1.
METHODS
Data collection
Survey data were collected in the Mexico City corporate office and the regional retail network of a global sportswear fashion brand. The company operates 37 branded stores in the country with a total headcount of approximately 500 employees. Preliminary interviews with the company's training manager provided a good overview of the firm's onboarding efforts and its clear intent to develop employee brand ambassadors among all personnel. Newly hired employees undertake a two-day orientation where the company and brand culture are thoroughly explained. It is followed by an additional day of on-the-job training. Training materials and workshops strongly emphasize the sports brand and, especially for store employees, the desired customer experience.
The survey instrument was hosted on a professional online platform (Qualtrics) to facilitate distribution to a dispersed workforce and enable respondents to answer using a PC or smartphone. The survey was anonymous to encourage candid responses from participants. The survey link was distributed through an email invitation sent by the training manager to all employees at the corporate headquarters and the retail stores. Within two weeks, 80 usable surveys were collected from corporate and 61 from the stores, representing a response rate of 51% and 25%, respectively, in line with other employee surveys administered by the company.
Measurement and scale validation
Previously published scales were used to measure all variables in the model. For brand- oriented training, two items from the scale by King and So (2015) are complemented with two from the knowledge dissemination scale by King (2010). Brand commitment is measured using four items from the scale by Kimpakorn and Tocquer (2010). Brand identification, with two items from the scale by Gammoh et al. (2018). Brand understanding, with four items from the brand knowledge scale by Xiong et al. (2013). Brand performance, with three items from Xiong et al.'s (2013) scale for brand consistent behavior.
The Appendix lists all the scales and items used in this research. The scales were translated to Spanish following a team approach (Harkness, 2003). The training manager reviewed all translated items and made minimal wording adjustments to ensure clarity and better match the organizational context.
Because data were collected with a single instrument using self-report measures, common method variance (CMV) can potentially affect results (Podsakoff et al., 2012). Accordingly, a marker variable that is theoretically unrelated to the study constructs was included in the survey instrument, specifically, three items measuring preference for solitary work (Ramamoorthy & Carroll, 1998), which meet the criteria for an "ideal marker" (Simmering et al., 2015).
Two statistical remedies were used to assess and control for CMV. First, the Harman single- factor test was used to run an exploratory factor analysis of all the items, with principal axis factoring extraction restricted to a single factor and non-rotated solution. The single factor extracted explained 42.5% of the covariance of the items, showing that no single factor explains the majority of the covariance, indicating CMV data contamination. Second, the construct level correction technique that Chin et al. (2013) proposed was applied by modeling a CMV control variable with the theoretically unrelated marker variable. This control variable predicted all endogenous latent variables in the final model.
The results with and without CMV controls show minimal changes in estimated path coefficients and R2 squared values and no significant changes. This finding confirms CMV does not pose an issue in this study.
As an initial assessment of the validity and reliability of the multi-item scales, a confirmatory factor analysis (CFA) was estimated using Mplus v. 8.3. Accordingly, Table 1 shows the means, standard deviations, and correlations of observed variables, and Table 2 shows the CFA results.
Table 1 Means, standard deviations, and correlations among observed variables
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | |
| 1. BRC1 | -- | ||||||||||||||||
| 2. BRC2 | .39 | -- | |||||||||||||||
| 3. BRC3 | .42 | .33 | -- | ||||||||||||||
| 4. BRC4 | .54 | .51 | .30 | -- | |||||||||||||
| 5. BRPF1 | .23 | .49 | .36 | .48 | -- | ||||||||||||
| 6. BRPF2 | .35 | .56 | .33 | .55 | .63 | -- | |||||||||||
| 7. BRPF3 | .45 | .48 | .34 | .68 | .61 | .55 | -- | ||||||||||
| 8. BU1 | .47 | .45 | .29 | .54 | .47 | .55 | .61 | -- | |||||||||
| 9. BU2 | .41 | .50 | .33 | .45 | .50 | .59 | .48 | .52 | -- | ||||||||
| 10. BU3 | .33 | .48 | .32 | .40 | .48 | .56 | .48 | .49 | .66 | -- | |||||||
| 11. BU4 | .38 | .40 | .30 | .53 | .52 | .53 | .58 | .62 | .60 | .67 | -- | ||||||
| 12. IDENT1 | .47 | .36 | .10 | .46 | .30 | .36 | .42 | .50 | .33 | .32 | .38 | -- | |||||
| 13. IDENT2 | .52 | .51 | .28 | .63 | .48 | .63 | .62 | .61 | .52 | .58 | .65 | .55 | -- | ||||
| 14. TRAIN1 | .34 | .29 | .18 | .23 | .24 | .21 | .28 | .46 | .33 | .29 | .30 | .39 | .31 | -- | |||
| 15. TRAIN2 | .27 | .29 | .12 | .23 | .29 | .27 | .41 | .30 | .23 | .28 | .22 | .24 | .33 | .49 | -- | ||
| 16. TRAIN3 | .60 | .34 | .26 | .29 | .23 | .28 | .33 | .42 | .36 | .36 | .32 | .45 | .36 | .51 | .49 | -- | |
| 17. TRAIN4 | .39 | .37 | .30 | .31 | .35 | .38 | .39 | .48 | .43 | .43 | .37 | .28 | .38 | .39 | .30 | .53 | -- |
| Mean | 4.09 | 4.66 | 3.51 | 4.60 | 4.42 | 4.50 | 4.62 | 4.48 | 4.57 | 4.51 | 4.46 | 4.38 | 4.41 | 3.88 | 4.16 | 3.99 | 4.15 |
| SD | 1.05 | .54 | 1.13 | .61 | .59 | .57 | .54 | .64 | .55 | .55 | .62 | .79 | .65 | 1.12 | .91 | .89 | .90 |
Source: Own elaboration.
The results in Table 2 show an adequate fit for the CFA model with χ²(df = 109) = 185.006, p = .000, CFI = .920, RMSEA = .070, SRMR = .063. Cronbach alphas above 0.70 suggest adequate scale reliability for the five latent constructs. A possible concern is that the average variance extracted (AVE) is slightly below the recommended level of 0.50 for the brand training and commitment scales. In contrast, AVE values above 0.50 indicate convergent validity for the remaining scales.
Table 2 Confirmatory factor analysis results
| Indicator | Loading |
| BRTRAIN (Alpha = .76, AVE = .46) | |
| TRAIN1 | .66 |
| TRAIN2 | .60 |
| TRAIN3 | .79 |
| TRAIN4 | .65 |
| BRUND (Alpha = .85, AVE = .60) | |
| BU1 | .75 |
| BU2 | .75 |
| BU3 | .77 |
| BU4 | .82 |
| BRCOM (Alpha = .69, AVE = .42) BRC1 | .65 |
| BRC2 | .68 |
| BRC3 | .45 |
| BRC4 | .78 |
| BRIDENT (Alpha = .70, AVE = .59) | |
| IDENT1 | .61 |
| IDENT2 | .89 |
| BRPERF (Alpha = .82, AVE = .59) | |
| BRPF1 | .72 |
| BRPF2 | .78 |
| BRPF3 | .80 |
Source: Own elaboration.
Since the hypothesized model will be estimated with the partial least squares (PLS) technique, which is designed to maximize the explained variance of the latent variables, a judgment on the convergent validity of brand training and brand commitment will be deferred until PLS results are available, while nevertheless making a note of the two low AVE's.
Hypothesized model estimation
Structural equation modeling (SEM) is needed to estimate the hypothesized model, which contemplates a system of eight hypotheses. While covariance-based SEM (CB-SEM) is the preferred technique for theory testing, it requires relatively large sample sizes and data compliance with the assumption of multivariate normality (Hair et al., 2022). SPSS analysis of kurtosis and skewness indicated a violation of multivariate normality in the dataset. This fact recommends using PLS as the model estimation technique because, unlike CB-SEM, it does not make distributional assumptions (Hair et al., 2022). Furthermore, this study is exploratory, applying an internal branding model to a previously unexamined context. Accordingly, PLS is a more appropriate estimation technique because it has greater statistical power than CB-SEM and, therefore, faces a lower risk of prematurely rejecting candidate constructs (Reinartz et al., 2009).
Table 3 Indicator loadings, scale Cronbach alphas, and Composite reliabilities
| Scale | Indicator loading | Cronbach alpha | Composite reliability |
|---|---|---|---|
| Brand training | 0.769 | 0.851 | |
| TRAIN1 | 0.772 | ||
| TRAIN2 | 0.700 | ||
| TRAIN3 | 0.840 | ||
| TRAIN4 | 0.755 | ||
| Brand understanding | 0.853 | 0.901 | |
| BU1 | 0.801 | ||
| BU2 | 0.825 | ||
| BU3 | 0.836 | ||
| BU4 | 0.868 | ||
| Brand identification | 0.709 | 0.869 | |
| IDENT1 | 0.828 | ||
| IDENT2 | 0.923 | ||
| Brand commitment | 0.739 | 0.834 | |
| BRC1 | 0.768 | ||
| BRC2 | 0.770 | ||
| BRC3 | 0.602 | ||
| BRC4 | 0.836 | ||
| Brand performance | 0.815 | 0.890 | |
| BRPF1 | 0.857 | ||
| BRPF2 | 0.853 | ||
| BRPF3 | 0.852 |
Source: Own elaboration.
The structural model was estimated using SmartPLS version 3.2.7 (Ringle et al., 2015). Following the recommendations of Hair et al. (2022), analysis of a PLS model starts with the assessment of the quality criteria for the measurement or outer model. Table 3 displays item loadings, Cronbach alphas, and composite reliabilities. All items achieved the recommended loading of 0.7 or above, except BRC3, with a loading of 0.602, which is considered acceptable for an exploratory study (Hair et al., 2022). The Cronbach alphas and the composite reliabilities exceeded the recommended value of 0.70, indicating acceptable internal consistency of the measurement scales (Hair et al., 2022).
The PLS quality criteria also require an assessment of convergent and discriminant validity (Hair et al., 2022). To this end, Table 4 displays the average variance extracted (AVE), the square root of the AVE (in italics on the main diagonal), and the correlations between the latent constructs.
Table 4 Average variance extracted and inter-construct correlations
| Variable | AVE | BRCOM | BRIDENT | BRPERF | BRTRAIN | BRUND |
| BRCOM | 0.561 | 0.749 | ||||
| BRIDENT | 0.769 | 0.675 | 0.877 | |||
| BRPERF | 0.730 | 0.715 | 0.652 | 0.854 | ||
| BRTRAIN | 0.590 | 0.529 | 0.510 | 0.467 | 0.768 | |
| BRUND | 0.694 | 0.675 | 0.688 | 0.745 | 0.560 | 0.833 |
Source: Own elaboration.
All PLS-estimated AVE values, including those of brand training and commitment, exceed the recommended 0.50 threshold, indicating acceptable convergent validity (Hair et al., 2022). Furthermore, in compliance with the Fornell-Larcker criterion, each of the elements on the main diagonal in Table 4 is larger than the respective row and column off-diagonal elements, which indicates adequate discriminant validity. In addition, the recently recommended HTMT criterion was also examined, and all values fell below the 0.90 threshold, confirming construct discriminant validity (Henseler et al., 2015). Therefore, with acceptable levels for the quality criteria of the PLS measurement model, the analysis can turn to the inner or structural model results (Hair et al., 2022), which are displayed in Figure 2.
The hypothesized model explains 64.5% of the variance of brand performance, which can be considered substantial (Chin, 1998). All path coefficients have the hypothesized positive signs, and most have a substantial magnitude. However, an additional procedure must be performed to establish statistical significance because PLS does not make distributional assumptions. A bootstrapping technique is used to generate confidence intervals for the path coefficients. Table 5 shows the procedure results using the recommended 5000 samples (Hair et al., 2022).
Table 5 Bootstrapping results and supported hypotheses
| Hypothesized path | Original Sample (O) | Standard Deviation (STDEV) | T Statistics (|O/STDEV|) | P Values | Hypothesis support |
|---|---|---|---|---|---|
| BRTRAIN → BRUND | 0.560 | 0.065 | 8.574 | 0.000 | H1 |
| BRTRAIN→BRIDENT | 0.183 | 0.085 | 2.153 | 0.031 | H2 |
| BRUND → BRIDENT | 0.585 | 0.066 | 8.829 | 0.000 | H3 |
| BRUND → BRCOM | 0.400 | 0.081 | 4.947 | 0.000 | H4 |
| BRIDENT → BRCOM | 0.400 | 0.084 | 4.738 | 0.000 | H5 |
| BRUND → BRPERF | 0.427 | 0.088 | 4.880 | 0.000 | H6 |
| BRIDENT → BRPERF | 0.130 | 0.083 | 1.558 | 0.119 | H7 X |
| BRCOM → BRPERF | 0.338 | 0.084 | 4.005 | 0.000 | H8 |
Source: Own elaboration.
The results show that all estimated path coefficients except the path from brand identification to brand performance are significant. Therefore, hypothesis 7 is not supported. Notwithstanding this, model results reveal a significant impact of brand identification on performance through the mediation of brand commitment. All other hypotheses in the study are supported.
As an additional test of robustness and taking advantage of the small-sample capabilities of PLS (Hair et al., 2012), the dataset was split into corporate (n = 80) and store (n = 61) employees, and the model was separately estimated for each subsample. The results for store employees were essentially the same as for the complete sample, with a marginal increase in the explained variance of brand performance (71.5%) and no change in the significance of path coefficients. For corporate employees, the direct path from brand training to brand identification became non-significant, making the effect of training on identification fully mediated by brand understanding. However, the significance of all other relationships remained unchanged.
DISCUSSION AND IMPLICATIONS
This empirical study of employees at a sportswear fashion firm tested internal branding relationships that have been previously validated in studies from various service industries (Buil et al., 2016; Du Preez et al., 2017; Huang & Lai, 2018; Murillo, 2022; Taku et al., 2022; Terglav et al., 2016). Survey results show strong support for nearly all of the hypothesized relationships (even with brand identification mediated by brand commitment), suggesting that internal branding practices, as implemented in fashion organizations, deserve greater research attention to build a holistic understanding of the profile and role of employee brand ambassadors in the fashion industry, and the best ways of recruiting, training and leading them, a topic which the fashion literature has yet to examine. Accordingly, and based on the results of this study, an Employee Brand Ambassador's model is advanced, as displayed in Figure 3.
The principal contribution of this research is to conduct the first theoretically grounded empirical examination of internal branding processes within a fashion firm, as indicated by a thorough review of the fashion and fashion marketing literature. The authors found this gap in the literature somewhat surprising, given that the industry, as opposed to the scholarly community, has long been aware of the importance of front-line employees who must bring the brand alive for customers (Pettinger, 2004). However, the literature review revealed that current mentions of brand ambassadors in the fashion literature focus on celebrity endorsers or social media influencers, leaving employees in the background. This omission in the research literature does not reflect the real-world practice of major fashion brands, which take care from initial orientation to equip new hires with detailed knowledge of their brand to enable them to portray the brand values in their interactions with customers (Pettinger, 2004).
The results of this study confirm industry practices and highlight critical managerial implications. First, the importance of spelling out during orientation and on-the-job training the differentiated brand promise of the fashion brand, which customers are familiar with through external brand marketing and expect to find clearly articulated by the company employees at branded stores. Second, it is essential to treat all company employees as brand ambassadors, not only those directly interacting with customers. It, in turn, translates into a strong company commitment to brand training and brand communications at both the corporate and operational levels. The researched company's training manager aptly summed this up: "Sports fashion brands seek to turn their retail collaborators into brand ambassadors because having an informed, self-confident, empathetic element, with a DNA related to the brand, increases the positive shopping experience."
Given the exploratory nature of the research, the study centered on assessing the effectiveness of brand training, the core practice in the internal branding toolkit. Results show a strong and positive impact of training on brand-aligned behavior through the mediation of two cognitive constructs-brand understanding and identification-and one affect-laden construct, brand commitment. This model explained a substantial proportion of the variance in brand performance. However, including additional theoretically grounded predictors in the model may improve results.
Within the growing body of internal branding theory, at least three concepts can contribute to future studies of fashion organizations. The first is brand-oriented recruitment. Studies in hospitality have found that values-based recruitment and selection can improve workforce identification with the values of the hospitality brand (Hurrell & Scholarios, 2014; King & So, 2015), an issue that resonates with fashion firms. The second concept is employee brand values fit, "the extent to which the employees perceive their values are consistent with the brand's values" (Xiong & King, 2015, p. 61).
Studies show a strong impact of brand values fit on brand motivation and performance (Xiong & King, 2015; Murillo, 2020) and additional psychological benefits for employees through enhancing their organization-based self-esteem (Murillo, 2019). In addition, brand values fit has been found to improve traditional human resource indicators such as job satisfaction and intention to remain (Du Preez et al., 2017; Murillo, 2020). The third concept is brand-oriented leadership, where previous studies have established the influence that supervisors and managers can have on employee attitudes and behavior toward the brand (Ayrom & Tumer, 2021; Terglav et al., 2016; Xiong, 2023), especially after employees are well past the onboarding and initial training stage (Murillo & King, 2019).
This study's limitations are identified to improve future research designs. The study is based on survey data collected with a single instrument. However, statistical procedures were adopted to detect and control for CMV, and results from two tests found no evidence of biased results. Another limitation is the relatively small sample size, which was further evidenced when the subsamples of corporate and store employees were examined separately. Future studies should ideally get large samples of both types of employees in fashion firms. Finally, the generalizability of these results is limited to fashion organizations with a culture (both organizational and national) similar to the one examined by this research.
