Antecedents and consequences of luxury brand engagement in social media

Abstract

Purpose

The purpose of this study is to analyse the antecedent (Consumer Involvement) and the consequences (Brand Connection and Brand Usage Intent [BUI]) of the three dimensions of Consumer Brand Engagement (CBE) (Cognitive Processing, Affection and Activation) in luxury brand engagement on Facebook.

Design/methodology/approach

Data were collected through an online questionnaire completed by fans/followers of luxury brands’ Facebook pages. The empirical study was conducted using structural equation modelling.

Findings

Consumer Involvement has a positive impact on the three dimensions of CBE (Cognitive Processing, Affection and Activation). This leads to the conclusion that Activation and Affection have an impact both on Self-Brand Connection and on BUI. Moreover, it was found that Cognitive Processing impacts only on BUI.

Practical implications

The results identified the factors which brand managers should focus on to increase CBE on Facebook.

Originality/value

This study is a pioneer, as it extends the consumer engagement model to the social media context in a hedonic and conspicuous consumption category which includes luxury brand products.

Propósito

Esta investigación analiza el antecedente (compromiso del consumidor) y las consecuencias (conexión con la marca e intención de uso de marca) de las tres dimensiones del compromiso del consumidor con la marca (procesamiento cognitivo, afectivo y activo) en el compromiso con la marca de lujo en Facebook.

Diseño/metodología/enfoque

Los datos se obtuvieron a través de un cuestionario online de los seguidores de varias páginas de marcas de lujo en Facebook. El estudio empírico se realizó utilizando un modelo SEM.

Hallazgos

El compromiso del consumidor tiene un impacto positivo en las tres dimensiones del compromiso del consumidor con la marca (procesamiento cognitivo, afectivo y activo). Esto lleva a concluir que los procesos activo y afectivo tienen un impacto tanto en la conexión con la marca como en su intención de uso. Además, se constató que el proceso cognitivo sólo afecta a la intención de uso de la marca.

Implicaciones prácticas

Los resultados identificaron los factores en los que los directores de marca deberían centrarse para aumentar el compromiso del consumidor hacia la marca en Facebook.

Originalidad/valor

Este estudio es pionero, dado que extiende el modelo de compromiso del consumidor al contexto de las redes sociales en una categoría de consumo hedónico y ostentoso que incluye a bienes y servicios como las marcas de lujo.

Citation

Brandão, A., Pinho, E. and Rodrigues, P. (2019), "Antecedents and consequences of luxury brand engagement in social media", Spanish Journal of Marketing - ESIC, Vol. 23 No. 2, pp. 163-183. https://doi.org/10.1108/SJME-11-2018-0052

1. Introduction

The last worldwide economic recession forced many companies to change their strategic and communication plans and to reduce their budgets. On the other hand, sales of luxury brands have increased, and this industry does not seem to have been as badly affected as many others (Cristini et al., 2017; Parrott et al., 2015).

In parallel, we have observed a strong growth in access to new technologies and the Internet, the widespread use of smartphones and apps and escalation in the use of social media, especially of networking websites (Cristini et al., 2017). Consequently, companies are now looking at social media differently by giving increasing emphasis to these platforms and integrating social media strategies into their overall communication strategies (Singhal, 2016).

The commitment to customer engagement on social media has shown such growth that companies aggressively acquire fans and followers on platforms, such as Facebook, by investing heavily in commercials on the platform (Lee et al., 2014).

Luxury brands came later to these platforms, compared to other brands, due to their unique features, such as exclusivity and high status (Kapferer and Valette-Florence, 2018). However, luxury brand companies eventually realised that social media can have a positive impact on brands and that they needed to be part of this new world, and thus began to explore ways to exploit these new media.

In this context, it is imperative to examine, quantitatively, the effects of social media on luxury brands in terms of the brand themselves and in terms of their relationships with their consumers (Kim et al., 2013; Gallant and Arcand, 2017). This study aims to bridge the academic gap in the research into luxury brand management in the social media context. In addition, this study aims to increase the understanding of what leads to consumer engagement with luxury brands on social media, especially Facebook, and the consequences of that engagement.

The model used in this research into social media engagement was originally created by Hollebeek et al. (2014), and has been adapted to the present study of Facebook pages and luxury brands. This model was chosen based on the call by the original authors to validate their scales in another country and context. This research aims to contribute to the management of luxury brands and to provide insights into how to increase e-engagement in the luxury brand market through the emotional and affective appeal of consumer brand relationships.

From the theoretical point of view, this research is pioneering by extending Hollebeek et al. (2014)’s model to the social media context in the hedonic conspicuous consumption category of luxury brand products.

The structure of this paper is as follows: it begins with a literature review; second, the hypotheses are formulated; third, the research design is presented; fourth, the results are presented; last, there is the discussion and consideration of the implications, limitations and suggestions for future research.

2. Theoretical background

2.1 Luxury consumer–brand management

Studies into luxury brands and their consumption have proliferated in the last years (Hung et al., 2011), with a particular focus on fashion brands. The survival and growth of these brands during the most recent global economic recession has aroused great curiosity (Hung et al., 2011) and resulted in increased research in the area.

To better understand what a luxury brand is, it is important to take as a starting point the etymology of the word. Luxury derives from the Latin word “luxus”, which means “indulgence of the senses”, regardless of cost (Nueno and Quelch, 1998).

In the literature review luxury is defined as the goods of a special brand whose simple use or display provides the owner with esteem (Wiedmann et al., 2007). They satisfy their owners’ psychological and functional needs (Wiedmann et al., 2007). It is in this context that the main distinctive feature of luxury goods comes into comparison with non-luxury, or counterfeit, products: luxury goods satisfy psychological and functional needs, the psychological benefit distinguishing them more (Vigneron and Johnson, 2004).

Values such as high status (Vigneron and Johnson, 2004), exclusivity and quality are traditionally associated with luxury brands (Jin, 2012; Atwal and Williams, 2009). However, due to changes that have occurred in consumer behaviour in western societies, a new sense and perception of luxury has emerged (Atwal and Williams, 2009). This “new luxury” has been defined as all products and services that have higher levels of quality than other products in the same category, but are neither tremendously expensive nor out of reach of the average consumer (Atwal and Williams, 2009). In fact, there is a growing global desire for luxury goods, reflecting an era of “democratization of luxury” (Tsai, 2005; Hennigs et al., 2012).

Luxury is understood to be composed of elements that appeal more to emotions than rationality. These are characterised by originality, creativity, skill, precision, emotional appeal, a prominent image, exclusivity, high quality and premium price (Okonkwo, 2009b). The consumer buying a luxury brand does so for personal pleasure, attracted by the exclusivity of the name, which, in turn, might be assumed to denote his/her social status (Munster, 2014).

The concept of rarity is intrinsic to luxury brands and forms the basis of a great contradiction they face. When brands increase their exposure and sales, they must at the same time maintain the perception of limited supply (Okonkwo, 2009b), a key factor in the management of luxury brands. On the other hand, “luxury brands engage consumers through the force of enchantment or charisma” (Kim et al., 2016, p. 304), created through storytelling in advertising; this is another key factor in luxury brand management.

There is a growing number of studies in the area of luxury brand management, but there is much yet to learn, essentially with regard to the motivations behind the intention to purchase luxury brands (Okonkwo, 2009a). This point is of great importance for the management of luxury brands and for their sales performance and customer engagement.

Social media have brought new challenges to the managers of luxury brands; hitherto they have been very reluctant to engage with the mass media involvement associated with the development of social networks (Okonkwo, 2009b; Dhaoui, 2014; Parrott et al., 2015). However, luxury brands now actively make use of social media in their advertising and marketing strategies (Kim and Ko, 2012; Dhaoui, 2014).

Although there is a great deal of academic literature on the importance of consumer engagement in social media, there appears to be limited academic research about social media marketing in the context of luxury brands (Chu et al., 2013; Juric et al., 2012; Dhaoui, 2014).

2.2 Social media: definitions and consumption

The concepts of user-generated content and social media are related to Web 2.0; they are often confused and need to be distinguished (Kaplan and Haenlein, 2010). Social media have their roots in the emergence of blogs, increasing access to high-speed Internet and the creation of social networks (Kaplan and Haenlein, 2010). In practice, they are a set of online applications built on the Web 2.0 base, which allows the creation and exchange of user-generated content (Berthon, Pitt, Plangger and Shapiro, 2012; Kaplan and Haenlein, 2010). The latter concept is categorised as all forms of content produced by consumers and available to the public (Kaplan and Haenlein, 2010).

Social media can take different forms, but is normally in text, image, video, audio and community/social media formats (Zhang, 2013, cited by Munster, 2014; Berthon et al., 2012). Social networks are the most recognised example of social media, consisting of platforms whose services connect friends, co-workers and organisations (Berthon et al., 2012; Munster, 2014). Specifically, social networking sites are applications that enable users to create and share information and send messages (Berthon et al., 2012; Kirtiş and Karahan, 2011) on the basis of social or professional interactions (De Vries, Gensler and Leeflang, 2012). Facebook, Instagram and LinkedIn are three examples of social networks. They have the capacity to leverage the power of relationships and collective knowledge (Berthon et al., 2012), created not only by companies, but also by consumers.

Social networks are some of the most popular forms of social media (Kaplan and Haenlein, 2010), and Facebook, in particular, is the preferred social network for consumers to engage with brands (Headstream, 2015).

In this world of social media, the consumer has complete control over the messages (s)he receives. It is (s)he who decides how to interpret them, whether (s)he considers them legitimate, or not, whether (s)he shares them, or not, and whether (s)he opens a discussion on them (Munster, 2014; Okonkwo, 2009b). From the perspective of companies, social media represent a revolutionary trend and it is certainly in their interests to be active in this area (Kaplan and Haenlein, 2010) to get closer to their customers.

The literature suggests that managers must recognise that social media are the most effective means at their disposal to communicate with the increasing numbers of consumers who spend a large part of their time online (Munster, 2014). In addition, social media enable companies efficiently to directly engage end-users in real time at a lower cost (Kirtiş and Karahan, 2011, Kaplan and Haenlein, 2010) than traditional media (Kirtiş and Karahan, 2011).

In the Facebook world, everything starts with the “like” button. When individuals put “taste” on a brand page, they are not only demonstrating their approval for the brand, but also to anyone in their “friends” network, that is, the Facebook network (Mariani and Mohammed, 2014). This process is referred to as electronic word-of-mouth (eWOM), one of the most important and effective marketing tools (Mariani and Mohammed, 2014).

In terms of the use of Facebook by brands, particularly retailers, the results have been very positive. According to previous studies, about 20 per cent of consumers are influenced in their purchasing decisions by Facebook brand pages (Hodis et al., 2015). It has also been verified that a greater level of engagement with brand pages is associated with a greater degree of loyalty. Obviously, loyalty has a very positive effect on consumer commitment (Hodis et al., 2015). It should be noted that the intensity with which consumers use these pages directly and indirectly affects loyalty by increasing engagement with the brand (Hodis et al., 2015).

2.3 Conceptual model and hypotheses

Interest in consumer engagement in the literature has evolved parallel to the growth of the Internet and new digital technologies and the tools (Sashi, 2012) included in Web 2.0 (Hollebeek, 2011). Marketers now realise that they have to be where their consumers and potential consumers are, and this is on social networking sites.

The concept of engagement is not recent. It has been of interest to professionals for some time and fundamentally for academics in disciplines such as sociology, psychology and organisational behaviour for the last decade (Hollebeek et al., 2014; Brodie et al., 2011; Bowden, 2009). According to the most recent literature review, there is no consensus as to its definition (Hollebeek, 2011); different interpretations are proposed according to the context, for example education or business.

The concept has been most studied in marketing literature (Hollebeek, 2011), based on its greater plurality and wealth of information. Consumer engagement, through its ability to enhance consumer loyalty, is increasingly viewed as a variable capable of increasing brand power (Brodie et al., 2013, Hollebeek, 2011; Bowden, 2009). More specifically, the consumer engagement process is presented as being related to, but distinct from, a final state of consumer loyalty (Hollebeek, 2011; Bowden, 2009).

In the marketing context, Relationship Marketing is more focussed on the concept of consumer engagement (Brodie et al., 2013; Kim et al., 2016). Thus, in this dimension, the academic focus considers engagement, loyalty (Bowden, 2009), trust and value creation as the drivers of engagement and the basis of the company-consumer relationship. (Gambetti and Graffigna, 2010).

Several research frameworks have been used to examine the concept of consumer engagement (Hollebeek et al., 2014; Vivek et al., 2014; Baldus, Voorhees and Calantone, 2015; Dessart, Veloutsou and Morgan-Thomas, 2016; van Doorn et al., 2010; Pansari and Kumar, 2017; Demangeot and Broderick, 2016). Dessart et al. (2016) developed a method of conceptualisation and operationalisation of consumer engagement in the context of online brand communities; they proposed seven sub-dimensions of consumer engagement and developed a new set of scales to use in all contexts.

Pansari and Kumar (2017, p. 295) defined customer engagement as “the mechanics of a customer’s value addition to the firm, either through direct or/and indirect contribution”. Their framework focussed on the customer engagement process, showing how customer engagement can be gained – through satisfaction and emotion – and then linked the direct and indirect contribution of customer engagement.

The Demangeot and Broderick (2016) framework focussed on website customer engagement underpinned by relational and communication knowledge. The authors defined customer engagement as “the process of developing a cognitive, affective and behavioural commitment to an active relationship with the website” (Demangeot and Broderick, 2016, p. 819).

In the present study we follow the Hollebeek et al. (2014) framework because there is little focus on luxury brands in previous studies into customer engagement. Moreover, the conceptual model proposed in the present study is already complex and introducing measures from the Demangeot and Broderick (2016) framework would make it even more complex.

Some studies suggest that consumer engagement can be explained by the emergence of interactive experiences and value creation in consumer–brand relationships (Brodie et al., 2011). In addition, it has been suggested that an engagement strategy is strategically imperative for the achievement of better company performance, not only at the level of competitive advantage but also of sales growth and profitability (Hollebeek et al., 2014; Tsai and Men, 2014; Brodie et al., 2011). These advantages are based on the fact that the consumers involved play a very important role in so-called viral marketing by providing recommendations for products, services and brands to other consumers (Hollebeek et al., 2014; Brodie et al., 2011).

Thus, consumer–brand engagement is an interactive process and can have different levels of intensity at different times (Brodie et al., 2013, Hollebeek, 2011). This phenomenon, therefore, reflects different stages of engagement (Brodie et al., 2013; Hollebeek, 2011). The starting point of this process is, in most cases, the consumers' need to seek information (Brodie et al., 2013, Hollebeek, 2011) from various platforms.

As for the consequences of consumer engagement, some authors cite as examples the concepts of trust (Hollebeek, 2011), commitment, emotional connection, empowerment and loyalty (Bowden, 2009). It should be noted that loyalty, commitment and empowerment are the most prominent concepts in the context of online brand communities (Brodie et al., 2013). The existing literature refers also to another consequence of consumer engagement, consumer satisfaction. This is the consumer's assessment of the performance of a product or service, and is seen as a consequence of engagement (Brodie et al., 2011, cited by Hollebeek et al., 2014).

In this context, Hollebeek et al. (2014), in their study of Consumer Brand Engagement (CBE) in social media, considered three generic dimensions related to the cognitive, emotional and behavioural nature of engagement: Cognitive Processing (CP), Affectivity and Activation.

The CP dimension is defined as “a consumer's level of brand-related thought processing and elaboration in a particular consumer/brand interaction” (Hollebeek et al., 2014, p. 154). The Affectivity dimension is the “degree of positive affectivity of the consumer related to the brand in a particular consumer/brand interaction” (Hollebeek et al., 2014, p. 154). Finally, the Activation dimension is the “level of energy, effort and time spent by the consumer with a brand in a particular consumer/brand interaction” (Hollebeek et al., 2014, p. 154).

It is important to highlight the concept of the brand adopted by Hollebeek et al. (2014), taking into account the concept of CBE, and not just Consumer Engagement. Thus, the brand is taken into account from an integrated perspective, which consists of the utilitarian, hedonic and symbolic aspects perceived by the consumer (Hollebeek et al., 2014).

The concept of consumer engagement is closely linked to the theory of relationship marketing (Ashley et al., 2011), where it is recognised that consumer behaviour is influenced by the experiences taking place in an increasingly complex environment. Thus, one can assume that the consumer engagement concept centres on specific interactive consumer experiences (Brodie et al., 2011).

Hollebeek (2011, p. 6) defined “customer brand engagement as the level of a customer’s motivational, brand-related and context-dependent state of mind characterised by specific levels of cognitive, emotional and behavioural activity in brand interactions”. Mollen and Wilson (2010, p. 5) defined “brand engagement” within a specific online context as “the cognitive and affective commitment to an active relationship with the brand as personified by the website or other computer-mediated entities designed to communicate brand value”. These authors suggested that a consumer’s brand engagement extends beyond mere involvement, as it encompasses an interactive relationship with the object of engagement, and is related to his/her perceived experiential value, in addition to the instrumental value obtained from interaction with a specific brand.

Moreover, the literature review suggests that customer engagement focuses on consumer satisfaction, because consumer satisfaction provides the brand with a higher value than its competitors’ brands (Sashi, 2012).

According to Hollebeek et al. (2014), Consumer Involvement (CI) is an antecedent of CBE. CI was explored by Zaichkowsky (1985) and Mittal (1995) and has been defined as “an individual’s level of interest and personal importance in relation to a focal object/decision in terms of his or her basic values, goals, and self-concept” (Hollebeek et al., 2014, p. 163); in this case, engagement with a specific brand. The term “engagement” has several definitions and measurements due to its different possible applications. In addition, it can be said that different people and different contexts result in different levels of engagement (Houston and Rothschild, 1978). Thus, it was necessary to create an engagement measuring method that could assess differences between people, objects and contexts.

Engagement is a multidimensional concept that includes cognitive, emotional and behavioural dimensions (Brodie et al., 2011; Zhang et al., 2016). On the other hand, “long-term involvement and engagement create brand commitment and, similarly, being committed to the brand can result in affective and enduring involvement and engagement”, as Parrott et al. (2015, p. 363) argued.

Hence, based on the study of Hollebeek et al. (2014), we suggest that CI is directly connected to the three CBE factors – more specifically, CI is an antecedent of CBE. In this context, the first three hypotheses are posited:

H1a.

CI through a luxury brand’s Facebook page has a direct positive impact on the CP of CBE.

H1b.

CI through a luxury brand’s Facebook page has a direct positive impact on the affection dimension of CBE.

H1c.

CI through a luxury brand’s Facebook page has a direct positive impact on the activation dimension of CBE.

As consequences of CBE, Hollebeek et al. (2014) selected Self-Brand Connection (SBC) and Brand Usage Intent (BUI).

A number of works have suggested that consequences of consumer engagement include trust and reputation (Casaló et al., 2009; Flavián et al., 2006; Hollebeek, 2011), satisfaction (Bowden, 2009), consumer value (Gruen et al., 2006; Schau et al., 2009) and loyalty (Schouten et al., 2007).

In addition, the literature review suggests that customer engagement is a set of measurable actions/behaviours that consumers exhibit on social media in response to brand-related content (Barger et al., 2016). Recent research has suggested that the way that consumers interact with social media can allow companies to analyse the engagement that these consumers have with their brands (Schivinski et al., 2016; Casaló et al., 2007).

SBC is a concept created by Escalas (2004) in the context of consumer brand and product use; consumers seek to create a desired self-image to present to others. SBC has been defined as “the extent to which individuals have incorporated brands into their self-concept” (Escalas, 2004, p. 170). Specifically, this concept is based on what a brand means to the consumer, in part as a consequence of narratives that the consumer builds over time and that incorporate the brand through mental associations.

Thus, SBC can be built based on the objectives, actions and results (perceived psychological benefits) contained in the narrative process of brands. It is assumed that, in individuals’ memories, stories tend to be related to the individual him/herself. So, when a brand tells a story, the individual might mentally create a connection between him/herself and the brand (Escalas, 2004). At the same time, consumers also value psychological and symbolic benefits, as these help them build the self-image, or self-identity, which they present to others. The choice of certain brands influences this whole scenario and coherence is created between the brands and consumers' self-image and expressed publicly or privately (Escalas, 2004).

SBC is not based on the discrepancy between user image and brand image; instead, it encompasses the degree to which a given brand is incorporated into the self-concept (Dwivedi et al., 2015). Consumers frequently engage in a process of defining their self-identities and desired self-images (Escalas, 2004). Certain products, luxury brands, have symbolic associations that are crucial to the consumer’s sense of self. These brands develop strong emotive bonds with consumers over time, which generally results in an increase in positive brand attitudes and behavioural intentions (Escalas, 2004).

From Hollebeek’s et al. (2014) perspective, SBC is a consequence of CBE, and thus related to its three constituent dimensions. Based on this premise, the following research hypotheses are posited:

H2a.

The CP dimension of CBE has a direct positive impact on the connection with the luxury brand/luxury brand’s Facebook page.

H2b.

The affection dimension of CBE has a direct positive impact on the connection with the luxury brand/luxury brand’s Facebook page.

H2c.

The activation dimension of CBE has a direct positive impact on the connection with the luxury brand/luxury brand’s Facebook page.

Hollebeek et al. (2014) proposed that BUI is an additional consequence of CBE; this was initially denominated “Overall Brand Equity” by Yoo and Donthu (2001). It has been defined as the “consumers' differential response between a focal brand and an unbranded product when both have the same level of marketing stimuli and product attributes” (Hollebeek et al., 2014, p. 163). This concept aims to address the notion of brand equity and to understand if consumers prefer a product with a brand name to an unbranded product which has the same features (physical attributes, quality and price). According to these authors, the difference in consumer response to a brand or an unbranded product can be described as the Brand Equity of the focal brand (Yoo and Donthu, 2001). This is the basis of BUI, used and reinforced by Hollebeek et al. (2014) as a consequence of CBE and its three constituent factors. Thus, the following hypotheses are formulated:

H3a.

The CP dimension of CBE has a direct positive impact on BUI/the luxury brand’s Facebook page.

H3b.

The affection dimension of CBE has a direct positive impact on BUI/the luxury brand’s Facebook page.

H3c.

The activation dimension of CBE has a direct positive impact on BUI/the luxury brand’s Facebook page.

3. Research design

To test the hypotheses we analysed luxury brands on Facebook. The reasons for choosing this networking website as the study object were very clear: Facebook is one of the most famous platforms and has the most users (Coutinho, 2014). With over 1.3 billion active users, this platform is a dominant player (Hodis et al., 2015).

The luxury brands chosen for the study were selected based on a financial evaluation taken from the Top 7 Best Global Luxury Brands 2014 Interbrand’s ranking[1]. We chose this ranking because Interbrand is one of the largest and more important brand-evaluation consultancies. Thus, the following brands, in decreasing order of value, were analysed: Louis Vuitton, Gucci, Hermès, Cartier, Prada, Tiffany and Co. and Burberry.

The conceptual model used in Figure 1 was adopted from Hollebeek et al. (2014) and demonstrates the connections between the studied variables. The research hypotheses are also presented.

3.1 Survey context and data collection

For the present study, keeping in mind the technological evolution represented by social media in the company–consumer relationship, we chose a quantitative method with a non-documentary indirect observation technique: the online questionnaire. Consequently, we designed seven online questionnaires with the same questions – only the object was different, based on the selected luxury brand.

The questionnaire used closed, mandatory questions, so that they would all be answered. The final section collected demographic information about the respondents.

To control for common bias, the present study followed Podsokoff et al. (2003). The respondents were unaware of the purpose and context of the research and knew nothing about the constructs under study. They were assured about the anonymity and confidentiality of their answers and that no judgments would be made based on their answers. Prior to online publication, the questionnaire was pre-tested on 25 subjects. The aim of this pre-test was to assess question formation and content, correct possible errors, check for unfamiliar words and to understand if the order of the questions was optimum. The subjects made a number of comments and necessary changes were applied.

The survey was available for Portuguese consumers on the Facebook social media brand sites. The luxury market in Portugal is stable in terms of consumer behaviour. The data were collected from Facebook users only. Three assumptions were made:

1. The individual was a follower/fan of one of the seven luxury brands’ Facebook pages.

2. The individual was a user of the brand selected.

3. The individual was aware of and/or had a positive image of one of the selected brands.

The data were collected between June and July 2015. A filter question was included to guarantee the first requirement to take part in the study, that is, that the participants were followers/fans.

From a total of 222 participants, 71 responded for Louis Vuitton, 41 for Burberry, 33 for Prada, 27 for Tiffany and Co., 22 for Gucci, 15 for Hermès and 13 for Cartier. It is important to note that, just as the largest number of people responded to Louis Vuitton in the questionnaire, this was the brand with the largest number of Facebook followers. The lowest number of responses was for Hermès and Cartier, which also have the smallest number of Facebook followers. The results also indicated that the majority of the respondents (69.8 per cent) were 25 to 50 years old and 92.3 per cent were women. Most of the respondents had a university degree (55.4 per cent) and were employed.

3.2 Measurement scales

We considered six variables, based on the Hollebeek et al. (2014) model:

1. CI;

2. CBE CP factor;

3. CBE affection factor;

4. CBE activation factor;

5. SBC; and

6. BUI (Table AI).

New scales were developed by Hollebeek et al. (2014) – these have not yet been applied in other studies. The scales were translated into Portuguese as accurately as possible. This was done to avoid deviations from the original context and meaning and to make them easier for the respondents to understand.

The latent variables were all measured by seven-point Likert-type scales, with (1) being “Totally Disagree” and (7) being “Totally Agree”. The CI scale was adapted, taking into account that the previous measurement was a semantic differential scale that included extremes with bipolar terms. Consequently, to meet the technical requirements of the Likert scale, only positive words were used for the assessments. The CBE Activation factor scale was amended by the inclusion of an extra item: “When I use social media associated with [brand], usually it’s the Facebook page.” This item was added to complement the previous item “When I use social media, usually it’s the Facebook page of [brand].” This, in addition, allowed us to determine whether the respondent was an occasional user or a fan/follower of the Facebook page.

3.3 Method

With the main goal of this research in mind – understanding CI’s impact on CBE and the impact of its 3 constituent factors on both SBC and BUI – we used structural equation modelling (SEM).

The present study used SEM to test the conceptual model and research hypotheses, using SPSS/AMOS 24 software and the procedures proposed by Anderson and Gerbing (1988). This procedure has two phases: first, a confirmatory factor analysis was used to test the measurement model; second, we estimated the hypotheses of the conceptual model by the maximum likelihood method. SEM was chosen because it facilitates the testing of hypotheses which propose complex relationships, defined in conceptual models, based on empirical measures. This is in line with Kaplan (2009).

4. Findings

Cronbach’s alpha was used to ensure the internal consistency and reliability of the scales. This procedure aims to demonstrate if the scales are credible and reproducible in different contexts, looking for their consistency (Marôco, 2010). In this context, Cronbach's alpha, which shows the average of all the scale items, is the most used coefficient. It ranges from 0 to 1, with values below 0.6 demonstrating unsatisfactory reliability (Malhotra, 2009). Kline (2011) stated that, if internal consistency is low, the content of the items will be too heterogeneous. Hair et al. (2010) suggested that the minimum threshold for Cronbach's alpha should be 0.7. The sample size for the present study was based on the rule that the number of respondents should be five times greater than the number of variables under study. However, some authors propose different minimum thresholds for SEM depending on the representativeness of the population. The most common procedure applied is maximum likelihood estimation (MLE). Some studies have suggested that, with MLE, samples with only 50 cases can provide stable and valid results (Hair et al., 2010). However, it should be borne in mind that these were different types of studies are carried out with less than ideal conditions.

For example, it has been suggested that one should think about recommended sample sizes in more absolute terms. Thus, the most commonly used dimension in studies using the structural equations model is 200 cases (Kline, 2011; Hair et al., 2010). However, a sample size of 200 may be too small if the model is very complex (Kline, 2011), which is not the case here. Hair et al. (2010) proposed some minimum sample size numbers based on characteristics of the measurement model. Thus, they argued that if a model contains “seven or fewer constructs, reasonable commonalities (0.5) and no unidentified constructs” (Hair et al., 2010, p. 662), the minimum sample size should be 150. As the present study examines six model constructs, there are high commonalities (above 0.6), and 222 individuals responded, it can be stated that the sample size is acceptable.

After the Exploratory Factor Analysis, a Confirmatory Factor Analysis (CFA) was carried out. It is a technique that “seeks to confirm whether the number of factors (or constructs) and the results of the variables (indicators) observed in them obey what is expected based on theory” (Malhotra, 2009, p. 725). Thus, the CFA specifies which variables define which construct (Malhotra, 2009), and at the same time allows us to confirm the measurement of the constructs used in the conceptual model (Hair et al., 2010). First, normality was assessed using skewness and kurtoses analyses and the values were above ± 2 and ± 7, respectively.

Based on the recommendations of most fit indexes used in academia to measure model goodness of fit (Hair et al., 2010), we can conclude that model has good fit: X²/df = 2.173, RMSEA = 0.073, CFI = 0.939, TLI = 0.929 and IFI = 0.940. So, the X²/df is lower than 3, the RMSEA is lower than 0.08 and CFI, TLI and IFI are above the 0.9 cut-off value.

According to Hair et al. (2010), CR should be greater than the AVE and the AVE should be greater than 0.5 to confirm convergent validity. The reliability constructs all present values greater than 0.9., except activation, which has a value of 0.893. We also undertook a confirmatory factor analysis to assess the measurement quality of the variables, using convergent and discriminant validity. The results indicate that the AVE is greater than 0.65 and that CR is greater than 0.9 for all constructs.

Discriminant validity can be confirmed through the analysis in Table I. We observed that the values of the square root of the AVE, presented diagonally in the matrix, are higher than the correlation values. Thus, the validity of the scales is proven.

We tested the validity of the research hypotheses using SEM.

The model fit is good: X²/df = 2.092, RMSEA = 0.070, CFI = 0.943, TLI = 0.934 and IFI = 0.943, in accordance with most common index used in the SEM approach (Hair et al., 2010). Based on the results shown at Table II, we can validate the following hypotheses:

H1a proposed a positive impact of CI on CP. This hypothesis was confirmed (γ = 0.740; p-value < 0.001). H1b proposed a positive impact of CI on the Affection factor. This hypothesis was also validated (γ = 0.35; p-value < 0.001). The loading of γ supports hypothesis H1c, – the impact of CI on the Activation factor; this hypothesis was also confirmed (γ = 0.248; p-value < 0.001).

Research hypotheses H2a, H2b and H2c, – the relations between the CP, Affection and Activation dimensions and SBC – were validated (β = 0.107; p-value = 0.085; β = 0.252; p-value = 0.002 and β = 0.566; p-value < 0.001, respectively).

Last, research hypotheses H3b and H3c were also confirmed – the impact of Affection and Activation on BUI was validated (β = 0.247; p-value = 0.028 and β = 0.503; p-value < 0.001 respectively). However, the impact of CP on BUI was not confirmed (β = 0.097; p-value = 0.255).

The indirect effects (standardized) of the antecedents of CI on the consequences of CBE reveal significant impact for both constructs, 0.475 for BUI and 0.577 for SBC. In addition, the most important mediator effect was with the affection construct for both outcome variables, follow by CP. CI has no indirect effect on SBC and on brand use intent through CBE (Table III).

5. Discussion

The first three research hypotheses related to the impact of CI on the three dimensions of CBE; they were validated with good results. These three positive inter-construct relationships were expected as they were proved by Hollebeek et al. (2014). We conclude that CI is a crucial antecedent of consumer engagement with brands – in this particular case, with luxury brands’ Facebook pages. This engagement is related to the level of the consumer’s interest in, and personal importance that (s)he gives to, the brand, when using its Facebook pages. Thus, involvement has a great impact on subsequent consumer engagement. This will, in turn, facilitate increased consumer engagement.

It should be noted that, of the three CBE factors, CI had the greatest impact on CP. However, in Hollebeek et al. (2014) it was not the most important mediator of CBE. To explain this difference one might distinguish between the objects under study and their basic functions. In the present study, the object was specific, luxury brands’ Facebook pages. Thus, the interest and personal importance given by consumers to these brand pages on this social network - that is, their involvement – has a great impact on thought, the cognitive side of engagement. This result is to the detriment, as noted, of the impact on the more emotional side of engagement, despite its overall positive impact. Therefore, the differences found between the results of the present study, about the impact of CI on the dimensions of CBE, and the results in Hollebeek et al. (2014), are acceptable.

The present study confirms the hypotheses about the mediator effects of CBE between CI and SBC and BUI, which is in line with Hollebeek et al. (2014). However, hypothesis H2a was accepted at a p-value inferior or equal to 10 per cent.

The relationship between CP and BUI was rejected. The present study confirms the conclusions of Hollebeek et al. (2014). CP was shown to impact on BUI based on the usage intent of luxury brands’ Facebook pages. Specifically, the cognitive and thinking dimension of CBE has an influence on the changes in the consumer’s usage of the brand. SBC was influenced more by the Activation and Affection factors, by order of importance. In Hollebeek’s et al. (2014) study, of the dimensions of CBE, Affection was found to exert the greatest impact on SBC. In the present study, Activation was found to be the dimension with the greatest impact on SBC. On the one hand, Affection is related to the positive feeling, or even happiness, felt towards a brand – in this case, towards the luxury brand’s Facebook page – that is created as the consumer interacts with it. Hence, Activation is related not only to time spent but also to the effort and energy dedicated by the consumer to these pages. Thus, the interaction time spent with luxury brands’ pages on Facebook leads to stronger relationships and to greater feelings of Affection. The more time and attention the consumer allocates to these pages, the more easily (s)he will introduce them into her/his self-concept – that is, (s)he will identify himself/herself with those pages and might even use them to present himself/herself to others.

In addition to the observation that Activation has great influence on SBC, we also confirmed the direct positive impact of Affection on SBC, as demonstrated in Hollebeek et al. (2014).

As for the consequences of BUI, we established the correlation between it and Affection. That is, if the consumer feels good when using a luxury brand’s Facebook page and there is a greater call to his/her emotional side, this will influence both the importance the consumer gives to the platform and his/her choice of that platform over others. This is in line with Hollebeek et al. (2014). Nevertheless, Affection was not found to exert a strong impact on BUI. Moreover, Activation was found to have a direct and strong impact on BUI. Thus, the time and effort allocated by the consumer to his/her interaction with luxury brands’ Facebook pages has a great impact on his/her choice of one platform over another. By visiting and interacting more frequently with luxury brands’ Facebook pages, the consumer favours these platforms over others with the same features and functions.

The present study diverges from Hollebeek et al. (2014) in that we argue that the CP of luxury Facebook pages is not an important antecedent of BUI. This result is due to luxury brand awareness. Moreover, when a consumer is a fan or a Facebook brand follower, (s)he already has knowledge about the brand. Thus, the present study emphasises the importance of affection and activation in the context of luxury BUI.

In accordance with previous studies, by visiting and interacting more often with luxury brands’ Facebook pages, the consumer prioritises this social network, even if others contain the same features and functions. Thus, hypothesis H3c was validated, also in accordance with Hollebeek et al. (2014).

In contrast to Hollebeek et al.’s (2014) study, the indirect effect of CI on SBC was mediated only for two of the three CBE mediators, CP and affection. CI has no indirect effect on either SBC or BUI through the CBE mediator effect.

6. Conclusion and implications

Presence on social networks is now mandatory for brands and companies worldwide. The reason is simple: the consumers are there. In fact, brand pages on social networks, such as Facebook, have quickly become the preferred platform for consumer-company interaction (Tsai and Men, 2013).

The revolution brought by social media to this consumer–brand relationship is profound, and goes much further that the technological evolution of the websites. More specifically, some publications have reported that consumers more frequently use social networks to search for information about a company, brand or product, than they use official websites (Dei Worldwide, 2008, cited by Tsai and Men, 2013).

Thus, brands must direct and invest their efforts towards social networks, which in themselves provide various advantages and opportunities. The social, participatory and community environment of these platforms makes them the perfect place for brands, in interactive and intimate ways, to communicate and engage with their consumers (Tsai and Men, 2013). In this way, they might activate and reinforce engagement, a crucial factor that involves the cognitive, emotional and behavioural side of consumer–brand interactions (Hollebeek et al., 2014). In addition, the added value of engagement lies in its contribution to the nature of consumer relations, which are increasingly interactive and experiential.

However, social networks can also have disadvantages for brands. The creation and diffusion of negative word-of-mouth about the brand is just one example (Brodie et al., 2013). Therefore, brand managers should take special care of how they contextualise their messages on social media and focus on building closer customer relationships (Tsai and Men, 2014).

Luxury brands are a rather special case for social media. They joined these platforms later in the light of their characteristics and many are still discovering the benefits of belonging to social networks. However, luxury brands no longer question whether they should be online; the pertinent question today is how they should present themselves online (Munster, 2014).

The present study emerged with the purpose of exploring one of the crucial elements of brand presence on social networks: CBE. Specifically, the purpose of the study is to understand the impact of CI (antecedent) on engagement with luxury brands’ Facebook pages, and to understand the impact of engagement on the two consequences of Brand Engagement, SBC and BUI, from the consumer’s point of view.

The results obtained proved the reliability and validity of the proposed conceptual model, and confirmed almost all of the research hypotheses.

The present study concluded that the consumer activation process in the context of luxury brands’ Facebook pages is the more important antecedent of the two constructs. CI, which translates into the personal importance given by the consumer (Mittal, 1995) to the Facebook page of the luxury brand, is a key antecedent of engagement with these pages. The more involved the consumer is, the greater will be his/her cognitive, emotional and behavioural engagement and relationship with the brand, and even his/her loyalty towards it.

CI has a higher impact on CP, which means that involvement reinforces the thinking of the consumer towards the brand’s Facebook page. That is, the more the consumer is involved with the page, the more his/her CP will increase, the more (s)he will think about it, and the stronger will be his/her engagement. This conclusion enhances the academic and managerial knowledge of luxury CBE. An individual, whether or not a luxury brand consumer, who interacts with a luxury brand’s Facebook page, will increase his/her engagement with the brand based on this interaction. The emotional and behavioural dimensions, in turn, have a smaller impact on CI; and the importance of the impact of the behavioural dimension on CI is less than the impact of the emotional dimension.

The consumer’s cognitive processes cause him/her to make a personal connection with the Facebook pages of the luxury brand, based on the psychological and symbolic benefits that (s)he derives from the brand and which helps to reflect his/her self-image (Escalas, 2004) – nor do they intend to use the Facebook page of the luxury brand over pages on other social networks. The present study shows the importance of the Activation process in the context of consumer luxury brand social media engagement. Thus, brands “act as symbols of personal fulfilment, provide self-esteem, allow each individual to differentiate and express individuality” Escalas (2004, p. 170). Therefore, it is the behavioural aspect of engagement that suggests that, if consumers spend more time and energy on the pages of these brands on Facebook, they are more likely to mirror their self-images and they are more likely to identify with the luxury brand.

The present study concludes that the time and dedication given by consumers to luxury brands’ Facebook pages influences their current, and probably future, intentions to continue to use the pages of this social network instead of those of other networks. It is further concluded that, in the context of luxury CBE, the CP dimension is not sufficiently important to link the consumer him/herself to the brand.

The next section examines the practical and theoretical contributions of the present study.

In managerial terms, it is crucial to understand the importance of CI on Facebook pages for creating and boosting CBE. This engagement is closer to the CP factor – the rational dimension of engagement. On the other hand, for CBE, the Activation factor better explains SBC via the luxury brands’ Facebook pages and BUI through the use of these pages.

We conclude that managers should create storytelling that engages the consumer, and should stimulate CP by sharing the stories on their Facebook pages - specifically, accurate and unambiguous messages. For this, managers should focus on the structure of their messages, so that the recipients observe, understand and respond positively to them. If they exercise some kind of influence, for what purpose they use the pages, among other possible issues.

To improve the proposed model it will be important to examine other possible antecedents and consequences of engagement in social media and to seek out the true motivations of the consequences of consumers’ use of luxury brands’ Facebook pages. Loyalty, satisfaction, trust and commitment have been identified as possible consequences (Brodie et al., 2013).

It would be very interesting to explore other luxury brands and social networks, for example, luxury car brands. In addition, it would be interesting to replicate this study with other social networks and to compare the differences to Facebook, for example with Instagram. This social network has grown exponentially over the last two years, essentially by attracting young people. Focussing on photography and videos, Instagram is a simple and much sought-after platform, for example, by photographers. Thus, brands, even luxury brands, are being attracted on a large scale to Instagram, which can generate up to 15 times the level of engagement of luxury brands compared to Facebook.

Figures

Figure 1.

Conceptual model and research hypotheses

Table I.

Constructs validity

CRb	AVEc	Constructs	CI	CP	AF	AC	SBC	BUI
0.94	0.65	CI	0.81a
0.92	0.78	CP	0.66*	0.88a
0.96	0.88	Affection (AF)	0.66*	0.72*	0.94a
0.90	0.68	Activation (AC)	0.58*	0.48*	0.72*	0.83a
0.94	0.69	SBC	0.65*	0.56*	0.73*	0.77*	0.83a
0.95	0.87	BUI	0.36*	0.48*	0.61*	0.64*	0.56*	0.93a

Notes:

^aSquare root of AVE in the diagonal;

*correlation is significant at level 0.01 (bilateral);

CR^b (Composite Reliability > 0.7); AVE^c (Average Variance Extracted > 0.5) and CR > AVE

Table II.

Results of SEM estimation

Hypotheses	Standardised estimates	t	p-value	Adjustment indicators
H1a: CI → CP (+)	0.740***	10.863	0.000	X2 = 736.283, df = 352, p-value = 0.000 RMSEA = 0.070 CFI = 0.943 TLI = 0.934 IFI = 0.943
H1b: CI → Affection (+)	0.355***	4.679	0.000
H1c: CI → Activation (+)	0.248***	3.448	0.000
H2a: CP → SBC (+)	0.107*	1.721	0.085
H2b: Affection → SBC (+)	0.252**	3.055	0.002
H2c: Activation → SBC (+)	0.566***	7.058	0.000
H3a: CP → Brand usage internet (+)	0.097	1.137	0.255
H3b: Affection → Brand usage internet (+)	0.247**	2.201	0.028
H3c: Activation → Brand usage internet (+)	0.503***	4.771	0.000

Notes:

***p-value < 0.001;

**p-value < 0.05;

*p-value < 0.1;

X² (Qui-square), df (degree of freedman); RMSEA (Root Mean Square Error of Approximation); CFI (Comparative Fit Index); TLI (Tucker–Lewis Index); IFI (Incremental Fit Index)

Table III.

Indirect effect (standardized)

Constructs	CI	CP	Affection
CP	0.000	0.000	0.000
Affection	0.338 [0.214; 0.481]a	0.000	0.000
Activation	0.321 [0.209; 0.489]a	0.332 [0.213; 0.532]a	0.000
Brand Usage Internet	0.475 [0.464; 0.769]a	0.178 [0.089; 0.388]a	0.278 [0.209; 0.574]a
SBC	0.577 [0.336; 0.602]a	0.209 [0.082; 0.350]a	0.349 [0.196; 0.499]a

Note:

^aBootstrapping 95% with 1000 iterations

Table AI.

Constructs	Items	Authors
CI	Not important – Important Boring – Interesting Irrelevant – Relevant Tedious – Exciting It means nothing to me – Means a lot to me Not appealing – Appealing Common – Fascinating Worthless – Valuable Not involving – Involving Not needed – Needed	Zaichkowsky (1994)
CP	Using Facebook gets me to think about Facebook I think about Facebook a lot when I’m using it Using Facebook stimulates my interest to learn more about Facebook	Hollebeek et al. (2014)
Affection	I feel very positive when I use Facebook Using Facebook makes me happy I feel good when I use Facebook I’m proud to use Facebook	Hollebeek et al. (2014)
Activation	I spend a loto f time using Facebook compared to other social media Whenever I’m using social media, I usually use Facebook Facebook is one of the brands I usually use when I use social media	Hollebeek et al. (2014)
SBC	Facebook reflects who I am I can identify with Facebook I feel a personal connection to Facebook I use Facebook to communicate who I am to other people I think Facebook helps me become the type of person I want to be I consider Facebook to be “me” (it reflects who I consider myself to be or the way that I want to present myself to others) Facebook suits me well	Escalas (2004)
Brand Usage Internet	It makes sense to use Facebook instead of any other brand, even if they are the same Even if another brand has the same features as Facebook, I would prefer to use Facebook If there is another brand as good as Facebook, I prefer to use Facebook If another brand is not different from Facebook in any way, it seems smarter to use Facebook	Yoo and Donthu (2001)

Note

1. www.bestglobalbrands.com/2014/ranking/#?filter=Luxury visited on the 7th of May of 2015.