Social Networking Brand Engagement using Creative Brand Content Experiences

In today‟s environment, societies are free to create and browse online content; marketers therefore to face vexing challenges in drawing expressive social media users to engage with their brands. This current group of users display postmodernism characteristics; i.e. need more for subjective experiences to achieve self-realization. Hence, the creation of consumer brand engagement among expressive social media societies is through content that should be able to provide these experiences. However, there is lack research study on this trend. The objective of this study was to evaluate passive experiential content that users perceived as novelty emotions (i.e. perceived creativity) that leads to intrinsic state of engagements (i.e. cognitive engagement and affective engagement) and to intentional engagement (the activation of willingness of brand clicking activities in social media). This, in turn, creates an advantage for the marketers because from this process if users engage the chances of building a long-term consumer brand engagement relationship is higher. This research study was done on 25-34 year-old Malaysian active expressive social media users who utilised expressive social media sites on a daily basis for at least 3 hours as the sample population from whom to collect data and administer a questionnaire survey with a still image as stimulus using social networking messaging platforms. Data analysis was conducted using IBM SPSS and IBM AMOS software and results showed positive significant relationship within all the 7 constructs which were functional appeal, emotional appeal, vividness, perceived creativity, cognitive engagement, affective engagement and intentional engagement.


Introduction
Advancements in technology & high-speed internet access has promoted the rise of expressive social media such as Facebook, Twitter, Instagram and etc. (Kotler, Hermawan & Setiawan, 2010;Gensler, Volckner, Liu-Thompkins & Wiertz, 2013). Ever since the global emergence of these platforms, there has been a constant increase in the number of users joining these platforms. Currently there are approximately 3.484 billion expressive social media users that translates to a 45% social media penetration rate in global population of 7.676 billion people (Kemp, 2019a).Expressive social media societies activities in these platforms actively utilize by social media users for personal usage which are to create, share, like, browse content (Gensler et al., 2013;Statista, 2018) and to connect with family and friends (Statista, 2018).
With a constant increase in the proportion of expressive social media societies, marketers believe that there is a need to tap on these platforms. This is with a view to build marketers" competitive advantage in the current market, by having brand content that is able to create consumer brand engagement (Bloomstein, 2012;Marketing Science Institute [MSI], 2014;MSI, 2016;MSI, 2018;Hollebeek & Macky, 2019) which in turn promotes brand activities" interaction with social media societies (The Star, 2018;Stelzner, 2018).Marketers believe that the foundation block of engagement begins with brand content consumption which involves reading the passive content, curation by clicking social media button and later co-creation in which in turn give rise to followers (Evans & Mckee, 2010;Syrdal & Briggs, 2018;Wang, Malthouse, Calder & Uzunoglu, 2019;Hollebeek & Macky, 2019) and creates long-term engagement in the form of a relationship that is able to indirectly drive organizational performance. For instance, for Naked Pizza, 60% sales came from its" 4000 Twitter followers Bodapati & Bucklin, 2010) allowing users to create content such as adding pictures, posting and consuming content, liking or sharing contents, or connecting with the online social networking platform community (Kaplan & Haenlein, 2010). It is named the "expressive social media" because it allows users to shout-out their thoughts and emotions not only with text but also with emoji (Trigg, 2014). This therefore offers hyper-reality; meaning it enables user to portray to others their self-image in a virtual world (Riva et al., 2016;Hamouda, 2012) and to enjoy the paradox of juxtaposition (i.e. opposite effect of the reality and virtual world) and allows the user to experience the unexpected, imaginative and creative (Firat, Dholakia & Venkatesh, 1995).These points show that the expressive social media has shifted to a postmodernism trend which provides affective characterization on the platform such as emoji for the users to create their own personal ID on the virtual platform. This affective characterization has led to high engagement rates among many users (Kemp, 2019a). A high engagement rate indicates the platform has kick-started the success in building a relationship with its" users.
The postmodernism marketing phenomenon is a break in thinking away from the modern, functional and rational (Bouagina & Triki, 2014) to focus more on the subjective (Sweetman, 2005;Shaw & Jones, 2005).For, societies" behavior no longer focuses on rationality and money only, but is moving towards individuals that are driven by continuously searching for feelings and sensations that captivate and amaze them (Addis & Podesta, 2005;Forlani, Pencarelli & Buratti, 2018). Such behavior makes such individuals less predictable societies (Batra & Kazmi, 2008). Thus, relationship marketers critique the extendibility of the traditional marketers" approach of the simplistic marketing mix (Addis & Podesta, 2005). This in reality is far more complex (Addis & Podesta, 2005) if there is subjective behavior in the process (Bray, 2008). This is followed by, experiential marketers who attack the issues of the consumer rationality, weakness of the traditional consumer behavior theory (Addis & Podesta, 2005) that oversimplifies marketing realities (Duhring, 2017) and encourages understanding the underlying psychological process for the behavior of postmodernism societies (So et al., 2015); the non-cognitive process including affective lenses (Pham, 2013;Jacoby & Morrin, 2015). Therefore, the attitudinal model that has emerged during the traditional marketing paradigm has limitation in the current landscape (Weilbacher, 2001;Kitchen, Kerr, Schultz, McColl & Pals, 2014) marketers have suggested experience and motivation as better concept (Pham, 2013;Gial, Zhang, Paul & Giala, 2018) to stabilize their brand relationship (Ranasinghe & Samarasinghe, 2019).
In short, the behaviour of societies has moved to a trend of postmodernism in expressive social media. In this study, the postmodernism marketing strategy is focused on humanized marketing and specifically the experiential marketing concept with a need for pragmatic, visual and feeling experiences, and literature inducing creative emotion, imaginative/curiosity (cognitive engagement) and positive affectionate emotion (affective engagement) to create intentional engagement to motivate the users. Therefore, marketers able to stabilize or strengthen relationship marketing with brand that can create long term engagement in future.
Integrated marketing communication perspective, display content; with more focus on functional element and gimmicks that complement the brand with a more customer-focused marketing communication strategy to nourish stakeholder relationships (Takalani, 2015;Belch & Belch, 2003;Smilansky, 2009). The content, named "creative ideas" focuses on Janusian concept (Goldenberg, Mazursky & Solomon, 1999), Yin & Yang concept (Blasko & Mokwa, 1986), hard & soft thinking (Blasko & Mokwa, 1988, informational and transformational strategy (Kim & Cheong, 2011), informational and emotional contents (Appelbaum & Haliburton, 1993), utilitarian and emotional elements (Plessis, 2015), was later named "creative execution strategy" concept. This concept focuses on marketers to enhance creative ideas in the physical form which can be named appeals and presentation (Sabharwal, 2018;Laskey,Day & Crask, 1989;Belch & Belch, 2003;Kivinen, 2014;Amira, 2015). And as time went by, due to marketers being too focused on the growing popularity of loyalty marketing, introduced loyalty card schemes with introduction of discount and financial benefits for repeated purchases, thereby moving back to traditional advertising with hard selling approach by making their brand akin to commodities (Harker & Engan, 2010;Smilansky, 2009).
In marketing literature, different schools of thought describe experience differently. Hedonic consumption schools view experience as fantasies, feelings and fun generated by individuals (Clarke, 2013) through hedonic activities, e.g. art (Schmit, 2009). Extraordinary experiences school focuses on experience that elicits feelings, absorption, imagination, is intrinsically rewarding, energizes, enthuses, enjoyable happening with high-risk adventure activities, e.g. rafting, camping (Loeffler, 2004;Celsi et al., 1993;Schouten et al., 2007;Arnould & Price, 1993). Marketing management schools focus on holistic experience (Zatori, 2013) and believe that the human mind works in holistic way (Carbone & Haeckel, 1994;Schmitt, 1999) and first conceptualized it as customer or consumer experience. Customer experience is how a company employs the experience module to provide consumer experience. Whereas consumer experience is how the individual perceives experience encounter that is delivered by experience providers (Schmitt & Zarantonello, 2013;Schmitt, 1999). These experiences include sensory, affective, creative, cognitive, physical and social-identity experiences (Schmitt, 1999). Furthermore, other researchers derived dimensions of sensorial, affective, physical, social and cognitive experience (Fornerino, Guizon-Helme &Gotteland, 2008), sensory, intellectual, cognitive, pragmatic, lifestyle and relational experience (Gentile, Spiller & Noci, 2007), sense, feel, think, act and relate experience (Chang & Chieng, 2006), sensory, intellectual/think, feel/emotional, and behavioral experience (Delgado-Ballester & Sabiote, 2015) and sensory, intellectual, social and pragmatic experience (Cachero-Martinez & Vaquez-Casielles, 2017). Then, when experience blended into branding literature to manage brand activities, the concept of experiential branding was derived; it consists of three experiential hybrids of sense-brand appeal, feel-brand appeal and think-brand appeal (Schmit, 1997) and later Brakus et al. (2009) from the perspective of consumer experience, derived the concept of "brand experience" as subjective, internal consumer responses (sensations, feelings and cognitions) as well as a behavioral response evoked by brand-related stimuli. Researchers agreed that the concept of brand experiences was the umbrella term for the marketing management school due to its broad construct (Skard, Nysveen & Pedersen, 2011;Schmitt & Zarantonello, 2013).
From then on, with the birth of internet, it was named "online brand experience" which mostly focused on providing online brand experience on a website focusing on attractiveness of cookies, variety visual displays, value of money, participating in events (Ha & Perks, 2005) and later used an extension of technology acceptance model to study the antecedent of online brand experience (Morgan- Thomas & Veloutsou, 2013). However, to date little, is known about the dimensions of online brand experience based on the humanistic frame experience as a subjective and inner phenomenon (Bujisic, Bilgihan & Smith, 2015). With the rising of expressive social media, Ashley & Tuten (2015) suggested the need for experiential appeals from her content analysis study and recently Simon et al. (2013) proposed an online brand experience in Facebook with dimensions of sensory, affective, cognitive, relational, usability and engagement and Tafesse(2016) suggested dimensions of sensations, cognitions, behaviours and social dimensions but without an empirical study. However, the study of content in expressive social media platform to date, mostly focuses on suitable type of content based on characteristics such as vividness of image (De Vries, Gensler & Leeflang, 2012;Sabate, Berbegal-Mirabent,Canabate & Lebherz, 2014;Luarn, Lin & Chiu, 2015;Coursaris,Osch & Balogh, 2016), functional (Coursaris et al., 2016;Kim,Kang, Choi & Sung, 2016), novelty content (Tafesse, 2015;Syrdal & Briggs, 2018), transformation (Tafesse & Wein, 2018), emotional (Lee, Hosanagar & Nair, 2018), image with text ( Kwok & Yu, 2013;Pancer, Chandler, Poole, & Noseworth, 2018;Lee et al., 2018;Trefzger, Baccarella & Voigt, 2016), photos (Hirvijarvi, 2017) by tracking on standardized metrics of engagement such as share, like, comment using interviewing techniques (Syrdal & Briggs, 2018), text mining (Kwok & Yu, 2013), content analysis (Swani,Milne & Brownm 2013;Sabate et al., 2014;Luarn et al., 2015;Coursaris et al., 2016;Trefzger et al., 2016 ),voting (Hirvijarvi, 2017), imagination of any brand (Kim et al., 2016), systemic analysis (Tafesse, 2015), coding (Tafesse & Wein, 2018) and calculation with Turk software (Pancer et al., 2018;Lee et al., 2018) which show that the trend of study is still in the exploratory phase and there is lack of focus on using an experiential content as a theoretical foundation. With that background, the study aimed to use an experiential content to create the process of consumer brand engagement.

Experiential Content
To fill in the gap of yet of empirical study of experiential content with engagement mentioned in section 2.2.1, in this section this study creates components of experiential content. According to Schmitt (1999) experience can only be communicated through experience provider. The experience provider used was the creative execution strategy concept which forms the foundation for this model because this strategy is the basic principles that underpin content in advertising and enhance creative ideas (Tafesse & Wein, 2017;Sabharwal, 2018).The concepts of creative execution strategy are "appeals" and "presentation" as mentioned in section 2.2. For this study, selection of which appeals and presentation to use was done through consideration of concept of experiential branding element and the study of suitable types of content in the expressive social media as mentioned in section 2.2.1. Hence the concept used in this study to research experiential content were functional appeal that provides the individual with pragmatic experience from the content (Cachero -Martinex & Vazquez-Casielles 2017), emotional appeal that provides the individual with feel experience (Whiting, 2009;Addis & Holbrook, 2001) and vividness which provides sense or aesthetic experience (Brakus et al., 2009;Schmitt, 1999;Schmitt, 1997;Devitt, 2017) from the content. We still included functional appeal because a brand item still consists of objective characteristics (Addis & Holbrook, 2001) for its" balance in the experiential content (Simon et al., 2013;Yilmaz, 2017). Think experience and act experience were not applicable to the experiential content because they are active experiences that happen after passive experience (i.e. brand content) (Pine & Gilmore, 1999). Relational and engagement experience were not included in the experiential content as these experiences are built after passive and active experience (Evans & Mckee, 2010;Syrdal & Briggs, 2018;Wang et al., 2019; Hollebeek & Macky, 2019).
Based on the above explanations, perceived creativity is an important construct for this study because it is an indication of the presence of the creative item and the presence of this reaction behaviour in societies is subjective and outside the realm of rational (Nystrom, 2000) which current postmodernism societies need internal lack of balance in life, constantly require novelty in everyday leisure activities to prevent them for boredom in the process to achieve self-realization (Gonzalez-Cutre et al., 2016).Moreover, according to the literature, (e.g. extraordinary experience literatures in section 2.2.1), an interaction with an experience requires this reaction for its process of consumer brand engagement to occur, for this engagement construct consists of characteristic of intrinsic motivation explained in section 2.4. However, the study of this construct in social media e.g. Lee & Hong, 2016;Tafesse, 2015;Syrdal & Briggs, 2018 mostly focuses on social media engagement metrics of output behaviour of share, like or comment. So, the construct of perceived creativity plays a role of another underlying subjective behaviour that creates the process of consumer brand engagement in this study.

Consumer Brand Engagement
According to the English dictionary, "engagement" refers to the period between a marriage proposal and the actual wedding, or partners entering into a contract (Stevenson, 2010). The first concept defined was used in the academic field of organizational behavior to refer to "employee or job or personal engagement" which focused on investing one"s intrinsic physical, cognitive and emotional values, when they attached these intrinsic values able to increase motivation to their job or company (Khan, 1990;Rich, Lepine & Crawford, 2010;Schaufeli, 2013).This concept also started to blossom in other fields e.g. the field of education psychology named it "student engagement" to refer to how students think, and their interests and behaviour at school (Fredricks, Blumnfeld & Paris, 2004).
In the marketing field, the term was basically derived from exchange schools that were based on generic or social exchange (Shaw & Jones, 2005) during the end stage of the service marketing management approach (Brodie et al. 2011;Sheth & Parvatiyar, 1995a;Sheth & Parvatiyar, 1995b;Javornik & Mamdeli, 2013), with marketers agreeing that services are intangible and required exchange beyond profit to include exchange of resources of belief, feeling and energy of both parties, this led to the emergence of customer-seller relationships (Kotler, 1972;Palmatier et al., 2017) with the concept named "co-production or relational engagement" (Heide & John, 1990). Then later, Brodie et al. (2011) named it "consumer engagement" which was defined as the psychological state that occurs with different consumer engagement levels of cognitive, emotional and behavioral in a context-dependent condition based on service dominant logic perspective. Hollebeek (2011b) named it "customer brand engagement" which was defined as the level of the customer motivation by specific levels of cognitive , emotional and behavior activities during brand interactions from the social exchange perspective.
It was then the technological revolution gave rise to the full focus of relationship marketing (Palmatier et al., 2017) due to new possibilities for customers" empowerment and activities (Javornik & Mamdeli, 2013). This word started to be actively used during internet era focus to have customers actively brand relationship is to have customers actively engage with frequently using the medium (Mollen & Wilson, 2010). Then, when expressive social media arose, the first focus was on customer engagement behaviour with this concept defined as customer behavioural manifestations such as blogging, word of mouth, participation etc. resulting from motivational drivers (Van Doorm et al., 2010;Verhoef, Reinartz & Krafft, 2010;Jaakkola & Alexander, 2014). Picked up from the general context, Hollebeek, Glynn & Brodie (2014) was the first person to define engagement in the social media context as "consumer brand engagement" which defined it as a consumer positively valance of cognitive, emotional and behavioural during brand interaction. This was followed by Solem (2015) who defined consumer brand engagement concept as the customer"s motivational and positive state of mind characterized by behavioural intentions, emotional and cognitive during brand interactions.
From then on, consumer brand engagement, as agreed by marketing researchers, is the only significant concept used as a metric for brand engagement in all contexts (Gambetti & Graffigna, 2010) and most marketing researchers agreed that this concept was multidimensional consisted of mind, heart and hand in an expressive social media context (Hollebeek et al., 2014;Dessart, Veloutsou & Thomas,2015;Solem, 2015) and in a general object or context (Brodie et al., 2011;Hollebeek 2011a;Hollebeek 2011b;Dwivedi, 2015). The mind-dimension which is also named cognitive engagement is described as thought processing (Hollebeek et al., 2014;Solem, 2016), and absorption (Dwivedi, 2015), and interest (Hollebeek, 2011b) of the consumer with the brand. The heart-dimension which is also named affective/emotional engagement is described as positive emotion (Hollebeek 2011a;Solem, 2016;Dessart et al., 2015), or positive related affect (Hollebeek et al. 2014) with the brand. The hand-dimension which is also named behavioural/behavioural intention engagement is described as the energy and time spent with the brand (Solem, 2015;Dwivedi, 2015).
On the other hand, some have argued that engagement of conative behavior is the consequences of dimension of cognitive and affective engagement (Malthouse & Calder, 2011).This was proposed by Higgins (2006) where engagement is a motivational force of attraction. And recently, a researcher conducted a literature review that explained cognitive and affective as a kind of internal state response which was a motivational experience known as antecedent of engagement (Javornik & Mamdeli, 2013) especially in online context suggested by Mollen & Wilson (2010) who reviewed past literature of online context found that for example in a e-learning context stimulation promotes interest and desire to engage for continuous learning (Jones, 1998).
Based on the above, the roots of engagement are from extended relationship marketing (Brodie, Glynn & Durme,2002) which is a relational construct its" dimension consists of characteristic of intrinsic motivation which is the process of pleasure and interest and later behavior that engages in the given activities with brand (Vallerand, 1997;Ryan & Deci, 2000a;Ryan & Deci, 2000b;Moore & Diehl, 2019). However, there is lack of research study on the relationship of multidimensional consumer brand engagement to verify it as an intrinsic motivation (e.g. Hollebeek et al., 2014) to support that intrinsic motivation is able to increase user brand relationship activities (Ranasinghe & Samarasinghe, 2019). Most studies focus on suitable content with social media engagement metric as mentioned in section 2.2.1.Hence, the study of the relationship within the multidimensional consumer brand engagement i.e. "cognitive, affective, intentional" engagement enables us to capture the underlying process of behaviour (So et al., 2015).

Research Framework and Hypotheses
This research model consists of seven constructs as shown in figure 1.
According to literature, when an audience interfaces with an environment, interface preliminary information of the content will be presented to the audience and later will create an interface assessment or appraisal (Oh et al., 2015;Moors, Ellsworth, Scherer & Frijda, 2013).Additionally, the creative literature, which is defined as that creative process produced through combinatory play (Zhanetta, 2011;Rothenberg, 1986) and brand literature described as combination of tangible and intangible of a product are considered to be creative (Kampylis & Valtanen, 2011). Feeling theorists also believed that to have emotional projection we need to have subjective feeling of touch (Shouse, 2005;Whiting, 2009: Moors, 2009Evers, Stock & Ridder, 2010). Experiential content for this research consists of functional appeal, emotional appeal and vividness that are described as providing the users with experiences of pragmatic, feel and sense/aesthetic for the processing of individual preliminary information and then production of perceived creativity. Perceived creativity, the short-lived neutral emotion describes the assessment or appraisals of the experiential content with three hypotheses were proposed: H1: Functional appeal is positively related to perceived creativity H2: Emotional appeal is positively related to perceived creativity H3: Vividness is positively related to perceived creativity Then, the interface assessment or appraisal shaped the process of engagement which begun with the intrinsic state of motivational behaviour of curiosity and interest (Oh et al., 2015;Moors et al., 2013). In the tourism, organization behaviour and artificial system literature supported that novelty behaviour plays an important role in building intrinsic motivation (Grant & Berry, 2011;Toyama & Yamada, 2012;Barto, Miroli & Baldassame, 2013). From the psychology literature, novelty can keep one motivated with intrinsic rewards of emotion and thinking (Csikszentmihalyi & Asakawa, 2016).Some explained this novelty behaviour as a neutral emotion is valance by emotions (Mellers, et al., 2012;Lindgreen & Vanhamme, 2005;Foster & Keane, 2015) which can be a good or bad (Lindgreen & Vanhamme, 2005) which is the reflection of novelty behaviour. And some say that novelty behaviour can elicit an instinct of curiosity (Reisensein, 2000;Eelen, Rauwers, Wottrich, Voorveld & Noort, 2016) of exploratory, interest and imagination (Reio & Choi, 2004;Kidd & Hayden, 2015) which some named "knowledge emotion" (Silvia, 2005;Kashdan, Rosse & Fincham, 2004). For this research, perceived creativity as the assessment or appraisal of the experiential content will shape the process of intrinsic state of engagement of cognitive engagement the curiosity behaviour and affective engagement the positive emotion and the two hypotheses were proposed: H4: Perceived creativity is positively related to cognitive engagement H5: Perceived creativity is positively related to affective engagement According to organizational behavior and psychology literature, the intrinsic motivation construct consists of characteristics of a curiosity-based behavior and pleasure of emotion that is accompanied by a willingness of action (Ryan & Deci, 2000a;Ryan & Deci, 2000b;Gagne, 2014). The traditional media/website literature supported the process of cognitive and affective behaviors an internal response in reaching a person"s response (Calder, Isaac & Malthouse, 2016;Mollen & Wilson, 2010). From the brain literature, the output behaviour response happens when a person is curious, and euphoria can increase dopamine in the brain which can heightened energy (Regan, 2011;Ostroff, 2016).The marketing literature also found that curiosity leads to behavior intention (Driessche, Vermeir, Pandelaere, 2013) and a brand passion promotes loyalty (Pouraza, Pare & Saniee, 2015). For this study, the behavioral intention engagement is the willingness to click on the social media button of like, share and comment. With that, the cognitive engagement and affective engagement can shape intention engagement with hypothesis proposed:

Research Method
This research used a descriptive-explanatory approach to describe the baseline characteristics of a targeted population study and to study the relationships between a set of different variables (Saunders et al., 2009) with a one shot treatment (Malhotra, 2010). This one shot treatment has been widely used by marketing researchers to understand how a context or the evaluation of an object lead to user further responses (Vaughan, Beal & Romaniuk, 2016;Taylor, Lewin & Strutton, 2011), e.g. a retail experience (Chang, Eckman & Yan, 2013), a Facebook brand still image as an advertisement response (Lee & Hong, 2016) and a hotel social media site to evaluate user response (Leung, Bai & Sathura, 2013). This is in accordance with the research focus for this study that aims to provide brand content experience assessed as creative that shapes the process of consumer brand engagement.

Sampling
Malaysia, with total population of 32.8 million (Statista, 2019a) has a 78% social media penetration rate. Its penetration rate is ranked among the top global 4 (Kemp, 2019a) and Malaysian social media users are predicted to increase from 24.1 million in 2019 to 26.1 million by 2023 (Statista, 2019b).The sample used was 25-34 year old Malaysian active expressive social media societies who accessed expressive social media sites for at least 3 hours daily. This age group selected aim to represent person who were living in era of advancement in technologies (Ting, Lim, DeRun,Koh & Sahdan, 2018) with the heaviest usage of expressive social media (Kemp, 2019b) and internet (Google Barometer, 2018).The population selected were also representative of the largest consumer segment (Dawn & Thomas, 2013). It was therefore anticipated that the economy would later experience a shift in spending power to this age group over the next two decades (Nielsen, 2015;Ting et al., 2018).
The sample size used was based on 10 to 20 participants per estimated parameter as the sufficient sample for the statistical power analysis (Kline, 2016). So, for this research, (49 parameters x 15 participants) giving us a sample size of 735 participants. We however, recruited 750 participants to enhance the precision of the sampling outcome (Malhotra, Kim & Patil, 2006).
The sampling technique was based on non-probability method because the sampling frame was challenging to define due to the unstructured nature of internet (Fricker, 2008) leads to an undefined population list of the online social network (Pina-Gracia& Gu, 2013; Pina-Garcia, Gershenson & Siqueiros-Gracia, 2016).So, a purposive sampling was used.

Measurements
Primary data was obtained using a two-part survey questionnaire.
The first part was used for pre-screening and to give an insight into the demographic characteristics and general behavior of the respondents. It had a total of 6 close-ended questions (i.e. age, gender, marital status, citizenship, type of expressive social site accessed each day, number of hours spent on these expressive social media sites daily) in a nominal scale.
In the second part, a stimulus was used as a treatment and 21 close-ended questions each with responses rated on 7-point Likert scale of extremely disagree to extremely agree, were used to capture the perceptions of the respondents. The treatment used was a still image from a Pinterest posting of an airline company named Cathay Pacific image as shown in appendix. The stimulus chosen for this study was measurable in each dimension of experiential content and creative criteria. First, the stimulus used a complementary colour i.e. 2 colours opposite the colour wheel i.e. blue (cold colour) and brown (warm colour) and created a high contrast in the still image, enlarged the number "5" on the center of the still image create an impact on the sense (Hauff, 2018;Bakar, Desa & Mustafa, 2015) and the "text" in white colour to balance the vibrance of rich colours used in the still image (Imtza, 2016). Next, for the feel experience, brown colour indicates security and protection for the psychology of reliability, and blue colour is a social sensory colour favourable to all age groups and sexes indicate provides feeling of trust (Imtza, 2016;Rider, 2009;Donovan, Agarwala & Hertzmann, 2011). The text & image clearly stated "accessible five times a day" and provided flight-timing on the right-hand side corner which was pragmatic evidence of 5 daily flights.
For the creative criteria, this still image selected, had been identified as the best airline online campaign in an internet advertising competition based on its advertisement creativity, design and copywriting from expert judges (Internet Advertising Competition [IAC], 2016). The airline was the chosen brand stimuli because it had been praised for its relative fluency in applying expressive social media as platform to improve customer experience and building relationships makings it a leader in social care (Social Bakers Report, 2014; Koch & Tritscher, 2017;NIIT Technologies, 2013).
The measurement items were adapted from related prior studies to align with the context of expressive social media. Experiential content consisted of three components i.e. functional appeal, emotional appeal and vividness. Functional appeal referred to the degree of information in the still image perceives as usable. This scale adapted from Taylor et al. (2011) with reliability of 0.87 had used as an informative scale in a general expressive social media context referenced by Cachero-Martinez & Vaquez-Casielles (2017) pragmatic experience scale from the retail context that had a reliability >0.7 for slight correction of items to suit the usable experience. Emotional appeal referred to the extent to which users are able to have a feeling when perceiving the intangible of the brand content from a still image; the scale was adapted from Delgado-Ballester & Sabiote (2015) mixture of emotions and feelings scale that had reliability 0.94 and referenced of Escalas & Stern"s (2003) feelings scale of a TV commercial that had a reliability 0.95 for slightly correction of items to suit the "feel" experience. Finally, vividness referred to the degree to which a user was able to have a keen sight of the beauty of the content. This was adapted from Delgado-Ballester & Sabiote"s (2015) scale of sensory brand experience that had a reliability 0.88. Perceived creativity construct is the degree to which individual evaluation of an experience produces novelty emotions; it has a scale of 0.903 and was adapted from Lee & Hong"s (2016) advertising creativity scale by dropping one of the items due to the item "intriguing" belongs to curiosity behaviour.
Next, consumer brand engagement consisted of three dimensions with cognitive engagement referred to the degree to which an individual produced in depth instinct of curiosity and was adapted from Chang, Tseng, Liang & Yan"s(2013) scale which measures curiosity behaviour in English learning and has a reliability of 0.90 and Hollebeek et al."s (2014) scale of cognitive processing in a LinkedIn site that has a reliability of 0.878 with item wording slight correction, affective engagement is the degree to which produces affectionate emotions and this was adapted from Hollebeek et al."s (2014) scale that measures affect emotion in a LinkedIn site with a reliability of 0.830 and intentional engagement refers to activation of users brand activities and was adapted from the scale of Solem (2016) that measures spending of energy on a brand on Facebook and has a reliability of >0.70.
Using adapted scales with slight correction of item-wording to suit the construct and context had to have their validity retested for its psychometrics properties (Heggestad et al.2019).Therefore, scale content validity was conducted by 5 expert panels (Shrotryia & Dhanda, 2019;Zamanzadeh et al., 2015) and pilot test conducted on 58 respondents (Viechtbauer et al. 2015).Additionally, exploratory factor analysis was conducted for parsimony to identify which item underlay which construct (Hair et al., 2006) and confirmatory factor analysis was performed to confirm the validity of the scale based on convergent and discriminant validity (Hair et al. 2014). Table 2 shows the finalized items of the scale.

Data Collection
An online self-administrated survey was conducted using Google Form tools. A link was sent to the respondents through expressive social media messaging sites to enable them to conveniently respond via a mobile telephone and laptop. Respondents completed the survey by clicking a mouse or keyboard to answer the questions that best represented their responses.

Pilot Test
From the results of the pilot test, the Cronbach Alpha for the scale of FA was 0.953, EA was 0.965, V was 0.929, C was 0.903, CE was 0.929, AE was 0.934 and IE was 0.928 which showed the consistency of the items in each scale. The respondents rated the clarity of the measurement items on a 4-point Likert scale (from very unclear, unclear, clear, to very clear), and 56 respondents rated the questions as clear and very clear with only 2 respondents rated unclear. With that, indicating more than half of the respondents found the questions clear and therefore usable for the main survey.

Pre-Test Analysis
A total of 750 questionnaires were sent out, 740 usable responses were returned (response rate of 98.6%), hence, non-response bias was not conducted (Roni, 2015). To minimize any potential common method variance in the data obtained from the self-reported survey with all the questions were collected in the same period of time using same method with the same sample (Podsakoff, MacKenzie & Podsakoff, 2003), the common method variance test was performed to assess for common method bias using Herman single factor test in IBM SPSS with a result of 42.482% which is less than 50% indicating no bias was present. To enhance the validity of the results, we further conducted a common latent factor method in IBM AMOS and the results showed that delta values ranging from -0.064 to 0.035 which was less than 0.5 exhibited no bias (Levin, Quach & Thaichon, 2019;Eichhorn, 2014).This indicated that the significance of the substantive relationship was not affected .

Demographic Profile
The respondents (25-34 years old Malaysian active expressive social media users) were mostly males and single. The majorities accessed expressive social media sites for 3 to 4 hours a day and were within the age group of 31-32 year-old. The expressive social media sites that was accessed daily by users was Facebook and most respondents logged in at least one account daily and with a minority logging in 4 accounts daily.

Measurement Model
Reliability Analysis was conducted using IBM SPSS. The internal consistency of the scale was established using Cronbach"s alpha measure (Reynaldo & Santos, 1999). Exploratory factor analysis was conducted using IBM SPSS to explore the number of factors with high loads of items (Reio & Shuck, 2014).This aimed to assist with the validity check of the scale for a cleaner measurement model for the purpose of confirmatory factor analysis (Watsons, 2017;Reio & Shuck, 2014;Hair et al., 2014). The values of these measures are given in Table 2 Cronbach Alpha measures in this study were all above 0.70 indicating that the consistency of the set of items consistently loaded onto the same factor (Hair et al. 2003) and factor loadings above 0.6 indicate that no items need to be deleted (Hair et al., 2006).   (Hu & Bentler, 1999).
Then, the measurement model was assessed for its composite reliability and construct validity using IBM AMOS. Composite reliability [CR] values are shown in Table 3 and range from 0.714 to 0.850 with all exceeding 0.700, suggesting good reliability (Fornell & Larcker, 1981) and thus confirm that the measurement items consistently represented the same latent construct.
Construct validity was tested using convergent validity, discriminant validity and nomological validity rules. Results in Table 3 shows that factor loadings of each item exceeded the recommended 0.50 threshold (Hair et al., 2006) and the averages variances extracted [AVE] were above 0.50 indicating that they exceeded the threshold (Hair et al., 2014) with an exception of the AVE of construct C that was 0.463 and therefore slightly below 0.50. The items in the unfulfilled of construct C were not deleted because there was no problem with measurement model fit and the CR for construct C was higher than 0.60 which met the rule of thumb of Fornell & Larcker (1981) and Chun, Wang, Wu & Wan (2013). These results indicated that most of the latent variables in this research accounted for more than half of the explained variance in each indicator. For, discriminant validity, AVE obtained for each construct was larger than the squared inter construct correlation [SIC] showed in Table 4, this results showed that the measures of a construct were not highly correlated with other constructs indicating they maintain its" distinctive from each other (Hair et al., 2014).Lastly, nomological validity was used to test if correlations of the construct were positively significant which indicates that these constructs are positively related and supports that the proposed model make sense (Hair et al., 2014).The results of these assessment showed that the measurement model is valid. were less than 0.85 with tolerance values of less than 1 and the variation inflation factor (VIF) values were within the range of 1.342 to 1.473, VIF <3 provides proves that multicollinearity problem does not exist (Ho, 2014;O"brien, 2007) and indicates that there is no overlap of information across these independent variables. These results are shown in Table 5.

Structural Model
In this stage, when two-headed arrows from confirmatory factor analysis to become single-headed arrow to form structural model, the model fit will decrease due to specifics of the relationship (Hair et al., 2014). The GoF test was used to test the structural model fit and the estimated relationship within unobservable variables in a structural theory was evaluated (Hair et al., 2014).  (Hu & Bentler, 1999) and [RMSEA<0.08]. Hence, the result showed that the GFI indices were not in the acceptable range indicates that the estimated model doesn"t reproduced data and explains the data well (Kline, 2016 (Hu & Bentler,1999).TLI does not meet the requirement in its baseline model (Kilne, 2016). When there were model fit issues, structural model loadings were checked and it was found out that C3 measurement items had a low factor loading of 0.398 led to the model fit issues. Therefore, C3 measurements items were deleted to improve the quality of the latent construct. After C3 deletion, the structural model improved as shown in figure 2. The new structural model was subsequently assessed for its model fit using the GoF test.  (Kline, 2016).

Figure 2. New Structural Model
Within the rules of structural equation modelling, residual error terms should be includes in any endogenous variables (Byrne, 2010).This was named square multiple correlation (R 2 ) for the purposes of providing an indication of the proportional (%) of contribution of an exogenous construct in estimating endogenous construct. R 2 values of 0.75, 0.50, or 0.25 for endogenous latent variables could be described as substantial, moderate, or weak respectively specifically in marketing research (Hair, Ringle & Sarstedt, 2011 Finally, the hypotheses of the path was tested; hypothetical model path relationships were all statistically significant at 0.001 level and strength of the relationship was above 0.02 varies from large, medium and small which provided clues that relationship strength was stable (Aiken & West, 1991). From the values shown in Table 6, of all the paths in the improved structural model, the relationship of C->CE and C->AE was the most significantly impactful of all relationships. This means a novelty reaction can certainly produce curiosity and positive emotion behaviour in a human being. And, based on the evaluation of brand content experience as creative by an individual, displays that EA->C had the most significant strength when compared to FA->C and V->C which implies that a content with feel experience has the most important role in creating novelty reaction whereas pragmatic experience and vividness plays a minor role in creating novelty reaction.

Discussions & Implications
In the context of social networking sites, with a focus on consumer lack of engagement in brand activities, this research proposed that identifying brand content experiences perceived by user as creative would initiate the process of consumer brand engagement. The study showed that all the seven hypotheses were proven indicating that all the seven relationships within the constructs were supported. The summary of the GoF results also indicated the empirical results were consistent with the proposed research model. This means that active expressive social media users indeed behave like postmodern users more focused on subjective elements which in turn kick-starts an engagement relationship with a brand.
In this aspect, whereas the study of the content marketing framework in social networking sites still in exploratory phase, with this empirical study using experience dimension showed that results of the proposed H1,H2 and H3 had positive significant relationships with perceived creativity. This helped to contribute to a great extent to the theoretical foundation which constitutes of the components of functional appeal, emotional appeal and vividness that are the basic passive experience components that form a brand content experience and produce novelty emotion which is perceived creativity. Therefore, this confirms that to create passive engaging experience, the content components need to consist of functional appeal, emotional appeal and vividness which all play an important role in helping to fill the gap of identifying a suitable content marketing framework needed in brand posting.
Following this, results of H4 and H5 showed positive significant strong relationships that imply the active experience of curiosity and positive emotions of the expressive social media societies is elicited from novelty emotion. This confirms that novelty emotions are an important appraisal of the passive experience from functional appeal, emotional appeal and vividness that convert passive experience to active experience (curiosity and positive emotion). Because previous studies on content marketing expressive social media literature are mostly focused on studies of novelty (eg: Tafesse, 2015;Syrdal & Briggs, 2018) with standard social media engagement metrics (share, like, comment) without studying the internal processes that provides the output behaviour. With the unpredictable behaviour of current societies (Batra & Kazmi, 2008), a strong positive relationship of H4 & H5 provide insight into marketing literature of what factors underlie the social media engagement metrics, the factors i.e. novelty emotion, curiosity and positive emotion which form the internal and intrinsic state that play utmost importance in contextualizing social media engagement metrics. Based on this, the research studies should more focus on the internal and intrinsic states of the user.
Finally H6 and H7 relationships in the research model, supported by the literature of psychology and organization behavior and results of H6 and H7 showed positive significant relationship and confirmed that multidimensional consumer brand engagement is an intrinsic motivation relation construct. This helped to enrich the concept of multidimensional consumer brand engagement and confirmed that multidimensional consumer brand engagement is an intrinsic psychological state of mind (i.e. cognitive engagement and affective engagement) that help stabilize the brand relationship, create frequent intentional engagement through clicking social media buttons of like, share and comment. This provides new knowledge to the literature i.e. the focus in social media marketing investment should not be focus on engagement rates. The results showed that the cognitive engagement and affective engagement are the ones that play a role in stabilizing social media engagement relationship to increase the current proportion of the expressive social media societies clicking and only in future, in enhancing probability of users becoming follower thereby building a long term consumer brand engagement kind of relationship For the industry wise, this model is able to help a marketing communication manager to visualize a solution for a brand to break into the expressive social media environment so as to generate exposure of their brand posting by emphasizing strategies of creating content using a combination of pragmatic experience, feel experience and sense experience. Moreover, the research findings of H2 showed that it had strongest significant relationship when compared to H1 and H2 accorded to the industry player approaches to create content with more emotional appeal which then only followed by pragmatic and sense experience for an engaging creative content. Also, this study provided knowledge to industry players on how to bring the consumer closer to its nutshell for engagements activities by focusing on creating activities that stimulate users" intrinsic state engagement so that the users can develop an association with the brand during the consumption process opposed to focusing on engagement rate.

Limitation & Recommendations
Although these research findings provide additional knowledge, but there were some limitations. Firstly, this was a cross-sectional study, therefore the research findings cannot explain the behavior of the respondents over of time. The adoption of purposive sampling method in this research also limits in generalization of study findings and making inference.
Future research may longitude this study and use probability sampling methods. This may also employ more than one or alternate brand still images or short videos meeting the criteria of experiential content as stimuli to test this research model. The research model can also be tested on a population from different age-groups of active expressive social media users and non-active expressive social media users.