Film or Film Brand? UK Consumers’ Engagement with Films as Brands

Many films are produced annually, but only a small number of films reach the state of being considered and identified by consumers as film brands. Film-brand identification is difficult to achieve, but it leads to engagement behaviours (e.g. repetitive viewing, positive word-of-mouth and purchase intention of relevant merchandise/franchise). To help filmmakers better develop films as brands and benefit from their brand status, this paper takes a consumer-centric approach to investigate how and why films are identified and engaged by consumers as brands. Using an abductive mode of reasoning, a consumer film-brand engagement framework was developed through qualitative data collected from 35 semistructured interviews and then validated using survey data with 1030 participants. This consumer film-brand engagement framework shows that film identity coherency drives film-brand identification through the mediation effects of popularity, sequels and emotional bonding, whilst marketing effort, iconic status, franchising/merchandising activities and timelessness are highlighted as key moderators, resulting in positive brand engagement behaviour. The paper sheds new light on film-branding literature by theoretically explaining and empirically showing a sequential and consolidated process, which consumers go through to identify and engage with films as brands, leading to several managerial and marketing implications for film-makers.


Introduction
The film sector is lucrative -in 2017, film market revenues exceeded £3.4 billion in the UK alone (British Film Institute, 2018). Nevertheless, the market is saturated and competitive (Gong, van der Stede and Young, 2011). Although many films are developed by film-makers as brands through a formula combination of various brand elements that consumers find appealing (Keller, 2008), most films are merely consumed as films, failing to be identified or engaged by consumers as brands.
Film-brand identification refers to the state/ condition in which a consumer considers a film as a brand, characterized by a consumer's recognition of a brand's symbolic meanings and the brand's ability to express and enhance a consumer's identity (Kim, Han and Park, 2001). It leads to posi-tive brand engagement behaviours, such as brand loyalty (Kim, Han and Park, 2001), increased buying intention (Ahearne, Bhattacharya and Gruen, 2005) and positive word-of-mouth (Tuškej, Golob and Podnar, 2013), all of which are likely to contribute towards favourable box office results and downstream revenues (Young et al., 2008).
Whilst the benefits of developing not only films but also film-brands are acknowledged, film branding is still a relatively new and underdeveloped field. Existing literature is predominantly conceptual and tends to discuss producing films as brands from filmmakers' and marketers' perspectives. For example, using production house logos, product placements and industry politics in developing branded texts (Grainge, 2007), employing culturally embedded brand entities within a film project into a brandscape (O'Reilly and 2 G. S. Kohli et al. Kerrigan, 2013) and increasing a film's marketability and playability that may affect consumers' film consumption journey (Gillespie, 2005;Hart, Kerrigan and vom Lehn, 2016;Kerrigan, 2010).
These studies overlook the fact that a brand's reality lies in the experience of its consumers (Topalian, 2003), consumers' brand identification (Tuškej, Golob and Podnar, 2013) and consumers' brand engagement (Gambetti, Graffigna and Biraghi, 2012). This purely managerial perspective on branding fails to help film-makers resolve their struggles in developing films that are identified and engaged by consumers as brands. Subsequently they lose out on user-generated wordof-mouth, repetitive viewing from fans, selling various merchandise goods and franchise opportunities and generating a dedicated fan base that follows and grows with the film brand over time.
To address this knowledge gap, this paper makes an initial attempt to explore how and why consumers would identify and engage with certain films as brands, rather than just consuming them as films. Taking on a consumer-centric approach to branding (Maney, Flink and Lietz, 2002;Wright, Stone and Abbott, 2002), two research objectives were identified: (1) to explore how and why consumers would identify certain films as brands, leading to the development of a conceptual framework (study 1); (2) to develop and validate a consumer film-brand engagement framework that elucidates the antecedents, outcomes and facilitators (e.g. mediators, moderators) of film-brand identification through a sequential process, illustrating how consumers engage with a film as a brand (study 2). This paper adopts a mixed-methods approach in order to gain a more elaborate understanding of the research problem, to triangulate the data sets so that they inform and clarify each other (Tashakkori and Teddlie, 1998) and extend the breadth and range of enquiry to accomplish the research objectives. Adopting an abductive mode of reasoning, it started with a qualitative study of 35 semi-structured interviews (study 1), followed by a quantitative survey of 1030 participants (study 2), to address the two research objectives.
By conceptually developing and empirically validating a novel consumer film-brand engagement framework, this paper contributes to filmbranding literature in several ways. In particular, by illustrating how consumers would engage with films as brands in a sequential process through identity coherency and film-brand identification to film-brand engagement, this paper explains how some films could be identified as brands and are being engaged by consumers as such, illuminating the consumer-centric route to film branding. Furthermore, by showing a film's popularity, sequels and emotional bond with consumers as mediators between film identity coherency and filmbrand identification, and proving the moderating effects of marketing efforts, timelessness, iconic status and franchise/merchandise, it highlights all the concepts that are relevant in consumers' engagement with films as brands. For film-makers and film marketers, this paper also provides effective brand engagement strategies in which consumers actively participate.

Film consumption journey and film branding
Consumers' film consumption journey can be classified into three stages: before, during and after viewing. Before committing to viewing a film, consumers' choice is dictated by a film's marketability and playability, communicated by the film's marketing efforts (Kerrigan, 2010;O'Reilly and Kerrigan, 2013). A marketable film has a unique selling point (in terms of differentiation), for example, an appealing combination of cast and crew, which enables the film to be promoted in a compelling manner (Kerrigan, 2010). Marketability could be influenced by the prominence of an actor's role (Albert, 1998) and a film's genre (Eliashberg, Hui and Zhang, 2014), age rating (Leenders and Eliashberg, 2011) and release window (Young et al., 2008). Marketability has a direct impact on a film's box office performance (Kerrigan, 2010). Consumers may consider a film's review (Hennig-Thurau, Marchand and Hiller, 2012), awards (Addis and Holbrook, 2010) and recommendations from their reference groups (Hennig-Thurau, Wiertz and Feldhaus, 2014) when evaluating a film's playability.
Recent research suggests that consumers' perception, selection and evaluation of films involves a nuanced process (Hart, Kerrigan and vom Lehn, 2016). Consumers may subjectively classify films based on their anticipated film experience, to screen out irrelevant 'noise', going beyond simplistic marketability and playability factors. Cooper-Martin (1991) found that consumers consider abstract global features of a film (e.g. awesome, moving, thoughtful, etc.), which refer to the entire 'product' rather than specific aspects. Hart, Kerrigan and vom Lehn (2016) found consumers' classification of a film to be based on a film's characteristics, the viewing environment, the time of the day, season, etc. Consumers consider these interrelated dimensions holistically and on a cumulative basis, simultaneously earning 'cultural capital', whilst retrospectively informing their future film selection (Hart, Kerrigan and vom Lehn, 2016). Nonetheless, when consumers struggle to comprehend what a specific film is about, they may become confused as to how to classify it and whether ultimately to watch it or not, which may be due to an incoherent film identity (Zhao, Ishihara and Lounsbury, 2013).
During actual film consumption, consumers' personal engagement with a film's narrative enhances cognitive and emotional immersion into a film's imaginary world, which is a crucial requisite for consumers' repeated enjoyment of a film (Batat and Wohlfeil, 2009). Such high involvement is possible because films are experiential and hedonistic (Bassi, 2010;Schmitt, 1999), thus consumers' evaluations of films are highly linked to their end goals and values (Blythe, 1997). They consume films holistically, based on the potential value arising from each film (Bassi, 2010;Mano and Oliver, 1993). For consumers, films' values consist of personal and social/community-driven dimensions: films allow escapism from worldly affairs; they aid consumers' identity building; offer assistance during rites of passage; inspire, educate, spread enjoyment and bring people together, thus widening network clusters (British Film Institute, 2011).
After viewing a film, consumers may consider film-related purchases and follow the film's premise onto different platforms (e.g. TV serials, novels and video games), due to their familiarity and fascination with the storyline and characters, which encourage them to explore deeply other related facets (Marshall, 2002). Consumers may also 'partake' in the process of co-producing works of art, like short films, as part of a cooperative network (Becker, 1982), made possible through communal relationships with film-makers. Such relationships result in global, passionate, fan-based communities, which are often formally institutionalized by film marketers at some point, to actively involve fans, thus amplifying a film's success and popularity (Muniz and O'Guinn, 2001). For example, May 4th is widely publicized as the official Star Wars day, on which consumers from all over the world reminisce about the franchise and indulge in cosplay events. Figure 1 summarizes consumers' experiences of the key stages in their film consumption journey, wherein consumers continuously and actively decode films and their relevant materials and myths before, during and after their film consumption journey. As part of the audience reception theory (Hall, 1980), when decoding, consumers can deviate from meanings and positions encoded by film-makers, refusing these altogether or negotiating them, whilst interacting with media texts as social subjects. The concept of spectatorship (Mayne, 1993) also refers to how film consumption and its associated myths are considered to be culturally significant and activities of a symbolic nature for consumers, who are found to have agency over media texts such as films (Hayward, 2018).
Film branding is a relatively new area, despite the sheer size of the film market. The film-branding literature is relatively unexplored and largely conceptual. It tends to explore how films could be conceived as brands from the perspectives of studios and marketers, and how branded entities within a film individually contribute towards the view of a film as a brand (Grainge, 2007;Kapferer, 2008;Keller, 2008). O'Reilly and Kerrigan (2013) explain how the ambiguity about a film brand's ownership, such as who has the authority to speak about and promote it -in a multi-stakeholder scenario, is a hindrance to the development of film-brand theories. Grainge (2007) labels film studios as corporate brands and details the internal film branding process via complex corporate structures. He also describes the commercialization of blockbuster franchises from the studios' and marketers' perspectives. Grainge (2007) is credited with exploring Hollywood's industry culture, the role of production house logos, product placements and industry politics in developing branded texts. Kapferer (2008) suggests that film brands cannot exist without a supporting product/service that embodies the brand, citing films made by Disney as successful examples. However, Keller (2008) argues that all films are brands at a basic level, and that some films can establish themselves as strong brands (e.g. Batman,  Austin Powers and Harry Potter) by combining various elements such as name, logo, symbol, brand characters, slogan, jingle, signage, etc. into a formula that consumers find appealing.
Rather than exploring how individual films could be branded, O'Reilly and Kerrigan (2013) investigate how branded entities within a film project are embedded within the broader cultural fabric, using the brandscape lens. Their conceptual framework is primarily applicable to mainstream, commercial films. Acknowledging consumers' contribution and involvement, O'Reilly and Kerrigan (2013) seek to explain how consumers understand individual brands in relation to one another, using an illustrative case study of the James Bond franchise. Preece, Kerrigan and O'Reilly (2018) apply the assemblage theory to explore the concept of brand longevity, focusing on consumers' engagement with serial brands over time. Again using the James Bond film brand, Preece, Kerrigan and O'Reilly (2018) explain how serial brands achieve longevity in evolving socio-cultural contexts, through social salience and ongoing consumer engagement to achieve brand longevity.
Overall, previous works predominately discuss films as brands from the perspective of film-makers and marketers, based on a handful of successful film brands as examples or taking a macro-level cultural approach to film branding. They have not taken account of consumer identification or engagement in the discussion of films as brands, thus failing to account for consumers' emotional and cognitive factors in explaining their identification and engagement with only certain films as brands. This ignorance highlights the problem that not all films produced are identified or engaged by consumers as brands; instead many are merely consumed as films.

Consumer-centric approach to branding
Brands can be developed and managed using different approaches: economic (Borden, 1964), identity (Kotler, 1997), consumer-based (Keller, 1993), personality (Aaker, 1997), relational (Fournier, 1998), community (Muniz and O'Guinn, 2001) and cultural (Holt, 2004;Kates, 2004). Heding, Knudtzen and Bjerre (2008) offer a detailed account of the seven approaches, comparing their benefits and drawbacks in brand management. Derived from the consumer-based and relational approaches, the consumer-centric approach is acknowledged by Maney, Flink and Lietz (2002) and Wright, Stone and Abbott (2002) to enable a micro-level focus on the consumer, in the context of a dyadic, one-to-one, consumer-brand relationship. It focuses on capturing and deploying consumer insights, in order to effectively market and better serve a brand's most promising prospects. This consumer-centric approach takes into account the consumer's cognitive and affective systems, without losing focus of the brand itself (Heding, Knudtzen and Bjerre, 2008), hence it is more versatile in applying to a variety of conceptual brand management tools (Grover, Halinski and Duxbury, 2016;Walsh et al., 2009).
Consumer engagement advances research on the consumer-brand relationship, as it describes the intensity of consumer participation and connection with a focal agent/object/activity in the relationship (Brodie et al., 2011;Vivek, Beatty and Morgan, 2012). Levels of engagement are determined by the consumer's overt behaviour (Kumar, 2013). In 2013, a Gallup study found that 'fully engaged' and 'engaged' customers lead to an increase of 23% and 7% in revenue; whilst groups of customers 'not engaged' and 'actively disengaged' accounted for falls in revenue of 1% and 13%, respectively (Kumar and Pansari, 2016). Consumer engagement is multi-dimensional, including: affect and attention (Brodie et al., 2011); enthusiasm (Calder, Isaac and Malthouse, 2013); enjoyment (Mollen and Wilson, 2010); behaviour (Gummerus et al., 2012); sharing (Van Doorn et al., 2010); learning and endorsing (Dessart, Veloutsou and Morgan-Thomas, 2016); cognition (Brodie et al., 2013); and absorption (Vivek, Beatty and Morgan, 2012). Nevertheless, the existing literature has not reached an agreement on the validity and reliability of consumer engagement's measurement scales; or on adopting multiple manifestations of engagement that stem beyond behavioural metrics (Dessart, Veloutsou and Morgan-Thomas, 2016). Hence, research is encouraged to discuss customerfocused engagement strategies, by focusing on understanding elements of consumer engagement from customers' perspectives (Vivek, Beatty and Morgan, 2012).
Engagement is inherently interactive and social (Dessart, Veloutsou and Morgan-Thomas, 2016); consumers' engagement varies in context -consumers may choose to engage with a variety of focal agent/object/activity concurrently but differ-ently (Gambetti and Graffigna, 2010). Therefore, research is required to understand the complete process of consumer engagement in a specific context, including the associated antecedents and consequences (Pansari and Kumar, 2017), in order to develop better managed consumer-brand relationships.
This paper focuses on consumer-brand engagement, which refers to positive consumerbrand interactions, of a cognitive, emotional and behavioural nature (Hollebeek, Glynn and Brodie, 2014). Previous consumer-brand engagement studies discuss how brands are engaged cognitively by consumers as part of their self-concepts (Sprott, Czellar and Spangenberg, 2009), to what extent affective advertising components elicit consumer engagement through feelings (Heath, 2007(Heath, , 2009) and behavioural manifestations of consumer-brand engagement beyond the initial purchase (Van Doorn et al., 2010). Consumer involvement, consumer participation and brand expressiveness drive consumer-brand engagement (Hollebeek, Glynn and Brodie, 2014;Leckie, Nyadzayo and Johnson, 2016), whilst consumerbrand engagement leads to increasing usage intent, self-brand connection, brand loyalty and brand equity (Bowden, 2009;Hollebeek, Glynn and Brodie, 2014;Leckie, Nyadzayo and Johnson, 2016).
Besides consumer-brand engagement, consumer-brand identification also plays a key role in consumer-brand relationships. Conceptualized as the extent to which a brand is able to express and enhance a consumer's identity through its symbolic meanings, consumer brand identification refers to the state/condition when a brand is being acknowledged and recognized by the consumer for its brand status (Kim, Han and Park, 2001). The sense of congruent sameness often plays a dominating role in brand identification (Bhattacharya and Sen, 2003;Tildesley and Coote, 2009) because of the overlap between a consumer's self-schema and the schema they hold for a brand (Carlson, Suter and Brown, 2008). Value congruity (Bhattacharya and Sen, 2003), perceived quality, self-brand congruity and innate innovativeness (Lam et al., 2013) drive consumerbrand identification, and lead to outcomes such as positive word-of-mouth and brand commitment (Tuškej, Golob and Podnar, 2013), brand preference (Tildesley and Coote, 2009) and brand loyalty (Kim, Han and Park, 2001). Nevertheless, existing 6 G. S. Kohli et al.  Grainge's work is credited with highlighting that branded texts such as films are developed via Hollywood's industry culture/politics, production house logos, product placements and that branding serves as a means of adding value and identity to a film as a commodity -Grainge does not specify exactly what a film brand is and how it is conceptualized -His work has also been criticized for underplaying the role of the consumer in the film-branding process -His work is also conceptual and has not been empirically tested Keller (2008) All films are brands once they are produced, from the marketer's perspective. Some film brands are stronger because they combine various branding elements (e.g. name, logo, signage, symbols, etc.) into a formula that consumers find appealing Keller's work is credited with acknowledging that films can exist as brands of varying strength and specifying some of a film brand's key brand elements -Keller does not state how exactly a film brand's various brand elements could be combined into a successful formula to appeal to consumers -He also assumes that all films will engage with consumers as brands, even though films are hedonic experiential goods, and some may therefore inherently just be consumed as mere films -His work is also conceptual and has not been empirically tested Kapferer (2008) Film brands cannot exist without a supporting product/service that embodies the brand Kapferer is credited with emphasizing the importance of a brand's tangibility, in order for consumers to better realize a brand's existence, from the marketer's perspective -Kapferer has not taken into account a brand's increasingly intrinsic intangible values -His work is also conceptual and has not been empirically tested O'Reilly and Kerrigan (2013) There are several branded entities within a film project (e.g. studio brand, character brand). These are embedded within the broader cultural fabric, using the brandscape lens, contributing towards the view of a film as a brand They have credited the consumers' role in the film-branding process by acknowledging them as co-creators of meaning as part of the brandscape -Their work is conceptual and has not been empirically tested -They have failed to consider that films are consumed holistically and that consumers do not consider the sum of a film's attributes but the potential value arising from their combination (Continued)  Consumer brand identification denotes a consumer's sense of sameness with a specific brand's symbolic meanings The study empirically validates the significance of a consumer's identification as a mediator between the impact of value congruity on forming committed brand relationships that are meaningful, as part of a psychological process -Their non-probability snowball sampling approach suggests that individuals with larger personal networks are oversampled and that their findings cannot be generalized -Their study only took into account consumers' brand identification processes with their favourite brands and increasing item means and correlations between them, thus perceived oneness with the brand is more likely in the first place Consumer brand engagement Hollebeek, Glynn and Brodie (2014) Consumer brand engagement refers to a consumer's positive brand-related activities (of cognitive, emotional and behavioural nature) during focal consumer-brand interactions Study is credited with identifying three consumer-brand engagement dimensions (cognitive processing, affection and activation). The study has also empirically validated consumer-brand 'involvement' as an antecedent and 'brand usage intent' as well as 'self-brand connection' as consumer-brand engagement's consequences -The study is cross-sectional and therefore only provides a snapshot of consumers' engagement with a brand at a specific moment in time Graffigna and Gambetti (2015) Consumer-brand engagement is a dynamic process with progressive relational phases in which consumers perceive a brand to be engaging when it is emotionally lived as a 'life-mate' Using a grounded theory approach, their study is credited with identifying distinctive characteristics and phases of consumer-brand engagement, which are: friendship, intimacy and symbiosis -Since the study took on a qualitative approach, their relatively small sample size has restricted the generalization of findings -Their conceptual framework has not been empirically validated -The study focused on the participants' favourite product brands, hence it is not clear if their conceptual framework is applicable to other brand categories and types of brand, as well as brands which consumers have varying levels of admiration for (Continued) The study has identified 'consumer involvement', 'consumer participation' and 'self-expressive brand' (inner self) as antecedents and brand loyalty as a consequence of consumer-brand engagement -The study is cross-sectional and therefore only provides a snapshot of consumers' engagement with specific mobile phone service provider brands consumer-brand identification literature is often restricted by the contextual scope of its findings (e.g. focusing on a specific firm in an industry or a particular product category). Some studies also struggle to distinguish between mere involvement and actual identification, by only investigating the participants' views on their favoured brands (Tuškej, Golob and Podnar, 2013). Table 1 summarizes the key concepts.

Methodology
To address the research question regarding how and why consumers would identify and engage with certain films as brands, rather than just consuming them as films, this paper takes on a mixedmethods approach using two studies and employs abductive reasoning to explore the relationship between theory and data. Table 2 describes the methodology in detail.

Study one
To explore why and how consumers only engage with certain films, but not all films as brands, 35 face-to-face semi-structured interviews were conducted in 2014 (see Appendix A in the online Supporting Information). The purposive sampling method was employed -only those consumers who self-classified themselves as active film viewers were invited, because participants' willingness to share their consumption experiences is crucial for collecting relevant information (Cappellini and Yen, 2013). Participants were recruited in the West London area, via posters on notice boards, social media platforms and the researchers' personal contacts, whilst also taking on a snowball approach (McDaniel and Gates, 2009).
Interviews were conducted over 4 months, ranging between 25 and 91 minutes, concluding when the data were found to have become saturated (Strauss and Corbin, 1998), specifically when participants discussed known themes, thus indicating that key dimensions, sub-dimensions and relationships identified amongst them were representative of the phenomenon's totality. Prior to interviewing, participants were given an information sheet, outlining the study's purpose, procedure involved (permission to record audio for transcribing), researchers' contact details, reassurance of feedback anonymity, data confidentiality and their rights and voluntary participation (Collis and Hussey, 2014). Please see the list of guideline questions in Table 3.
All the interviews were audio recorded and transcribed. Initial themes within the findings were then gathered using thematic analysis (Spiggle, 1994), thus categorizing the data in a manner that is relevant to the research focus and to aid theoretical understanding (Bryman and Bell, 2015). Core or more abstract themes which have emerged, such as 'emotional bonding' and 'popularity' and subthemes, such as 'self-congruity' and 'emotional attachment' (which contribute towards encapsulating the core themes) were highlighted for coding purposes in NVivo (see Appendix B in the online Supporting Information for how these were discovered). Table 4 shows the 11 key concepts that emerged from the interview findings, their definitions and an exemplar quote for each. The following discusses how the 11 constructs are connected, starting from a film identity's coherency to film-brand identification, through possible mediators and moderators, leading to four favourable engagement behaviour outcomes.

Mixed-method approach
This paper adopts a mixed-method approach, rather than a multi-method approach (Creswell and Plano Clark, 2011), by incorporating a combination of methods from different research paradigms (qualitative and quantitative). In doing so, the researchers are able to: (1) extend the breadth and range of enquiry; (2) use the results of Study one (qualitative) to inform and develop Study two (quantitative) -for example, possible cause/effect via hypothesis development and measurement decisions; (3) further clarify and validate the results of Study one in Study 2; (4) utilize the strengths of one method to offset the weaknesses of another (Greene, Caracelli and Graham, 1989;Tashakkori and Teddlie, 1998). Furthermore, a mixed-method approach also enables methodological triangulation, whereby measurement, sampling and procedural bias are minimized, providing the results complement each other (Denzin, 1989).

Abductive mode of reasoning
This paper adopts an abductive mode of reasoning, rather than a deductive or inductive approach. The most common issue associated with deduction is the notion that the process strictly relies on the logic of theory and hypothesis testing, with difficulties arising as to which theory to select for testing (Bryman and Bell, 2015;Mantere and Ketokivi, 2013;Wilson, 2006). On the contrary, inductive reasoning is notorious for its inability to enable the development of theory, regardless of the size and depth of empirical data (Creswell and Plano Clark, 2011). Abductive reasoning is pragmatic, as it requires researchers to make logical inferences and build theories about the world (Bryman and Bell, 2015). It enables us to unravel the puzzling mystery as to why not all films produced by film-makers with the intention of being branded are engaged by consumers as brands. Also, it explains how some films acquire brand status without an initial intent by their producers to do so. This involves identifying conditions that would make the phenomenon less puzzling and therefore 'easier' or more logical to comprehend, thus pinpointing the 'best' explanation plausible to interpret the phenomenon (Mantere and Ketokivi, 2013).

Research context
The UK was chosen as the research setting because it is the world's fourth largest film market, with film revenues reaching £3.4 billion in 2017 and with more than 5 billion film viewings annually across various platforms (British Film Institute, 2018).

Study one
Using semi-structured interviews, Study one addresses objective one, which is to explore how and why consumers would identify certain films as brands. The findings derived from the qualitative interview data led to the development of our conceptual framework using an interpretivist orthodoxy. From an ontological perspective, a constructionist position was deemed suitable for Study one. Relating to the concept of postmodernism, constructionism assumes that meanings constructed during social interaction are ephemeral (Bryman and Bell, 2015), since they are dependent on both time and place, as is also evident in discourse analysis (Potter, 1996). Since films have now been releasing for more than a century and across the world, and the perception of a film as a brand can essentially be considered subjective in nature, the purpose of this study is not to determine which film is a film and which film is a film brand, but to better understand how consumers identify and engage with films as brands and which factors are fundamental to comprehend how the consumer film-brand engagement process functions effectively. We took part in a process known as 'dialectical shuttling', which involves engaging forwards and backwards with the social world, the literature and the empirical data (Atkinson, Coffey and Delamont, 2003;Schwartz-Shea and Yanow, 2012). Thus, in the process of developing this consumer film-brand engagement framework (Study one) and finalizing it through quantitative validation (Study two), we remain open to the idea of being 'surprised' by data, and not being confined by confirming previous understandings of how consumers engage with a film as a brand (Bryman and Bell, 2015).

Study two
Study two addresses objective two, which is to validate the consumer film-brand engagement framework that elucidates the antecedents, outcomes and facilitators (e.g. mediators, moderators) of film-brand identification through a sequential process, illustrating how consumers engage with a film as a brand, using quantitative data collected through a questionnaire survey. A consumer survey questionnaire was deemed appropriate as a deductive, conclusive and confirmatory means to measure how widespread are the quantifiable attitudes, motivations and behaviours identified in the preliminary Study one (Brace, 2013;Wilson, 2006).

Film identity coherency.
A coherent film identity was regarded as the starting point by participants in the film-brand engagement journey. It refers to a film identity's elements, including actors, characters, storyline, mise-en-scene, production house, etc. and collectively, how meaningful these are to the consumers. The findings indicate that film identity coherency helps participants experience a film more deeply, interpret it with more meaning, perceive it as more popular/successful and bond affectively due to more simplified comprehension. This echoes previous works that when consumers perceive a film's identity to be coherent (Cooper-Martin, 1991), they are also less likely to experience 'category confusion' (Zhao, Ishihara and Lounsbury, 2013), thus simplifying their subjective classification of films and interpretation of the meaning of the films (Hart, Kerrigan and vom Lehn, 2016) and engaging with the film as a brand (Kapferer, 2012). Contrary to O'Reilly and Kerrigan (2013), our findings show that elements such as actors, directors, production houses, etc. don't necessarily dictate a film's brand status, but instead collectively constitute a film's identity. This is because experiential goods are not simply considered as the sum of their attributes; instead, consumers value the potential arising from their combination holistically (Bassi, 2010 Popularity. When asked why certain films are regarded as brands, popularity, fan-based community, sequels and emotional bonding were identified as agents that help translate film identity coherency into subsequent film-brand identification. In particular, participants referred to the level of awareness a film gathers within their social/professional circles as popularity (e.g. how fre-quently a film is discussed, how widespread knowledge is about a film's storyline/characters, etc.). Interestingly, participants also expressed that a film's popularity is not necessarily dictated by a film's box office results but how it is felt subjectively (Bassi, 2010), and this sense of a film as popular helps them further identify it as a brand. For example: Titanic is a film brand because people talk about it, it's word-of-mouth . . . if it's a big discussion in society, if it's been referred to in the news, in society in general, if people keep referring to a film and comparing it, they are keeping it alive . . . basically to keep it circulated within society, it almost becomes like common knowledge, you expect people to know about the film. [Participant 24] Based on the findings, we propose: H1a: Popularity mediates the effect film identity coherency has on consumer's film-brand identification.
Fan-based communities. Participants also agreed that fan-based communities help promote films as brands, as fan-based communities promote a collective sense of belonging that facilitates their pro-activeness in identifying and engaging with films as brands, as part of something bigger. This process of communal relationship amongst consumers, film-makers and fellow viewers (Batson, Clark and Mills, 1993) explains why sometimes,  (Holt, 2002) 'One of a kind, a new variant of a kind, so the first Spaghetti Westerns were iconic, nobody had ever seen anything like them before, and you'd wanna see the next one, and the next one, and the next one, and so it became a brand. The first of the Spaghetti Westerns was amazing, the first time I saw it, it was very bleak, every time I see it again, I see the comedy in it, it becomes funnier and funnier, but yeah, in order for a film to become a brand, I think it has to be iconic, to earn its brand status' [Participant 29] Emotional bonding The feelings consumers have towards a particular film, on an affective level (Heath, 2009  fan-based communities are formally institutionalized by film marketers (Muniz and O'Guinn, 2001). Our findings show that fan-based communities are recognized for their contribution towards how widely accepted a film becomes in society, whilst community members' projection of passion for films acts as a seal of approval for a film's stature as a brand: Twilight, they have a lot of really passionate fans, who passionately promote the film! That's why they are film brands. [Participant 10] Based on the findings, we propose: H1b: Fan-based community mediates the effect film identity coherency has on consumer's filmbrand identification.
Sequels. Sequels are also identified as a key mediator because sequels help explain and elaborate a film's identity further and deeper to participants, thus increasing the likelihood that they identify the film as a brand. Our findings extend Preece, Kerrigan and O'Reilly's (2018) work on serial brands, revealing how a film's sequels contribute towards its brand status, rather than brand longevity. For example: Star Wars -the storyline of the original trilogy is wide enough and rich enough to allow prequels, sequels and origin stories to be made . . . This makes it almost immortal because you can have a factory that churns out movies on a regular basis that are based on the events of the original trilogy. It keeps the brand alive and fresh. [Participant 16] Based on the findings, we propose: H1c: Sequels mediate the effect film identity coherency has on consumer's film-brand identification.
Emotional bonding. During the interviews, participants conveyed strong emotions for some films they identified and engaged with as brands, which were triggered by profound thoughts and deep exploration of a film's various elements. Participants expressed their emotional attachment and passion for selected characters/stories they relate to and admire and the meanings these hold in their lives. Our findings echo Graffigna and Gambetti (2015), showing that consumers' emotional bonding can contribute as a predictor to consumerbrand engagement due to films' hedonic, highinvolvement nature, and their ability to integrate viewers' intellectual and emotional spheres (Bassi, 2010). Therefore we propose that besides having a coherent identity that assists consumers' desires for protagonism and self-achievement (Gambetti, Graffigna and Biraghi, 2012), when a film manages to build a deep, intimate, authentic, emotional and friendship-like relationship with consumers (Graffigna and Gambetti, 2015), the film is more likely to be identified as a film brand. Thus, we propose: H1d: Emotional bonding mediates the effect film identity coherency has on consumer's film-brand identification.
Marketing efforts. The findings also highlight four possible moderators: marketing efforts, timelessness, franchise/merchandise and iconic status; these may intensify participants' identification and engagement with films as brands. From the marketers' perspective, brands are to a large extent established through investments in media cul-ture, advertising/promotional campaigns, product placements, sponsorship and co-branding initiatives (Janson, 2002;Kellner, 1995) to enhance their marketability (Kerrigan, 2010). Such marketing efforts are not unnoticed -they are acknowledged by participants in the form of trailers, posters, promotional events, advertising campaigns, etc. preand post-launch and are appreciated as a means to assist participants' identification and engagement with films as brands. The findings show that when a film is supported with more marketing efforts, the relationships from film identity coherency to popularity, sequels, fan community and emotional bonding are likely to be strengthened. This is because a film's marketing effort could further promote the film's identity coherency, thus increasing its prominence (Kerrigan, 2010) and effect on perceived popularity, film sequels, fan community events and emotional bonding. Therefore, the following are proposed: H2a: Marketing efforts moderate the effect of film identity coherency on popularity. H2b: Marketing efforts moderate the effect of film identity coherency on fan-based community. H2c: Marketing efforts moderate the effect of film identity coherency on sequels. H2d: Marketing efforts moderate the effect film identity coherency has on emotional bonding.
Merchandise/franchise. The findings show that merchandise and franchise availability act as suggestive cues for the film's popularity and brand status, especially for films that participants are less attached to or haven't seen. This may be because merchandising goods' tangibility acts a token of the film; an embodiment that consumers can physically touch, feel and interact with (Kapferer, 2008 The findings extend previous works (Gambetti, Graffigna and Biraghi, 2012;Kapferer, 2008), showing that a brand cannot exist for consumers without a supporting product/service embodying the brand, thus placing the brand into reality and acting as a brand evaluation source. Whilst franchise and merchandise provide brands with firm reference points, to be embedded and consumed in consumers' daily lives, we argue that franchise/merchandise availability strengthens the impact of perceived popularity on subsequent consumers' film-brand identification, leading to the development of: H3: Franchise/merchandise moderate the effect popularity has on consumer's film-brand identification.
Timelessness. During the interviews, participants alluded to how a film may evolve into a film brand over time, rather than circulating momentarily and fading away. For instance, this may be achieved when a film reinforces its popularity status by continuously circulating within society and therefore staying 'alive', by being passed on through generations, etc. Participant 16 recalled: Star Wars has become a strong brand over time; its appeal has endured for over 30 years to widespread audiences, the young and the old, that means the passion and enthusiasm for these movies can be passed between generations. Many people, including myself, were introduced to the Star Wars universe by members of the previous generation . . . it is timeless. [Participant 16] Participants suggested that when a popular film becomes timeless, it would not be affected by the passage of time and/or changes in taste, fashion or trends. Instead, the meanings associated with a film would change over time and in accordance with their life stages (British Film Institute, 2011). The finding echoes the work of Preece, Kerrigan and O'Reilly (2018) on the 55-year-old James Bond serial brand, which has stayed relevant yet true to its values over time. Thus, we propose: H4: Timelessness moderates the effect popularity has on consumer's film-brand identification.
Iconic status. Participants discussed iconic status as those films that go against the norm, are thought provoking, immortalized in popular culture, lead the way and offer something entirely new/original -relative to that era in time. Examples provided include Gone with the Wind for its remarkable music, Avatar for its revolutionary 3D experience and Sholay for being the first true Indian western shot on 70 mm. Such iconic status explains why some films are recognized by far more people than those who have actually viewed them (British Film Institute, 2011). Whilst existing literature (Holt, 2002;Preece, Kerrigan and O'Reilly, 2018) suggests that a film's sequels contribute towards its subsequent iconic status, our findings show that iconic status could strengthen the effect of sequels on filmbrand identification. Based on the above discussion, we propose: H5: Iconic status moderates the effect sequels have on consumer's film-brand identification.

Film-brand identification and engagement.
Participants discussed their identification of a film brand as a 'state' in which they recognize a film to be not just a film but also a brand. Clearly this brand status is not spontaneously granted for all films, for example: Not every film is a brand, you could even say most films are not brands, they are only movies! [Participant 11] During the interviews, it became apparent that participants were more passionate about films they identified as brands than those that were merely films to them. Repetitive viewing, loyalty and advocacy for the brand, and purchase intention for brand-related merchandise, are discussed as the key behavioural measures that consumers use when identifying and engaging with films as brands. The findings echo existing works, showing that once a film is identified as a brand, consumers would engage in brand support behaviour. These include increased buying intention of various merchandise (Ahearne, Bhattacharya and Gruen, 2005), willingness to accept a brand extension and follow the brand (Kim, Han and Park, 2001;Lee et al., 2013), repetitive G. S. Kohli et al. viewing (Lacher and Mizerski, 1994) and subsequent word-of-mouth (Tuškej, Golob and Podnar, 2013). Therefore, we propose: H6: Consumers' film-brand identification has a direct effect on film-brand engagement. Figure 2 summarizes Study one's key findings and hypotheses in the film-brand engagement framework. Table 5 provides an overview of the proposed hypotheses, with exemplar quotes.

Study two
In order to empirically validate the proposed framework, we conducted a questionnaire survey including three steps: pre-test, pilot test and main data collection. Table 6A explains the steps taken to ensure validity and reliability in detail. We followed Churchill's (1979) guidelines to ensure content and face validity of all constructs and scales. The questionnaire contained 11 constructs, derived from Study one. We adapted existing measurements for eight constructs and developed the operationalization for film identity coherency, fanbased community and timelessness, using findings derived from Study one (see Appendix C in the online Supporting Information for an overview of how these were developed).
The questionnaire required respondents to answer a series of questions with their most recently viewed film in mind. This approach allows film choices to be randomized, hence further minimiz-ing interviewer bias (Collis and Hussey, 2014). The final set of constructs and corresponding items are presented in Appendix D (Supporting Information), listing each construct's conceptualization and operationalization. All items developed use seven-point Likert scales (ranging from 'Strongly disagree' to 'Strongly agree') with an option of 'Don't know' to capture any lack of awareness.
The final sample represents the UK population well, with 48.7% being male and 51.3% female. 88.7% were 'white' British participants and 11.3% were ethnic minorities. The sample also adequately represents the UK's key regions and has a good mix of educational qualifications and film viewing frequency, demonstrating validity and representation of the UK population (see Appendix E in the online Supporting Information).

Study two findings.
To validate the proposed conceptual framework and to examine the hypotheses, data analysis was performed using three main techniques: exploratory factor analysis (EFA), confirmatory factor analysis (CFA) and the SPSS process macro. EFA and CFA together determine the construct reliability and validity of the proposed framework, whilst the SPSS process macro examines the proposed mediation and moderation effects.
With the emergence of three new constructs, conducting EFA first was deemed appropriate to detect any underlying structure and discover which items belong to which construct (Hair et al., 2010), thus purifying the measures (Churchill, 1979).

Pre-test
Three marketing academics and one film industry practitioner were invited to scrutinize the questionnaire and to give feedback, to ensure that all the items were suitably modified for films. Upon satisfactory verification, another 30 participants were given the questionnaire, using convenience sampling. No errors were reported, some minor amendments were made after receiving feedback from respondents (e.g. the definition of 'iconic' was added and the layout was revised to improve reader-friendliness).

Pilot study
A pilot study was conducted with 200 participants through convenience sampling of university students in a UK university, to assess the reliability of the measurements. The results of the pilot study indicated that all constructs have a high level of internal consistency (Cronbach's α > 0.70, ranging from 0.74 to 0.96).

Main data collection
In order to assist with the main UK nationwide data collection, a market research company was approached. A quota sampling strategy was employed based on the 2011 UK Census, in relation to various demographics (e.g. age, gender, ethnicity, region and level of education) (McDaniel and Gates, 2009). We received 1136 responses, wherein 106 were omitted due to incompletion or excessive missing answers (Hair et al., 2010), leading to a final sample of 1030.

Fan-based community
This construct was eliminated since there were some items related to 'fan-based community' which cross-loaded with 'marketing efforts'. This cross-loading could be explained through the notion that a 'fan-based community' may contribute to a film's pre-release buzz (Schneider, 2012) if there is an existing fan following, such as the case with sequels (Preece, Kerrigan and O'Reilly, 2018) or when a film has been adapted from another creative cultural industry (O'Reilly and Kerrigan, 2013).
Brand following This construct was eliminated because some items were dropped due to lack of convergent validity, with items loading <0.50 (Hair et al., 2010). Conceptually, the construct is intertwined with 'repetitive viewing', since 'repetitive viewing' could be regarded as a facet of one's loyalty for a film, because it is an act of ritual that symbolizes one's admiration for a film (Hart, Kerrigan and vom Lehn, 2016), which was also reiterated in our qualitative findings.
Principal component analysis with Promax rotation was performed on all 11 constructs, comprising 59 items using the pilot data. Items were retained if (1) they loaded 0.50 or more on a factor, (2) they did not load more than 0.50 on two factors and (3) they had a Cronbach's alpha loading of more than 0.7 (Hair et al., 2010). Fanbased community and brand following were eliminated during EFA. Please see Table 6B for further explanation.
The results reveal that nine constructs (41 items in total) were retained from the original 11. These nine constructs (eigenvalues > 1) explain 79.1% of the total variance. Harman's single-factor test was then conducted to identify any common method variance. All variables were included in the analysis, constraining the factors to be extracted to one, using an unrotated solution. 38.8% of the variance was explained by a single factor, which is below the 50% threshold, confirming that this model does not suffer from any common method bias issue (Podsakoff et al., 2003). CFA was then conducted in AMOS version 25 with 1030 respondents, to assess the measurement model's validity before mediation and moderation tests (Hair et al., 2010). The full measurement model fits the data well (χ² [732] = 3448.57, p < 0.001; χ²/df = 4.71; IFI = 0.93; TLI = 0.92; CFI = 0.93; RMSEA = 0.06). This model was significant, which is common with larger data sets (Hair et al., 2010). Constructs' reliability was tested using both composite reliability and Cronbach's alpha, and they were all above the recommended level, as shown in Table 7. The correlation (covariance) among the constructs is also acceptably low, ranging from 0.14 to 0.77; AVE is >0.5 (Fornell and Larcker, 1981) and loading coefficients are all >0.50 (Hair et al., 2010), thus meeting convergent and discriminant validity. A further test to ensure the adequacy of discriminant validity was performed successfully, since the square root of each construct's AVE is greater than the bivariate correlations (coefficients range from 0.77 to 0.93, p < 0.01) and there are also no substantial cross loadings with standardized residuals all <0.28 (Byrne, 2001). Thus, the assessment results support the adequacy of discriminant validity of the measurement model (Table 7). Mediation. The mediation test was performed using the SPSS process macro, as recommended by Preacher and Hayes (2008). Film identity coherency was found to be a significant predictor of  all three hypothesized mediators (p ࣘ 0.001). Popularity, sequels and emotional bonding were also found to be significant predictors of film-brand identification (see Figure 3). In all three cases, the introduction of the mediators has led to a reduction in the effect sizes of the original direct effects, whilst all three confidence intervals are also not equal to 0, which suggests that mediation is taking place (Preacher and Hayes, 2008). Popularity and sequels were found to partially mediate, since the IV → DV total effect is significant and so is the direct effect (IV → M → DV) after the introduction of these two mediators. Interestingly, after the introduction of the emotional bonding mediator, the direct effect has become insignificant, indicating full mediation (Baron and Kenny, 1986;Hair et al., 2010;Preacher and Kelley, 2011;Zhao, Lynch and Chen, 2010). Since paths 'a' and 'b' were significant for all mediators, mediation analysis was further conducted using the bootstrapping method with bias-corrected confidence estimates, obtaining a 95% confidence interval of indirect effects with 5000 bootstrap resamples (Preacher and Hayes, 2008). The findings reveal that popularity (β = 0.40, p ࣘ 0.001, PM = 0.36, abcs = 0.11) and sequels (β = 0.53, p ࣘ 0.001, PM = 0.24, abcs = 0.09) are both partial mediators, whilst emotional bonding reveals the most impact as a full mediator (β = 0.06, ns, PM = 0.90, abcs = 0.29). See Table 8 for the mediation results.
Moderation. Subsequently, the moderation effect of marketing efforts, franchise/merchandise, timelessness and iconic status is tested using the SPSS process macro model no. 1 (Hayes, 2013), as shown in Figure 4.
The interaction effects between marketing efforts and film identity coherency were examined against popularity (H2a), sequels (H2c) and emotional bonding (H2d). The findings show that film identity coherency (β = 0.55; p ࣘ 0.001) and marketing efforts (β = 0.22; p ࣘ 0.001) accounted for a significant amount of variance in a film's popularity (R² = 0.46; p ࣘ 0.001). The interaction term (Aiken and West, 1991) resulted in a significant negative effect (β = −0.05; p ࣘ 0.01) and demonstrated a significant proportional change in the variance of a film's popularity ( R² = 0.004; p ࣘ 0.01). This indicates that when marketing efforts are increased, the effect a coherent film identity has on a film's popularity is marginally lessened, highlighting the substitutional effect of marketing efforts, thus H2a is partially supported. This result may be counterintuitive; however, it suggests that marketing effort could be employed as a feeble attempt to rescue a film's popularity due to its lack of identity coherency.
Film identity coherency (β = 0.20; p ࣘ 0.001) and marketing effort (β = 0.42; p ࣘ 0.001) also accounted for a significant amount of variance in sequels (R² = 0.19; p ࣘ 0.001). However, the in-teraction term between marketing efforts and film identity coherency was found to be insignificant, thus having no moderation on the relation between film identity coherency and sequels ( R² = 0.00; ns; β = 0.004; ns) and rejecting H2c. A plausible reason for this insignificance may be that consumers are already familiar with a film's premise if it has more than one part. Thus, a film's marketing efforts may not have any effect on strengthening the relation between film identity coherency and its sequels.
G. S. Kohli et al. Film identity coherency (β = 0.45;p ࣘ 0.001) and marketing efforts (β = 0.24; p ࣘ 0.001) also accounted for a significant amount of variance in emotional bonding (R² = 0.31; p ࣘ 0.001). The interaction term between marketing efforts and film identity coherency was also found to significantly strengthen the effect of film identity coherency on emotional bonding ( R² = 0.006; p ࣘ 0.01; β = 0.07; p ࣘ 0.01). H2d is therefore confirmed, proving marketing efforts' positive moderating effect on the relation between film identity coherency and emotional bonding.
Popularity (β = 0.15; p ࣘ 0.001) and franchise/merchandise (β = 0.70; p ࣘ 0.001) accounted for a significant amount of variance in film-brand identification (R² = 0.56; p ࣘ 0.001). The results showed that the franchise/merchandise interaction term ( R² = 0.002; p ࣘ 0.01; β = 0.05; p ࣘ 0.01) significantly and positively affects the relationship between popularity and film-brand identification. This affirms H3 and proves that the relation between popularity and film-brand identification could be further strengthened by franchise/merchandise activities.
Finally, film-brand identification's impact on film-brand engagement is positive (β = 0.56; p ࣘ 0.001), showing that once consumers identify a film as a brand, they will engage in brand support behaviour.

Discussion
By taking on the novel approach of developing and validating a film-brand engagement framework, derived from consumer insights, this paper contributes to existing film-branding literature in several ways. Firstly, the film-brand engagement framework explains how consumers would interpret, identify and engage with films as brands in a sequential order, illuminating the routes by which films are identified and engaged by consumers as brands, taking the consumer-centric perspective to film branding. It is important to note that not all factors in the framework are mandatory to facilitate consumer-brand engagement; the more a film can incorporate, arguably the higher the chances that consumers will go on to identify and engage with such films as brands.
The model starts with film identity coherency as the antecedent. It shows that when a film communicates its identity cues in a clear, considered and careful manner to ensure all features are coherently composed to complement each other aesthetically and meaningfully, the film will have a higher chance of being considered as a brand, reflecting Kapferer's (2012) thoughts on the effectiveness of coherency to differentiate amongst the competition. Film identity coherency then leads to film-brand identification through popularity, emotional bonding and sequels. Filmbrand identification is key in the consumer filmbrand engagement framework because it resembles the state/condition wherein a film is considered not only a film, but also a brand, achieving the desired brand status (Hollebeek, Glynn and Brodie, 2014). By specifically revealing how brand coherency drives consumer-brand identification through possible routes of popularity, sequels and emotional bonding, the finding depicts a clear formula for the development of film brands, extending Keller's (2008) work.
The results also confirm that film-brand identification drives film-brand engagement, which focuses on the behavioural dimensions of engagement, such as positive word-of-mouth, repetitive viewing and purchase intention of film brandrelated merchandise. The finding confirms previous works on the association between brand identification and brand engagement (Hollebeek, Glynn and Brodie, 2014;Leckie, Nyadzayo and Johnson, 2016), but shows brand identification as  a precursor to film-brand engagement, rather than a consequence. This highlights the importance of brand identification in terms of film consumption, showing that consumers would identify films as brands first before their enactment of brand engagement behaviour.
Besides the main constructs, the results also confirm the importance of popularity, sequels and emotional bonding with consumers as mediators between film identity coherency and film-brand identification, showing that a film's popularity, number of sequels and the emotional bonding that consumers have with the film can all lead to better film-brand identification. With its full mediating effect, emotional bonding has emerged as the most powerful mediator. This suggests that feeling emotionally connected to films is the most effective way for consumers to identify films as brands -be it moral value with regard to the storyline, characters, etc. which they can resonate with (Oatley, 1999), or through self-congruity (Bagozzi, Batra and Ahuvia, 2014). The findings extend the previous debate on brand engagement to film branding (Gambetti, Graffigna and Biraghi, 2012;Graffigna and Gambetti, 2015).
Besides mediators, the results also reveal the importance of marketing efforts, timelessness, franchise/merchandise and iconic status as four moderators in explaining consumers' film-brand identification and engagement. The results show that marketing efforts can significantly strengthen the relationship between identity's coherency and emotional bonding, as well as substitute the effect of identity's coherency on popularity, providing empirical support on the use of marketing efforts to enhance a film's marketability and playability (Kerrigan, 2010). Nevertheless, although significant, marketing efforts' moderation effect is relatively small, meaning that rather than spending lots of the post-production marketing budget on promoting films, it is critical that film identity coherency is adhered to in the first place.
Furthermore, iconic status is found to significantly strengthen sequels' impact on film-brand engagement, showing that when a film's storyline extension potential is backed up with an iconic status, it has a much higher chance of being identified as a brand, offering empirical support to the discussion of Holt (2004) and Young et al. (2008). The sense of timelessness and franchise/merchandise initiative further enhance the effect of popularity on film-brand identification, showing that when a film is popular and deemed timeless or has franchise/merchandise goods available, the film is more likely to be recognized as a brand. This highlights the value of creating films that are classic, which can be enjoyed by generations regardless of the passage of time (Preece, Kerrigan and O'Reilly, 2018), and supports the argument of Kapferer (2008) whereby a brand cannot exist for consumers without a supporting product or service that embodies the brand. The tangibility of merchandise and franchise (e.g. theme parks, TV shows, novels, video games, etc.) adds another dimension for the consumer to interact and resonate with a film as a brand, since it allows them to showcase their passionate behaviour and admiration amongst their immediate social circles and wider community of fans. Table 10 lists the managerial implications for production houses, film-makers and film marketing agencies, derived from the film-brand engagement framework.

Limitations and future research directions
Despite the contributions, this paper has its limitations. Firstly, film is considered as a cultural entity, therefore it is important for future studies to validate the consumer film-brand engagement framework using data from outside the UK to enhance its generalizability. Cross-cultural research is also recommended to further investigate whether consumers' identification and engagement with films as brands would differ across cultures. Furthermore, film-branding literature is still in its infancy.
Whilst our framework shows how and why consumers could identify and engage with films as brands, leading to favourable brand-engagement support behaviour, it has not discussed the possible 'dark side' of film brands. For example, consumers' obsessions with a film brand may turn into abuse and bullying, if the creative choices made by film-makers are at odds with the consumers' views and beliefs of the brand. This is illustrated by the heated criticism of Ben Affleck's casting as Batman in Batman versus Superman: Dawn of r Use a film's identity coherency and capability to bond emotionally with the chosen target audience as two key criteria, when evaluating and allocating a film's promotional resources, in order to position films as brands in the minds of consumers.
r Develop innovative merchandising and franchising concepts (e.g. theme parks, theatre productions or activities) unique to a film's premise, to strengthen a film's popularity and view as a film brand.

Film-makers
r Build a strong/coherent film identity by including: identifiable characters with depth of exploration; a strong message; consistent creative themes and references to past involvement and nostalgia (if adapted from an existing media text, e.g. Harry Potter).
r Nurture an emotional bond during the target audience's film consumption journey; gain a deeper understanding of them/their values and cater for these coherently in the storyline, character selection, plot arrangements, music choice and mise-en-scene, etc.
r Apply the consumer film-brand engagement framework to be in a better position to secure funds to greenlight film projects with potential investors and attract additional capital through negotiating financing deals and downstream contracts.

Film marketing agencies
r Clarify what the production house wants to achieve (i.e. increase a film's popularity, build an emotional bond with the target audience, or highlight a film's unique selling point) with the allocated marketing budget before implementing promotional activities.
r For films that are already released, promote their iconic status/timelessness and take advantage of cultural industries and populist worlds as facilitators of the process.
r Anticipate and embrace that a film will pass through time, be subject to cultural forces, as well as that consumers have become co-creators of the brand.
r Maintain a sense of timelessness to sustain authenticity, by not always managing the brand with vested commercial interests; empower consumers to lead the way in co-creating the brand, going against the norm and consistently offering something new.
Justice, or the departure of director Zack Snyder during the making of Justice League, both contributing to below-target box-office performances and returns on investment (Bianco, 2018). Future research is recommended to explore the adverse consequences of films as brands; after all, not all films produced by film-makers are intended to take up branded status. Furthermore, not all films will succeed in engaging their consumers as brands and enjoy engagement-support behaviour.
Gurdeep Singh Kohli is a scholarship recipient (BSc) and has been awarded a PhD in Marketing from Brunel University London. He has published in the Marketing Review and has had conference papers accepted for the Academy of Marketing and the European Marketing Academy. He currently teaches Business Management at a FE/HE institution and is a Visiting Lecturer in Marketing at Brunel University London. His research interests include film marketing/branding, experiential goods, iconic brands and consumer-brand relationships. Dorothy A. Yen is a Reader in Marketing at Brunel University London. She takes on a consumercentric approach to marketing and branding. Dorothy is particularly interested in exploring how culture affects human behaviour, in both b2b and b2c domains. She has led several special issues in