Consumers' intention to follow brands on social media : the moderating roles of the type of platform and the type of brand

CONSUMERS’ INTENTION TO FOLLOW

BRANDS ON SOCIAL MEDIA: THE

MODERATING ROLES OF THE TYPE OF

PLATFORM AND THE TYPE OF BRAND

MSc. in Business Administration - Digital Business Track

ABS, UvA

Mayke Oosterveld

11410981

23 June 2017

Final version

Dr. J. Y. Guyt

STATEMENT OF ORIGINALITY

This document is written by Mayke Oosterveld who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

TABLE OF CONTENTS

DIFFERENT FORMS OF SOCIAL MEDIA ... 9

MOTIVATIONS FOR CONSUMERS TO FOLLOW BRANDS ON SOCIAL MEDIA ... 12

HYPOTHESES ... 17

INTERACTION BETWEEN DRIVERS AND TYPE OF PLATFORM ... 17

INTERACTION BETWEEN DRIVERS AND TYPE OF BRAND ... 21

METHODOLOGY ... 24

DATA COLLECTION AND SAMPLING METHOD ... 24

SAMPLE ... 25 PRE-TEST EXPERIMENT ... 26 RESEARCH DESIGN ... 26 MEASUREMENT ITEMS ... 27 CONTROL VARIABLES ... 28 RESULTS ... 29

EXTREME AND MISSING VALUES ... 29

DESCRIPTIVE STATISTICS ... 30

RELIABILITY ANALYSES ... 30

CORRELATION ANALYSIS ... 31

TESTING THE HYPOTHESES ... 33

DISCUSSION ... 39

ACADEMIC CONTRIBUTION AND MANAGERIAL IMPLICATIONS ... 41

LIMITATIONS AND FUTURE RESEARCH ... 42

CONCLUSION ... 43

REFERENCES ... 44

LIST OF TABLES

Table

Page

1

Means, standard deviations, and correlations...32/33

2

Hierarchical regression model of intention to follow brand...36/37

3

Outcomes of the hypotheses...40

LIST OF FIGURES

Figure

Page

1

Research design...24

2

Results...39

CONSUMERS’ INTENTION TO FOLLOW BRANDS ON SOCIAL

MEDIA: THE MODERATING ROLES OF TYPE OF PLATFORM

AND TYPE OF BRAND

ABSTRACT

In today’s world, social media play a significant role for both users and brands. Besides, social media altered the way people do business and communicate with brands, such as consumers following brand pages on social media. The uses and gratifications theory is a common theory in examining the antecedents of consumers’ intention to do so. However, the roles the type of platform (i.e., Facebook or Twitter) and the type of brand (i.e., utilitarian or hedonic) play are still unclear, while these variables might affect consumers’ motivations to follow a brand on social media. To fill this gap, this study examines these roles by means of an experiment embedded in a survey that was distributed online. The results suggest that both types of platform strengthen the influence of information seeking, although to a greater extent for Twitter compared Facebook. Furthermore, the type of platform does not seem to play a moderating role on the relationship between self-expression, entertainment seeking, social-interaction seeking, and incentive seeking and consumers’ intention to follow a brand. Regarding the type of brand, no moderating role could be found on the drivers of consumers’ intention to follow brand pages. This research helps to enrich and improve uses and gratifications theory, and proves helpful for marketers to better segment and target customers for brand pages.

Key words: social media marketing, intention to follow brand, uses and gratifications theory,

INTRODUCTION

In today’s world, social media play a significant role for both users and brands (Mancuso and Stuth, 2014). Social media altered the way people do business and communicate with brands, and brands appear to take notice of this change (Kaul et al., 2015). This new media landscape can significantly impact consumers’ behaviour on these social media regarding brands (Chu, Cheng and Sung, 2016). Besides, through social media marketers are able to interact and communicate with customers in various ways, while they enable customers to have online conversations about brands and products (Chi, 2011; Fogel, 2010; Smith, Fischer and Yongjian, 2012).

Social media continue to grow. For example, Facebook is expected to grow its user base from 1.43 billion monthly active users in 2016 to 1.87 billion monthly active users in 2020 (eMarketer, 2016). Although Twitter’s growth pace is slightly decreasing, its user base is expected to grow from 286.3 million monthly active users in 2016 to 368.8 million monthly active users in 2020 (Statista, 2015). Social media’s growth in prominence in the past years was in tandem with technological innovations, such as smart devices and inexpensive high-speed broadband Internet (Lamberton and Stephen, 2016). Consequently, more people around the world are connected through social media platforms.

Many authors have focused on consumer motivations to engage with brand on social media in general (e.g., Azar, Machado, Vacas-de-Carvalho and Mendes, 2016; Kwon, Kim, Sung, and Yoo, 2014). The uses and gratifications approach is common in research on motivations for consumers to follow brands on social media (e.g., Muntinga, Moorman and Smit, 2011; Saridakis, Baltas, Oghazi, Hultman, 2016). Specifically, self-expression, information seeking, entertainment seeking, social-interaction seeking, and incentive seeking are found to influence consumers’ intention to follow brands on social media. To date, however, no article has focused on brand-specific aspects such as the type of brand (i.e., hedonic or

utilitarian) in combination with the type of platform (i.e., Facebook or Twitter), nor do they control for brand-specific factors (i.e., brand likeability and brand commitment), while these aspects might influence consumers’ drivers to follow brands on social media (Phua, Jin and Kim, 2017). Besides, brands can gain significant advantages when consumers like or follow their pages (Chu et al., 2016; Xie and Lee, 2015; Zwass, 2010), such as stimulating word-of-mouth (Chu et al., 2016), building brand communities (Lamberton and Stephen, 2016), and creating consumer perceptions (Simmons, Conlon, Mukhopadhyay and Yang, 2011). Consequently, this leads to the following research questions:

I. How does the type of platform influence consumers’ motivations – self-expression, information seeking, entertainment seeking, social-interaction seeking, and incentive seeking – to follow a brand on social media?

II. How does the type of brand influence consumers’ motivations – self-expression, information seeking, entertainment seeking, social-interaction seeking, and incentive seeking – to a follow a brand on Facebook and Twitter?

From an academic perspective, the results of this paper aim to provide insights in the factors driving consumers’ behavioural intention to follow brands on two social media platforms: Facebook and Twitter. Moreover, it specifically examines if these drivers differ across product category and the type of brand, while controlling for brand likeability and brand involvement. For marketers and social media managers, this paper aims to contribute through a better understanding of consumers’ liking behaviour and the factors that influence this. In this way, they will be better able to segment and hence target consumers for their social media activities.

LITERATURE REVIEW

Social media

According to Kaplan and Haenlein (2010), the concepts social media, User Generated Content (UGC) and Web 2.0 are often used interchangeably, while in fact they do not imply the same thing. Hence, in order to provide clarity onto what definition is used for the purpose of this article, these constructs are explained briefly. Social media refer to “Internet-based applications that build on the ideological and technological foundations of Web 2.0 and that allow the creation and exchange of User Generated Content” (Kaplan and Haenlein, 2010, p.61), covering areas such as social networking sites (e.g., Facebook), blogs and microblogs (e.g., Twitter), content collaborative projects (e.g., Wikipedia), virtual worlds (e.g., Second Life) and online forums (Alves, Fernandes and Raposo, 2016). Web 2.0 slightly deviates from social media, as it implies “a platform whereby content and applications are no longer created and published by individuals, but instead are continuously modified by all users in a participatory and collaborative fashion” (Kaplan and Haenlein, 2010, p.61). Finally, UGC can be referred to as all types of media content which is created by users of social media and is publicly available (Kaplan and Haenlein, 2010). In order for content to be categorized as UGC, it needs to meet three requirements (Organization for Economic Co-operation and Development [OECD], 2007); these are: (1) the user-created content has to be made publicly available, whether this is on a website that is publicly available or on a social media platform, (2) the user has to put creative effort into generating the content or adapting existing content to create new content, and (3) the content has to be created independently of professional practices and routines (OECD, 2007), which implies that it is not created in a market nor institutional context. However, it is beyond the scope of this paper to elaborate on these requirements further.

Compared to traditional media, such as TV and radio, social media offered the opportunity for companies to engage with their customers, allowing for two-way conversations

rather than one-to-many monologues (Fournier and Avery, 2011), as well as enabled marketers to communicate and target their customers in different ways (Lamberton and Stephen, 2016). These social media platforms also affected consumers’ behaviour across different market settings, as “the ubiquity of social media has changed how buyers share information with each other and interact with brands” (Lamberton and Stephen, 2016, p.146).

However, despite marketers’ efforts and desires to utilize social media to their full potential, it became present that these platforms were created for people, not for brands (Fournier and Avery, 2011; Berthon, Pitt, Plangger and Shapiro, 2012). Social media is about connecting, sharing and collaborating with other people in a collective conversation rather than selling products and advertising them (Fournier and Avery, 2011; Kaplan and Haenlein, 2010). According to Farshid, Plangger and Nel (2011), for brands to develop the most effective strategy for social media, brand managers would be recommended to develop a profound understanding of social media, its various forms, and its specific channels within each form. As such, brands need to recognize that in order to reach potential customers they have to mix these elements (e.g., connecting, sharing, collaborating) in their message to commercially exploit its potential.

Different forms of social media

Within social media, five different forms can be identified, being Video-Sharing Websites, Picture-Sharing Websites, Blogs, Micro-Blogs, and Social Networking Websites (Mills, Botha and Campbell, 2015; Mills and Plangger, 2015; Parent, Plangger and Bal, 2011).

So-called Video-Sharing Websites, such as YouTube, allow their users to upload and share videos, as well as watch them (Mills and Plangger, 2015). Another important feature of these websites for its users is the possibility to generate text-based content (e.g., commenting on videos), which emphasizes the social character (Babić Rosario, Sotgiu, de Valck and

Bijmolt, 2016; Tang, Fang and Wang, 2014). Somewhat similar to these websites are Picture-Sharing Websites, where instead of videos, still pictures are uploaded and consumed (Mills and Plangger, 2015). An example of these kinds of websites is Pinterest. It is difficult for organizations on Pinterest to determine who their active users are, as it is “kind of a guessing game with the overall strategy” (Howard, Mangold and Johnston, 2014, p.659).

Blogs, which is an abbreviation for web logs, refer to “websites maintained and written by individuals, but hosted and technically owned by an organization that provides access to web space and a content management system” (Mills and Plangger, 2015, p.524). Some blogs resemble personal diaries, while others are more sophisticated and focused on a particular subject (Mills and Plangger, 2015), such as on the fashion industry or politics. Wordpress, Blogger, and LiveJournal are among the most popular blogging sites (Mills and Plangger, 2015). Micro-Blogs are essentially the stripped-down versions of blogs, and imply “social networking services that enable users to send and read very short messages, which are usually restricted by the number of characters in the message” (Berthon et al., 2012, p.263; Mill and Plangger, 2015). The most popular and best known Micro-Blog is Twitter, through which its users can ‘tweet’ each other. These short, text-based messages are shown to everyone following the author as well as on the author’s profile page, and the tweets cannot be longer than 140 characters (Berthon et al., 2012). Twitter may be considered among the leading social media for disseminating information (Davis and O’Flaherty, 2012), and as such can enhance communication campaigns (Waters and Williams, 2011), as well as strengthen viral-based marketing campaigns (Fogel, 2010; Jansen, Zhang, Sobel and Chowdury, 2009). Because of these abilities, Twitter is used for brands to boost conversations with consumers or to facilitate word-of-mouth communication with target customers (Chu et al., 2016; Fogel, 2010; Jansen et al., 2009).

Social Networking Websites refer to “channels on which users can create an individual ‘profile’ page, find and add friends and contacts, send messages and update their personal profiles to notify friends, contacts or colleagues about themselves” (Berthon et al., 2012; Mills and Plangger, 2015, p.525). Facebook is the most popular Social Networking Website, with an expected user base of 1.82 billion users in 2020 (eMarketer, 2016). This site offers its users the possibility to use other social media technologies as well, since posting a ‘status update’ is similar to micro-blogging, posting ‘notes’ is similar to blogging, and posting videos and pictures resembles features of Video- and Picture-Sharing Websites (Mills and Plangger, 2015).

Of these social media forms, Facebook and Twitter yield the best results regarding influencing consumers’ attitudes toward brands (Kim and Ko, 2012; Kumar and Mirchandani, 2012). Brands need to ascertain their presence on these social media (Alves et al., 2016; Porter and Donthu, 2008; Xie and Lee, 2015). Therefore, this study focuses on Facebook and Twitter, and hence it is useful to explore them briefly. Both platforms provide several opportunities for companies, such as building and maintaining consumer-brand relationships, building brand communities, capturing market intelligence, advancing customer engagement, servicing consumers, driving consumers to websites, and generating leads (Lamberton and Stephen, 2016). Following brands on Facebook and Twitter may have an impact on consumers’ sharing behaviour regarding brand-related information and their intention to purchase products of these brands (Chu et al., 2016), as it increases customers’ exposures to brand-related content posted on brand pages or shared between customers’ social connections (Xie and Lee, 2015). As such, companies can commercially utilize the value that is created from subsocial network relationships (Xie and Lee, 2015; Zwass, 2010).

Motivations for consumers to follow brands on social media

Uses and gratifications (U&G). The uses and gratifications approach to online media

is a functionalist viewpoint that focuses on the users of social media (Muntinga et al., 2011). In contrast to communicator-centric perspectives that examine media effects, the U&G research tradition studies media effects from the perspective of an individual user (Aitken, Gray and Lawson, 2008). As opposed to study how media affect consumers, U&G has been used to explore the drivers of consumers for using media (Katz, 1959; Katz, Blumler and Gurevitch, 1975). Although U&G originates from traditional media, it is applicable for social media research, since traditional as well as social media necessitate users to actively participate, while U&G assumes consumers to actively and selectively use media (Eighmey, 1997; Ruggiero, 2000), and this approach has been proven successful in social media research (e.g., Sheldon, 2008; Quan-Haase and Young, 2010) and in brand-related social media research (Azar et al., 2016; Kwon et al., 2014).

According to McQuail, Blumler and Brown (1972), four main categories of motivation to select media can be distinguished: personal identity, diversion, personal relationships, and surveillance. In recent literature on U&G slightly different category labels have been used, although the content of the categories is similar: personal identity, integration and social interaction, information, and entertainment (Calder, Malthouse and Schaedel, 2009; Muntinga et al., 2011). The different motivations consist of sub-motivations, as described below.

Personal identity. This motivation variable focuses on the self (Saridakis, Baltas, Oghazi, Hultman, 2016), and Nov (2007) found personal identity to be a driver of online engagement. Personal identity covers three sub-motivations: self-expression, self-assurance, and self-presentation. Consumers can use a brand to shape or express their personality, and/or identity, which constitutes the first aspect of this motivation: self-expression (Muntinga et al., 2011). In this way, brands are used to extend one’s personality (Saridakis et al., 2016).

Furthermore, brands can be used to gain positive feedback from other people, and as such receive recognition and get self-confidence (Lampel and Bhalla, 2007; Muntinga et al., 2011), constituting the second sub-motivation. Finally, constituting the third sub-motivation, consumers can use brands in order to present their personality to others by “contributing to brand-related content” (Muntinga et al., 2011, p.29). Shao (2009) studied self-expression in social media use. He defined self-expression as “the expression of one’s own identity, especially one’s individuality” (Shao, 2009, p.14). For the purpose of this study, personal identity will be referred to as Shao’s definition of self-expression, since this definition covers all aspects of personal identity.

Information seeking. Park, Kee and Valenzuela (2009) found consuming information as an incentive itself to affect internet usage. Sub-motivations for this motivation include knowledge, inspiration, surveillance, and pre-purchase information (Saridakis et al., 2016). The knowledge aspect implies that users can benefit from other users’ expertise and knowledge by consuming information that is related to brands, while inspiration denotes users to consume brand-related information in order to get inspired by brands or products (Muntinga et al., 2011; Saridakis et al., 2016). Furthermore, the surveillance aspect refers to “observing and staying updated about one’s social environment” (Muntinga, 2011, p.27). Finally, people can read product reviews on brand communities to facilitate decision-making about brands or products (Saridakis et al., 2016).

Entertainment seeking. Entertainment seeking includes several media-related gratifications, such as emotional relief, relaxation, passing time, and enjoyment (Muntinga et al., 2011; Saridakis et al., 2016). For example, consumers might interact online to escape from everyday life (Courtois, Mechant, Marez and Verleye, 2009). The need for entertainment has been found as a motive to select media (McQuail et al., 1972), and specifically as a motivation to interact online (Park et al., 2009; Shao, 2009). Most U&G researchers studying social media

refer to entertainment as one overall motivation – implying no sub-motivations are included (e.g., Park et al., 2009; Shao, 2009).

Social-interaction seeking. Social-interaction includes multiple gratification, all less related to the self and more to other people (McQuail et al., 1972). This motivation variable consists of sub-motivations, such as helping and seeking support regarding brand-related questions, connecting with others (e.g., joining a brand community to get in touch with like-minded people) (Daugherty, Eastin and Bright, 2008; Kaye, 2007), and getting a sense of belonging (e.g., joining a brand community to establish ‘us vs. them’ regarding other brand users) (Boyd, 2008; Muntinga, 2011; Saridakis et al., 2016).

Incentive seeking. Besides the four main categories of motivation as identified by McQuail et al. (1972), incentive seeking is another important motivation variable for using social media for brand-related purposes (Hennig-Thurau, Gwinner, Walsh and Gremler, 2004), and implies that people might “expect to gain some kind of future reward” (Muntinga et al., 2011, p.21). For this, different types of rewards can be distinguished, relating to personal wants (e.g., gaining peer recognition) (Hars and Ou, 2002), to jobs (e.g., peer recognition) (Hars and Ou, 2002), or to financial incentives (e.g., winning a prize or money) (Wang and Fesenmaier, 2003).

Brand-specific constructs. Besides individual-level motivations for interacting with

brands online, consumers might be influenced by brand-specific factors as well (e.g., Kwon et al., 2014; Logan, 2014). In other words, if consumers do not like a brand, this might inhibit them from following this brand on social media, even though their individual-level motivations (e.g., incentive seeking) are high. Two main factors in this context have been identified: brand commitment and brand likeability.

Brand commitment. Brand commitment refers to “an emotional or psychological attachment to a brand within a product class” (Beatty and Kahle, 1988, p.4). In other words,

brand commitment is about positive feelings related to the attachment towards a particular brand, and has been found as a motivation for engaging and interacting with brands online (Wolny and Mueller, 2013). Brand commitment and brand involvement are often used interchangeably, but can, however, be considered two different constructs (Beatty and Kahle, 1988). Whereas brand commitment is focused on a particular brand, brand involvement “addresses a general level of interest or concern in an issue (i.e., with a product class or a purchase in that product class) without reference to a specific position or choice (i.e., brand)” (Beatty and Kahle, 1988, p.4).

Brand likeability. Customers either like a brand or not. The perceived likeability of a brand can be defined as the psychological factor influencing customers’ reactions towards this brand (Nguyen, Melewar and Chen, 2013). Byrne and Rhaney (1965) suggest that customers are specifically involved with brand likeability when they find a brand attractive. Sung, Kim, Kwon and Moon (2010) found brand likeability to be a significant driver of consumers’ future intention to participate in brand communities.

Type of brand. Brands can be classified in different product categories, such as goods

or services (Holbrook and Hirschman, 1982). However, in previous research on consumer behaviour product categories are most often classified as hedonic or utilitarian (e.g., Batra and Ahtola, 1991; Roy and Ng, 2012), due to the “distinct characteristics of each of these categories” (Chen, Kim and Lin, 2015, p.210). Hedonic products have intangible, subjective features that create pleasure or enjoyment, while utilitarian products have objective, tangible features that produce utility and functional benefits (Chaudhuri and Holbrook, 2002). It should be noted, however, that products can have both types of features as well as no types of features at all (Addis and Holbrook, 2001), but the relative balance between the two is of interest here. Hence, this classification scheme is more like a continuum than a dichotomy, since products

can be placed along a continuum ranging from hedonic to utilitarian products (Chaudhuri and Holbrook, 2002).

Type of platform. As described earlier, this study focuses on Facebook and Twitter as

social media platforms. Both platforms provide the ability to reach several other platform users at once. However, the ability to create microblogs, one of Twitter’s main features, implies a unique form of communication due it its undirected nature (Buechel and Berger, 2016). Unlike Facebook, Twitter offers the possibility to display messages to potentially everyone, in which every user of the network can freely decide to respond to this message or not. This facilitates communication (Bargh and McKenna, 2004; Forest and Wood, 2012), making it suitable for situations in which interaction with others or potentially brands is desired (Buechel and Berger, 2016). Hence, there might be a difference between the findings of this research for Facebook and Twitter users for the hypothesized relationships.

Intention to follow brand. Customers can follow brands on Facebook or Twitter by

clicking ‘like’ or ‘follow’, respectively (Bunker, Rajendran and Corbin, 2013; Kwon et al., 2014). These click options allow customers to broadcast their preferences and interests, hence showing friends who they follow and/or support; a relationship can be affirmed publicly this way (Bunker et al., 2013). As such, the number of ‘likes’ of a brand on Facebook or Twitter may indicate the degree of popularity (Bunker et al., 2013). Turner and Shah (2011) found customers to like companies on social media, to broadcast their association with these companies to others. A like may also indicate future behaviour, such as word-of-mouth (Kim, Shim and Ahn, 2011). In turn, word-of-mouth may substantially impact customers’ perceptions about products (Simmons et al., 2011). Besides, in a research among Twitter users in which a sample of 12,000 respondents were surveyed, and in which the results were combined with actual behaviour data, 54 percent of the respondents reported that after they have seen brand mentions in tweets, they have taken action regarding this brand (Midha, 2014). For example,

23 percent of the respondents reported to visit the website after being exposed to brand mentions, while 20 percent reported to visit the brand’s Twitter page or search for the brand online (Midha, 2014). Either ‘liking’ a brand on Facebook or ‘following’ one on Twitter increases the likelihood of seeing brand mentions, and hence it is important to understand the factors influencing consumers’ intention to like or follow brands on these social media sites.

HYPOTHESES

Interaction between drivers and type of platform Self-expression

Previous research assumes that when people have a need for self-expression, they like to present their inner self or ‘true’ identity to others (Bargh and McKenna, 2004; Goffman, 1959; McKenna and Bargh, 1999), and to get others to know them like they know themselves (Shao, 2009). This can be operationalized on social media through activities that allow to show oneself and to show how important one is, such as clicking like or following a brand (Shao, 2009). This gratification need may be satisfied by following a brand, as friends are able to see this, and allows users to show the importance of who there are and what they do (Shao, 2009).

Gao and Feng (2016) found self-expression to be a driver of interaction with brands on both microblogs and Social Networking Sites (SNSs), such as Twitter and Facebook. Muntinga et al. (2011) studied consumers’ online brand-related activities (COBRAs) and worked towards a typology. Three different types of COBRA can be distinguished: content consumption, content contribution, and content creation. Following a brand page can be classified as content contribution, since it involves a higher level of activity of consumers than only passively consuming brands’ content (Muntinga et al., 2011). Self-expression is found to drive both content contribution and content creation (Muntinga et al., 2011; Saridakis et al., 2016; de Vries, Peluso, Romani, Leeflang and Marcati, 2017). This implies that self-expression might

drive consumers to follow a brand on Facebook or Twitter as well, since this activity adheres to content contribution. Finally, Muntinga et al. (2011), and Gao and Feng (2016) found self-expression to be one of the two most important motivations for consumers to follow brands on social media. Prior literature suggests that self-expression need is a consistent driver across different social media channels. Hence, the following hypothesis is proposed:

H1: Self-expression behaviour will positively influence consumers’ intention to follow a

brand to an equal extent for Facebook and Twitter.

Information seeking

Althaus and Tewskbury (2000) found information seeking behaviour to remain consistent across different forms of media. For example, people who often read newspapers ‘offline’ tend to extend their information seeking behaviour ‘online’. However, in social media research, the role of information gratifications sought by users is not clear and might be different for Facebook and Twitter. On the one hand, Agrifoglio, Black, Metallo and Ferrara (2012), and Witkemper, Lim and Waldburger (2012) found users to exhibit information seeking behaviour on Twitter, while Raacke and Bonds-Raacke (2008) and Bonds-Raacke and Raacke (2010) found the same behaviour among Facebook users. On the other hand, Langstedt (2013) suggests that users of social media are more passively consuming information rather than actively communicating information. This might imply that information seeking does not lead consumers to follow brands on Facebook or Twitter, as they need to perform an activity (e.g., click ‘like’) before they can passively consume information. However, Kwon et al. (2014) found information seeking to be a motivation for consumers to follow brands on Twitter, while Azar et al. (2016) found the same relation for Facebook users, and Saridakis et al. (2016) for both Facebook and Twitter.

Logan (2014) studied information seeking as a driver of users to follow brand pages on both Facebook and Twitter. Specifically, she found information seeking to significantly influence brand following intention for Twitter users, but not for Facebook users. Besides, Shimp, Wood and Smarandescu (2007) found users to follow brands pages on Twitter because of its informative features. Gao and Feng (2016) found the information seeking need to be better satisfied by microblogs (e.g., Twitter) than by SNSs (e.g., Facebook). Prior literature suggests that Facebook and Twitter might be used for different information gratification purposes regarding following brands. Therefore, the following hypothesis is proposed:

H2: Information seeking will positively influence consumers’ intention to follow a brand to a

greater extent for Twitter compared to Facebook users.

Entertainment seeking

Previous literature on social media provides that many social media users use SNSs to relax and to be entertained or to ‘kill time’ (Shu and Chuang, 2011, p.29). More specifically related to brands, entertainment can be a motivation to join brand pages to be entertained (Dholakia, Bagozzi and Pearo, 2004; Rohm, Kaltcheva and Milne, 2013), and to have fun and to be amused (Enginkaya and Yilmaz, 2014). Madupu and Cooley (2010) identified entertainment to drive consumers to participate in brand communities, as well as Sung et al. (2010) did. Entertainment is found to be one of the main motivations for consumers to join these communities as well (Chi, 2011; Dholakia et al., 2004). However, Kwon et al. (2014) did not find entertainment to be a gratification motivation to follow brands on Twitter. Besides, Saridakis et al. (2016) did not find entertainment to drive consumers to contribute to brand content for both Facebook and Twitter (e.g., following brands). Azar et al. (2016) found entertainment gratifications to drive consumers to follow brand pages on Facebook, while Gao and Feng (2016) find the same

relationship to be consistent across Facebook and Twitter users, although to a greater extent for Twitter compared to Facebook users. As findings in previous are not consistent, the following hypothesis will be proposed:

H3: Entertainment seeking will positively influence consumers’ intention to follow a brand to

an equal extent for Facebook and Twitter users.

Social-interaction seeking

Social-interaction seeking is found to be an important driver for brand page users to contribute to brands’ content, such as following brands (Gummerus, Lilijander, Weman and Philström, 2012; Muntinga et al., 2011). Moreover, according to Jahn and Kunz (2012), establishing social interaction in brand pages is of utmost importance to attract users to these pages. Whereas Saridakis et al. (2016) did not find social-interaction seeking to influence consumers’ content contribution for Facebook and Twitter (e.g., following brand pages), Muntinga et al. (2011) did. Kwon et al. (2014) found social-interaction seeking behaviour to drive consumers to interact with brands on Twitter. Finally, Gao and Feng (2016) found social-interaction seeking gratifications to be better satisfied by SNSs (e.g., Facebook) than by microblogs, such as Twitter. Given that Facebook is a platform that is more socially oriented compared to Twitter, the following hypothesis is proposed:

H4: Social-interaction seeking will positively influence consumers’ intention to follow a brand

Incentive seeking

Incentive seeking behaviour (i.e., remuneration) was not found to drive consumers to contribute to brands’ content (the second COBRA type) for Facebook nor Twitter users (Muntinga et al., 2011). Other authors did find incentive seeking behaviour to drive consumers’ content consumption on these platforms, but no relation between incentive seeking behaviour and content contribution was found (Saridakis et al., 2016). However, Kwon et al. (2014) found incentive seeking behaviour to be a main driver of consumers’ intention to follow brands on Twitter, while Azar et al. (2016) found reward motivation (e.g., time savings or deals) to drive consumers’ interaction with a brand on Facebook. Besides, consumers can join brand pages to attain economic gratifications (e.g., discounts) or to join competitions on Facebook (Gummerus et al., 2012), or to gain material benefits on social media (Martins and Partício, 2013). Finally, according to Tsai and Men (2013), consumers’ main motivation for following brands on Facebook is to gain economic benefits or other incentives. Hence, the following hypothesis is proposed:

H5: Incentive seeking will positively influence consumers’ intention to follow a brand to an

equal extent for Facebook and Twitter users.

Interaction between drivers and type of brand

A brand can be used by customers as a way to express themselves to others and to the society (Aaker, 1999; Belk, 1988). Often, customers develop a sense of affinity with a brand and will purchase the brand if it expresses their self-concept (Alex and Joseph, 2012). Consumers tend to personify brands based on perceived human traits that can be attributed to it, so as to establish a congruency fit between the brand’s personality and their own personality (Aaker, 1997; Geuens, Weijters and de Wulf, 2009; Grohmann, 2009; Sirgy, 1982). When such a fit occurs,

it provides consumers a sense of ease which plays an important role in self-expression (Aaker, 1999; Sirgy, 1982). Malär, Krohmer, Hoyer and Nyffenegger (2011) suggest that the effects of the congruence between the self-concept and a brand’s personality may be more important for brands with hedonic value as compared with utilitarian brands for which other functionalities are important (Alex and Joseph, 2012). Hence, the following hypothesis is proposed:

H6: Self-expression behaviour will influence consumers’ intention to follow a brand on

Facebook or Twitter to a greater extent for hedonic brands compared to utilitarian brands.

As suggested by Hirschman and Holbrook (1982), consumers experience hedonic products differently than utilitarian products. With hedonic brands, consumers attain benefits such as sensory and emotional gratifications, while consumers gain functional gratifications and utility with utilitarian brands. The underlying nature of social media, being socially oriented, “may predispose consumers to expect brand-related information on Facebook pages to tap into their emotions” (Chen et al., 2015, p.211; Schulze, Schöler and Skiera, 2014). This suggests that consumers, while evaluating a hedonic brand, will rely more on their affective feelings. However, Dijksterhuis, Smith, van Baaren and Wigboldus (2005) found the same behaviour among consumers evaluating utilitarian products, which implies that consumers unconsciously rely less on the functional aspects of these products. Therefore, it is expected that the type of brand (i.e., hedonic) will moderate the relation between information seeking behaviour and intention to follow a brand.

H7: Information seeking behaviour will influence consumers’ intention to follow a brand on

Hedonic brands relate to affective feelings and emotions (Addis and Holbrook, 2001; Holbrook and Hirschmann, 1982). Besides, one might express oneself in order to get “the sense of being cut off from down-to-earth sights, sounds, textures, and emotions of everyday existence” (Toffler, 1980, p.404), which relates to escapism – one of the expressions of entertainment seeking behaviour. Utilitarian brand benefits refer to “the functional and practical benefits of consumption” (Alex and Joseph, 2012, p.79), while hedonic brand benefits refer to “aesthetic, experiential, and enjoyment-related benefits” (Alex and Joseph, 2012, p.79). Chitturi, Raghunathan and Mahajan (2007) suggest that after consumers meet a specific level of functional benefits in a brand, they search for hedonic benefits. This suggest that after consumers perceive a certain threshold of functional value is attained by a brand, they will take hedonic value into consideration. Hence, the following hypothesis is proposed:

H8: Entertainment seeking behaviour will influence consumers’ intention to follow a brand on

Facebook or Twitter to a greater extent for hedonic brands compared to utilitarian brands.

The motivation for social-interaction may drive consumers to follow brands on social media (e.g., Kwon et al., 2014; Muntinga et al., 2011), mainly to connect with others, and to help others with using the brand. However, the moderating role of the type of brand is still unclear. In addition, as reward motivation (i.e., incentive seeking) relates to “the degree to which the community members want to gain utilitarian rewards through their participation in the community” (Baldus, Voorhees and Calantone, 2015, p.981), it could be expected that incentive seeking behaviour will affect consumers’ intention to follow brands more strongly for utilitarian brands. However, the possible moderating role of the type of brand cannot be established on prior literature. Consequently, the moderator role of the type of brand on these relationships (i.e., social-interaction seeking and incentive seeking) is to be found empirically.

As can be seen in Figure 1, a conceptual model is drawn upon the proposed hypotheses.

FIGURE 1: RESEARCH DESIGN

METHODOLOGY

Data collection and sampling method

According to Edmonson and McManus (2007), theories in management research can be placed along a continuum. This continuum comprises three states of theory, starting from mature, to intermediate, to nascent. Mature theory proposes “well-developed constructs and models that have been studied over time with increasing precision by a variety of scholars, resulting in a body of work consisting of points of broad agreement that represent cumulative knowledge gained” (Edmonson and McManus, 2007, p.1158). On the other hand, nascent theory represents

“tentative answers to novel questions of how and why, often merely suggesting new connections among phenomena” (Edmonson and McManus, 2007, p.1158). In between, intermediate theory can be positioned, presenting “provisional explanations of phenomena, often introducing a new construct and proposing relationships between it and established constructs” (Edmonson and McManus, 2007, p.1158). The current state of research on social media may be considered nascent. However, the examined constructs that have been long studied. As such, a quantitative research method is justified. An explanatory research design was used in order to gain insights on the proposed research questions (Zikmund, Babin, Carr and Griffin, 2013). For this, an experimental approach was employed. In order to collect data, a web-based survey was used. A 2 (Facebook vs. Twitter) x 2 (hedonic brand vs. utilitarian brand) between subject factorial research design was used, which resulted in four different surveys. As such, this study aimed to maximize variance in both the independent as well as moderator variable. Randomly, surveys were assigned to respondents. Finally, a pre-test was performed in order to ensure maximum variance. By means of this pre-test, errors in the design, such as vagueness, can be detected, as well as the instrument could be finalized and refined (Blumberg, Cooper and Schindler, 2014).

Sample

The population of this study consists of Dutch Facebook and Twitter users. Due to time and budget constraints, convenience sampling will be used, which is a form of non-probability sampling of which the samples are unconstraint (Blumberg et al., 2014). Although this method cannot control precision, it is the best possible way to conduct research given the aforementioned constraints. Snowball sampling will be used as well, which implies that after an identified sample subjects filled in a survey, he or she will be asked to identify similar subjects (Blumberg et al., 2014).

At least 50 responses per survey (200 respondents in total) are necessary to ascertain that the sample is large enough for analysis. To do so, the surveys were distributed on Facebook and Twitter. However, this makes it difficult to determine the response rate. In order to increase the response rate, two times 50 euro were raffled among respondents who indicated they like to enter the raffle. The survey was build using Qualtrics and the data was collected in a three-week period around May 2017 .

Pre-test experiment

Logos of 2 (utilitarian vs. hedonic) brands were pretested with 31 respondents. For this, the top 50 brands in The Netherlands in 2016 was studied, as to ascertain the brands are well-known. Ten brands were pre-selected to be included in the pre-test: Albert Heijn, KLM, Philips, Heineken, Gall & Gall, Aegon, Nationale Nederlanden, TomTom, Delta Lloyd, and KPMG. As can be seen in Appendix A and B, to confirm hedonic or utilitarian value of the brands, four items were asked on a seven-point bipolar scale, in which a score of 1 indicates low hedonic value or low utilitarian value, and 7 indicates high hedonic value or high utilitarian value, respectively (Alex and Joseph, 2012; Batra and Ahtola, 1991).

There was a statistically significant effect of brands on levels of utility, F(9, 160) = 50.96, p <.05. Bonferroni post-hoc tests revealed that the perceived level of utility was significantly higher for TomTom compared to Delta Lloyd (p = .02), compared to Heineken (p < .05), and compared to Gall & Gall (p < .05). There was no statistically significant difference between all other brands. Therefore, TomTom, Delta Lloyd, Gall & Gall, and Heineken were included in the questionnaire.

Research design

four different Facebook posts and four different tweets were created, as if they belonged to either a utilitarian brand (i.e., TomTom or Delta Lloyd) or a hedonic brand (i.e., Gall & Gall or Heineken). The posts and tweets included a neutral image related to King’s Day in The Netherlands and a text with “Happy K8ing’s Day! #kingsday2017”. Respondents were evenly but randomly assigned to one of the surveys. Respondents were asked to look at it and after this answer the questions related to their intention to follow the brand, brand likeability, and brand commitment. Then, respondents were asked to their general motivations to follow brands on either Facebook or Twitter, after which they were asked about social media use, and demographic variables.

Measurement items

In order to measure the constructs used in the hypotheses, established scales from brand-related social media use studies were used (see Appendix C, E and E). When necessary, the measurement items were adjusted to this specific study. Since the constructs do not have “a common scale” (Blumberg et al., 2014, p.192), which means that it is not possible to measure the constructs in commonly accepted numbers, they were measured using a seven-point Likert scale. This scale ranged from 1 (“strongly disagree”) to 7 (“strongly agree”). Since this study is carried out in The Netherlands, the measurement items were translated into Dutch by means of the translation-back translation approach (Brislin, 1970).

Regarding the independent variables, self-expression (a = .89) adapted from Gao and Feng (2016), information seeking (a = .87) adapted from Sung et al. (2010), entertainment seeking (a = .85) adapted from Kwon et al (2014), and Sung et al. (2010), and incentive seeking (a = .83) adapted from Sung et al. (2014) were measured using a 5-item scale. Six items were used to measure social-interaction seeking (a = .79) adapted from Kwon et al. (2014). The brand-specific constructs brand likeability (a = .92) adapted from Sung et al. (2010), and brand

commitment (a = .77) were measured with a 7-item and 5-item scale, respectively. The independent variable, intention to follow a brand on Facebook or Twitter (a = .93) adapted from Logan (2014) was measured with a 3-item scale.

Control variables

First, the questionnaire asked respondents if they either have Facebook (“yes” or “no”) or Twitter (“yes” or “no”), in order to determine their qualification to take part in this research. As described above, the different surveys were distributed evenly, but when respondents answered no on the first type of platform that was being asked about (e.g., Facebook), they were redirected to the other type of platform (e.g., Twitter). When both questions were answered with “no”, the respondent did not satisfy the conditions to participate in this research, and he or she was thanked for his or her time. In addition, as a manipulation check, respondents were asked to indicate whether a brand was utilitarian or hedonic on a seven-point bipolar scale ranging from 1 (“utilitarian”) to 7 (“hedonic”). At the end of the survey, respondents were asked for some demographic variables, such as gender, age and education level. Also, respondents were asked to indicate variables related to their social media use, such as Facebook or Twitter duration, minutes using per day, number of brands following, and frequency to log on to Facebook or Twitter. An overview of the control variables can be found in Appendix F.

Plan of analysis

The analyses were performed by means of SPSS. First, unnecessary information (e.g., recorded date, IP address) was removed from the dataset. Second, the variables were labelled and the dataset was checked for extreme values and outliers. Whenever these occurred, the corresponding data would have been removed. Next, no recoding of counter-indicative items was necessary, since all constructs were measured in the same direction. In addition, descriptive

statistics were determined. Fifth, reliability analyses were performed in order to check the internal consistency of the items. When the items had a high reliability, they were computed into a new variable in order to represent the whole variable in one. Also, before testing the hypotheses, a correlation analysis was done in order to test whether the constructs correlated.

Before the statistical test was determined, the measurement levels of the variables were evaluated. Since the independent and dependent variables were measured on a 7-point Likert scale, these variables could be categorized as interval level. In order to test the hypotheses, a multiple hierarchical regression analysis was performed. For this, the control variables were included as covariates to determine whether these influenced the independent variable.

RESULTS

Extreme and missing values

First, unnecessary information was removed from the dataset. Also, the constructs were labelled. The dependent and the independent variables were measured by using a 7-point Likert scale. As such, it was assumed that possible outliers on these scales were meaningful. Consequently, no extreme values were found for these variables. However, three control variables, including age, minutes per day and total duration, were measured on a continuous scale. As such, it could have been possible for respondents to make a typing error. By looking at the boxplots of these variables, no extreme values were found. Furthermore, the sample consisted of 346 respondents. Of this sample, 343 (99.1%) respondents met the requirements to fill in the questionnaire, and from the remaining amount 86 (35.4%) responses were incomplete. Hence, these were listed as missing values and were eliminated from the dataset, which resulted in a sample with 257 respondents. Since all constructs were measured in the same direction, no recoding of counter-indicative items was necessary. Finally, two dummy

variables were created, one for the type of platform (“0 = Facebook”, “1 = Twitter”), and one for the type of brand (“0 = utilitarian”, “1 = hedonic”).

Descriptive statistics

The sample consisted of 257 respondents. Of these, 100 (38.9%) were male, whereas 157 (61.1%) were female. The average age of the respondents was 32.3 (SD = 13.2); the youngest respondent being 16 years old, and the oldest one being 88 years old. Furthermore, the education level of the respondents was relatively high, since 150 (58.4%) of them had a university degree, 70 (27.2%) of them had a Dutch HBO degree (‘University of Applied Sciences’), 31 (12.1%) of them had a Dutch MBO degree (‘community college’), and 6 (2.3%) of them had a high school degree. Regarding the experiment conditions on platforms, 138 (53.7%) of the respondents were assigned to Facebook, while 119 (46.3%) of them were assigned to Twitter. For the type of brand, 129 (50.2%) of the respondents were assigned to a utilitarian brand, whereas 128 (49.8%) of them were assigned to a hedonic brand. With respect to the manipulation check, the respondents’ answers on whether they perceived the brand to be utilitarian or hedonic corresponded with the actual type of brand.

Reliability analyses

Reliability analyses were performed in order to check for internal consistency in the measurement items. Intention to follow brand (α = .930), self-expression (α = .884), information seeking (α = .838), entertainment seeking (α = .873), social-interaction seeking (α = .871), incentive seeking (α = .905), brand likeability (α = .921) and brand commitment (α = .851) had high reliability, since the Cronbach’s Alpha’s (α) were higher than .7. In addition, since all items were above .3, all of them have a good correlation with the total score of the scale. When eliminating items, the reliability would not substantially be affected (D<.10).

Consequently, five new variables were computed by combining the items into new variables with an average value: self-expression total, information seeking total, entertainment seeking total, social-seeking total, and incentive seeking total.

Correlation analysis

A correlation analysis was performed in order to test whether the variables correlate. The results of this analysis can be seen in Table 1. As can be seen from this table, number of brands following on Facebook or Twitter, age, brand likeability, brand commitment, self-expression, information seeking, entertainment seeking, social-interaction seeking, and incentive seeking correlated significantly (p < .01) with intention to follow brand, as well as Facebook or Twitter duration did (p < .05). Specifically, brand likeability (r = .62), brand commitment (r = .50), and incentive seeking (r = .44) had a strong linear relationship with intention to follow brand, whereas number of brands following (r = .30), self-expression (r = .28), information seeking (r = .32), entertainment seeking (r = .26), and social-interaction seeking (r = .39) correlated medium. Furthermore, Facebook or Twitter duration (r = .13), and age (r = -.17) correlated weakly with intention to follow brand. Finally, minutes using per day (r = .03), frequency to log on to Facebook or Twitter (r = -.02) and education level (r = -.01) were not found to significantly correlate with intention to follow a brand on Facebook or Twitter.

TABLE 1

MEANS, STANDARD DEVIATIONS, AND CORRELATIONS

(CONTINUED)

M SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14

1. Facebook/Twitter

duration 6.43 2.87

2. Minutes using per

day 45.15 47.36 .12* 3. Frequency to log on 4.89 1.68 .23** .38** 4. Number of brands following 2.19 1.54 .14* .12 .20** 5. Age 32.26 13.08 -.21** -.18** -.13** -.33** 6. Education level 1.58 .81 -.13* -.01 .03 .00 .28** 7. Brand likeability 2.85 1.38 .17** .09 .01 .26** -.22** -.06 (.92) 8. Brand commitment 3.03 1.30 .12 .12* .03 .24** -.21** -.02 .77** (.85) 9. Self-expression 3.46 1.45 .15* .09 .12 .24** -.22** .03 .28** .32** (.88) 10. Information seeking 4.07 1.31 .15* .22** .18** .28** -.25** .06 .31** .37** .41** (.84) 11. Entertainment seeking 4.33 1.37 .11 .13* .02 .31** -.44** -.02 .36** .33** .39** .50** (.87)

TABLE 1 (CONTINUED)

MEANS, STANDARD DEVIATIONS, AND CORRELATIONS

** Correlation is significant at the .01 level (2-tailed) * Correlation is significant at the .05 level (2-tailed)

M SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 12. Social-interaction seeking 3.39 1.34 .15* .13* .07 .28** -.25** .02 .36** .41** .69** .50** .38** (.87) 13. Incentive seeking 3.75 1.58 .21** .17** .12* .30** -.41** -.21** .42** .40** .50** .48** .41** .52** (.91) 14. Intention to follow brand 2.29 1.28 .13* .03 -.02 .30** -.17** -.01 .62** .50** .28** .32** .26** .39** .44** (.93)

Testing the hypotheses

A hierarchical multiple regression analysis was conducted to test whether self-expression, information seeking, entertainment seeking, social-interaction seeking, and incentive seeking drive consumers’ intention to follow a brand on Facebook or Twitter, after controlling for gender, education level, age, Facebook or Twitter duration, minutes using per day, frequency to log on, and number of brands following. In addition, this regression analysis was performed to test if the type of platform, and the type of brand moderated these relationships. For this, ten new variables were computed to display the interaction effects (see variables with termed “interaction” in Table 2). Finally, driving ability of brand likeability, and brand commitment on intention to follow brand was investigated. The outcomes of this hierarchical multiple regression analysis can be found in Table 2.

In the first part of the hierarchical multiple regression analysis, seven drivers – control variables - were entered: gender, education level, age, Facebook or Twitter duration, minutes using per day, frequency to log on, and number of brands following. This model was significant F (7, 251) = 4,41; p < .001, and explained 10.9% of the variance in intention to follow brand. Only number of brands was found to significantly influence intention to follow a brand t (251) = 4.17; p < .001, although the effect is relatively medium (effect = .27, SE = .05).

In the second part of the analysis, the control variables, as well as brand likeability, brand commitment, self-expression, information seeking, entertainment seeking, social-interaction seeking, incentive seeking, the social-interaction effects of the type of platform (Facebook or Twitter) with the independent variables, and the interaction effects of the type of brand (utilitarian or hedonic) with the independent variables were entered. This model as a whole explained 48.2% of the variance in intention to follow brand and was significant F (26, 232) = 8,30; p < .001. After controlling for Facebook or Twitter duration, minutes using per day, frequency to log on, numbers of brands following, gender, age, and education level, the

independent variables as well as the moderators explained an additional 37.3% of the variance in intention to follow brand (R2 Change = .37; F (19, 232) = 8.78; p < .001). Brand likeability was found to positively influence intention to follow a brand (effect = .48, SE = .07), and its effect is significant, t (232) = 6.08; p < .001. However, brand commitment was not found as a driver of intention to follow a brand (p = .881).

The regression coefficient for the independent variable self-expression is .10, and was

found to be statistically significant t (232) = .96; p < .01. However, no interaction effect of self-expression and the type of platform was found, since its influence (effect = .02) is not significant (p = .94). This means that the relation between self-expression and intention to follow a brand does not depend on the type of platform. As such, self-expression affects intention to follow a brand to an equal extent for Facebook as well as Twitter users. Consequently, H1 is supported.

As can be seen from Figure 2, the regression coefficient for the interaction effect of information seeking and the type of platform is .18 and is statistically significant, t (232) = 1.9; p < .05. This means that the influence of information seeking on intention to follow depends on the type of platform. The conditional effects indicated that the relationship between information seeking and the intention to follow brand as significant for both Facebook users (effect = .18, SE = .09, p < .01) as well as Twitter users (effect = .605, SE = .10, p < .001). As such, information seeking positively influences intention to follow a brand for both Facebook and Twitter users, although to a greater extent for Twitter users. Consequently, H2 is supported.

The regression coefficient for entertainment seeking was found to be .08. However, its effect is not found to be significant (p = .41). In addition, the regression coefficient for the interaction effect of entertainment seeking and the type of platform is .08, but not statistically significant, t (232) = 83, p = .720. Consequently, H3 is not supported.

TABLE 2

HIERARCHICAL REGRESSION MODEL OF INTENTION TO FOLLOW BRAND

(CONTINUED) R R2 R2 change B SE ß t Control variables .33 .11*** Gender -.07 .16 -.03 -.42 Education level .04 .10 .03 .43 Age -.01 .01 -.08 -1.23 Facebook/Twitter duration .05 .03 .10 1.59

Minutes using per day .00 .00 .02 .24

Frequency to log on -.08 .05 -.11 -1.57

Number of brands following .23 .05 .27*** 4.17

Model variables .69 .48*** .37***

Gender -.10 .13 -.04 -.75

Education level .07 .08 .04 .83

Age .00 .00 .03 .54

Total Facebook/Twitter duration .00 ,03 .01 .16

Minutes using per day -.00 .00 -.02 -.38

Frequency to log on -.07 .05 -.09 -1.34

Number of brands following .12 .05 .14* 2.55

Brand likeability .45 .07 .48*** 6.08

Brand commitment .01 .08 .01 .15

Self-expression .10 .10 .10** .96

Information seeking .16 .09 .18 1.74

TABLE 2 (CONTINUED)

HIERARCHICAL REGRESSION MODEL OF INTENTION TO FOLLOW BRAND

* Significant at p < .05 ** Significant at p < .01 *** Significant at p < .001 R R2 R 2 change B SE ß t Social-interaction seeking .17 .08 .18* 2.18 Incentive seeking .12 .08 .15* 1.54

Dummy Facebook Twitter .07 .61 .02 .11

Interaction Facebook Twitter *

Self-expression .01 .16 .02 .08

Interaction Facebook Twitter *

Information seeking .33 .17 .42* 1.94

Interaction Facebook Twitter *

Entertainment seeking .08 .09 .08 .83

Interaction Facebook Twitter *

Social-interaction seeking .17 .08 .18 2.18

Interaction Facebook Twitter *

Incentive seeking .12 .08 .15 1.54

Dummy utilitarian hedonic .16 .48 .06 .32

Interaction utilitarian hedonic *

self-expression .07 .12 .12 .59

Interaction utilitarian hedonic *

Information seeking .13 .11 .20 1.21

Interaction utilitarian hedonic *

Entertainment seeking .01 .11 .01 .05

Interaction utilitarian hedonic *

Social-interaction seeking .01 .09 .01 .12

Interaction utilitarian hedonic *

The regression coefficient for social-interaction seeking is .18, and is statistically significant, t (232) = 2.18; p < .05. However, the type of platform was not found to moderate the relationship between social-interaction seeking and intention to follow a brand (effect = .18, p = .35). This means that social-interaction seeking influence intention to follow a brand, and does not depend on the type of platform. Consequently, H4 is not supported.

The regression coefficient for incentive seeking is .15, and statistically significant, t (232) = 1.54, p < .05. However, no interaction effect for incentive seeking and the type of platform was found, since its effect (.15) was not statistically significant (p = .20). This means that incentive seeking drives consumers to follow a brand, but its effect does not depend on whether these consumers are Facebook or Twitter users. Consequently, H5 is supported.

The regression coefficient for the interaction effect of self-expression and the type of brand is .12. However, it is not found to be statistically significant (p = .554). As described before, self-expression was found to statistically significantly influence intention to follow a brand, (effect = .1; t (232) = .96; p < .01). This means that self-expression drives consumers’ intention to follow a brand on Facebook or Twitter, and its effect does not depend on the type of brand. As such, the effect of self-expression on the intention to follow a brand is equal for utilitarian brands as well as hedonic brands. Consequently, H6 is not supported.

The regression coefficient for the interaction effect of information seeking and the type of brand is .20, and is not found to be statistically significant (p = .228). In addition, the regression coefficient for information seeking was .18, and is not statistically significant (p = .08). This means that information seeking does not drive intention to follow a brand, whenever a utilitarian or hedonic brand is concerned. Consequently, H7 is not supported.

The regression coefficient for the interaction effect of entertainment seeking and the type of brand is .01, which is a small effect, but it is not statistically significant (p = .96). Furthermore, entertainment seeking does not drive consumers’ intention to follow a brand on

Facebook or Twitter, since its effect (.08) is not statistically significant (p = .41). Consequently, H8 is not supported.

The results of the multiple regression analysis and the related outcomes for the hypotheses were summarized and can be found in Table 3.

FIGURE 2 RESULTS

TABLE 3

OUTCOMES OF THE HYPOTHESES

DISCUSSION

The aim of this study was to investigate what factors drive consumers’ intention to follow a brand on Facebook and Twitter. The research model incorporated four variables from uses and gratifications theory, including self-expression, information seeking, entertainment seeking, and social-interaction seeking. Based on previous literature, incentive seeking was included as

No. Hypotheses Outcomes

H1 Self-expression behaviour will positively influence consumers’ intention

to follow a brand to an equal extent for Facebook and Twitter. Supported

H2

Information seeking will positively influence consumers’ intention to follow a brand to a greater extent for Twitter compared to Facebook users.

Supported

H3 Entertainment seeking will positively influence consumers’ intention to

follow a brand to an equal extent for Facebook and Twitter users. Not supported

H4

Social-interaction seeking will positively influence consumers’ intention to follow a brand to a greater extent for Facebook compared to Twitter users.

Not supported

H5 Incentive seeking will positively influence consumers’ intention to

follow a brand to an equal extent for Facebook and Twitter users. Supported

H6

Self-expression behaviour will influence consumers’ intention to follow a brand on Facebook or Twitter to a greater extent for hedonic brands compared to utilitarian brands.

Not supported

H7

Information seeking behaviour will influence consumers’ intention to follow a brand on Facebook or Twitter to a greater extent for hedonic brands compared to utilitarian brands.

Not supported

H8

Entertainment seeking behaviour will influence consumers’ intention to follow a brand on Facebook or Twitter to a greater extent for hedonic

a driver as well. Furthermore, the model incorporated the proposed moderating roles of the type of platform and the type of brand, as well as brand likeability and brand commitment.

First, it was proposed that self-expression, entertainment seeking, and incentive seeking would influence consumers’ intention to follow a brand to an equal extent for Facebook and Twitter users. Regarding these constructs, a moderating effect of the type of brand could not be found. This means that whether the platform concerns Facebook or Twitter does not affect the strength of the effect of self-expression, entertainment seeking and incentive seeking on consumers’ intention to follow a brand, which is consistent with previous studies (e.g., Gao and Feng, 2016; Saridakis et al., 2016). Interestingly, however, entertainment seeking was not found to be a primary motivator for consumers to follow a brand. Whereas entertainment seeking has been found as one of the main drivers of content consumption of brands (e.g., watching videos related to a brand, reading comments on brand pages) (Muntinga et al., 2011; Saridakis, 2016), it does not necessarily drive consumers’ content contribution (Saridakis, 2016). A plausible reason for this might be that content contribution (i.e., following a brand page on social media) requires a greater level of activity (Muntinga et al., 2011). That is, entertainment seeking as a motivator might lead consumers to consume brand-related content, but does not provoke them to follow this brand. In addition, the results suggest that information seeking drives consumers’ intention to follow a brand to a greater extent for Twitter compared to Facebook users, which is also consistent with previous literature (e.g., Gao and Feng, 2016; Logan, 2014). This means that when consumers have a high information gratification motivation, they are more likely to follow brands on Twitter than on Facebook, although the direction of the effects is positive for both groups of users. A moderating role of the type of brand on the effect of social-interaction seeking and intention to follow a brand could not be found. That is, whether the type of platform is Facebook or Twitter does not impact the strength of the effect of social-interaction seeking on consumers’ intention to follow a brand. This outcome contradicts most other literature (e.g.,