Master thesis, 30hp
Customer engagement behavior on social media brand communities
A quantitative study regarding engagement behavior, perceived benefits, and relationship outcome on different social media platforms
Author: Sarah Sjöqvist Supervisor: Mosad Zineldin Examiner: Anders Pehrsson Date: Spring, 2015
Subject: Marketing
Acknowledgements
The author would like to thank professor Mosad Zineldin for guidance throughout the writing process, the examiner professor Anders Pehrsson for important pointers during
the seminars.
The author would also like to express gratitude towards the company Wakakuu and their brand manager for a good and important collaboration when distributing the questionnaires. Without their spared time and help this study would not been possible to
implement.
Also a special thank you to fellow classmates and my family who has been providing great support and guidance throughout the writing process.
Linnaeus University, Växjö Spring 2015
__________________________________
Sarah Sjöqvist
Abstract
Keywords: Customer engagement, relationship benefits, relationship outcome, loyalty, trust, satisfaction, social media, brand communities, Facebook, Instagram, Pinterest.
Background: Social media has provided both companies and customer with new opportunities. Customers are increasingly integrating social media into their daily lives and companies has noticed these new traditional medias and started to take advantage of them through brand communities. The new behavior occurring on brand communities is what research call customer engagement behavior and goes beyond transactional behavior. However, customer engagement has not been fully researched on different social media platforms. The most researched platform to date is Facebook. And with the rapid growth of social media and the constant development of new platforms it is of importance to understand customer engagement behavior on different social media platforms to further being able to adapt to each unique platform.
Purpose: This study aimed to investigate the frequency of customer engagement behavior and its affect on perceived relationship benefits and ultimately, relationship outcomes. This based on three different social media platforms where one company were present with brand communities and then compare the outcome of each platform with each other.
Hypothesis: 𝐻1 = The frequency of customer engagement behavior leads to perceived relationship benefits of engaging in a brand community.
𝐻2 = Customer perceived relationship benefits have a positive effect on relationship outcomes.
Methodology: Cross-sectional online questionnaires distributed on three different social media platforms – Facebook, Instagram, and Pinterest.
Analysed using linear regressions.
Findings: The findings indicates that the frequency on which a customer engage in engagement behaviors showed no statistical significance on Facebook, however, the frequency of reading messages, visiting the brand community, and purchasing products did show statistical significance on Instagram. Furthermore, the perceived relationship benefits that showed significance for both Facebook and Instagram was practical and economic benefits. While on Facebook social enhancement was considered an important indicator for relationship outcome and entertainment benefits was considered important on Instagram.
Table of content
1. Introduction ______________________________________________________________________ 1 1.1 Background ____________________________________________________________________ 1 1.2 Problem discussion ______________________________________________________________ 1 1.3 Purpose _______________________________________________________________________ 3 1.4 Research questions ______________________________________________________________ 3 1.5 Delimitations __________________________________________________________________ 4 1.6 Report structure ________________________________________________________________ 4 2. Conceptual framework _____________________________________________________________ 5 2.1 Theories ______________________________________________________________________ 5 2.1.1 Brand communities on social media _____________________________________________ 5 2.1.2 Customer engagement ________________________________________________________ 7 2.1.3 Relationship benefits and outcomes of customer engagement on brand communities ______ 10 2.2 Model: concepts and relations ____________________________________________________ 12 2.3 Hypothesis ___________________________________________________________________ 12 3. Method _________________________________________________________________________ 15 3.1 Research design _______________________________________________________________ 15 3.1.1 Sample ___________________________________________________________________ 16 3.2 Questionnaire and measures ______________________________________________________ 17 3.3 Measures – operationalization ____________________________________________________ 18 3.4 Data collection ________________________________________________________________ 18 3.5 Quality criteria ________________________________________________________________ 19 3.5.1 Validity and reliability ______________________________________________________ 19 3.6 Data analysis __________________________________________________________________ 20 4. Analysis and results _______________________________________________________________ 22 4.1 Facebook _____________________________________________________________________ 22 4.2 Instagram ____________________________________________________________________ 26 4.3 Hypothesis ___________________________________________________________________ 30 5. Discussion _______________________________________________________________________ 32 6. Conclusions and implications _______________________________________________________ 36 6.1 Theoretical contributions ________________________________________________________ 36 6.2 Managerial implications _________________________________________________________ 37 6.3 Limitations and suggestions for further research ______________________________________ 38 References ________________________________________________________________________ 39 Appendices _________________________________________________________________________ I Appendix 1 Questionnaire, Facebook, Swedish version _____________________________________ I Appendix 2 Questionnaire, Facebook, English version ____________________________________ IV Appendix 3 Questionnaire, Instagram, Swedish version __________________________________ VII Appendix 4 Questionnaire, Instagram, English version ____________________________________ X Appendix 5 Questionnaire, Pinterest, Swedish version __________________________________ XIII Appendix 6 Questionnaire, Pinterest, English version ___________________________________ XVI Appendix 7 Correlation analysis, tables ______________________________________________ XIX
1. Introduction
1.1 Background
Since introduced, Internet has grown rapidly and generated new opportunities for customers. Searching for information, communication with others, and expressing feelings are now easier than ever through social media (Tsimonis & Dimitriadis, 2014).
The integration of Internet into customers’ everyday life is constantly increasing and the phenomenon is often referred to as social media or Web 2.0. Arguably Web 2.0 is transforming people’s individual and group behavior and this ultimately affect the structure within the marketplace (Kietzmann et al., 2011). According to Constantinides and Fountain (2008) the definition of Web 2.0 is (p. 232) “a collection of open-source, interactive, and user-controlled online applications expanding the experiences, knowledge and market power of the users as participants in business and social processes.” Social media is highly interactive platforms, which employs both mobile and web-based technologies where the users can share, co-create, and adjust user- generated content (Kietzmann et al., 2011).
Furthermore, social media does not only provide customers with new opportunities but for companies as well. These new media platforms are increasingly replacing traditional media like TV, magazines, and radio, further, the buzz of social media and its marketing opportunities seems limitless (Bruhn et al., 2012). Wirtz et al. (2013) note that there has been an increase of online brand communities on social media during the last decade.
Brand communities are interactive pages where the company can share information, spread the history and culture of the brand, and provide customers with assistance (Laroche et al., 2012). These communities has emerged in order to facilitate relationship marketing and maintaining long-term relations with its customers through social media (Wirtz et al., 2012). Unlike traditional media, social media allows the individual to become the content-creator, and customers engage with companies in terms of sharing, liking, and posting within the individual’s own personal social network. The way organization-related content is being distributed, created and used has thus changed (Men & Tsai, 2014).
1.2 Problem discussion
As previously stated, brand communities on social media provide both customers and companies’ with new ways to engage with each other. Companies’ wishes to engage
with loyal customers, learn from and about them, influence their community members’
perception of the brand, and spread information. Hence, customer engagement is fundamental within brand communities (Gummerus et al., 2012).
Customer engagement has attracted the interest of consultants and managers within different industries and companies worldwide. The interest has grown equally to the growth of Internet and Web 2.0. It is the interactive nature of social media that has increased the potential for companies to better serve and understands the needs of their customers (Sashi, 2012). Moreover, research have conduced studies in order to better understand, define, and build upon this phenomenon (e.g. Sashi, 2012; Gummerus et al., 2012; Men & Tsai, 2014; Hollebeek et al., 2014; Bunker et al., 2013; Bowden et al., 2014; Bitter et al., 2014; Kabadayi & Price, 2014). Sashi (2012) argue that companies both within the private and public sector are striving to better connect with their customers. They wish to develop a high level of customer engagement and hence, establish an intimate and long-term relationship with its customers. By utilizing new technologies like social media, companies will connect with both existing and potential customers and understand them better. It has been argued that Internet is considered to be essential in building customer engagement (Sashi, 2012).
Gummerus et al. (2012) argue that customer engagement can be looked upon as a behavioral manifestation and this type of behavior is considered to be a consequence of social media and the way people communicate with each other. Customer engagement includes both customer-to-company interactions as well as customer-to-customer interactions. Furthermore, this new behavior involves all communication through brand communities and other social media and includes firm-related behavior that did not exist before social media (e.g. customers writing positive or negative products reviews online) (Gummerus et al., 2012). There are researchers who have investigated the notion of customer engagement on social media (Sashi, 2012; Gummerus et al., 2012; Men &
Tsai, 2014; Hollebeek et al., 2014; Bunker et al., 2013; Bitter et al., 2014; Kabadayi &
Price, 2014). However, the majority of the studies focus on customer engagement behavior on the social media platform Facebook (Gummerus et al., 2012; Men & Tsai, 2014, Bunker et al., 2013; Bitter et al., 2014; Kabadayi & Price, 2014). Social media is evolving rapidly and new platforms are constantly emerging (Gummerus et al., 2012).
Several researchers argue that there is a gap within the existing literature focusing on customer behavior and engagement within brand communities across different social
media platforms (Gummerus et al., 2012; Bitter et al., 2014; Men & Tsai, 2014;
Hollebeek et al., 2014; Sashi, 2012).
It is of interest to further investigate the differences in customer engagement between Facebook brand communities and brand communities on other social media platforms.
Thus, examining if measurements used for measuring customer engagement on Facebook is applicable on other platforms as well (Gummerus et al., 2012). Men and Tsai (2014) further argue that it is important to explore other social networking sites other than Facebook. They suggest platforms like Twitter, Instagram, and LinkedIn.
Conducting research on more platforms will contribute in getting a better understanding of the effects of customer engagement on brand communities on social media. They also emphasize on further studies regarding the growing number of mobile audience and emerging mobile social media tools like Snapchat (Men & Tsai, 2014). Malthouse and Calder (2011) also argue that customer engagement can only be comprehended through customer experience and that these experiences are context-dependent. Hence, investigating customer engagement behavior across different platforms is important in order to further understand the phenomenon of engagement and is important for companies who wishes to establish brand communities on different platforms.
1.3 Purpose
This study aims to investigate the frequency of customer engagement behavior and its affect on perceived relationship benefits and ultimately, relationship outcomes. This based on three different social media platforms where one company is present with brand communities and then compare the outcome of each platform with each other.
1.4 Research questions
What affect does the frequency of engaging in engagement behaviors have on perceived relationship benefits?
What affects does customer perceived relationship benefits have on relationship outcomes?
1.5 Delimitations
This study aims at investigating customer engagement behavior on different social media platforms, the study will be delimited to one company’s brand communities on following social media platforms - Facebook1, Instagram2, and Pinterest3.
1.6 Report structure
This paper will be structured as follows. First, a literature review of previous research within customer engagement on social media is presented. Second, the main concepts will be presented within the theoretical chapter; customer engagement, Web 2.0, brand communities on social media, relationship benefits of customer engagement on brand communities, and relationship outcomes. Third, the methodological approach and process will be explained. Fourth, the empirical findings will be analyzed using mediation analysis in SPSS and then the results derived from the different social media platforms will be compared. Last, the findings will be concluded with managerial and theoretical implications, limitations and recommendations for further research.
1Facebook – Social media platform where people can share and connect with friends and family (Facebook, 2015)
2Instagram – Social media platform focused on visual storytelling (Instagram, 2015)
3Pinterest – Social media platform where you find ideas to your project, interests created by people like yourself. You use visual bookmarks called pins when you find something you like on the web or Pinterest. (Pinterest, 2015)
2. Conceptual framework
2.1 Theories
2.1.1 Brand communities on social media
Since the emergence of Web 2.0 researchers have been interested in this particular phenomenon and how companies can take advantage of the new emerging opportunities (Sashi, 2012). Social media provide companies with distinctive opportunities to foster their relationship with its customers and at the same time attract new ones through brand communities (Laroche et al., 2013). Social media channels are inexpensive and user-friendly, attracting a large number of users and this makes it an interesting platform for companies (Tsimonis & Dimitriadis, 2014). According to Nair (2011) social media include platforms like Facebook, Twitter, and YouTube to mention some of the most well-known sites. Furthermore, Nair (2011) define social media as (p.45) “online tools where content, opinions, perspectives, insights, and media can be shared. Some people create content, while others lurk, observe, or disseminate content. As its core, social media is about relationships and connection between people and organizations.” Social media is allegedly new to the business world and there are a variety of different social media platforms to consider for companies. However, even though blogs, podcasts, widgets, wikis, video logs, and mashups are different expressions of social media, serving different purposes, they all create experience on the Internet (Nair, 2011).
Hollebeek et al. (2014) state that research has highlighted the dynamics of focal consumer – brand relationships, particularly the notion of consumer brand engagement (CBS, this concept is previously referred to as customer engagement). Furthermore, brand communities established on social media have showed to have positive effects on customer relationships with the company and thus, have positive effect on brand trust, and further brand loyalty (Laroche et al., 2013). Relationship building with customers is embedded within both marketing concepts related to meeting customers’ need and the marketing orientation concept, which emphasizes on meeting these needs by providing superior value to the competitors. Moreover, the interactive nature of social media is allegedly likely to distort the roles of the seller and customer by inviting customers to take part in the value creation. Furthermore, customer engagement requires the establishment of commitment and trust in a buyer-seller relationship (Sashi, 2012).
Accordingly, trust exists (Sashi, 2012 p. 259) “when one party has confidence in an exchange partner’s reliability and integrity”. Nair (2011) argue that companies often fail
to recognize their presence on social media as more than just another media-outreach program – they should handle it as a completely new platform. Social media has changed the way we communicate with others and it is also important to recognize the differences between different social media platforms (Nair, 2011).
Muniz and O’Guinn (2001) define brand communities as (p.412) “specialized non- geographic bound community based on a structured set of social relationships among admires of a brand. It is specialized because at its center is a branded good or service.
Like other communities, it is marked by shared consciousness, rituals and traditions, and a sense of moral responsibility.” However, in order to understand a brand community one must look at community as its own concept. Community is considered to be a core construct within social thought and has been a large topic of interest for researchers since the nineteenth century. Moreover, it is important to recognize that communities are no longer limited by geography due to new communication technologies (Muniz & O’Guinn, 2001).
Within a brand community the common and shared interest among the members are the brand, those communities are premised on diversity and appears to be communities of limited liability. There are relatively stable groupings within brand communities with somewhat strong degrees of commitment. Members of a brand community feel connection to the brand, but more importantly strong connection to other members (Muniz & O’Guinn, 2001). Additionally, Benedikt and Werner (2012) argue that an effect of membership prolongation leads to brand loyalty intentions and that customers who were active within online communities of a brand had stronger brand commitment than customers who were not member of the community. Brand communities have also shown to be a successful tool in increasing sales and the main driver of this is the sharing of information (Benedikt & Werner, 2012).
Social media management is strongly shaped by service-dominant logic and inherently implies a customer and relation-oriented view. Thus, if the social media brand page does not regularly deliver value for its members, the members will leave the page.
Benedikt and Werner (2012) argue that statistics for followers and likes might not be affected but the customer engagement will decrease. Furthermore, it has been stated that companies cannot themselves deliver value or experience on a brand page without the assistance of the community members (Benedikt and Werner, 2012). Hence, customers are part of the co-creation process within brand communities. Co-creation occurs when
customers participate in spontaneous behaviors such as helping other customers, make suggestions regarding how the consumption experience can be improved and helping service providers are all aspects of co-creation and hence, customer engagement behaviors (van Doorn et al., 2010).
2.1.2 Customer engagement
According to Sahsi (2012) customer engagement is a topic that has had an emerging interest during the last years. The increasing interest has paralleled with the development of Internet and the new tools that has emerged with it – Web 2.0 (Sashi, 2012). There has also been a conceptual shift from product-centric organization to a customer-centric organization and the management of customer relationships has thus become a top priority among companies. Due to the increasingly networking society customers can easily interact with other customers and this non-transactional customer behavior has become more important for companies when developing their strategies (Verhoef et al., 2010). Kabadayi and Price (2014), Bitter et al. (2014), Men and Tsai (2014), Gummerus et al. (2012), Sashi (2012), Hollebeek et al. (2014), and van Doorn et al. (2010) all emphasize on the importance of understanding customer engagement within social media settings as well as recognizing the opportunities for companies to extract value from their customers.
Hollebeek et al. (2014) highlight the outcomes of increased levels of customer brand engagement and that it might lead to superior organizational performance results for instance, growth in sales, brand referrals, reduction of costs, and collaborative product development process between customers and company. Van Doorn (2010) additionally states that it is important for companies to fully understand the impact of customer brand engagement since the digital world constitutes of a broad audience with immediacy breadth.
van Doorn et al. (2010) define customer engagement behavior as (p. 254) “behaviors that go beyond transactions, and may be specifically defined as a customer’s behavioral manifestations that have a brand or firm focus, beyond purchase, resulting from motivational drivers.” Verhoef et al. (2010) however, argue that the transactional side of the relationship is important for companies, but that ignoring non-transactional behaviors may lead to lost opportunities (WOM and co-creation) for companies and thus affecting cash flows. Therefore, overlooking customer engagement may lead to customers being valued inadequately (Verhoef et al., 2010). Gummerus et al. (2012)
state that customer engagement is often used in order to express the highest form of loyalty but that it contains all kinds of behaviors since it is a behavioral manifestations, not only behaviors that characterize high degrees of loyalty. Similar to Sashi (2012), Gummerus et al. (2012) also argue that social media provide customers with platforms where they can co-create value with companies and engage in behaviors like participating in online discussions, search for information, and commenting.
van Doorn et al. (2010) further argue that customer engagement also encompasses customer co-creation and that this involves customer participation in the creation of the core offering itself. Thus, co-creation occurs when customers participate in spontaneous behaviors that customize the customer-to-brand experience. They further argue that behavior as helping other customers, make suggestions regarding how consumption experience could be improved and helping service providers are all aspects of co- creation and hence, customer engagement behaviors (van Doorn et al., 2010).
Furthermore, van Doorn et al. (2010) argue that the general measurement of brand engagement builds on the concepts of self-schema theory and attachment theory and refer to Sprott et al. (2009) defining brand engagement as (p. 92) “an individual difference representing consumers’ propensity to include important brands as part of how they view themselves”. Moreover, Verhoef et al. (2010) notes that customer engagement consists of multiple online-behaviors as blogging, customer ratings, word of mouth et cetera. They further state that customer engagement is affected by firm initiatives, customer characteristics, and environmental/contextual factors (Verhoef et al., 2010).
Companies aspire to have relationships with its customers both with the goal of meeting the customer’s needs and provide customers with superior value in relation to competitors. Customer engagement requires establishment of trust and commitment in the customer-company relationship and trust only exists (p. 259) “when one party has confidence in an exchange partner’s ability and integrity” (Sashi, 2012). Furthermore, van Doorn et al. (2010) argue that the most significant factor affecting customer engagement is of attitudinal decedents. These include trust, customer satisfaction, brand commitment, brand attachment, and brand performance perceptions, thus, high or low levels of these factors can lead to engagement. Individual customer characters and tendencies can also affect the level of customer engagement, these characteristics may influence the customers decision making and cognitive process in a foreseeable way to
affect resulting behavior. Furthermore, customer resources (time, money and effort) may also affect the level of customer engagement. It has been argued that customers are most likely keen to evaluate cost and benefits of engaging in a certain behavior (van Doorn et al., 2010).
Brodie et al. (2011) note five fundamental propositions in defining the conceptual domain4 of customer engagement. The fundamental propositions derive from the literature synthesis and used to further define customer engagement. The first proposition include customer engagement as a reflection of a psychological state, which occur when interactive customer experience with a pivotal agent (brand, product or company) within a specific service relationship. Engaged customers may thus experience confidence in a certain brand, they might even believe in the integrity and pride of the brand and feel passion for it. Furthermore, engagement objectives might include certain products or services, specific communication (e.g. advertisement), or specific channels of communication (Brodie et al., 2011). Additionally, Men and Tsai (2014) found that customers who are deeply engaged with the company’s Facebook page tended to be more trusting of the company and feel more satisfied.
Second, Brodie et al. (2011) argue that customer engagement occur within dynamic, interactive process of relationships that co-creates value. Furthermore, customer engagement processes may vary between long- and short-term, with customer engagement levels varying in complexity over time. The nature of customer engagement process implies that interactions with a pivotal engagement object customer engagement might recur at different levels, over time, and across interactions. Third, customer engagement is a central role within nomological network of relationships. Hence relational customer engagement originators include both participation and involvement and consequences may include trust, commitment, self-brand associations, emotional connections, and loyalty. Fourth, customer engagement is a multidimensional concept to a context. Especially the importance of cognitive, behavioral, and emotional dimensions of customer engagement may vary between the different settings in which customer engagement is being observed. Fifth, customer engagement occurs within a specific set of situations generating different levels of engagement. Customer engagement is individual and context-dependent concept that can be observed at different levels of intensity and complexity. Engagement states can range between non-engaged to highly
4A “conceptual domain” defines the scope and delineation of a concept Brodie et al. (2010, p. 257)
engaged, and it is important to understand the contextual nature of engagement (Brodie et al., 2011).
2.1.3 Relationship benefits and outcomes of customer engagement on brand communities
Gwinner et al. (1998) note that in order to establish and maintain a relationship between two parties, both must feel that they gain something from each other. Furthermore, Gummerus et al. (2012) state that trust, satisfaction, commitment, and loyalty are consequences of customer engagement (Gummerus et al., 2012). They further argue that customers experience other relationship benefits other than becoming more loyal and satisfied with a brand by engaging in brand communities. They proposed that by engaging in different behaviors customers would receive different relationship benefits, for instance entertainment. Customers engage in numerous behaviors that will strengthen their relationship with a brand, this goes beyond traditional loyalty measures like intended behaviors, frequency of visit, and purchasing behavior. The most common online form in which customers engage with companies is social media and these platforms are considered to be particularly suitable for developing customer relations (Gummerus et al., 2012).
Kaplan and Haenlein (2010) notes that brand communities in social media share three characteristics. Brand communities enable social presence in terms of physical, graphic, and acoustic contact (Kaplan & Haenlein, 2010). The goal of any communication on brand communities is to avoid uncertainty and reduce ambiguity. Brand communities on social media are also closely related to the concept of self-presentation and individuals desires to have control over the impression they give others. Moreover, self-disclosure is considered to be an important part of relationship building and occur especially over social media platforms like Facebook. Accordingly this is an indicator that customers may gain social benefits from engaging in brand community behaviors (Gummerus et al., 2012).
Moreover, Gummerus et al. (2012) focused on following relationship benefits in their study (p. 860); practical benefits, social benefits, social enhancement, entertainment, and economic benefits. Additionally, van Doorn et al. (2010) state that customer engagement has consequences for all shareholders involving the customer, brand/company as well as customers of competitors. The most basic level of consequences of customers includes cognitive, attitudinal, and behavioral consequences.
Looking at the behavioral perspective, customers’ successful customer engagement behavior will lead to more engaged customers and they will participate more in customer engagement behavior actions. However, if unsuccessful this can cause customers to change to different engagement strategies (Gummerus et al., 2012). Van Doorn et al. (2010) further state that successful customers will actively contribute to content at the community and thus influencing customer equity. (van Doorn et al., 2010) Additionally, customer engagement is considered to relate to several brand relationship outcomes, like trust, satisfaction, affective commitment, and loyalty (Gummerus et al., 2012).
Customer satisfaction and loyalty emerge in several ways; Brodie et al. (2013) discovered that the participants in their netnographic study conveyed loyalty towards a brand and its community by expressing satisfaction and recommending the brand to others. They additionally found that customers felt empowered by engaging in brand communities in terms of themselves having influence over companies and other customers. Customers also felt connected to the other community members and they felt the need of help others within the community after receiving help themselves.
Additionally, the study identified trust and commitment as customer engagement outcomes (Brodie et al., 2013). van Doorn et al. (2010) notes that customer engagement behaviors are affected by context-based factors resulted from P.E.S.T.-aspects5 of society. Competitors’ actions can also create strong contextual force in affecting customer engagement (van Doorn et al., 2010). Brodie et al. (2013) also state that customer engagement is a context-dependent physical state with different levels of intensity. The customer engagement process is largely affected by customers’ needs of information and is a highly interactive process in which customers interact with each other and the company (Brodie et al., 2013). Thus, loyalty can be further strengthen by engaging in a brand community and customer satisfaction also has a positively influence by customers response of enjoyment, excitement, and pleasure of using the community (Gummerus et al., 2012).
Customer engagement behaviors also has consequences for companies, both financial and reputational consequences has been discovered within the field. Customer engagement behaviors such as referral behavior, word-of-mouth and actions of
5P.E.S.T. = Political/Legal, Economic/Environmental, Social, and Technological aspects (van Doorn, 2010, p. 258)
spreading and generating information (e.g. blogging) have showed to have an effect on purchase behavior of existing and new customers. Furthermore, reputational consequences emerge from engaged customers who will co-create value and participate in brand communities. Highly engaged customers can be crucial when it comes to sources of knowledge, they will help the company with activities ranging from generating new ideas of design to development of new products (van Doorn et al., 2010).
2.2 Model: concepts and relations
This study will look at following customer engagement behaviors - read messages, use
“like” option, write comments and purchase products (transactional behavior) and further see how perceived benefits (practical benefits, social enhancement, entertainment benefits and economic benefits) are affected by these behaviors. The next step is to see how customers’ engagement behaviors and their perceived benefits affects outcomes of loyalty, trust and satisfaction. This study also takes three control variables into consideration, which type of social media platform the customer interacts with the company, customer’s age and gender. Relations and hypothesis are presented in figure 1, conceptual model.
Figure 1. Conceptual model
2.3 Hypothesis
Due to an increasingly networking society where customers can interact with other customers it is important for companies to look into this non-transactional customer behavior (Verhoef et al., 2010). Customer engagement behavior is defined as (p. 254)
“behaviors that go beyond transactions, and may be specifically defined as a customer’s behavioral manifestations that have a brand or firm focus, beyond purchase, resulting from motivational drivers.” Furthermore, it has been argued that customer engagement encompasses customer co-creation in terms of spontaneous behaviors that customize the customer-to-brand experience (van Doorn et al., 2010). Brodie et al. (2011) note five fundamental propositions when defining the customer engagement and they argue that customer engagement reflects a psychological state that occurs when a customer interact
with a company. Engagement objectives could be certain products, communications, or specific channels of communication. Moreover, reading messages, use “like” options and writing comments can be seen as engagement behaviors (Brodie et al., 2011;
Gummerus et al., 2012). Gummerus et al. (2012) argued that customer engagement extends beyond transactional behavior, however, they found that transactional behavior (i.e. purchasing products) had positive impact on some of the perceived relationship benefits (social benefits and entertainment). Thus, it can be argued that transactional behavior is also part of customer engagement behavior (Gummerus et al., 2012). This study proposes that the frequency on which a customer engage in an engagement behavior leads to perceived relationship benefits.
𝐻1 = The frequency of customer engagement behavior leads to perceived relationship benefits of engaging in a brand community.
𝐻1𝑎 = The frequency of reading messages on brand communities will lead to perceived relationship benefits.
𝐻1𝑏 = The frequency of use the “like” option on brand communities will lead to perceived relationship benefits.
𝐻1𝑐 = The frequency of writing comments on brand communities will lead to perceived relationship benefits.
𝐻1𝑑 = The frequency of purchasing products will lead to perceived relationship benefits.
In order to establish and maintain relationships between two parties, both must feel that they gain something from each other (Gwinner et al., 1998). Furthermore, it has been argued that customers can perceive following benefits from engaging in a brand community on social media; practical benefits, social enhancement, entertainment benefits, and economical benefits. Brand communities are closely related to the concept of self-presentation and it is an indicator that customers may gain social benefits from engaging in community behaviors. Moreover, customers can seek social enhancement deriving from the feeling of being useful and recognized within the community.
Practical benefits including informational benefits that include getting feedback and ask questions. Entertainment benefits are related to customers having fun. Economic benefits refer to customers feeling that they gain discounts, time savings, or take part in competitions (Gummerus et al., 2012). Based on hypothesis 1 there is an expectation of a positive relationship between customer engagement behaviors and perceived relationship benefits. Brand communities on social media have showed to have positive
effects on customer relationships with the company and thus, have positive effects on brand trust, and further brand loyalty (Laroche et al., 2013). It has also been stated that satisfaction is a consequence of customer engagement (Gummerus et al., 2012). Hence, this study proposes that loyalty, satisfaction, and trust are consequences of the perceived benefits that customers experience on brand communities on social media.
𝐻2 = Customer perceived relationship benefits have a positive effect on relationship outcomes.
𝐻2𝑎 = Practical benefits will have a positive effect on relationship outcomes.
𝐻2𝑏 = Social enhancement will have a positive effect on relationship outcomes.
𝐻2𝑐 = Entertainment benefits will have a positive effect on relationship outcomes.
𝐻2𝑑 = Economic benefits will have a positive effect on relationship outcomes.
3. Method
3.1 Research design
This study was based on deductive, quantitative research. Deductive research is applied when data is created through theory, whilst its counterpart inductive research is when theory is collected from data. Accordingly to deductive research, hypothesis has been stated and they are based on what is already theoretically known within a certain area (Bryman & Bell, 2011).
In order to test the theoretical model proposed in chapter 2.2 Model: concepts and relations primary data was collected using cross-sectional online questionnaires distributed through three different social media platforms – Facebook, Instagram, and Pinterest. Customer engagement, perceived benefits, and relationship outcome where measured using different sub-concepts based on theoretical frameworks illustrated in table 1 below. The sub-concepts were based on previous studies within the field conducted by Weman (2011) and Gummerus et al. (2011).
Table 1. Main- and Sub-Concepts
Main concept Sub-concepts
Customer engagement
Read messages Use “like” option Write comments Purchase products
Perceived benefits
Practical benefits Social enhancement Entertainment benefits Economic benefits
Relationship outcome
Loyalty Trust Satisfaction
Furthermore, Cross-sectional online sample questionnaires were chosen as the data collection method in order to make inferences regarding the studied population at one point in time. This type of study is typically used in order to sort out causal effects of one (or more) independent variable(s) upon a dependent variable at a given point in time (Bryman & Bell, 2011). Furthermore, previous research has used quantitative methods and questionnaires as a tool for measuring customer engagement on social media (Gummerus et al., 2012; Men & Tsai, 2014; Bitter et al., 2014; Kabadayi & Price, 2013;
Bunker et al., 2013; Weman, 2011).
In order to measure customer engagement on brand communities on different social media platforms a company (Wakakuu6) was contacted which is active on Facebook, Instagram, and Pinterest. Customer engagement in this study was measured using one company and their customers on different platforms, this in order to avoid analyzing customer engagement among different companies on different platforms and hence, get misleading and non-comparable answers. The distribution of the questionnaires was dependent upon collaboration with a company since the company needed to post links to the questionnaires through their different platforms.
3.1.1 Sample
The population of the study was all of Wakakuu’s followers on each social media platform. Thus, the whole population for Facebook was about 30 000 people, Instagram 11 000 people, and Pinterest 137 people7. Probability sampling was selected on random basis, hence, each unit of the population had a chance of being selected (Bryman &
Bell, 2011). Furthermore, sampling occurs in three steps. (Yin, 2009) First the population is defined, second the sampling frame should be identified, where the sampling should take place and how to reach these people. In this study the sampling occurred through the different social media platforms, the people that followed Wakakuu and decided to answer the questionnaire was part of the sample. The third and final step include determination of sample size, this choice should be guided based on resource limitations such as time and money (Malhotra & Birks, 2003). In order to decide upon a sample size one must decide how much error to allow. Hence, a confidence interval should be determined. A confidence interval determines how high or low than the population mean one’s sampling mean is prepared to fall. In this study a confidence level of 90% was accepted with a margin of error of 5%. Thus, the sample size for each platform was calculated and presented in table 2 below together with the result of each platform (Bryman & Bell, 2011). However, due to the limited time-span of a week that each questionnaire was distributed and the large amount of noise that exists on the platforms the author accept a smaller outcome than the calculated sample size.
6Wakakuu is an online- and offline-store offering high-fashion clothes with high quality, founded 2011 in Sweden (Wakakuu, 2015).
7The numbers of followers are as of 6th of May 2015.
Table 2. Sample size and outcome
Platform Population Sample size Outcome % of sample
Facebook 30 000 269 167 62%
Instagram 11 000 265 437 165%
Pinterest 137 92 0 0%
3.2 Questionnaire and measures
The sample questionnaires were developed in two stages. First, relevant literature was reviewed for existing scale items measuring customer engagement. Measures were found and the author got access to a questionnaire used by Gummerus et al. (2012) and Weman (2011) measuring customer engagement on a gaming brand community on Facebook. The process of acquiring access to the questionnaire involved the author to contact Johanna Gummerus and she further stated that their questionnaire based on a previous questionnaire by Weman (2011). Weman (2011) in turn based his questions on following authors - Dholakia et al. (2004), Gwinner et al. (1998), and Ouwersloot and Odekerken-Schröder, (2008). The questionnaires in this study (see appendix 1-6) were based on previous questionnaire used by Weman (2011) and Gummerus et al. (2011).
Furthermore, the variables measuring perceived benefits and relationship outcome was considered interval and measured using a multi-item Likert measures on a five-point scale ranging from ‘totally disagree’ (1) to ‘totally agree’ (5). The variables measuring customer engagement behavior was considered ordinal. Ordinal variables are variables whose categories can be ranked without an equal distance across the range. Hence, the distance between visiting ‘daily’ and ‘once a week’ is not the same as the difference between ‘once a week’ and ‘once a month’. However, one can see that visiting ‘daily’ is more frequent than ‘once a week’ et cetera (Bryman & Bell, 2011). Further, the variables measuring customer engagement behavior ranged from ‘daily’-‘more seldom than once a month’, ‘often’-‘never’, and ‘daily’-‘I do not purchase products from Wakakuu’. Interval variables are variables where the distance between the different categories is identical across the range. The questionnaire also included two control variables - age and gender. Furthermore, the questionnaires were distributed in Swedish in order to facilitate the answering process for the respondents following the Swedish brand Wakakuu.
3.3 Measures – operationalization
Table 3. Operationalization
Concept Definition of concept
Type of scale and its construction
Sub-concept Items used
Customer engagement behavior
“Behaviors that go beyond transactions, and may be specifically defined as a customer’s behavioral
manifestations that has a brand or firm focus, beyond purchase, resulting from motivational drivers.”
(van Doorn et al., 2010)
4-point scale where at one end you have daily/often and the other never/less than monthly.
Ordinal variables
Read messages Question 1-2
Use the “like”
option
Question 3
Write comments Question 4
Purchase products Question 5
Perceived relationship benefits
In order to establish and maintain
relationships between two parties, both must feel that they gain something from each other (Gwinner et al., 1998)
5-point Likert scale where
(1) Strongly disagree
(5) Strongly agree.
Scale variables
Practical benefits Question 6-8
Social enhancement Question 9
Entertainment benefits
Question 10-11
Economic benefits Question 12-14
Outcome of customer engagement on brand communities
Engaging in brand communities can strengthen loyalty and customer satisfaction has also shown to have a positive influence by customers’ response of enjoyment of using the community
(Gummerus et al., 2012). Trust is also considered to be an engagement outcome (Brodie et al., 2013).
5-point Likert scale where
(1) Strongly disagree
(5) Strongly agree.
Scale variables
Loyalty Question 15-18; 21- 22
Trust Question 19-20
Satisfaction Question 23-25
3.4 Data collection
The data collection-phase involved two different stages. First, the author contacted the company Wakakuu and explained that there was going to be three different questionnaires, one for each platform. In order to retrieve as large number of respondents possible the author and Wakakuu settled upon a raffle of a designer clutch that would be given to one of the respondents. Second, the three questionnaires were created using Google Drive and subsequently the links to the three questionnaires were
handed to Wakakuu who later posted the different links on respective platform along with information regarding the raffle. The data collection process spanned between the 6th to the 13th of May 2015. The 12th of May a reminder was posted on all social media platforms informing Wakakuu’s followers that the 13th is the last day to fill in the questionnaire and hence, have a chance to win the clutch. Ultimately the questionnaires generated 167 responses from Facebook, 440 responses from Instagram, and 0 responses from Pinterest. Due to the low response rate from Pinterest this social media platform was not included in the analysis. After evaluating the Pinterest platform response rate with the company the most likely answer was that Wakakuu was not as active on Pinterest as they were on Facebook and Instagram, hence one possible explanation to the low response rate.
3.5 Quality criteria
When evaluating business research, there are two essential criteria to look at – validity and reliability. Validity refers to the degree in which a measurement instrument measure what it is supposed to measure, and reliability refers to the stability of that measurement. Though, a concept cannot be valid without being reliable, but can be reliable without being valid (Bryman & Bell, 2011).
3.5.1 Validity and reliability
There are several ways in which to establish measurement validity, one way is to let other people review the questionnaire. Since the questionnaire used in this study was based on two previous studies the questions are considered to be valid and hence, measure what they are supposed to measure due to previous research. Moreover, a correlation analysis was conducted in order to ensure contract validity. Pearson’s correlation coefficient was used and the result is presented in appendix 7. Furthermore, the result showed that the concepts did not measure the same things and hence, are considered valid (Bryman & Bell, 2011). The validity is further supported by theoretical framework and previous research using the same parameters (Gummerus et al., 2012;
Weman, 2011). A concept cannot be valid without being reliable hence, the questionnaires are considered reliable due to their validity (Bryman & Bell, 2011).
Furthermore, reliability was tested using Cronbach’s alpha for both dependent and independent variables and the results were presented in table 4 below.
Table 4. Reliability test - Cronbach’s Alpha
Concept Cronbach’s Alpha
Cronbach’s Alpha Instagram Customer engagement
behavior
0.610 0.566
Perceived relationship benefits
0.728 0.782
Relationship outcome 0.823 0.837
The result of the reliability test indicated that all variables apart from Customer engagement behavior on Instagram were within the accepted level of <0.60. However, 0.50 is an accepted level of Cronbach’s alpha, even if it is considered poor. It can thus be concluded that the items within the questionnaires are reliable in that they were derived from two previous studies and thus, considered valid (Bryman & Bell, 2011;
Hair et al., 2014).
3.6 Data analysis
First a factor analysis was conducted in order to define underlying structure among the variables within the analysis. This type of analysis is used in order to determine whether groups of indicators have a tendency to group together into clusters (factors). (Hair et al., 2014) Furthermore, looking at the unrotated (for cases where only one component were found) and the rotated factor analysis, question 17 and 24 were omitted from the linear regression due to their non-coherence with the other variables, these two questions were negatively worded as opposed to the others which was positively worded. Hence, these two questions where not part of the subsequent linear regression analysis.
In order to analyze the gathered data and test the hypothesis linear regressions where carried out and summarized in four tables inspired by tables used in Devine (2010).
Linear regression analysis was chosen due to the fact that the study aimed to investigate the relationships between customer engagement perceived benefits and perceived benefits relationship outcomes and due to the classification of the variables. When deciding upon an analysis method one should look at the measurement of the dependent variable and in both cases the dependent variables (perceived benefits and relationship outcomes) were classified as interval, hence, a linear regression was considered as suitable analysis method (Bryman & Bell, 2011). Multiple linear regression measure if there is a statistical linkage between a dependent variable (𝑌1 = perceived relationship benefits; 𝑌2 = relationship outcomes) and two or more independent variables (𝑥1 = Read
messages, 𝑥2 = Use “like” option, 𝑥3 = Write comments, 𝑥4 = Purchase products; 𝑧1 = Practical benefits, 𝑧2 = Social enhancement, 𝑧3 = Entertainment benefits, 𝑧4 = Economic benefits). (Bryman & Bell, 2011; Hair et al., 2014) The formula for the two linear regression analysis carried out in this study where constructed as follows.
𝑌1 = 𝛽0+ 𝛽1𝑥1+ 𝛽2𝑥2+ 𝛽3𝑥3+ 𝛽4𝑥4 𝑌1 = Perceived relationship benefits
𝑥1 = Frequency of reading messages 𝑥2 = Frequency of using “like” option 𝑥3 = Frequency of writing comments 𝑥4 = Frequency of purchasing products
𝑌2 = 𝛽0 + 𝛽1𝑧1+ 𝛽2𝑧2+ 𝛽3𝑧3+ 𝛽4𝑧4 𝑌2 = Relationship outcomes
𝑧1 = Practical benefits 𝑧2 = Social enhancement 𝑧3 = Entertainment benefits 𝑧4 = Economic benefits
4. Analysis and results
4.1 Facebook
Table 5. Hypothesis testing if customer engagement behavior on Facebook brand communities leads to perceived benefits. Analysis method: Linear regression.
Model 1 Control
Model 2 𝑥1
Model 3 𝑥2
Model 4 𝑥3
Model 5 𝑥4
Model 6 All
Constant 2.906 3.225 3.090 3.294 3.276 3.604
Control variables
Age -0.206***
(0.043)
-0.204***
(0.043)
-0.214***
(0.043)
-0.202***
(0.043)
-0.205***
(0.043)
-0.204***
(0.043) Customer
engagement behavior 𝐻1𝑎 The frequency of reading messages on brand communities will lead to perceived relationship benefits
-0.162**
(0.068)
-0.128*
(0.071)
𝐻1𝑏 The frequency of using the
“like” option on brand communities will lead to perceived relationship benefits
-0.081 (0.052)
-0.021 (0.057)
𝐻1𝑐 The frequency of writing comments on brand communities will lead to perceived relationship benefits
-0.122**
(0.55)
-0.086 (0.061)
𝐻1𝑑 The frequency of purchasing products will lead to perceived relationship benefits
-0.111 (0.080)
-0.039 (0.084)
𝑅2 0.121 0.151 0.134 0.146 0.131 0.169
Adjusted 𝑅2 0.115 0.140 0.123 0.136 0.120 0.143
Std. Error of the Estimates
0.697 0.687 0.694 0.688 0.694 0.685
F-value 22.637*** 14.535*** 12.634*** 14.069*** 12.346*** 6.554***
Degree of freedom (df) Regression
1 2 2 2 2 5
Sig. *p<0.10, **p<0.05, ***p<0.001, N=167
S.E. (Standard Error) is present within parenthesis for each of the independent variables.
Table 5 shows the result from the linear regression between customer engagement behavior and perceived benefits on the social media platform Facebook. Hypothesis 𝐻1𝑎- 𝐻1𝑑 was tested and the control variable of age showed to be significant when measuring the relation between engagement behaviors and perceived benefits on a high level of p<0.001. Hence, age has an impact on customer engagement behaviors effect on perceived relationship benefits. Only 𝐻1𝑎 were shown to be significance, even though the significance level was low (p<0.10). Hence, reading messages and visiting a brand community on Facebook was the only predictor for perceived benefits (practical benefits, social enhancement, entertainment benefits, and economic benefits). It can thus be argued that use “like” options, writing comments, and purchasing products have no significant impact on the perceived benefits of engaging in a brand community.
Furthermore, looking at the adjusted 𝑅2 for the overall model indicates that all of the variables measuring customer engagement explain14.3% of perceived benefits. This is a fairly small number and indicates that there are other measures of customer engagement not present in the tested model that explains the relationship between customer engagement and perceived relationship benefits. The F-value measure the statistical significance of the regression equation as a whole and it has been argued that the ‘rule of thumb’ measuring F-value is that all values above 4 is significant (Hair et al., 2014).
Looking at table 4 it is evident that all models are significant and hence, the regression equations are highly significant on a p<0.001-level.
Moreover, looking at the changes in 𝑅2 in table 5 the changes for model 1 and 6 were statistically significant, while the other changes did not show any significance (however, model 4 and model 6 shows that hypothesis 𝐻1𝑐 is partially supported).
