1
MSc. Business Administration - Marketing Management
‘Are engaged fans within an online brand community also the
most loyal in terms of behavioral loyalty’?
University of Groningen Faculty of Economics and Business
2 Management Summary
Researchers showed that online brand community participants have high levels of customer engagement behaviors with the firm’s product(s) and brand(s) (McAlexander, Koenig and Schau, 2002), are motivated to help other customers (Bagozzi and Dholakia, 2006), are very loyal and actively recruit others to the community (Algesheimer, Dholakia and Herrmann, 2005). Within literature it is questioned whether these fans within an online brand community, who are treated as engaged customers, are behavioral loyal? In other words behavioral loyalty means consumers’ repurchase behavior or intension of specific brand (Kumar and Shah, 2006). Algesheimer, Borle, Dhoakia and Singh (2010), showed in their research that an increase in participation and an increase in activities within an online brand community led to a lower share of wallet and decreased sales of products. Therefore the problem statement of this research is: ‘Are engaged fans within an online brand community loyal in terms of behavioral loyalty?
According to Lee, Kim and Kim (2011); Sprott, Czellar and Spangeberg (2009) and Algesheimer et al. (2010), the most common engagement behaviors within an online brand community consist of:
- Brand engagement: how a fan of a brand can identify himself with a brand (Sprott et al., 2009);
- Community engagement: how addicted to or active a fan is within an online brand community (Algesheimer et al., 2010) and;
- Community identification: whether a fan can identify himself with the online brand community (Lee et al., 2011).
The contribution of brand engagement, community engagement and community identification on behavioral loyalty will be measured within this research.
Onishi and Machanda (2012) showed in their research that the relationship between customer engagement behaviors and behavioral loyalty is much stronger in the beginning phase than in de end phase of brand building process. Therefore this research finds out whether the duration of the relationship moderates this relationship negatively.
4 Management Summary 2 Table of Contents 4 1. Introduction 6 1.1 Problem Statement 8 1.2 Relevance 9
1.3 The Structure of this research 9
2. Literature 10
2.1 Online brand community defined 11
2.2 The importance of behavioral loyalty within this research 12 2.3 The relationship between behavioral loyalty and customer engagement
behaviors. 13
2.3.1 How do brand engagement behaviors influence behavioral loyalty? 15 2.3.2 How do community engagement behaviors influence behavioral loyalty 16 2.3.3 How do community identification behaviors influence behavioral loyalty 17 2.4 Do engaged fans within an online brand community become more
behavioral loyal over time? 18
3. Methodology 19
3.1 Data collection 19
3.1.1 innocent introduction 21
3.2 Characteristics of the sample 22
3.3 Measurement of variables 22
3.3.1 Validity 25
3.3.2 Internal consistency reliability 25
3.4 Analysis 26 4. Results 27 4.1 Descriptives 27 4.2 Model 1 28 4.2.1 Hypotheses h1, h2 and h3 29 4.3 Model 2 30 4.3.1 Hypothesis h4 31
5. Discussion and Conclusions 32
6. Implications 34
6.1 Theoretical implications 34
5
7. Limitations and Future research 37
References 39
6 1. Introduction
Companies are becoming more interested in strategies to involve customers into their products or brands. By involving the customer, companies gain value through the variety of practices that they perform online and offline. Traditionally, companies have tried to reach and build up relationships with customers through offline direct marketing activities, like reward programs and public relations. A Nielsen (2011) research shows that Americans for example spend more than 23 percent of their time online on what is called ‘online brand communities’, for example Facebook or blogs. That is why companies try to involve their customers via online channels and therefore many companies are spending more of their marketing budgets on online marketing programs. There has been a groundswell of interest among marketers in organizing these customer online brand communities (Belk and Tumbat, 2002; Johnson, 2004; Nail, 2005). Due to the popularity of concepts such as consumer empowerment, collaboration, and customer-led marketing (Evans and Wolf, 2005; Prahalad and Ramaswamy, 2004; Selden and MacMillan, 2006). Customers are able to join local fan communities on the firm’s website or participate in an online brand community and can communicate with one another or with company employees, as well as post pictures and stories about the brands’ marketing campaigns.
7 Doorn, Lemon and Mittal, 2010). All the fans within an online brand community are judged to have a certain level of engagement, since they took effort to become a member of the online brand community. But until now, no definition of customer engagement behaviors within an online brand community is given. According to van Doorn et al. (2010), the behavioral manifestations within an online brand community must have firm focus. Within literature three know behavioral manifestations within an online brand community are several times described. According to Lee, Kim and Kim (2011); Sprott, Czellar and Spangeberg (2009) and Algesheimer, Borle, Dhoakia, and Singh (2010), the following concepts are the most common behavioral engagement behaviors within an online brand community:
- Community engagement behaviors, such as word of mouth, which are the level of motivation or participation of community members within the community; - Brand engagement behaviors, which can be defined as including brands as a part of the self-concept;
- Customer identification behaviors, which are the individual perceptions of actual or symbolic belongingness to a group.
So far, many positive outcomes result from engagement behaviors within an online brand community. Research has shown that these findings can be questioned.
Within a research of Algesheimer et al. (2010), where they examined the consequences of online brand community membership on loyalty, brand community members became more selective and conservative in actual purchasing behaviors which is leading to a null or negative effect of community participation on individual-level buying volume, product listings, average amount spend by buyers and revenue earned by sellers. Furthermore, Libai (2011), concludes that customer engagement behaviors are directly related to all the new ways in which customers can interact, but do not result in a more intense purchase behavior.
8 engagement behaviors within their online brand communities will lead to profits and extra buying behaviors. After all, the creation and maintenance of an online brand community is a large investment for a company.
To summarize, online brand communities are created to foster the engagement behaviors of customers and fans of a brand. Theory concludes that these engagement behaviors automatically lead to loyalty. But not every engaged fan within an online brand community is loyal and may even cost a company/brand more money than it will yield. It is important for both companies and literature to know whether fans of an online brand community who, according to literature are mentioned to be engaged, can be judged as behavioral loyal as well. In other words behavioral loyalty means consumers’ repurchase behavior or intension of specific brand (Kumar and Shah, 2006). As companies heavily invest in the creation and maintenance of online brand communities, it is beneficial to investigate whether these engaged fans within the online brand community are profitable. Although prior research exists on customer engagement (Bowden, 2009), little empirical research has been done on the buying behavior of engaged fans within online brand communities.
1.1 Problem Statement
So far, there are few empirical studies on customer engagement behaviors in general, particularly in social media, even though customer engagement behaviors have been recognized as a key research priority of the Marketing Science Institute (Bolton, 2011, p. 272). Consequently, there is little known about customer engagement behaviors of fans within online brand communities, or about the relationship between the customer behavioral engagements and purchase behavior. The main purpose of this research is to determine whether the fans of a brand, which seem to demonstrate customer engagement behaviors within an online brand community, are behavioral loyal as well.
Therefore, the main question of this research is:
‘Are engaged fans within an online brand community also the most loyal in terms of behavioral loyalty’?
The sub-questions that can be asked are: - What is an online brand community?
- To what extent are behavioral loyalty and customer engagement behaviors related? - To what extent are the different forms of customer engagement behaviors related to
9 - Do engaged fans within an online brand community become more or less behavioral
loyal over time?
1.2 Relevance
Within this paper it will be investigated how different forms customer engagement behaviors within an online brand community affect the level of behavioral loyalty of a company or a brand. This research contributes to the literature in several ways. It will determine how the fans within an online brand community behave. Fans might show different levels of the different customer engagement behaviors within an online brand community (Algesheimer et al., 2005), but do these different levels create more loyal behavior in terms of purchasing. Not only the recent developments in the marketing realm that customers should be evaluated by their purchase behavior (Kumar et al., 2010). This research will also include whether the duration of the relationship influence the contribution of the different engagement behaviors to behavioral loyalty (Onishi and Machanda, 2012). A model will be developed that will provide companies more insight into the relationships between the engagement behaviors within online brand communities and the behavioral loyalty of customers (fans). Afterwards, a conclusion can be drawn about the effectiveness of an online brand community, in a way that companies are able to spend their marketing budget online or offline more effectively. This simplified model can be used by both companies and researchers to evaluate the level of engagement and behavioral loyalty of online brand community member For example overspending to increase loyalty will not end up with increased behavioral loyal customers (Kumar and Reinartz, 2006). Moreover it will be proven that on average the value of one fan in terms of their customer engagement behaviors, is not as high as has been researched before. Within this research the fans of a brand will be sent an extensive questionnaire in order to determine whether the online brand community members, who show customer engagement behaviors, can be classified as behavioral loyal customers. The empirical study will be performed on the smoothie producer innocent.
1.3 The structure of this research
10 2. Literature
Based on the reviewed literature and the hypotheses build in this chapter, the conceptual model below, figure 1. is developed. This model reflects all theoretical relations between constructs and defines the relationships that will be researched. First of all the online brand community engagement behaviors will be introduced.
Within an online brand community fans show customer engagement behaviors. According to Lee et al. (2011); Sprott et al. (2009) and Algesheimer et al. (2010) the most common customer engagement behaviors within an online brand community consist of
- Brand engagement , how a fan can identify himself with a brand (Sprott et al., 2009); - Community engagement, how addicted someone is to the online brand community
(Algesheimer et al., 2010) and;
- Community identification, if someone can identify himself with the online brand community (Lee et al., 2011).
Onishi and Machanda (2012), showed in their research that relationship between customer engagement behaviors and behavioral loyalty is much stronger in the beginning phase than in de end phase of brand building process. Therefore this research finds out whether the duration of the relationship moderates this relationship negatively.
Effects of engagement behaviors on loyalty are discussed within literature, but results are mixed (Franke and Shah, 2003; Algesheimer et al. 2010). Due to the argues, research must find out whether engaged fans within an online brand community are loyal to the brand in terms of purchasing behavior (behavioral loyalty).
11 Figure 1. Conceptual Model
2.1 Online brand community defined
To define online brand communities, online brand communities are groups of users and admirers of a brand who engage jointly in group actions to accomplish collective goals and/or to express mutual sentiments and commitments (Bagozzi and Dholakia, 2006, p. 45). Examples of online brand communities are online blogs or a Facebook page. As has been mentioned before, online brand community are an important channel to increase engagement behaviors of fans (Packard and Pattabhiramaiah, 2012).
The popularity of online brand communities started by creating stronger relationships between company, brand and customers through the offline brand communities (Muniz and O’Guinn, 2001). Before online brand communities existed, fans of a brand could become a member of offline brand communities. An offline brand community can be defined as a “group of consumers with a shared enthusiasm for the brand and a well-developed social identity, whose members engage jointly in group actions to accomplish collective goals and/or express mutual sentiments and commitments” (McAlexander et al., 2002). According to Muniz and O’Guinn (2001) strong consumer and brand relationships are known to produce positive outcomes of offline brand communities for both customers and the company/brand such as increased satisfaction or loyalty. These research mention in their researcher that online brand communities share many outcomes with offline brand communities.
12 Online brand communities are perceived differently. Members of online brand communities use online brand communities to create and read content such as reviews, recommendations and opinions and they can help each other (by making purchasing decisions) online (Nambisan and Baron, 2009). Furthermore, online brand communities differ in the wider geographical spread (they are not stick to one specific location where fans should meet), are more approachable, more interactive and easier to join. As Muniz and O’Guinn (2001), explained in their research that offline brand communities have proofed that the relationship between brand and customer do have positive outcomes. In contrast to offline brand communities, online brand communities have different types of users and fans of their brand based on how strong their ties, such as customer loyalty and customer satisfaction, to the brand and to the other community members are (Algesheimer, Dholakia and Herrmann, 2005).
2.2 The importance of behavioral loyalty within this research
Let’s first introduce the concept loyalty. Theoretically loyalty is based on the following construct; According to Kumar and Reinartz (2006) the underlying construct of loyalty begins with product performance, service performance and employee performance. All of these lead to customer satisfaction, satisfaction leads to loyalty and when customers are loyal the profits will increase. Loyalty has been made by costs, but an overinvestment of cost into loyalty does not pay off. Moreover too many investments in loyalty do not have a high return when people are already satisfied. Kumar and Shah (2006) describe two alternate forms of loyalty: behavioral loyalty and attitudinal loyalty. Behavioral loyalty means consumers’ repurchase behavior or purchase intension of specific brand. Attitudinal loyalty means consumers’ sense of specific products or services.
13 just like to be a member of an online brand community because it is possible (Algesheimer et al., 2010).
2.3 The relationship between customer engagement and behavioral loyalty
How are behavioral loyalty and the customer engagement behaviors within an online brand community related? Van Doorn et al. (2010) for example point out that the term ‘‘engagement’’ is behavioral in nature and they state that ‘‘customer engagement’’ goes beyond transactions and is specifically defined as a customer’s behavioral manifestation toward a brand or firm, beyond purchase, resulting from motivational drivers.’’ These motivational drivers (of customer engagement), can be both positive (i.e. posting a positive brand message on a blog) and/or negative (i.e. organizing public) actions against a firm (van Doorn et al., 2010). One can envision the different ways in which a customer can interact or ‘‘engage’’ with the firm, purchasing from the firm naturally arises (Kumar et al., 2010). Customer value to the firm (or the behavioral loyalty of a customer) is therefore driven by the nature and intensity of ‘‘customer engagement behaviors’’ regarding the firm (and its product/service offerings).
The effects of engagement behaviors within an online brand community on behavioral loyalty are mixed in literature. No direct conclusions can be made yet. On one hand Woisetschlager, Hartleb and Blut (2008) did show in their research that offline marketing programs have a positive influence on relationships between the customer and the brand, between the customer and the firm, between the customer and the product in use, and among fellow customers. They showed an effect of the online brand community participation on word of mouth, brand image and community loyalty. An effect on brand loyalty was not shown and the buying behaviors showed negative effects. Fans within the online brand community were active on the page, showed customer engagement behaviors and they had the intention to purchase, but the actual purchase lay down.
14 level bidding volume, product listings, average amount spent by buyers, and revenue earned by sellers. A unique aspect of the online brand communities which has been investigated in the above study is that it exists to “make markets.” Thus, most of the important marketing mix elements (such as product and price) are a function of the actions of independent agents rather than actions of the firm. By increasing customer engagement, a company would expect an increase of loyalty of the fans, but according to the studies above this is not usual. According to Kumar and Reinartz (2006) not every customer is influenced by these investments to increase loyalty. Customers’ loyalty to a product or service by repeated purchases can also be due to their natural preferences, products’ features and benefits or other marketing programs and plans of the firm.
On the other hand successful online brand communities can lead other members to engage voluntarily in various online brand community engagement behaviors such as membership intention, recommendation, and active participation. Online brand communities in terms of the social psychological processes are channels that motivate consumers to participate in different engagement behaviors within an online brand community (Lee et al,. 2011).
Kim, Eyehook, Qualls and Choi (2008) showed that online brand community commitment is a driver of brand commitment. They showed that online brand community participants possess stronger brand commitment than consumers who are not members of the online brand community as well. Adjei, Noble and Noble (2010) verified in a netnography and experimental approach that online brand communities are successful tools for increasing sales. Firms that set up such communities expect to increase customer engagement and eventually loyalty (Fournier and Lee, 2009; Porter and Donthu 2008). The expectation is that this increased engagement and/or loyalty will lead to better economic outcomes for the brand, as exemplified by predictions that firm sponsors of online communities will be “richly rewarded with peerless customer loyalty and impressive economic returns” (Hagel and Armstrong, 1997, p. 2). Using surveys and self-report data, previous academic research has reported an increase in purchase intention among online brand community members (Algesheimer et al., 2005; Porter and Donthu, 2008). Other researchers have shown that enabling consumer interactions in a firm-sponsored online community is one of seven factors linked to increase future purchase intention and willingness to pay a price premium on online retailers (Srinivasan, Anderson and Ponnavolu 2002).
15 by social intelligence company Syncapse, studied more than 2,000 Facebook users who had liked a brand, by taking into account factors such as product spending, loyalty, propensity to recommend and brand affinity to determine the value of a Facebook fan. They stated that Facebook fans spend more money not only on the brands they liked ($116 more per year than nonfans), but also within the brand's sector 43 percent more, despite not having a higher income than nonfans. Those fans are also 18 percent more satisfied with their brands than non-friends and 11 percent more likely to continue using the brand than non-friends. The increase in average fan value is driven by fans' tendencies to be super-consumers. Not only do they tend to be brand users first, they spend more, engage more, advocate more and are more loyal (Smith, 2013).
On one hand a positive relationship exists between customer engagement behaviors within online brand communities and behavioral loyalty, but on the other hand not every engaged fan of an online brand community does show a certain level of loyalty. The customer engagement behaviors within online brand communities are discussed by different authors. According to Lee et al. (2011); Sprott et al. (2009) and Algesheimer et al. (2010) the most common customer engagement behaviors within an online brand community consist of brand engagement, (Sprott et al,. 2009) community engagement, (Algesheimer et al., 2010) and community identification, (Lee et al., 2011). The following chapters will discuss the effects of these different constructs on behavioral loyalty.
2.3.1 How do brand engagement behaviors influence behavioral loyalty?
16 suggested that people vary in their likelihood to engage in identity building and expression through online brand communities (Muniz and O’Guinn, 2001).
The influence of brand engagement on behavioral loyalty is investigated in literature. The branding literature provides grounds for the assertion that the potential for a brand to play a self-defining role as a component of a brand’s equity (Keller, 1993). This additional equity may lead to increased brand loyalty and more inelastic demand (Kapferer, 2008; Keller, 1993). Within the research of Sprott et al. (2009) they posit that high brand engaged customers within online brand communities are associated with more positive brand attitudes and purchase intentions for the highly priced products. The following hypothesis is:
H1: Brand engagement behavior contributes positively to behavioral loyalty
2.3.2 How do community engagement behaviors influence behavioral loyalty?
Community engagement behaviors are engagement behaviors an online brand community member shows to others (Lee et al., 2011). According to Algesheimer et al. (2010) community engagement is the level of motivation or participation of community members within the community.
According to Lee et al. (2011) the most important and common community engagement behavior is word of mouth (WOM). Word of mouth communications can be defined as “informal communications directed at other consumers about the ownership, usage or characteristics of particular goods and services and/or their sellers” (Westbrook, 1987). Arndt (1967) defined word of mouth as “oral, person-to-person communication between a perceived non-commercial communicator and a receiver concerning a brand, a product or a service offered for sale”. These human interactions (e.g., referrals, observation of product, service owners/adopters, etc.) play an important role in the diffusion of products and services and in the marketing of products and services. The rise in popularity of the online environment offers significantly increased opportunities for WOM. In addition, the growth of social networking sites has allowed users to broaden the scope of their connections with others by allowing them to build and maintain a network of friends for social or professional interaction and to share ideas with others (Trusov, Bucklin and Pauwels, 2009).
17 Kumar et al. (2010) mention in contrast to Algesheimer et al. (2010) that customer behavior through customers’ influence via WOM on other acquired customers as well as on prospects, minimizes buyer remorse to reduce defections and encourages increased share-of-wallet of existing customers. This WOM activity persuades and converts prospects to customers. This is also supported by Srinivansan, Anderson and Ponnalova (2002), they showed that community member who are influenced by WOM, have a higher willingness of paying a premium price than other buyers. But Algesheimer et al. (2010), showed that people were questioning their buying decisions and it took a longer period of time before people made buying decisions, with the help of recommendations of others online.
Overall a positive influence of community engagement behaviors can be found on behavioral loyalty. Researchers not only believe that high customer community engagement is necessary for future growth, they also believe that low customer community engagement is detrimental to success, both due to lost sales or opportunities and negative WOM (Businessnewsdaily, 2013). Online brand community fans can not only be engaged with a firm, they can be disengaged with a firm as well. Moreover, the majority of executives stated that both low and high engaged customers within an online brand community provide frequent feedback about products and services (EIU, 2007). That is why it can be concluded that community engagement behaviors do affect behavioral loyalty positively, unless fans show high or low community engagement behaviors.
H2: Community engagement behavior contributes positively to behavioral loyalty 2.3.3 How do community identification behaviors influence behavioral loyalty?
18 Bagozzi and Dholakia (2006) explain that when ‘one’s social identity’ within the online brand community increases and participation in group activities is fostered, a greater involvement with the brand occurs, which should promote the assimilation of the brand’s images into one’s identity. This is in line with Escalas and Bettman’s (2003), reasoning of the influence of reference groups on consumers’ connections to brands. They show that brand use of a reference group is a source of brand associations and those consumers develop a self-brand connection. When their reference group has a strong self-brand connection and their reference group has a strong association with the brand, there might be a strong connection between the reference group and the consumer’s self-concept. As online brand communities of admired brands are important to valued groups or to the individual, it can be stated that the individual’s identification with the online brand community has a strong and positive influence on their intentions regarding behavioral loyalty.
H3: Community identification behavior contribute positively to behavioral loyalty
2.4 Do engaged fans within an online brand community become more behavioral loyal over time?
According to literature a positive relationship exists between relationship duration and relationship quality. Customers make series of purchases over time, they face increasingly high costs when switching to a new supplier and will therefore come to view their commitment level to a particular supplier as relatively stable (Jackson, 1985). Besides high switching cost, relationship quality is build by relationship marketing attributes like trust, commitment, communication and conflict handling and empathy on relationship quality and customer loyalty. The relationship quality is build over time; companies need to invest time to increase the quality (Lee, Hung and Hsu, 2007). Research has shown that a moderating effect of relationship age on the effect that relational constructs, such as trust, satisfaction, and commitment, have on relationship outcomes. Longer relationships are qualitatively different from shorter ones (Grayson and Ambler, 1999). Longer relationships are accompanied by the relationship quality characteristics.
19 rely more on brand performance in the later phase than in the beginning of the brand building process. In the beginning phase customer engagement behaviors such as brand engagement, community engagement and community identification may be more important. In a later phase when fans already know the characteristics of the brand and do have already brand associations, the actual performance of the brand is more important instead of the influential engagement behaviors of other fans. Verhoef, Franses and Hoekstra (2002), suggest more or less the same in their research, but in an offline environment. They stated that in a later phases of the relationship, customers are more experienced and are more aware about long-lasting relationships and the relationship quality characteristics. The relationship quality is build over time. (Lee, Hung and Hsu, 2007).
It can be stated that in a later phase of relationships with fans, who are already familiar with a brand, have both associations with the brand and higher switching costs, are less influenced by the engagement behaviors of others within an online brand community. To conceptualize the relationships between the engagement behaviors, relationship duration and behavioral loyalty, a moderating effect of relationship duration on the relationship between the customer engagement behaviors and behavioral loyalty can be found. Unfortunately no literature about the relationships do exist. That is why the only following hypothesis can be made:
H4: Within an online brand community relationship duration contributes negatively on the relationship between customer engagement behaviors and behavioral loyalty
3.Methodology
Within this chapter the research methods of this research will be discussed. 3.1 Data collection
The hypotheses of this research are derived from existing studies within the literature. To measure the different constructs, scales will be developed and an highly structured questionnaire is made. Questions are carefully chosen or crafted, sequenced and precisely asked to every respondent of the online brand community of innocent.
20 post, since they are not a fan of the innocent Facebook page. Fans can click on the link showed in the Facebook post, in which respondents are asked for a favor, to give their opinion on every construct that has been measured within the Dutch questionnaire (appendix 1). The stimuli is shown within figure 2.
In order to measure the causation of the independent variables on the dependent variable, the variables will be manipulated by different questions of engagement behaviors and behavioral loyalty on the Likert-scales within the questionnaire. Respondents were not informed about the goal of the research in advance. Questions about behavioral loyalty are asked in the beginning of the questionnaire. All the respondents are judged to be engaged, since they took the effort to like the Facebook page in the past. The independent variables are judged to be the cause and the dependent variable to be the effect of the cause. This fits the main goal of the research to find out if engaged fans are behavioral loyal.
In order to draw a profile of the fans, their age and gender are asked. The first question is ‘what is your age’ the second one is ‘what is your gender’. Also questions such as ‘how long they are a fan’ and their ‘level of education’ will be asked. The experiment will only take place in the Netherlands and mainly concerns participants who live in the large urban cities. This is checked through a function on Facebook that gives an overview of cities where mainly the fans are located. The complete experiment will be analyzed by different methods in SPSS that give the final results of the research. Participants can complete the questionnaire within 3 minutes.
21 The target population of this research consists of fans within the online brand community Facebook of innocent. According to Malhotra (2007), the target sampling size for a study like this is 200 respondents. As this study tests four conditions each condition was set to have a minimum of 50 respondents. The sampling technique used in this research is the ‘convenience sampling’ technique, which is stated by Malhotra (2007) to make use of respondents that are in the right place at the right time. Using convenience sampling, snowball sampling is likely to occur. An example of this is, after filling in the questionnaire, fans liked the post by pressing on the like button online, these likes are seen on the Facebook page of other fans who liked the innocent Facebook as well.
3.1.1 Innocent introduction
22 building process? Do they buy the brand week after week and are they actively promoting the brand to others and are they encouraging others to buy the brand?
3.2 Characteristics of the sample
Totally 263 fans responded to the questionnaire. Some questionnaires were double, unfinished or had too many missing values. These questionnaires were corrected or deleted. In the end 227 questionnaires were useful for this research. The Facebook page of innocent has 7723 fans, 263 of them filled in the questionnaire, that makes a response rate of 263/7723*100 = 3.4%. Totally 25 men and 202 women respond on the questionnaire. The average age of the respondents is 30.6 years old. The education level of the respondents was high; at least 173 out of the 227 respondents have an education level of ‘higher education’ or higher. The respondents were asked how they became a member of the innocent Facebook page. The most effective channel to create membership increase is via Facebook (advertising), thereafter due to friends, thereafter the innocent wrapping, thereafter the ‘Big Knit’ campaign, thereafter the website of innocent and finally the Newsletter.
3.3 Measurement of variables
Operationalization of the independent variables. In the conceptual model three independent variables can be found: To measure the different constructs, these variables must be defined. Three different types of engagement behaviors within an online brand community are measured by the different constructs of Sprott et al. (2009); Algesheimer et al. (2005) and Lee et al. (2011).
To measure brand engagement a validated scale of Sprott et al. (2009) is used. They developed a brand engagement scale of self-concept (BESC). The brand engagement will be measured by eight questions on a 7-point Likert scale. The questions that will be asked are: (a) I have a special bond with the brands that I like; (b) I consider my favorite brands to be a part of myself; (c) I often feel a personal connection between my brands and me; (d) Part of me is defined by important brands in my life; (e) I feel as if I have a close personal connection with the brands I most prefer; (f) I can identify with important brands in my life; (g) There are links between the brands that I prefer and how I view myself; (h) My favorite brands are an important indication of who I am.
23 ‘strongly disagree’ to ‘strongly agree’. Specifically, the items asked subjects to indicate how likely they would be to participate in the following online brand community activities in future: (a) providing new information about the brand to other people; (b) actively participating in the online brand community’s activities; (c) supporting other members of the online brand community; (d) saying positive things about the online brand community to other people; (e) recommending the online brand community to anyone who sought their advice about the brand; (f ) encouraging other people to use the brand in future; and (g) not hesitating to refer other people to the brand.
Community identification will be measured by the validated scale of Lee et al. (2011). Community identification is measured on a six 7-point Likert-type items, ranging from ‘strongly disagree’ to ‘strongly agree.’ Specifically, the six items asked the participants to indicate how likely or unlikely they would be to: (a) belong to the online brand community; (b) identify themselves with the community; (c) be attached to the community; (d) see themselves as a part of the community; (e) be related with the community; and (7) get involved in the community.
Operationalization of the dependent variables. Behavioral loyalty has been measured by the developed scale of Algesheimer et al. (2005) of behavioral loyalty intentions. Behavioral loyal intentions are measured on a three 7-point Likert scales questions, ranging from strongly agree to strongly disagree. Questions that are asked are: (a) I intend to buy this brand in the near future; (b) I would actively search for this brand in order to buy the product; (c) I intend to buy other products of this brand. In order to measure the current purchasing behavior some questions of Keller (1993) are added to the questionnaire. These questions are: (d) I buy innocent since the day I know it exist (e) I bought innocent several times last month (f) if this brand were not available, it would make little difference to me if I had to use another brand.
Operationalization of moderating variable. The moderating variable is the duration of the relationship of fans within the online brand community. This will be measured by a multiple choice question. They can choose a category and these categories are: 0-3 months, 3-6 months, 6-12 months or 12 months or longer.
24
Relationship duration
#Months #Fans #Percentage #Cumulative
0 – 3 13 5.3 5.7
3 – 6 39 17.2 22.9
6 – 12 83 36.6 59.5
12> 92 40.5 100
Total 227 100
Table 1. Relationship duration.
Within figure 3. an overview can be found of the questions that will be asked in the questionnaire to manipulate the independent and dependent variables specified on innocent.
Construct: Questions: Authors:
Brand Engagement (a) I have a special bond with innocent; Sprott et al.
(b) I consider innocent to be a part of myself; (2009). (c) I often feel a personal connection between innocent and
me;
(d) Part of me is defined by innocent in my life; (e) I feel as if I have a close personal connection with
innocent;
(f) I can identify with innocent in my life; (g) There are links between innocent and how I view myself;
(h) innocent is an important indication of who I am.
Online community engagement
(a) I provide new information about innocent to other people;
Algesheimer et al. (2005). (b) I actively participate within the online brand community’s
activities of innocent;
(c)I support other members of the online brand community
of innocent;
(d)I say positive things about the online brand community of
innocent to other people;
(e)I recommend the online brand community of innocent to anyone who sought their advice about the brand; (f )I encourage other people to use innocent in future; (g)I do not hesitate to refer other people to innocent.
Community (a)I belong to the online brand community of innocent; Lee et al.
Identification (b) I identify myself with the community of innocent; (2011).
(c) I am attached to the community of innocent; (d) I see myself as a part of the community of innocent; (e) I am related with the community innocent; (f) I get involved in the community of innocent.
Relationship Duration
(a) How long are you a member of the innocent Facebook
page?
25 (b) I would actively search for innocent in order to buy the
product; al. (2005);
(c) I intend to buy other products of innocent;
(d) I buy innocent since the day I know it exist; Keller (1993). (e) I bought innocent several times last month;
(f) If innocent is not available, I would actively search for
innocent.
Figure 3. Questions of the questionnaire 3.3.1 Validity
In order to capture the validity, all constructs are derived from literature. The measurements of the constructs have been done by already validated scales within the literature. In order to overcome the translation mistakes from English to Dutch, the questions are translated to Dutch and afterwards back in English, to see if the translation was consistent. This has been done by seven test-persons. Furthermore, in order to get honest answers, questions about the buying behaviors are asked in the beginning of the questionnaire. A pretest has been done under a test-group, to see whether all questions were understandable and if they could easily be answered. Selection of the pretest and the final test group is randomly, the only criterion was being a member of de innocent Facebook.
3.3.2 Internal consistency reliability
Before testing the hypotheses, a reliability check for the questionnaire must be done. The Cronbach’s Alpha’s will be determined for every construct: brand engagement, community engagement, community identification and behavioral loyalty. Cronbach’s Alpha is the average of all possible split-half coefficients resulting from different ways of splitting the scale items (Malhotra, 2007). The value of this coefficient varies from 0 to 1 and as stated by Malhotra (2007). The outcomes of the Cronbach’s Alpha need to be higher than 0.6. All outcomes under 0.6 will be deleted from the research
26
Main construct Sub-constructs Number of measurements CA Brand engagement 8 0.918 Community engagement behaviors 6 0.857 Community member identification 7 0.932 Relationship duration Length of relationship 1 Ordinal data Behavioral loyalty 2 0.895 Current Purchase behavior 3 0.848 Behavioral Loyalty intentions 2 0.800
Table 2. Internal consistency 3.4 Analysis
To measure the effect of the independent variables on the dependent, a multiple regression technique will be used. According to Malhotra (2007) multiple regressions involve a single dependent variable and two or more independent variables. The reason why a regression analysis is used, is because it gives a simple and easy readable overview of the conceptual model, whereby interaction effects are easily measurable. Two regression models will be conducted. Within model one the effect of the three independent variables on the dependent variable will be explained. Since this is the most important part of the research. A separate model will be conducted to include the effect of the moderating variable. The second regression model will include the interaction effects.
The regression formulas will be:
= Dependent variable Behavioral Loyalty = Independent variable = Moderating variable Z1=Relationship Duration
Whereby the formulas of the different hypotheses are:
27 And
By adding the interaction effect of relationship duration on the relationships between the engagement behaviors and behavioral loyalty, increases the chance that multicollinearity will exists. A major disadvantage of a multiple regression analysis is, that it is complicated by the presence of multicollinearity and this means that the inter-correlations are high between independent variables. For example, for the relationship duration four categories are made, for every independent variable three new variables will be made in SPSS. High correlation can exist between these new variables and the independent variables used in the research. The Tolerance-value of below 0.1 or a VIF-value of higher than 10 is considered to be too much multicollinearity and will invalidate the results of the analysis (Field, 2005).
4. Results
This chapter will present the results of the research that has been conducted. The results will lead to the acceptation or rejection of the hypotheses stated in the literature review. Two multiple regressions are conducted. The first regression model includes the direct effects of the independent variables on the dependent variable. The second regression model includes the interaction effects.
4.1 Descriptives Community Identification Brand Engagement Community Engagement Behavioral Loyalty Mean 3.2989 4.1096 3.9185 3.7656 St. Devation 1.377 1.27 1.09 1.57 Table 4. Descriptives of construct.
28 mean of almost 3.8. This means that on average people filled in average or partly disagreed to behavioral loyalty. By plotting a normal distribution graph of the means and standard deviations of the different engagement behaviors, the graphs were steep, which explains that 95% of the respondents are within a small area on the scale. The graph of behavioral loyalty was more flat.
After mentioning the means, the question arise, what does it mean? The averages of all constructs are approximately 4, which mean average. If average is explained, respondents respondent to the questionnaire between partly agree and partly disagree. Brand engagement within the fan base shows the highest average followed by community engagement. Fans within an online community are slightly more brand engaged than community engaged. But overall fans who are mentioned to be engaged within an online brand community, are after all more or less average engaged. Moreover the score of behavioral loyalty is disappointing
4.2 Model 1
Table 4. gives a statistical overview of the direct effects of the independent variables on the dependent variable.
Variable: Behavioral Loyalty Significance VIF
ß-value P-value Community Identification 0.087 0.278 2.286 Community Engagement 0.248 0.004*** 2.286 Brand Engagement 0.218 0.005*** 1.73 Relationship Duration 0.029 0.621 1.012 Model overview R2 0.246 Adjusted R2 0.232 F-score 18,113 P-value (significance) 0.000*** * = Significant on 90%, α<.10 ** = Significant on 95%, α<.05 *** = Significant on 99%, α<.01
29 4.2.1 Hypotheses h1, h2 and h3
A multiple regression was performed by regressing the different independent variables on behavioral loyalty. The R2 (the measure that shows the unique contribution of the explained variance in the outcome) is 0.246 that indicates that 24.6% of the variability in behavioral loyalty is explained by the different independent variables. The overall F-score, which calculates the overall model fit is 18,113 and indicates that the independent variables have a strong positive effect on behavioral loyalty (p= 0.000: p<0.05). Within this model no high correlation exists between the different constructs, al VIF values are above 0.1 and under 10. The first construct was brand engagement: Brand engagement behaviors contribute to behavioral loyalty positively (Hypothesis h1). It can be concluded that hypotheses h1 is significantly supported (ß= 0.218, p= 0.005).
The second construct was community engagement: Community engagement behaviors contribute to behavioral loyalty positively (Hypothesis h2). It can be concluded that hypotheses h2 is significantly supported (ß= 0.248, p= 0.004).
The third construct was community identification. Community identification behaviors contribute to behavioral loyalty positively (Hypothesis h3). The influence of community identification behaviors on behavioral loyalty is not significantly supported within this research (ß= 0.087, p= 0.278).
30 4.3 Model 2
Within the total model that included all main and interaction effects, the tolerance-value of VIF-value higher than 10 and is considered to be too much multicollinearity and invalidated the results. In order to overcome multicollinearity, the moderating effects of relationship duration on the relationship between every single independent variable and the dependent variable, were isolated from each other. Three single regression models are made to find out if there are any significant results.
Variable:
Behavioral
Loyalty Significance VIF
ß-value P-value
Brand Engagement 0.392 0.000*** 1.147 Relationship Duration Dummy 1 * Brand
Engagement 0.128 0.196 2.683
Relationship Duration Dummy 2 * Brand
Engagement 0.193 0.109 3.86
Relationship Duration Dummy 3 * Brand
Engagement 0.122 0.199 2.475 Model overview R2 0.191 Adjusted R2 0.177 F-score 13.057 P-value 0.000*** Community Engagement 0.431 0.000*** 1.16 Relationship Duration Dummy 1 * Community
Engagement 0.11 0.249 2.57
Relationship Duration Dummy 2 * Community
Engagement 0.1 0.406 4.07
Relationship Duration Dummy 3 * Community
Engagement 0.097 0.313 2.60 Model overview R2 0.214 Adjusted R2 0.200 F-score 15.035 P-value 0.000*** Community Identification 0.337 0.000*** 1.250 Relationship Duration Dummy 1 * Community
identification 0.102 0.315 2.671 Relationship Duration Dummy 2 * Community
identification 0.141 0.256 4.011 Relationship Duration Dummy 3 * Community
identification 0.109 0.273 2.576
Model overview
31 Table 5. Isolated moderating effect 4.3.1 Hypothesis h4
The models within table 5. show the interaction and main effects. A multiple regression was performed on regressing the influence of brand engagement behaviors on behavioral loyalty moderated by relationship duration.
The first variable brand engagement has a R2 of 0.191 that indicates that 19.1% of the variability in behavioral loyalty is explained by the independent variable brand engagement, moderated by relationship duration. The overall F-score, which calculates the overall model fit is 13.057, which is significant (p=0.000: p<0.05). No high correlation exists between the different construct, all VIF values are above 0.1 and under 10. No significant results show up for brand engagement. It can be concluded that there are no differences between groups. The second variable is the influence of community engagement behaviors on behavioral loyalty moderated by relationship duration. The R2 is 0.214 that indicates that 21.4 % of the variability in behavioral loyalty is explained by community engagement moderated by relationship duration. The overall F-score, which calculates the overall model fit is 15.035, which is significant (p=0.000: p<0.05). No high correlation exists between the different construct, all VIF values are above 0.1 and under 10. The moderated relationship between community engagement and behavioral loyalty including the interaction effect of the relationship duration is not significant. It can be concluded that there are no differences between groups.
The third variable is community identification. The R2 (the measure that shows the unique contribution of the explained variance in the outcome) is 0.146 that indicates that 14.6 % of the variability in behavioral loyalty is explained by community identification moderated by relationship duration. The overall F-score, which calculates the overall model fit is 9.451, which is significant(p= 0.006: p<0.05). No high correlation exists between the different construct, all VIF values are above 0.1 and under 10. The moderated relationship between community identification and behavioral loyalty including the interaction effect of relationship duration is not significant. It can be concluded that there are no differences between groups.
32 Overall h4 cannot be supported due the insignificant results of all interaction effects. Within model 1. hypotheses h1 and h2 are significant.
5. Discussion and Conclusions
Within this chapter, the conclusions to the problem statement will be drawn based on the results which are presented in chapter five. By doing so, the main goal of the research is to evaluate if the engaged fans within an online brand community such as Facebook are behavioral loyal. Based on research that is already done within literature, researchers showed that online brand community participants have high levels of customer engagement with the firm’s product(s) and brand(s) (McAlexander et al., 2002), are motivated to help other customers (Bagozzi and Dholakia, 2006), are very loyal and actively recruit others to the community (Algesheimer et al., 2005). For companies opportunities to increase customer engagement behaviors is the most important reason to develop online brand communities (Franke and Shah, 2003). The engagement behaviors within an online brand community do consist of three different common definitions. According to Lee et al. (2011); Sprott et al. (2009) and Algesheimer et al. (2010) they consist of brand engagement, how a fan can identified themselves with a brand (Sprott et al., 2009) community engagement, how addicted a fan is to the online brand community (Algesheimer et all., 2010) and community identification, whether a fan can identify him or herself with the online brand community (Lee et al., 2011). With significant results of h1 and h2, it can be stated that both customer engagement behaviors such as brand engagement and community engagement within an online brand communities contributes to behavioral loyalty positively.
33 the amount of money invested in an online marketing strategy for example the costs of setting up an online brand community and the continuation of it.
Even though it can be agreed on Algesheimer et al. (2010) that community engagement behaviors within an online brand community are educational and that fans who will search for reference before buying the product might slow or discourage the buying process. Recent research undermines the effect of community engagement behaviors, for example word of mouth or other non-transactional behavior may create opportunities, like sales (Verhoef et al., 2010). For example within a Facebook page, fans can create community engagement behaviors together with other members. Concluded from the research can be that fans that show higher community engagement behaviors, are more willing to show behavioral loyalty. An online brand community is not the channel that directly leads to extra sales, but gives the possibility that it can create a channel for openly word of mouth and interactions, which can increase community engagement behaviors. Both high and low community engaged fans are useful, because it is not only believed that high customer community engagement is necessary for future growth, it is also believed that low customer engagement is detrimental to success, both due to lost sales or opportunities and negative WOM and can be used for different purpose, such as the co-creation of the brand and marketing program (EIU, 2007) By good management of community engagement behaviors within the online brand community, behavioral loyalty will be contributed positively.
Another way to increase behavioral loyalty within an online brand community is through brand engagement behaviors. The results of the research show significant results for the relationship between brand engagement behaviors and behavioral loyalty. Fans have the propensity to include relevant brands as important parts of themselves. By associating specific characteristics of the brands, for example marketing programs or trademarks of the brands, brands become a part of the self-concept. The fans have the propensity to share their brand associations with others within an online brand community, for example, friends, family or strangers. They can use their relationship with the brand to enlarge the own self-concept or brand building, which is showed online to others (Sprott et al., 2009). Moreover it can be agreed on Sprott et al. (2009) that high brand engaged customers within online brand communities are associated with more positive brand attitudes and purchase intentions. By showing others the associations, it can increase the brand awareness and buying recommendations for others as well.
34 (Field, 2005). After conducting other regression analysis to overcome multicollinearity, no significant results showed up. According to that there is no moderating effect of membership duration on the relationship between the customer engagement behaviors and behavioral loyalty. Concluded from the research, no differences between long-term member and short-term fans within Facebook are found. This is not in line with the research of Onishi and Manchanda (2012). They stated that the influences of the different customer engagement behaviors are stronger for short-term fans, because they cannot rely on the brand performance yet.
To conclude, overall a positive relation between engaged fans within an online brand community and behavioral loyalty can be found. The general level of both brand and community engagement within an online brand community and behavioral loyalty was not spectacular. Concluded can be that other researches overestimate this effect, by suggesting that an engaged Facebook (fan) leads to enormous returns. Within this research it is proven that Facebook is useful due to the significant hypotheses, but must be seen as a channel that can create opportunities such as brand engagement behaviors and community engagement behaviors. Customers can create value from a firm through the sharing of positive news and opinions with others and this has the potential to affect engagement behaviors within an online brand community of both senders and receivers. It can be concluded that such sharing influences attitudes, ideas, behaviors and decision making. All though this must be transformed into additional buying behaviors, but these (additional) buying behaviors lie behind and can be improved by an ideal management of the online brand community, where both brand and community engagement behaviors are fostered.
6. Implications
6.1 Theoretical implications
The results of this research show that engaged fans within an online brand community affect behavioral loyalty positively. This is in line with the theory as already mentioned and showed by other studies: customer participation in online brand communities positively affects loyalty and strengthens relationships (Algesheimer et al., 2005). There is a big but, this article nuance this relationship.
35 significance of hypotheses h1and h2, the results of this research show significant results for community engagement and brand engagement, which contributes to behavioral loyalty positively. Further research can be done on this new model if other factors within an online brand community influence behavioral loyalty.
The reason why this research has been done is to give an overview of the behavioral loyalty of fans within an online brand community, who are treated as engaged customers. It showed the level of two forms of customer engagement of single fans and the level of behavioral loyalty. Both levels were disappointing. This research is quit unique due to the fact that this article is not directly too enthusiastic about online brand communities or thinks that an online brand community is obligatory for a company/brand. All thought this research did not quantify the worth of every single fan, it did make an overview of the general purchase behaviors of the engaged fans within an online brand community. This model can be used by companies or researchers to give an indication about the level of engagement and behavioral loyalty within an online brand community.
This research does not find support for the moderating variable relationship duration, which is not in line with Onishi and Manchanda (2012). An explanation can be, that due to low behavioral loyalty, it can be concluded that respondents did not buy products often. They have less brand experience and probably are not aware of relationship quality characteristics, such as trust and commitment. Concluded from the insignificant community identification, brand engagement and community engagement seem to be more important behaviors within an online brand community.
The effect of word of mouth seems to be very important within research, since the effect of word of mouth is huge within online brand communities; fans are reviewing and recommending the brand to all kind of people and within different channels online. Mainly word of mouth is separated variable, within this research word of mouth is part of community engagement behaviors and not a single independent variable.
6.2 Managerial implications
The research highlights the effect if engaged fans within online brand communities influence positively behavioral loyalty. In line with the results of the research it is valuable for marketing managers to manage and set up an online brand community such as Facebook for their brand in order to create the right channel for increased customer engagement behaviors and to spend their marketing budget properly.
36 intensive buying behavior at all. The research shows that customer engagement behaviors such as brand and community engagement do influence behavioral loyalty positively, but the average level of engagement and buying behaviors are average within the online brand community that has been researched here. Marketing managers must be creative to come up with ideas that increase brand engagement and community engagement in order to increase buying behavior. He or she has to prioritize the social media marketing investment to make them happy. Moreover they have to feel appreciated and find ways to talk about the brand and share their opinions (BusinessNewsDaily, 2013).
The results show substantial variation among consumers in their level of engagement with their ‘favorite’ brand within an online brand community. Prior research has suggested (Schau and Gilly, 2003), that a favorite brand plays a self-defining role only for some engaged fans within an online brand community. Try to find these fans and focus on them, since they have more favorable brand attitudes and have higher purchase intentions, which can actually influence the final purchase behaviors.
Even though not every engaged member within an online brand community is behavioral loyal, but the value of both higher and lower engaged customers within an online brand community can be used for different purposes. Higher engaged customers within an online brand community can be used for future growth, extra sales or increased brand equity. Lower engaged customers within an online brand community can be used for providing frequent feedback about products and services (EIU, 2007). Another non purchase-related way that both lower and higher engaged customers can create value for a firm is through their participation in new product development processes, co-creation, and their willingness to provide feedback for innovations and improvements to existing products and services via online brand communities. An online brand community serves as a platform for such collaboration with customers providing opportunities to easily offer suggestions and input to the firm (Sawhney, Verona and Prandelli, 2005).
Recent research also undermines the effect of word of mouth within the community engagement behaviors that may create opportunities, like sales (Verhoef et al., 2010; Kumar et al., 2010). The financial value can be increased when there is some existence of word of mouth within an online brand community. Marketing managers should create an online brand community that is open for comments, reviews or other forms of community engagement behaviors, because such an environment creates opportunities for sales.
37 exceptional brand engagement and community engagement behaviors of the online brand community member towards additional and extraordinary buying behaviors. If the company is not able to do that, it could cost them more money than the company will yield.
7. Limitations & Future research
This research shows some limitations. First of all, the results can only prospect short term results. The existence of Facebook is only for a short period of time yet. Long term effects should be measured over time.
Second, within in this research attitudes and behaviors are measured towards behavioral loyalty. The attitudes of the fans on Facebook of innocent were measured on one specific moment in the middle of a large campaign.
Third, this research lacks an extensive pretest. A pretest has been done within a small group of fans. An extensive pretest was not possible, because the fans would be asked for an additional request in a short period of time; it would influence the response rate after all.
The results of this research cannot be generalized. To get a better overview of the influence of an online brand community on behavioral loyalty, different online brand communities of different brands should be examined. It can be suggested that an online brand community of low-involved products have a different function than high-involved products. Moreover there might be difference of the function of an online brand community within a business-to-business market or a business-to-business-to-consumer market.
In order to get a better comparison of the effects of engagement behaviors of fans within an online brand community, a comparison should be made with a group of people who are not a fan. This group was difficult to find, because it was impossible to find a group of people who were brand aware of innocent, due to no offline customer database of fans. Moreover another questionnaire should be created.
38 Future research can find out which kind of engagement behavior within an online brand community has the strongest effect on behavioral loyalty. Moreover a research can be done if community engagement has an influence on brand engagement. It can be proposed, that through being engaged with the community, a fan is more likely to be engaged with the brand as well. Moreover it is likely that through brand engagement a fan will be community engaged as well.
