Online Brand Communities:
Brand Engagement and Brand Loyalty in the Company-led
Online Brand Community Environment
The Impact of Company-led Online Brand Communities on Brand Loyalty.
Master thesis author:
Julia Nicole Florine Alkemade (11407883)
Under the supervision of:
Dr. Hüseyin Güngör
MSc in Business Administration – Marketing Track
Amsterdam Business School
University of Amsterdam
Statement of Originality
This document is written by Julia Alkemade who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Table of Contents
ABSTRACT ... 4
1. INTRODUCTION ... 5
2. LITERATURE REVIEW ... 7
2.1ONLINE BRAND COMMUNITIES ... 7
2.2CUSTOMER EXPERIENCE AND BRAND ENGAGEMENT ... 9
2.3BRAND LOYALTY ... 11
2.4ONLINE BRAND COMMUNITIES, BRAND ENGAGEMENT AND BRAND LOYALTY ... 13
3. THEORETICAL AND MANAGERIAL CONTRIBUTION ... 15
3.1THEORETICAL CONTRIBUTION ... 15
3.2MANAGERIAL CONTRIBUTION ... 16
4. CONCEPTUAL FRAMEWORK ... 18
4.1HYPOTHESIS 1:BRAND ENGAGEMENT AND BRAND LOYALTY ... 19
4.2HYPOTHESIS 2:ONLINE BRAND COMMUNITY AND BRAND ENGAGEMENT ... 20
4.3HYPOTHESIS 3 AND 4:ONLINE BRAND COMMUNITY AND BRAND LOYALTY ... 21
5. METHODOLOGY ... 23
5.1SAMPLE AND DATA COLLECTION ... 23
5.2SURVEY ... 24
5.3VARIABLES AND MEASUREMENTS ... 25
5.3.1 Company-led online brand community ... 25
5.3.2 Brand engagement ... 25 5.3.3 Brand loyalty ... 26 5.3.4 Control Variables ... 27 6. RESULTS ... 28 6.1DEMOGRAPHICS... 28 6.2RELIABILITY ANALYSIS ... 29 6.2.1 Brand engagement ... 29 6.2.2 Brand loyalty ... 30
6.2.3 AH Customers and AH Members in comparison to entire dataset ... 30
6.3HYPOTHESIS TESTING ... 31
6.3.1 Hypothesis 1: Brand engagement and brand loyalty ... 31
6.3.2 Hypothesis 2: Online brand community and brand engagement ... 33
6.3.3 Hypothesis 3: Online brand community and brand loyalty ... 34
6.3.3 Hypothesis 4: Online brand community, brand engagement and brand loyalty ... 35
6.4DIFFERENCES IN NPS AND BRAND LOYALTY... 36
7. DISCUSSION ... 39
7.1THEORETICAL IMPLICATIONS ... 39
7.2PRACTICAL IMPLICATIONS ... 41
7.3LIMITATIONS AND SUGGESTIONS FOR FURTHER RESEARCH ... 42
8. CONCLUSION ... 44 REFERENCES... 45 APPENDIX ... 47 APPENDIX A ... 47 APPENDIX B ... 50 APPENDIX C ... 51 APPENDIX D ... 51
Abstract
Online brand communities become more and more important as managers believe in the fact that an online brand community can enhance engagement with the brand. Companies increasingly build their own hosted online brand community, as they are concerned about the loss of control in customer-initiated online brand communities. This research uses data that is collected within an online brand community of a giant Dutch supermarket and among general grocery shoppers (i.e. non-community members) of this supermarket chain. To confirm the managers’ belief, brand engagement and brand loyalty of the community members and non-community members are compared to reveal whether brand engagement and brand loyalty increases when consumers are member of an online brand community. Moreover, the relationship between an online brand community membership and brand loyalty is explained with brand engagement as moderator. The results indicate that there is a significant positive relationship between brand engagement and brand loyalty for both the grocery shoppers of the Dutch supermarket and members of the community. However, there is no significant relationship between a membership of an online brand community and brand engagement and brand loyalty respectively. Finally, brand engagement does not moderate the relationship between a membership of an online brand community and brand loyalty. Taken together, these findings shed light on the importance of own-hosted online brand communities within the Dutch Fast Moving Consumer Goods (FMCG) industry.
1. Introduction
The rise of the Internet has encouraged companies to build online brand communities because these non-geographic limited communities help marketing managers to reach consumers worldwide and at any time (Jung, Kim and Kim, 2014; Zaglia, 2013). Based on the belief that consumers who join the community become more engaged with the firm or products and the increasing economic activity with the firm, online brand communities have become a growing concept (Manchanda, Packard and Pattabhiramaiah, 2015). Both small and large companies (e.g. Coca Cola, Apple, Starbucks and Nike) build online brand communities to improve their customers’ visits and purchases (Wu, Huang, Zhao and Hua, 2015).
An online brand community is an internet-based platform where people share and expose the same interest and enthusiasm in a specific brand, share values and norms and form relationships with other members of the community (Nambisan and Watt, 2011; Sicilia and Palazon, 2008). Online brand communities are initiated by the company (i.e. company-led) or by consumers (i.e. consumer-led) (Kang, Shin and Gong, 2016). Companies use online brand communities for long-term relationship-building purposes with consumers whereas consumer-initiated online brand communities are created for informational purposes (Jung et al., 2014). Several companies have decided to build their own-hosted brand community, as they are concerned about the loss of control and strong negative reactions of customers in customer-led communities hosted by a third-party (e.g. Facebook) (Manchanda et al., 2015).
As 50 per cent of the top 100 global brands make use of an online brand community, these communities become more and more important in their businesses (Manchanda et al., 2015). For example, the giant Dutch supermarket chain Albert Heijn (AH) introduced a few years ago its own AH Community platform (AH Community, n.d.). Within this community, members can interact with each other, ask questions, help each other with these questions and share their opinions about new products or initiatives (Ibid.) However, when being a member
of the AH Community, does this mean that this consumer does groceries at AH supermarkets and comes back frequently? In other words, will the member of the online brand community be more loyal to the brand when there is a high level of engagement with the brand? And are these community members more loyal to the brand than non-community members? These are questions that companies face nowadays.
Although online brand communities have been broadly researched in the literature, the focus of these researches is online brand community engagement, which focuses on the feeling of belonging to an online brand community (Zheng, Cheung, Lee and Liang, 2015). Little attention is paid to brand engagement and brand loyalty among online brand community members, which focuses on the brand. Yet there is research to suggest more attention to the power of brand engagement in predicting actual behaviour (e.g. brand loyalty in terms of repurchase of the preferred brand) in online brand communities (Zheng et al., 2015).
To have a better understanding of these concepts and their relationships, this research will investigate the relationship between the membership of a company-led online brand community, brand engagement and brand loyalty. Brand engagement will be used as a moderator in the relationship between a company-led online brand community membership and brand loyalty to identify if brand engagement increases brand loyalty.
The brand is the central focus of this research rather than the feeling of belonging to an online brand community itself (i.e. online brand community engagement).
In the following paragraphs, the literature is discussed and criticized to explore shortcomings. Subsequently, the contribution of this research, conceptual framework, methodology, results, discussion, limitations and suggestions for further research are discussed.
2. Literature review
This section reviews the existing literature about online brand communities, brand engagement and brand loyalty. The purpose of this review is to define online brand communities, brand engagement and brand loyalty and criticise the current literature to identify a literature gap, which is the setting for this research.
2.1 Online brand communities
The literature paid a lot of attention to online brand communities. Wu et al. (2015) studied the relationship between online brand communities and purchase frequency whereas Jung et al. (2014) studied the relationship between attitude towards the online brand community and the consequences of this attitude (i.e. brand trust and revisit intention) based on marketer-created and consumer-created online brand communities.
Many researches have defined the concept online brand community. Table 1 shows several definitions of a brand community followed by definitions of an online brand community. Based on the definitions of brand communities in the literature, a brand community is defined as a not geographically limited group of people who have the same interest and enthusiasm in a specific brand based on social relationships (Muñiz and O’Guinn, 2001; Cova and Pace (2006) in Sicilia and Palazon, 2008). Applying this definition to the online context, online brand communities are a company-initiated internet-based platform where people share and expose the same interest and enthusiasm in a specific brand and form social relationships which link consumer with other consumers and with the brand (Luo, Shang, Hu and Wang, 2016; Nambisan and Watt, 2011; Sicilia and Palazon, 2008).
Author Definition
Muñiz and O’Guinn (2001). ‘A brand community is a specialized community with no geographical limitations, and is based on a structured set of social relations among people who like particular brands.’
Cova and Pace (2006) in Sicilia and Palazon (2008).
‘A brand community is defined as any group of people that possess a common interest in a specific brand and create a parallel social universe rife with its own myths, values, rituals, vocabulary and hierarchy.’
Fuller, Jawecki and Muhlbacher, 2007 in Brogi (2014).
‘An Online Brand Community is a brand community that take place in a virtual setting in which the members’ interaction is primarily Internet-mediated.’
Nambisan and Watt (2011). ‘Online communities constitute a social space wherein relationships and
ties are formed among the members and a common set of values and norms are established and shared.’
Sicilia an Palazon (2008). ‘Virtual brand community as a group of individuals with common
interests in a brand who communicate to each other electronically in a platform provided by the company which supports the brand.’
Muniz and O’guinn (2001); Luo, Shang, Hu and Wang (2016).
‘Online brand community is not only a physical place, but, in reality, is a set of social relationships linking a brand with its customers, customers with other customers, which constitute a basic framework of community relationships.’
Table 1 Definitions of (online) brand communities
Online brand communities are divided in consumer-led communities and company-led. Consumer-led communities are formed by consumers who share their experiences and information about a particular brand or product (Kang, et al., 2016). The members of these communities are strongly tied to each other as they create their own content and Word-of-Mouth (i.e. WOM) within the online brand community (Kang, et al., 2016). Reasons for consumers to become a member of a brand community are developing knowledge and skills related to the brand, entertainment and socialization (Zaglia, 2013).
Company-led communities are initiated by the company and encourage positive consumer attitudes towards the brand and products or services provided by the company (Kang et al., 2016). Reasons for companies to use online brand communities are building long-term relationships with consumers and increase economic activity (Jung et al., 2014;
Manchanda et al., 2015). While Jung et al. (2014) only identified consumers for building relationships, Wirtz et al. (2013) mentioned employees, consumers of competing brands and market experts. This implies that relationships can be built with anyone that is interested and active in the market and the brand. Moreover, due to the interaction in the online brand community between companies and consumers, companies have access to consumer insights (Wirtz et al., 2013). This information might give insight in how consumers perceive and experience (new) products and services for further improvements (Kumar, Chattaraman, Neghina, Skiera, Aksoy, Buoye; Henseler (2013) in Wirtz et al., 2013).
Thus, companies use an online brand community to build long-term relationships with their consumers, which will in turn increase economic activity.
2.2 Customer experience and brand engagement
Customer experience, offline as well as online, is frequently discussed in the literature (Brakus, Schmitt, & Zarantonello (2009) in Kim and Yu, 2016; Nambisan and Watt, 2011; Rose, Hair and Clark, 2011). These different researches discuss antecedents and consequences of customer experience. Several consequences of customer experience are customer satisfaction, repurchase intentions (Rose et al., 2011), positive attitudes towards the product and the company (Nambisan and Watt, 2011) and consumer engagement (Brodie et al., 2013). The latter is increasingly used in the academic marketing literature since 2005 (Ibid.).
Customer experience involves an interaction between the consumer and an aspect of a brand or organization, which elicits a personal reaction towards the brand (e.g. rational, emotional, sensorial, physical and spiritual) (Gentile, Spiller and Noci, 2007). In an online context, this interaction is based on interactions in an online brand community that evokes a
personal reaction (Nambisan and Watt, 2011). This personal reaction can result in brand engagement and brand loyalty (Bowden (2009) in Brodie et al., 2013; Kang et al., 2016).
According to Hollebeek, Glynn and Brodie (2014), customer brand engagement (CBE) is the consumer’s positive balanced brand-related cognitive, emotional and behavioural activity during or related to focal consumer or brand interactions. Previous research of Hollebeek found three themes, which are related to these activities (Hollebeek, 2011). Firstly, immersion relates to ‘strongly focusing on’ and ‘being absorbed in’ and corresponds to cognitive processing. Therefore, immersion is defined as the levels of brand-related thought processing and elaboration in particular brand interactions (Hollebeek et al., 2014).
Secondly, passion relates to being obsessed towards and/or loving the brand, which corresponds to the emotional activity in brand interactions (Ibid.). Passion is defined as the degree of the consumer’s positive brand-related affect in particular brand interactions (Ibid.).
Lastly, activation relates to behavioural activity. Activation is the consumer’s level of energy, effort and time spent on a brand in particular brand interactions (Ibid.). This latter theme aligns with Keller (2013, p. 348) who defined active engagement as ‘the extent to which consumers are willing to invest their own personal resources – time, energy, money – on the brand, beyond those resources expended during purchase or consumption of the brand’.
In conclusion, brand engagement can be divided in three types of activities: cognitive processing, affection and activation. These activities are consequences of an interaction with the brand. However, these activities do not involve consumption or repurchase activities of the brand.
2.3 Brand loyalty
Different researches have identified the consequences of brand engagement such as satisfaction, loyalty, commitment and emotional connection (Chan and Li (2010) in Brodie et al., 2013; Brodie et al., 2013). In these researches, brand loyalty is the consequence of online brand community engagement rather than the consequence of brand engagement. For this research, loyalty is highlighted as a consequence of brand engagement.
Brand loyalty is conceptualized into behavioural and attitudinal loyalty. Behavioural loyalty involves consumer’s continuous repurchases of products and services from the same brand (Zheng et al., 2015) and the amount of category volume attributed to the brand (Keller, 2013). In other words, how often do consumers purchase the brand and how much do they purchase of this brand? Attitudinal loyalty involves willingness of being in a relationship with the brand and feeling committed to the brand (Zheng et al., 2015).
Table 2 presents different definitions of brand loyalty over the years. Most of these definitions refer to behavioural loyalty (i.e. (intended) repurchases of the brand) (Oliver, 1997; Keller, 2013; Solem, 2016). However, the relationship-building aspect of brand loyalty is highlighted is several definitions as well (Aaker, 1996, p.320; Keller, 2013; Khan and Mahmood in Maheshwari, Lodorfos and Jacobsen, 2014). According to Keller (2013), loyalty is part of brand resonance in building Customer-Based Brand Equity (CBBE) and involves the nature of the customer’s relationship with the brand and feeling in synch with the brand. True brand resonance indicates that consumers have a close relationship with the brand, want to be actively involved with the brand and share experiences with others (i.e. high level of loyalty) (Ibid.). These aspects of a high degree of loyalty generate benefits for companies such as positive WOM (i.e. share experiences with others) and reduced marketing costs as it is less costly to retain consumers than gaining them (Aaker, 1996; Chaudhuri and Holbrook (2001) in Maheshwari et al., 2014).
Author Definitions
Solem (2016). ‘Brand loyalty denotes an intended behaviour in relation to the brand
and/or its services.’
Aaker (1996, p. 320). ‘The behavioural willingness to consistently maintain relations with a
particular brand.’
Oliver (1997). ‘Customer loyalty is an actual and profound behaviour that consistently
supports and repurchases preferred products or services, despite competitors’ marketing efforts to change their behaviour’
Keller (2013). ‘Loyalty is the repeat purchase rates and the extent to which customers
seek out brand information, events and other loyal customers.’
Khan and Mahmood (2012, p. 33) in Maheshwari, Lodorfos and Jacobsen (2014).
‘Brand loyalty can be defined as the customer’s unconditional
commitment and a strong relationship with the brand, which is not likely to be affected under normal circumstances.’
Table 2 Definitions of brand loyalty
Another measure for loyalty is the Net Promoter Score (NPS), which indicates the firm-customer relationship. According to Reichheld (2003), this is a strong indicator of loyalty as consumers act as reference, which can affect their reputation. Therefore, consumers refer only if they feel intense loyal to a brand. The NPS is measured by the simple question ‘How likely is it that you would recommend Company X to a friend or colleague?’ and is ranked on a 0 to 10 Likert-scale (Ibid.). Respondents who rate the company from 0 to 6 are ‘Detractors’ and extremely unlikely to recommend the company, rating 7 or 8 are ‘Passively Satisfied’ and 9 and 10 are ‘Promoters’ who are extremely likely to recommend the company (Ibid.). To calculate the NPS, the percentage of ‘Promoters’ is subtracted by the percentage of ‘Detractors’ and gives a score between -100% and 100% (Ibid.). When the NPS is 75% or more than 80%, the company has world-class loyalty (Ibid.). To keep this growth sustainable and profitable, more promoters and fewer detractors need to be created which increases the NPS (Ibid.).
In conclusion, the existing literature shows different views on brand loyalty. Whereas behavioural loyalty is about repurchase behaviour, the NPS involves the referral behaviour of consumers. These definitions of loyalty are used in this research.
2.4 Online brand communities, brand engagement and brand loyalty
Online brand communities, brand engagement and brand loyalty are broadly defined and explained concepts in the literature. Nevertheless, these researches paid a lot of attention to customer engagement within the online brand community (i.e. online community commitment) rather than brand engagement, which focuses on the brand (Kang et al., 2016; Zheng et al, 2015). Moreover, these researches examine the antecedents of consumer engagement rather than the consequences such as loyalty, commitment and participation (Kang et al. 2016). Therefore, more attention is suggested to the consequences of brand engagement, particularly the power of brand engagement in predicting actual behaviour (e.g. behavioural loyalty) (Kang et al., 2016; Zheng et al, 2015).
Further, the literature rarely makes a distinction between company-led and consumer-led communities. The majority focuses on online brand communities, which are initiated by consumers or does not make a distinction between company-led communities and consumer-led communities (Brodie et al., 2013; Luo et al., 2016). As these online brand communities are used for different purposes by consumers and companies, these communities need to be researched separately. In addition, several companies have expressed their concerns about the loss of control and strong negative reactions of consumers in customer-led communities hosted by a third-party (Manchanda et al., 2015). Therefore, more insights about company-led online brand communities might help companies to consider an own-hosted online brand community by showing the importance of company-led online brand communities.
This research examines the relationship between the membership of a company-led online brand community, brand engagement and as a consequence brand loyalty within the Dutch FMCG industry. The relationship between brand engagement and brand loyalty is identified among members of the AH Community (i.e. AH Members) and non-members of the AH Community (i.e. AH Customers) respectively. Additionally, the moderating effect of
brand engagement on the relationship between the membership of the company-led online brand community and brand loyalty is identified. Lastly, the difference between the average level of brand loyalty and the NPS of AH Customers and AH Members is clarified. For all these relationships, the focus is brand engagement rather than online community engagement that members experience.
This research aims to respond the following question:
‘What is the effect of a company-led online brand community on brand loyalty in terms of intended repurchases of the brand? And is this relationship moderated by brand engagement?’
3. Theoretical and managerial contribution
3.1 Theoretical contribution
Online brand communities are becoming increasingly important for companies, as the community becomes a key aspect of Web-based initiatives and facilitate long-term relationships with consumers (Jung et al., 2014; Nambisan and Watt, 2011). The literature rarely makes a distinction between company-led and consumer-led communities and focuses mainly online brand communities, which are initiated by consumers (Brodie et al., 2013; Luo et al., 2016). This research contributes to the concept company-led online brand community which is a company-hosted website or platform rather than a social media-based platform that is hosted by a third party.
Since 2005, engagement is increasingly used in the academic marketing literature (Brodie et al., 2013). The literature paid a lot of attention to the antecedents of consumer engagement rather than consequences (Kang et al., 2016; Zheng et al., 2015). This research focuses on the brand and therefore the power of brand engagement is highlighted. The power of brand engagement is highlighted in the relationship with a membership of a company-led online brand community, brand loyalty and brand engagement functions as moderator in the relationship between a membership of a company-led online brand community and brand loyalty. The relationship between brand engagement and brand loyalty is identified to confirm previous research (Brodie et al., 2013; Nawaz and Usman, 2011).
Loyalty is viewed from two different perspectives, behavioural loyalty (i.e. repurchases of the brand) and referral behaviour (i.e. NPS). The difference in the level of behavioural loyalty and referral behaviour between members and non-members of the community will show the impact of online brand communities on loyalty. The NPS contributes to the existing literature by measuring actual reference behaviour of consumers in the FMCG industry.
The combination of a membership of a company-led online brand community, brand engagement and brand loyalty will generate new insights, as previous research focuses mainly on online community commitment (Kang et al., 2016; Zheng et al, 2015). Moreover, this research focuses on the outcomes of the brand (i.e. brand engagement and brand loyalty) among consumers and community members within the Dutch FMCG industry.
3.2 Managerial contribution
As 50 per cent of the top 100 global brands make use of an online brand community, these communities become more and more important for marketing managers (Manchanda et al., 2015). Therefore, marketing managers are considering an own-hosted online brand community for their brand or company. Moreover, the boost of the Internet facilitates worldwide growing accessibility to consumers, which implies that the non-geographic limitation of online brand communities helps marketing managers to reach consumers worldwide at any time (Zaglia, 2013).
The focus on company-led online brand communities in this research, rather than consumer-led online brand communities, will generate a deeper understanding of this type of community for marketing managers in the Dutch FMCG industry, especially for AH. Company-led online brand communities might encourage brand engagement and contribute to brand loyalty. This means that a company-led online brand community will build a strong and tied relationship between the consumer and the brand, which is fundamental for long-term relationship building (i.e. brand resonance) (Keller, 2013). Additionally, this research will show if brand engagement will lead to more intended repurchases among members of an company-led online brand community (i.e. AH Members). Based on findings in previous research, brand loyalty can result in decreasing marketing costs and positive WOM (Aaker, 1996; Chaudhuri and Holbrook (2001) in Maheshwari et al., 2014). If brand engagement
functions as a moderator in the relationship between the membership of a company-led online brand community and brand loyalty, these findings can become highly significant for marketing managers within the Dutch FMCG industry.
Finally, loyalty is examined from different perspectives. The level of behavioural loyalty will show marketing managers the intended repurchases whereas the NPS will show consumers’ active and actual referral behaviour. The NPS will gives insights in whether AH consumer, including AH Customer and AH Members, recommend doing groceries at the Dutch supermarket chain AH.
4. Conceptual framework
This section visualizes the conceptual framework used for this research and explains the relationships between the variables in the framework (see Figure 1). The model shows that the direct relationship between brand engagement and brand loyalty is expected to be positive (H1a, H1b, H1c). This relationship is identified to confirm previous research (Brodie et al., 2013; Nawaz and Usman, 2011). The second relationship (H2a, H2b, H2c) expects that a company-led online brand community has a positive direct effect on brand engagement. The third relationship expects a positive direct relationship between a company-led online brand community and brand loyalty (H3). This latter relationship is expanded by brand engagement as moderating variable (H4). Therefore, H4 expects that a company-led online brand community has a positive effect on brand loyalty under the condition of brand engagement. Within this framework, the control variables age and gender are incorporated in each relationship. However, it is expected that there is no effect of these variables on the relationships. H3 (+) H1a (+), b (+), c (+) H4 (+) Company-led Online Brand Community Membership Brand Loyalty Intended repurchase of the preferred brand
Figure 1 Conceptual framework
Brand Engagement
Affection Activation H2a (+), b (+), c (+)
4.1 Hypothesis 1: Brand engagement and brand loyalty
Existing literature focuses on the antecedents of customer engagement and less on the consequences such as loyalty (Kang et al. 2016). Brodie et al. (2013) found a positive relationship between engagement and loyalty in terms of referral and purchase behaviour of consumers. They showed that a high level of consumer engagement increases consumers’ recommendations of the brand to others and arouses actual positive behaviour towards the brand, which will result in behavioural loyalty (i.e. repurchases of the brand). Brodie et al. (2013) examined this relationship within a company-led online brand community by observing the communication and interactions (i.e. posts). They focused on what consumers say in the community rather than their attitude towards the brand on cognitive, emotional and behavioural level.
Nawaz and Usman (2011) found a significant weak relationship between brand commitment and loyalty in the telecom sector. This implies that the relationship between brand engagement and brand loyalty is dependent on the industry in which the brand operates. Therefore, this relationship needs to be examined in a different industry.
To examine if the relationship between brand engagement and brand loyalty holds within the Dutch FMCG industry, the CBE concept of Hollebeek et al. (2014) is used as this concept focuses on the brand and the customer’s investments and motivation towards the brand (i.e. attitude). Cognitive processing, affection and activation deepen into the cognitive, emotional and behavioural motivation of consumers respectively (Hollebeek, 2011; Hollebeek et al., 2014). These three themes are tested on social media platforms such as Facebook, LinkedIn and Twitter, which are third-party platforms. This research will investigate the positive relationship between brand engagement (i.e. affection and activation respectively) and brand loyalty among AH Customers and AH Members. Only the affection (i.e. emotional) and activation (i.e. behavioural) aspects of CBE will be used because within
this FMCG industry consumers frequently purchase products from the same brand or supermarket. It is therefore expected that they feel emotionally engaged with the brand or supermarket chain. The activation aspect of CBE is used because the membership of the community and interactions within the community implies that a consumer wants to put behavioural effort in the brand beyond consumption. The extra effort is therefore expected to be a good predictor for repurchases.
Thus, this research will investigate if the relationship between affection and brand loyalty, activation and brand loyalty and overall brand engagement and loyalty holds among AH Customers, who are not members of the AH Community, with the following hypothesis (see Figure 1):
H1a: There is a positive relationship between affection and intended repurchases of the brand. H1b: There is a positive relationship between activation and intended repurchases of the brand.
H1c: There is a positive relationship between brand engagement and intended repurchases of the brand.
4.2 Hypothesis 2: Online brand community and brand engagement
Company-led online brand communities decrease the distance between consumers and the brand. Lee, Kim and Kim (2011) found that there is a positive relationship between company-led online brand communities and online brand community commitment. As this research focuses on brand engagement, it is expected that this relationship hold for company-led online brand communities and brand engagement as well.
Moreover, becoming a member of an online brand community indicates that consumers want to put more effort in the relationship with the brand and stay up-to-date about the developments of the brand (Zaglia, 2013). The CBE aspects affection and activation are aligned with these findings as they refer to the emotional closer relationship and active behaviour towards the brand beyond consumption, respectively. Therefore, it is
expected that the membership of a company-led online brand community has a positive relationship with affection and activation.
Wirtz et al. (2013) state that a membership of an online brand community contributes to creating brand engagement. However, this direct relationship has not been examined yet in the literature. To examine this relationship, again affection and activation will be explained according to the following hypothesis (see Figure 1):
H2a: There is a positive relationship between a company-led online brand community and affection. H2b: There is a positive relationship between a company-led online brand community and activation. H2c: There is a positive relationship between a company-led online brand community and brand engagement.
4.3 Hypothesis 3 and 4: Online brand community and brand loyalty
Company-led online brand communities are used for building long term relationship with consumers (Jung et al., 2014). Zheng et al. (2015) found evidence for the positive relationship between online community commitment and brand loyalty among Facebook users, both attitudinal and behavioural. This relationship is examined on a third-party platform (i.e. Facebook) rather than a company-led online brand community. Moreover, participation in online brand communities has a significant positive influence on purchase frequency (Wu et al., 2015). The purchase frequency can be aligned to behavioural loyalty as consumers are repurchasing the product or service of the brand. However, this participation is based on promotion-focused customers, which are seeking for benefits and rewards to participate in the company-led online brand community (Ibid.). This research focuses on the membership of a company-led online brand community without promotional purposes or incentives to become a member of the community.
Wirtz et al. (2013) state that a membership and participation in the online brand community enhance brand engagement and brand loyalty, however this relationship is not empirically tested. Thus, being a member of the company-led online brand community might therefore positively influences brand loyalty in terms of intended repurchase behaviour. This direct relationship between online brand communities and brand loyalty might be influenced by brand engagement, as there is a positive relationship between brand engagement and brand loyalty (Brodie et al., 2013; Nawaz and Usman, 2011).
In this research, the focus is on company-led online brand communities as companies are increasingly investing in building their own-hosted online brand community (Manchanda et al., 2015). This research will test if the membership of a company-led online brand community has a positive relationship with brand loyalty under the condition of being engaged with the brand rather than the community. The following hypotheses will be tested (see Figure 1):
H3: There is a positive relationship between a company-led online brand community and intended repurchases of the brand.
H4: Brand engagement moderates the positive relationship between a company-led online brand community and intended repurchases of the brand.
5. Methodology
For this research, the data is collected via a survey among consumers of the Dutch supermarket chain AH and the AH Community. This company is the largest supermarket company in The Netherlands and still growing. A few years ago, they introduced the AH Community. This company-initiated community consists of approximately 5000 active members and provide members to interact with other members, ask questions, rate (new) ideas within the AH Community and help other members (AH Community, n.d.). To become a member of this community, consumers sign up with their Facebook account. Both AH consumers and the members of the AH Community are used for this research as these consumers are frequently encountered with AH and do their groceries at this supermarket on a daily basis.
5.1 Sample and data collection
The overall sample population is based on company-led online brand communities within the FMCG industry in The Netherlands. The sample consists of AH consumers which can be divided in AH Customers and AH Members. AH Customers includes consumers who are not a member of the AH Community whereas AH Members are members of the AH Community. There are approximately 5000 members within the AH Community who can participate in the survey.
Cross-sectional data about consumer’s feelings of belonging to and the relationship with the brand is collected according to a survey in two different groups. To distinguish between AH Members and AH Customers, a control question is included (i.e. ‘Are you a member of the AH Community?’). For both groups, the participants are unknown and the survey is accessible for everyone. Therefore, it cannot be assumed that the participants who
who are member and participate via Facebook or LinkedIn. However, it is expected that most respondents who find the survey via the AH Community are community members. With the distinction between these type of AH consumers, a difference between the average level brand loyalty and NPS of AH Members and AH Customer can be revealed.
For AH Customers, there is a focus on the direct relationship between brand engagement and brand loyalty and the survey is distributed in the online environment via Mail, Facebook and LinkedIn (Appendix A). These consumers all had an equal chance to participate in the survey as convenience sampling is used. The survey was open on May 1, 2017 and closed when approximately 120 respondents were collected.
For AH Members, the survey was placed on the homepage of the AH Community platform (Appendix A). Every member had access to the link of the survey and the survey was open from May 17, 2017 until June 9, 2017. In this case, non-probability quota sampling is used because AH Members are used within the entire company-led online brand community population. There is one limitation within this environment as many members are not active and cannot be forced to participate. Consequently, there is a chance that the sample of AH Members is smaller than the sample of AH Customers.
5.2 Survey
The same survey is used for both AH Customers and AH Members (Appendix A). The survey focuses on doing groceries at the supermarket chain AH rather than a specific branded product of AH. Some questions are eliminated to make the survey doable for both groups. For example, AH Customers might not be interested in participating in the survey or are in a hurry while doing their groceries. Therefore, the survey is as short as possible to make it attractive for consumers to participate in.
The survey was build with the online survey platform Qualtrics that is licensed to the University of Amsterdam. For both groups, the survey was accessible via a link created by Qualtrics.
As shown in Appendix A, the survey is distributed in Dutch. Because the measurement scales used were in English, a translation process was needed. First, the survey was translated from English to Dutch. Afterwards an independent person reviewed this translation and translated the questions back to Dutch (Appendix B).
The actual distributed survey for both AH Customers and AH Members can be found in Appendix A.
5.3 Variables and measurements
5.3.1 Company-led online brand community
Company-led online brand community is measured by the membership of the AH Community. The control question (i.e. ‘Are you a member of the AH Community?’) assigned each participant to one of the groups, AH Customers or AH Members. Participants who answered ‘yes’ are assigned to the group AH Members and participants who answered ‘no’ became part of the AH Customers group.
5.3.2 Brand engagement
Brand engagement is measured by the validated CBE concept (Consumer Brand Engagement) of Hollebeek et al. (2014). CBE involves cognitive processing, affection and activation (Ibid.). This scale is validated on a seven-point Likert scale anchored with ‘strongly disagree’ (1) through to ‘strongly agree’ (7). The correlations between cognitive processing, affection and activation are significant (r=0.856 (COG-AFF), r=0.816 (AFF-ACT), r=0.711 (ACT-COG), p<0.05) and the scale reliability has a Cronbach’s alpha α = 0.825 for cognitive processing, α = 0.907 for affection, α = 0.894 for activation and α = 0.933
for the overall CBE scale (Ibid.). This scale is reliable and consists of 10 items of which one is ‘I think about [brand] a lot when I am using it (Table 3) (Saunders, Lewis and Thornhill, 2009).
For this research, only the affection (i.e. emotional) and activation (i.e. behavioural) aspects of CBE are used as AH consumers do their groceries almost every day, come back to this supermarket chain and therefore feel emotionally engaged with doing groceries at AH. The activation aspect of CBE is used because the membership of the community and interaction within the community implies that consumers want to put effort in the brand and show active behaviour beyond consumption of the brand. This behavioural activity might be a good predictor for repurchases of the brand. Both affection and activation are rated on a seven-point Likert scale anchored with ‘strongly disagree’ (1) through to ‘strongly agree’ (7).
5.3.3 Brand loyalty
Brand loyalty is measured by the scale used in Hur, Ahn and Kim (2011) because this research focuses on the intended behaviour of repurchase of the preferred brand. This scale is reliable with a Cronbach’s alpha α = 0.75 which is above the threshold of 0.7 (Saunders et al., 2009). The scale consists of 3 items of which one is ‘I will repurchase [brand] in the year to come’ (Table 4). However, for the reliability analysis only the first two items are included. All items of the scale are rated on a seven-point Likert scale anchored with ‘strongly disagree’ (1) through to ‘strongly agree’ (7).
At the end of the survey, an additional question measures the NPS to examine loyalty in terms of referral behaviour. The NPS consists of the question ‘How likely is it that you would recommend Albert Heijn to a friend or colleague?’ and is ranked on a 0 (i.e. not likely at all) to 10 (i.e. highly likely) Likert scale (Reichheld, 2003). The NPS is calculated by subtracting the percentage of ‘Detractors’ from the percentage of ‘Promoters’. Afterwards, the NPS of both AH Customers and AH Members can be compared.
5.3.4 Control Variables
The control variables are age and gender. This demographic information is gathered by asking for gender and age. It is expected that the majority of the respondents are female as women do more often groceries than men.
Age is divided in four subgroups, younger than 21, from 21 to 40, from 41 and 60 and older than 60. It is expected that the two subgroups between 21 and 60 are the largest group of respondents.
For both control variables, it is expected that there is no influence of these variables on the relationships between company-led online brand community, brand engagement and brand loyalty.
Table 3 Items of cognitive processing, affection and activation
Notes: (a): items ignored in reliability test Table 4 Items of brand loyalty
Cognitive processing Affection Activation
Using [brand] gets me to think about [brand].
I feel very positive when I use [brand].
I spend a lot of time using [brand], compared to other [category] brands.
I think about [brand] a lot when I am using it.
Using [brand] makes me happy. Whenever I am using [category], I usually use [brand].
Using [brand] stimulates my interest to learn more about [brand].
I feel good when I use [brand]. [Brand] is one the brand I usually use when I use [category].
I am proud to use [brand] (0.795).
Brand loyalty
1. I will repurchase [brand] in the year to come (0.90). 2. I would love to use [brand] continuously (0.89).
6. Results
In this section, the results of this research are presented. Before any analysis in SPSS is executed, a reliability analysis is done followed by the explanation of the calculation of brand engagement. The hypotheses are tested via a correlation analysis and a regression analysis in which the interaction effect of company-led online brand community and brand engagement is included. To identify differences between the two groups, an ANOVA analysis is done to identify the differences in the average level of brand loyalty and the NPS of both groups.
The data is divided into two datasets. For the first hypothesis, the dataset includes only the data of AH Customers as hypothesis 1 focuses on the direct relationship between brand engagement and brand loyalty. The second data set involves both AH Customers and AH Members and is used for hypothesis 2, 3 and 4.
6.1 Demographics
Table 5 shows the general descriptive statistics of the dataset including both AH Customers and AH Members. The total response rate of AH Customers is 138 of which 8 respondents did not finish the survey and 5 respondents are member of the AH Community. These 13 respondents will be excluded which result in a total number of 125 respondents.
For AH Members, the response is 38 of which 2 respondents did not finish the survey and 4 respondents are not members. These respondents are excluded from the dataset and the 5 respondents from the AH Customers survey, who are member of the AH Community, are included which result in a total number of 37 respondents. This small sample of AH Members (n=37) in comparison to AH Customers (n=125) is one of the limitations of this research.
As expected, the majority of the respondents are female (67.9%) and the largest age groups are 21-40 years (61.1%) and 41-60 years (31.5%) (Table 5).
Variables Frequency Per cent Total 162 100% AH Members 37 22.8% AH Customers 125 77.2% Age 162 100% ≤ 20 years 5 3.1% 21-40 years 99 61.1% 41-60 years 51 31.5% ≥ 61 years 7 4.3% Gender 162 100% Male 52 32.1% Female 110 67.9%
Table 5 Descriptive statistics of the overall dataset
6.2 Reliability analysis
A reliability analysis of the entire dataset (i.e. AH Customers and AH Members) is executed for the concept brand loyalty to confirm previous findings of Hur et al. (2011) and for brand engagement because the measurement scale of brand engagement of Hollebeek et al. (2014) involves only affection and activation. The overall brand engagement scale is calculated by all items of affection and activation.
6.2.1 Brand engagement
The scale reliability of affection has a Cronbach’s alpha α = 0.835, activation has a Cronbach’s alpha α = 0.817 and the overall brand engagement has a Cronbach’s alpha α = 0.819. The results show that all scales are reliable because the Cronbach’s alpha values are above the threshold of Cronbach’s alpha α = 0.7 (Saunders et al., 2009). Therefore, the measurement scales of affection and activation of the CBE concept of Hollebeek et al. (2014) are validated again. Although brand engagement consists solely of affection and activation, the scale of brand engagement is reliable. The corrected total-item correlations indicate that all items for affection, activation and brand engagement have a good correlation with the total
score of the scale respectively (i.e. all above 0.30). This means that there is a highly significant discrimination between the questions of each variable scale. Additionally, none of the items would substantially affect reliability if they were deleted.
6.2.2 Brand loyalty
The scale reliability of brand loyalty is measured with the first two items of the scale as measured by Hur et al. (2011). The Cronbach’s alpha for brand loyalty is α = 0.867 which means that this scale is reliable (Cronbach’s alpha α > 0.7) (Saunders et al., 2009). The Cronbach’s alpha α = 0.867 is higher than the Cronbach’s alpha found by Hur et al. (2011) (α = 0.75) and confirms the reliability of the scale with the first two items. The corrected total-item correlations indicate that the first two total-items are highly significantly discriminated as the values are above 0.3. Item 3 has a corrected total-item correlations value far below 0.3 (i.e. 0.098) and would substantially change the reliability of the scale into an unreliable scale with Cronbach’s alpha α = 0.484 (Saunders et al., 2009). Therefore, item 3 of this scale is excluded in the reliability analysis, which is suggested by Hur et al. (2011).
6.2.3 AH Customers and AH Members in comparison to entire dataset
A reliability analysis is executed for the data set of AH Customers and AH Members separately. To distinguish between AH Customers and AH Members, a case selection is used to exclude AH Members in the analysis of AH Customers and AH Customers in the analysis of AH Members. Again, the reliability of brand engagement is calculated with all the items of affection and activation. The results of these analyses are presented in Appendix C, Table 11.
For AH Customers, the Cronbach’s alpha of affection, activation, brand engagement and brand loyalty are lower than the Cronbach’s alpha of the entire dataset. Again, the measurement scales are reliable. None of the items of activation, affection and brand engagement would substantially affect reliability if they were deleted and the items of each
scale show highly significant discrimination as the values of the total-item correlations are above 0.3. However, deleting item 3 of the brand loyalty measurement scale substantially affects the reliability of this scale. Again, the corrected total-item correlations indicate that the first two items are highly significantly discriminated as the values are above 0.3.
For AH Members, the reliability analysis shows higher Cronbach’s alphas of the measurement scales affection, activation, brand engagement and brand loyalty than the Cronbach’s alphas of the entire dataset. Moreover, the corrected total-item correlations indicate that all items for affection, activation and brand engagement have a good correlation with the total score of the scale respectively (i.e. all above 0.30). Additionally, none of the items would substantially affect reliability if they were deleted.
Brand loyalty show a remarkable difference between the AH Members dataset and the entire dataset. When all three items are included in the reliability analysis, the analysis shows a Cronbach’s alpha of α = 0.646 which is higher than the Cronbach’s alpha α = 0.484 of the entire dataset and closer to the threshold rule of 0.7 (Saunders et al., 2009). When adjusting the threshold to 0.6, all the items of brand loyalty form a reliable scale.
In conclusion, all measurement scales are reliable and validated among members of a company-led online brand community and consumers within the Dutch FMCG industry.
6.3 Hypothesis testing
6.3.1 Hypothesis 1: Brand engagement and brand loyalty
The first hypothesis is tested by a Pearson-correlation analysis to identify whether there is a relationship between brand engagement and brand loyalty among AH Customers (i.e. non-members of the AH Community).
Table 6 displays descriptive statistics and correlations of the data representing AH Customers. Brand engagement is calculated by all the items of affection and activation,
which result in a new mean value of brand engagement. The results show that there is a significant positive relationship between activation and affection (r = 0.36; p<0.01). Moreover, there is a significant positive relationship between affection and brand loyalty (r = 0.60; p<0.01), activation and brand loyalty (r = 0.43; p<0.01) and brand engagement and brand loyalty (r = 0.62; p<0.01) as proposed in H1a, H1b and H1c. The relationship between activation and brand loyalty and brand engagement and brand loyalty are strong (r > ±0.5) whereas the relation between activation and brand loyalty is less strongly correlated (r<0.50) (Field, 2013). Therefore, activation has less impact on brand loyalty than affection and brand engagement. These results show support for hypothesis 1a, 1b and 1c. In other words, there is a strong positive relationship between affection and brand loyalty and between brand engagement and brand loyalty and a less strong positive relationship between activation and brand loyalty among AH Customers.
Additionally, Table 6 shows a significant positive relationship between affection and gender (r = 0.23; p<0.01) (Ibid.). This indicates that gender has an moderate positive effect on affection (r<0.50) (Ibid.).
Brodie et al. (2013) examined the relationship between brand engagement and brand loyalty by observing the communication and interactions within the online brand community and Nawaz and Usman (2011) found a significant weak relationship between brand commitment and loyalty in the telecom sector. The results of this research confirm the findings of Brodie et al. (2013) with different measurements (i.e. attitudes) and the findings of Nawaz and Usman (2011) within a different industry. However, for the FMCG industry the relationships are significantly stronger than the telecom sector. These results show that the relationship between brand engagement and brand loyalty is industry dependent.
Variables M SD 1 2 3 4 5 6 1. Gender 1.69 0.47 1 2. Age 2.39 0.60 -0.80 1 3. Affection 0.49 1.02 0.23** -0.16 1 4. Activation 4.93 1.41 0.07 -0.55 0.36** 1 5. Brand Engagement 4.91 0.98 0.19* -0.12 0.82** 0.83** 1 6. Brand Loyalty 4.45 0.91 0.04 -0.01 0.60** 0.43** 0.62** 1
**. Correlation is significant at 0.01 level (2-tailed). *. Correlation is significant at 0.05 level (2-tailed). Table 6 Correlation Matrix AH Customers
6.3.2 Hypothesis 2: Online brand community and brand engagement
The second and third hypotheses are tested by a new correlation analysis to identify whether there is a relationship between the membership of a company-led online brand community and brand engagement and between the membership of a company-led online brand community and brand loyalty.
Table 7 displays descriptive statistics and the correlations between the variables. The results are based on the dataset that includes both AH Customers and AH Members. To distinguish between AH Members and AH Customers, dummy variables are created. Again, brand engagement is calculated by all the items of affection and activation, which result in a new mean value. Again, there is a positive relationship between affection (r = 0.64; p<0.01), activation (r = 0.47; p<0.01) and brand engagement (r = 0.65; p<0.01) with brand loyalty respectively and confirms previous research from Brodie et al. (2013) (Table 7).
Hypothesis 2a, 2b and 2c suggest that there is a positive relationship between the membership of an online brand community and affection, activation and brand engagement respectively. The results in Table 7 show no support for these hypotheses, as the correlation values (r) are not significant. In other words, there is no significant positive relationship between the membership of an online brand community and affection (r = 0.07; p>0.05), activation (r = -0.01; p>0.05) and brand engagement (r = 0.04; p>0.05) respectively.
It is striking that gender (r = -0.04; p>0.05), age (r = -0.06; p>0.05) and activation (r = -0.01; p>0.05) are the only variables that show a weak negative relationship with company-led online brand community (Table 7). This is in contrast with hypothesis 2b that proposed a positive relationship between a membership of a company-led online brand community and brand engagement. However, it cannot be assumed that there is a negative relationship as the results are not significant.
6.3.3 Hypothesis 3: Online brand community and brand loyalty
Hypothesis 3 proposed that an online brand community has a positive effect on brand loyalty. Table 7 shows that there is no significant positive relationship between the membership of a company-led online brand community and brand loyalty (r = 0.09; p>0.05). Therefore, it cannot be assumed that the membership of a company-led online brand community has impact on brand loyalty.
Although the results of AH Customers (Table 6) show a significant moderate positive relationship between gender and affection, Table 7 shows a significant weak positive relationship between gender and affection among AH Members (r = 0.18; p<0.05) (Table 7). The relationship between gender and affection is stronger for AH Customers (r = 0.23; p<0.01) than AH Members (r = 0.18; p<0.05).
In conclusion, the membership of an online brand community shows no positive relationship with brand engagement and brand loyalty, which indicate that an online brand community is not essential for creating brand engagement or brand loyalty within the FMCG industry. Consequently, the results do not support the managers’ belief of the impact of a company-online brand community on brand engagement and brand loyalty.
Variables M SD 1 2 3 4 5 6 7 1. Gender 1.68 0.47 1 2. Age 2.37 0.62 -0.08 1 3. Affection 4.94 1.02 0.18* -0.14 1 4. Activation 4.93 1.45 0.05 0.05 0.41** 1 5. Brand Engagement 4.93 1.01 0.14 -0.05 0.83** 0.85** 1 6. Brand Loyalty 4.49 0.93 0.07 -0.02 0.64** 0.47** 0.65** 1 7. Membership 0.23 0.42 -0.04 -0.06 0.07 -0.01 0.04 0.09 1
**. Correlation is significant at 0.01 level (2-tailed). *. Correlation is significant at 0.05 level (2-tailed). Table 7 Correlation Matrix Community Members
6.3.3 Hypothesis 4: Online brand community, brand engagement and brand loyalty
To test the moderating relationship, a hierarchical regression analysis is executed including the membership of the AH Community, brand engagement and the interaction effect between these two variables on brand loyalty. As the dataset involves AH Customers and AH Members, dummy variables are created to include solely AH Members in the hierarchical regression analysis. The interaction effect is calculated by multiplying AH Members and brand engagement. The hierarchical regression analysis includes two models. The first model includes brand engagement, membership, gender and age on brand loyalty. The second model includes the same variables as model 1 and the interaction effect of membership and brand engagement. A hierarchical regression analysis is executed to show if the interaction effect in model 2 will change the variance in brand loyalty of model 1. Moreover, the control variables gender and age are included of which no effect in this relationship is expected.
Table 8 presents the results of the two models of the regression analysis. The first model is statistically significant F (4,157); p<0.001 and explained 43.2% of the variance in brand loyalty. The second model explains no additional variance in brand loyalty (R2=0.433).
This indicates that the interaction effect has no significant impact on the variance in brand loyalty. In both models, brand engagement is the only significant predictor of brand loyalty,
engagement increases for one, brand loyalty increases for 0.66 and 0.63 respectively. Moreover, the interaction effect shows no significance and therefore hypothesis 4 is not supported. It cannot be assumed that brand engagement moderates the relationship between a membership of a company-led online brand community and brand loyalty. A company-led online brand community membership might not be as crucial for creating brand loyalty according to the data gathered among AH Members.
R R2 R2 Change B SE ß t Model 1 0.66 0.43*** 0.43 Brand Engagement 0.60 0.06 0.66*** 10.76 Membership 0.14 0.14 0.06 1.03 Gender -0.04 0.12 -0.20 -0.32 Age 0.03 0.09 0.02 0.29 Model 2 0.66 0.43 0.01 Brand Engagement 0.58 0.07 0.63*** 8.76 Membership -0.24 0.64 -0.11 -0.37 Gender -0.03 0.12 -0.02 -0.28 Age 0.02 0.09 0.01 0.21 Interaction (Membership*Brand Engagement) 0.08 0.13 0.18 0.60
Note: Statistical significance: *p<0.05, **p<0.01, ***p<0.001 Table 8 Hierarchical Regression Model of Brand Loyalty
6.4 Differences in NPS and brand loyalty
To identify differences between AH Customer and AH Members, the NPS and average level of brand loyalty (i.e. behavioural loyalty) are analysed and compared. The NPS is calculated manually whereas the average level of brand loyalty is analysed with a one-way ANOVA analysis to identify a significant difference between the two groups.
Table 9 shows the number of ‘Detractors’, ‘Passively Satisfied’ and ‘Promoters’ of each group. The NPS, for both groups, is calculated by subtracting the percentage of ‘Detractors’ from the percentage of ‘Promoters’. The NPS for AH Members is 5.4 which is positive, however low. The NPS of AH Customers is -15.3 which is even lower than the AH Members and negative. A positive NPS is considered as good (Reichheld, 2013). The NPS of AH Members is positive which means that they recommend AH for grocery shopping. However, for AH Customers the NPS is negative which indicates that there are more ‘Detractors’ than ‘Promoters’ who are unlikely to recommend grocery shopping at AH. Therefore, there is much to improve and identify how this low NPS can be improved.
Grade Frequency Per cent
MEMBERS 37 100% Detractors (3,6) 6 16.2% Passively Satisfied (7,8) 23 62.1% Promoters (9,10) 8 21.6% NON-MEMBERS 125 100% Detractors 34 27.4% Passively Satisfied 76 60.5% Promoters 15 12.1%
Table 9 Net Promoter Score
Lastly, brand loyalty of AH Customers and AH Members is compared with a one-way ANOVA test. The results of the Levene’s test shows that there is a difference between the variances in the population as this test of homogeneity is insignificant, F(1,159); p>0.05). In addition, the graph of this analysis shows that AH Members have a higher average brand loyalty level (4.64) in comparison to AH Customers (4.46) (Appendix D, Figure 2). However, the one-way ANOVA analysis shows no statistically significant differences between the group means on brand loyalty F(1,159) = 1.04; p>0.05 (Table 10). Therefore, it cannot be
Remarkably, the higher mean score of brand loyalty of AH Members is based on a smaller sample (n=37) than the sample of AH Customers (n=125). This might indicate that when the sample of AH Members is significantly higher, the difference will be significant.
SS DF MS F Sig.
Between Groups 0.896 1 0.90 1.04 0.31
Within Groups 136.464 159
Total 137.360 160
7. Discussion
This study aimed to empirically test the relationship between the membership of a company-led online brand community and brand loyalty, which is moderated by brand engagement. Specifically, it first tested the direct effects of brand engagement on brand loyalty and the membership of a company-led online brand community on brand engagement. Next, the relationship between the membership of a company-led online brand community and brand loyalty was analysed, followed by the moderating effect of brand engagement on this relationship. The theoretical and managerial implications of the foregoing findings are discussed followed by the limitations and suggestions for further research.
7.1 Theoretical implications
This research expands the existing literature with new insights about online brand communities, brand engagement and brand loyalty within the Dutch FMCG and supermarket industry. Previous research has not made a distinction between the different community types (Brodie et al., 2013; Luo et al., 2016), has not examined the direct relationship between online brand communities, brand engagement and brand loyalty respectively (Wirtz et al., 2013) and has highlighted solely online brand community commitment (Brodie et al., 2013).
This research contributes to the literature as it focuses on company-led online brand communities rather than consumer-led online brand communities. Moreover, the direct relationships between company-led online brand communities, brand engagement and brand loyalty are examined among community members and consumers of the giant supermarket chain AH. These outcomes (i.e. brand engagement and brand loyalty) are focused on the brand rather than online brand community commitment.
The results of this research validated the reliability of the different measurement scales of Hollebeek et al. (2014) and Hur et al. (2011) and confirm the findings of Brodie et
al. (2013) and Nawaz and Usman (2011). As Nawaz and Usman (2011) found a weak relationship between brand engagement and brand loyalty within the telecom sector, this research shows a significant strong relationship between brand engagement and brand loyalty within the Dutch FMCG industry. The use of a different real business context reinforces the empirical results and show that the relationship between brand engagement and brand loyalty is dependent on the industry.
As the literature focuses mainly on the antecedents of brand engagement, the relationships of this research highlight the power of brand engagement among grocery shoppers and within a company-led online brand community environment. Brand engagement and brand loyalty show a strong positive relationship among both members and non-members of a company-led online brand community. Specifically, affection, as aspect of brand engagement, results in more intended repurchases of the brand (i.e. grocery shopping at AH supermarket). As these relationships are significant, the membership of an online brand community does not have a significant impact on brand engagement and brand loyalty, respectively. Thus, the importance of company-online brand communities in the literature can be questioned.
Moreover, brand loyalty is viewed from two different perspectives, behavioural loyalty (i.e. repurchases of the brand) and referral behaviour (i.e. NPS). These two different types of brand loyalty contribute to the literature by focussing on two different aspects of consumer behaviour. The difference in the level of behavioural loyalty and referral behaviour between members and non-members of a company-led online brand community give new insights in the impact of online brand communities. The results show a higher average level of behavioural loyalty and NPS for community members than consumers.
