Escolar Documentos
Profissional Documentos
Cultura Documentos
Master Thesis
Author: Justina Malciute
Advisor: Polymeros Chrysochou
Aarhus University
Business and Social Sciences
August 2012
Abstract
A great interest in the concept of customer engagement has emerged along with the rise of
online social media during the past few years. Marketing practitioners were the first ones
attempting to define and understand the potential outcomes of customer engagement.
However, due to a lack of scholarly interest and empirical support, the nature of customer
engagement has remained rather vague and its presupposed capability to enhance customer
relationships still uncertain. The aim of this study is to bridge this gap by proposing a
conceptual model of customer brand engagement in the context of online social media
platforms and conducting an empirical analysis. Drawing on the overview of academic
literature and the results of a quantitative online consumer study, the paper delivers a
thorough investigation of the concept and offers empirical evidence of its impact on the
ultimate business performance. The most important findings of this study suggest that both
customer brand relationship related factors and online social media platform related factors
can influence the level of customer engagement, which in turn will influence the level of
behavioral loyalty and the spread of word-of-mouth communication. Thus, this paper is an
important contribution to academic marketing literature in the field of customer
engagement, which still remains mostly conceptual or qualitative, and provides useful
managerial insights for marketing practitioners.
Keywords: customer engagement, brands, social media, customer relationships, brand
loyalty, word-of-mouth.
Table of contents
1. Introduction.................................................................................................................. 1
2. Literature review .......................................................................................................... 3
2.1 Conceptual foundations....................................................................................... 3
2.2 Engagement conceptualizations in social science and management literature ... 5
2.3 Engagement conceptualizations in the marketing literature ............................... 6
2.4 Conceptual relationships ................................................................................... 10
2.5 Engagement in the online social media context ................................................ 14
2.6 Problem statement ............................................................................................. 16
2.7 A conceptual model of customer brand engagement on online social media
platforms ................................................................................................................. 17
3. Methodology............................................................................................................... 20
3.1 Data collection .................................................................................................. 20
3.2 Measurement of constructs ............................................................................... 23
3.3 Statistical analysis ............................................................................................. 28
4. Results ........................................................................................................................ 30
4.1 Descriptive analysis .......................................................................................... 30
4.2 Measurement reliability and validity ................................................................ 33
4.3 Model estimation results ................................................................................... 37
4.4 Moderation effects ............................................................................................ 38
5. Discussion and implications ....................................................................................... 41
5.1 Implications for marketing theory ..................................................................... 42
5.2 Managerial implications .................................................................................... 43
5.3 Limitations and future research ......................................................................... 45
6. Conclusion .................................................................................................................. 46
References ...................................................................................................................... 48
Appendix 1: Online Questionnaire ................................................................................. 52
Appendix 2: Top Facebook Pages, Worldwide, Food & Drink Brands ......................... 62
ii
List of figures
Figure 1. Conceptual model of customer engagement behavior .................................... 14
Figure 2. Conceptual model of customer brand engagement on online social media
platforms ......................................................................................................................... 19
Figure 3: Fan distribution based on engagement level (N1=112) .................................. 31
List of tables
Table 1: Characteristics of the respondents (N1=112, N2=307) .................................... 23
Table 2: Construct measurement items, sources and scale reliabilities .......................... 26
Table 3: Means, standard deviations and results of t-test for equality of means (N1=112,
N2=307) .......................................................................................................................... 31
Table 4: Means, standard deviations and results of t-test for equality of means in
behavioral brand loyalty of high and low engaged fans (N1a=56, N1b=56) ................. 32
Table 5: Reliability and validity measures for first-order latent constructs (N1=112) .. 33
Table 6: Average variance extracted and squared correlations between first-order latent
constructs (N1=112) ....................................................................................................... 35
Table 7: Reliability and validity measures for second-order latent construct of customer
brand relationship related antecedents (N1=112) ........................................................... 36
Table 8: Estimated weights and variance inflation factors for formative dimensions of
second-order latent construct of online social media platform related antecedents
(N1=112) ........................................................................................................................ 36
Table 9: Results and direct effects of the structural path model (N1=112) .................... 37
Table 10: Results of the two-stage PLS approach for estimating moderating effects
(N1=112) ........................................................................................................................ 40
iii
1. Introduction
Engage or die is the new marketing catchphrase, which emerged as a result of the rise of
social media in the past few years (Nelson-Field & Taylor, 2012). The practitioners from
various industries caught on to it and the topic quickly became of great interest. Numerous
business conferences, seminars, discussion forums, blogs, commentaries and white papers
were suddenly talking about the concept of customer engagement, which did not really
exist in the marketing literature before (Brodie, Hollebeek, Juri, & Ili, 2011a).
The rules of engagement are new to the marketers and require some major changes in
the conventional marketing thinking. It is no longer a monologue dictated by the firm
through a commercial, print ad or a corporate website. The emergence of new media
provides businesses with an opportunity to start a two-way digital conversation with the
audiences and makes it almost effortless for an individual customer to talk back and also
talk to each other (Deighton & Kornfeld, 2009). The new media channels such as YouTube,
Wikipedia, Facebook, Twitter or MySpace gave a voice to the customers and made it
possible for them to create and easily share their own web content. In other words, each
individual has now the opportunity to become a media producer, an author, a reviewer, or
engage in many other kinds of behaviors that can be consumed by others on the Internet.
Thus, instead of generally being the ones to talk brands have now become the ones mostly
talked about.
The businesses gradually came to realization that they have to change their way of
looking at the customer, and the concept of engagement appeared to be the key to success.
The rationale behind this assertion is the prevailing conception of customer engagement as
a way to create deeper and more lasting customer brand relationships (Kumar et al., 2010).
And even though the traditional media still plays the major role in reaching the customer,
the companies are increasingly using the new social media channels for managing their
customer relationships. Research showed, that social media has emerged as a valuable tool
widely employed by businesses and even 54% of executives of consumer goods companies
participating in a recently conducted survey said that social media was central to their effort
to engage consumers in 2011(WARC, 2012a). Hence, even though no single theory exists
1
on how customer engagement on social media works, marketers have been actively
pioneering the field. Almost every brand today has an established profile on the mainstream
social media platforms such as Facebook, Twitter or Google+. Others have also turned to
more novel platforms such as Instagram, Pinterest or Foursquare. There are multiple
different ways and tools that businesses can use in order to engage their customers.
However, despite all the effort the levels of customer engagement resulting on the
platforms of social media suggest that the conventional marketing knowledge lacks the
ability to explain and guide the marketers throughout the process. One recent practitioner
study of the most popular brands on Facebook has discovered that less than 5% of brands
were able to attract repeated fan visits to their page within a 30 day period, meaning that
under one in 20 fans in a month chose to return to the brand page more than once (WARC,
2012b). On the other hand, the proportion of Facebook fans who not only visit the fan page
but also engage with it was found to be even lower. Only 1% of customers observed in
another study were found to actually engage with the brand after initially becoming a fan
on Facebook (Creamer, 2012). Hence, given the entire struggle that businesses are going
through trying to engage their customers, the inevitable question arises is it worth it?
Some of the biggest brand owners such as Coca-Cola, Unilever and Ford who already
managed to establish a large fan base are still attempting to define the potential return on
investment from using Facebook and expect that it will take at least a couple more years
until the value of fans is established (WARC, 2012c). Thus, the brands are willing to take a
leap of faith building on the core premise of social media paradigm, which suggests that
brands need to engage their customers in order to sustain growth. Yet, the link between the
effects of engagement and business performance remains tenuous and fails to explain the
return in real terms (Nelson-Field & Taylor, 2012). Not surprisingly, the concept of
engagement on social media platforms has also received criticism and is sometimes even
referred to as an air of the early-dot-com hype (Baker, 2009), given that its effectiveness
and consequences to the brand are still largely uncertain.
The buzz of social media along with the dilemma of the newly emerged concept of
customer engagement among the practitioners has also started attracting the interest of
marketing scholars. The Marketing Science Institute has listed customer engagement as one
of the research priorities for the period of 2010-2012 recognizing the lack of conceptual
2
frameworks and methods for understanding this concept (MSI, 2010). Hence, making use
of the new media opportunities requires a deeper knowledge of how customers engage with
the different types of media and what it ultimately means for the brand. This study attempts
to develop a conceptual model of customer brand engagement on online social media
platforms by reviewing the existing marketing literature concerning the concept and
subsequently refining it through empirical analysis in order to help marketers better
understand how the process of customer engagement works in this increasingly complex
landscape of social media.
2. Literature review
While the notion of engagement is not new in the literature of various academic disciplines,
it has only emerged in the field of academic marketing relatively recently. Before 2005
there were very few academic articles in the field of marketing which have mentioned the
term engagement (Brodie et al., 2011a). Since then the term has gained popularity.
However, despite the significant practitioner interest evolved during the last decade, there
have only been a few systematic scholarly attempts to define the concept, its distinctiveness
from the more traditional relational concepts like participation or involvement, and, finally,
the conceptual roots of customer engagement.
meaning that creating superior value in cooperation with the customer becomes a source of
competitive advantage for the firms. To date, a set of 10 foundational S-D logic premises
have been established building on marketing relationships characterized by customers
interactive service experience (Vargo & Lusch, 2008):
1.
2.
3.
4.
5.
(Kumar et al., 2010). The paper of Kumar et al. (2010) introduces a new metric for
customer valuation, where they include both the value from transactional and the nontransactional behaviors and, therefore, disagree with the view of van Doorn et al. (2010).
Hollebeek (2011b) presents the concept of customer brand engagement and defines it as
the level of an individual customers motivational, brand-related and context-dependent
state of mind characterized by specific levels of cognitive, emotional and behavioral
activity in direct brand interactions, where the focus lies on the interactions between a
specific subject (the customer) and the focal object (brand). The cognitive activity refers to
the level of engrossment or concentration towards a brand, whereas the emotional and
behavioral activities reflect the level of an individuals pride or inspiration and the level of
energy expressed while interacting with the brand, respectively (L. D. Hollebeek, 2011b).
Just like Bowden (2009), Hollebeek (2011b) also suggests that customer brand engagement
contributes to developing customer loyalty by focusing on conceptualizing the positively
valenced expressions of customer brand engagement. In her other works Hollebeek (2011a)
further explores the concept of customer brand engagement and, by utilizing qualitative
research methods, identifies the key themes of customer engagement behavior: immersion,
passion and activation. This implies that the level of customers brand-related concentration
(immersion), positive affect (passion) and the level of energy put in particular brand
interactions (activation) together represent just how much the customer is prepared to exert
cognitive, emotional and behavioral investments while interacting with the focal brand (L.
Hollebeek, 2011a).
Mollen and Wilson (2010) elaborate on the concept of engagement from the perspective
of online consumer experience. Building on the findings from e-learning and online
marketing literature, the authors suggest that a consumers experiential response to a
website or some other computer-mediated entity comprises three experiential states
including perceived interactivity, telepresence and engagement. In particular, engagement
is defined as a cognitive and affective commitment to an active relationship with the brand
as personified by the website or other computer-mediated entities designed to communicate
brand value and is suggested to comprise the dimensions of active, sustained, cognitive
processing, attainment of instrumental value (relevance and utility), and experiential value
(emotional congruence) (Mollen & Wilson, 2010).
Another conceptualization addressed in the literature is the brand engagement in selfconcept (Sprott, Czellar, & Spangenberg, 2009). The construct suggests that consumers
vary in their tendency to possess brand related schemas, meaning that differences exist in
consumers tendency to engage brands in their self-concepts and, therefore, also in their
brand-related behaviors. Sprott et al. (2009) develop a scale to measure the self-brand
connections in individuals. However, the concept has been criticized for failing to fully
capture the interactive nature of customer engagement (Brodie et al., 2011a).
Engagement has also been conceptualized as a state of sustained attention, which can be
characterized by full absorption and involvement as well as being fully occupied or
engrossed in something (Higgins & Scholer, 2009). Higgins & Scholer (2009) also
recognize that individuals can be engaged on different levels of intensity and suggest that
the more a person is engaged, the more intense will be the experience of the motivational
force. This means that a more engaged individual will experience the positive target more
positively and the negative target more negatively in the pursuit of his goal. Thus, the
authors express considerations towards both positive (e.g. attraction) and negative (e.g.
repulsion) expressions of engagement.
Brodie et al. (2011a) have derived the main themes prominent in the literature
concerning customer engagement and developed a set of five fundamental propositions,
which consequently provide the basis for the suggested general definition:
Customer engagement (CE) is (1) a psychological state that occurs by virtue of
interactive, co-creative customer experiences with a focal agent/object (e.g. brand) in
focal service relationships. It occurs (2) under a specific set of context-dependent
conditions generating differing CE levels; and (3) exists as a dynamic, iterative process
within service relationships that co-create value. CE plays (4) a central role in a
nomological network governing service relationships in which other relational concepts
(e.g. involvement, loyalty) are antecedents and/or consequences in iterative CE
processes. It is (5) a multidimensional concept subject to a context- and/or stakeholderspecific expression of relevant cognitive, emotional and/or behavioral dimensions.
Unlike most other reviewed conceptualizations, Brodie et al. (2011a) suggested a
definition that can be applicable in a wide range of contexts. Furthermore, a generic
expression of the dimensions (cognitive, emotional and behavioral) comprising the
engagement concept allows for it to encompass any context-specific expressions of the
customer engagement. However, this particular conceptualization has also received
criticism for being too broad and exposing to the danger of confounding the behaviors,
which are potentially caused by engagement, and all other behavioral indications
(Malthouse & Calder, 2011). A comment on Brodies et al. (2011a) conceptualization also
suggests that the interactive and co-creative nature of experiences should not imply that
engagement requires a high level of overt activity. Malthouse & Calder (2011) point out
that engagement can arise not only from active behaviors such as e.g. blogging, but simply
receiving communication can also be viewed as interactive and co-creative, as long as these
experiences are immersive. Finally, Brodies et al. (2011a) definition also addresses the
issue of differentiating customer engagement from other relational concepts and suggests
that they represent the potential antecedents and/or consequences embedded in the iterative
process of service relationships.
the customer towards a product based on the perceived importance and/or general interest
in the purchase process (Bowden, 2009). Mollen & Wilson (2010) identify three major
differences between engagement and involvement. First of all, the definition of
involvement indicates that it requires a consumption object (e.g. product category). Second,
involvement refers to a more passive allocation of mental resources and unlike engagement
does not encompass an active relationship with the consumption object. Finally,
engagement not only requires the attainment of instrumental value through relevance and
utility, but also a certain degree of emotional bonding, which can be achieved through
pleasant and satisfying experiences.
Besides involvement, Bowden (2009) also compares the customer engagement process
with and delineates the distinction from the concepts of commitment and loyalty.
Commitment often encompasses some sort of psychological attachment, where a customer
views a specific commitment object as the only acceptable choice alternative. Thus,
commitment generally means that unlike in the case of involvement, a customer is not
simply interested in an issue, but rather holds an actual attitudinal position. Loyalty is also
known to comprise an attitudinal element. However, it is most often evaluated in the
behavioral manner, e.g. the intention to repeat a purchase. Commitment and loyalty are
often considered as highly related concepts. Nevertheless, the effects of the two may yield
different behavioral outcomes. It has been discovered that due to attitudinal attachment
brand-committed customers are actually less likely to switch brands than the brand-loyal
customers (Bowden, 2009).
Mollen & Wilson (2010) also discuss the constructs of interactivity, flow and
telepresence in relation to the online brand engagement. However, these are depicted as a
process, where interactivity is assumed to be an antecedent of telepresence, which
consequently is an antecedent of engagement. There is no consensus about the definition of
interactivity in the literature, so the authors propose their own definition, which suggests
that interactivity is an experiential phenomenon, which describes to what degree
customers perceive the communication as two-way, controllable and responsive to their
actions. The construct of flow is viewed as a cognitive state, which asserts when
individuals are so involved in an activity, that it makes them forget everything else.
11
12
rapport and value co-creation in particular have been noted as of high relevance in service
contexts and Web 2.01 settings, which can be characterized by human interactive forms.
Van Doorn et al. (2010) introducing the concept of customer engagement behaviors
offer a somewhat different perspective of the potential antecedents and consequences, and
present a useful theoretical framework for research in this area (see Figure 1). As already
mentioned in the previous section, the authors suggest a conceptual model, which examines
different types of motivational drivers and outcomes of customer engagement behaviors.
The antecedents are divided into three major groups and include not only customer-based,
but also firm-based and context-based factors. The model implies that not only customer
related factors, such as attitudes, goals, resources and perceptions, but also the
characteristics of the brand and the firm together with the different aspects of contextual
environment can have just as much impact on customer engagement behaviors. Though,
some of these factors may not necessarily elicit a direct effect. The model also indicates
that the factors can interact with each other and moderate the effect of other particular
factors on customer engagement behaviors. The consequences considered in the model
include the effects on the customer, the firm and other constituents (e.g. the customers of
other products and brands).
To sum up, the literature reviewed explores different sub-forms of engagement and
offers a variety of conceptualizations. Yet, even though the topic has received considerable
attention among the practitioners (Cheung, Lee, & Jin, 2011), the lack of consensus in the
academic literature suggests that the concept of customer engagement is still understood in
a rather unsystematic way.
Web 2.0 is a collection of open-source, interactive and user controlled online applications expanding the
experiences, knowledge and market power of the users as participants in business and social processes.
(Constantinides & Fountain, 2008)
13
Customer-Based
Customer
Satisfaction
Trust/commitment
Identity
Consumption goals
Resources
Perceived
costs/benefits
Firm-Based
Brand
characteristics
Firm reputation
Firm
size/diversification
Firm information
usage and
processes
Industry
CUSTOMER
ENGAGEMENT
BEHAVIOR
Valence
Form/modality
Scope
Nature of impact
Customer goals
Cognitive
Attitudinal
Emotional
Physical/Time
Identity
Firm
Financial
Reputational
Regulatory
Competitive
Employee
Product
Others
Consumer
welfare
Economic surplus
Social surplus
Regulation
Cross-brand
Cross-customer
Context-Based
Competitive
factors
P.E.S.T.
o Political
o Economic/
environmental
o Social
o Technological
richness and the reach of their communication. That is, a rich dialogue with a customer
requires personal interaction and physical proximity, which means that there is only a
limited number of customers that the firm can communicate with in the most effective
manner. Internet, however, allows the firms to overcome these constraints and reach a
much larger number of customers without having to lose on the richness of the
communication too much.
The emergence and rise of new social media channels in the recent years enabled the
customers to increasingly participate in the new forms of customer/firm interaction
processes. Discussion forums, chat rooms, email, bulletin boards, blogs and social networks
are just some of the tools facilitating interactive customer experiences, that may eventually
also foster the development of customer engagement with the specific brands (Brodie et al.,
2011b). Hollebeek (2011b) also recognizes the importance of customer engagement in the
so called Web 2.0 applications, which are designed in a way that enables them to aggregate
the information from their user base in order to expand their content as well as value
(Wilkins, 2007). Some practitioners even refer to customer engagement as the Holy Grail in
the context of online marketing (Mollen & Wilson, 2010). One of the main reasons behind
the suggested importance of the concept lies in the definition of Web 2.0 and the fact that
this kind of setting would not persist without the user-generated content, which in turn
requires users to be engaged in the new media. Not surprisingly, this specific sub-form of
engagement has also gained attention among the researchers. For instance, Cheung et al.
(2011) have initiated a study exploring customer engagement in online social platforms.
The authors of the research-in-progress paper have defined it as the level of a customers
physical, cognitive, and emotional presence in connections with a particular online social
platform. The conceptual model developed suggests that customer engagement in an
online social platform is a construct comprising vigor (level of energy and mental
resilience), absorption (level of concentration and engrossment) and dedication (sense of
significance, enthusiasm, inspiration, pride and challenge) towards the online social
platform, which are driven by involvement and social interaction. The consequences
reflected in the model exhibit the authors belief that customer engagement will have a
positive effect on online social platform participation and word-of-mouth communication
about the platform (Cheung et al., 2011). The study by Cheung et al. (2011) is expected to
15
contribute highly to the existing knowledge about social media engagement by providing a
validated measurement scale for customer engagement in online social platforms. However,
the research is still in progress and no results have been published to date. Thus, even
though the new media present a number of significant opportunities and challenges for both
researchers and practitioners (Hennig-Thurau et al., 2010), most of the existing research is
primarily conceptual or qualitative (Cheung et al., 2011).
16
Hence, the main objective of this study is to bridge this gap by conceptualizing
customer brand engagement on online social media platforms and answering two important
research questions:
1.
What drives the customer to engage with brands on online social media
platforms?
2.
18
INVOLVEMENT
SATISFACTION
COMMITMENT
TRUST
CUSTOMER BRAND
ENGAGEMENT ON
ONLINE SOCIAL
MEDIA PLATFORMS
CONSEQUENCES
BEHAVIORAL
BRAND LOYALTY
EMOTIONAL
WORD-OF-MOUTH
COGNITIVE
INVOLVEMENT
PARTICIPATION
TELEPRESENCE
EASE OF USE
3. Methodology
3.1 Data collection
In order to collect the data and test the proposed model of customer brand engagement
on online social media platforms an online survey was conducted using a convenience
sample of Facebook2 account holders. With 901 million active monthly users Facebook is
currently worlds largest online social network (Facebook, 2012) and a highly relevant
platform for this study. Among many various online services offered by Facebook, there is
also something called Facebook Pages. Facebook Pages are public profiles meant to
promote brands, products, artists, web sites or organizations. Once registered Facebook
users visit a Page, they are able to 'become fans' by clicking on the 'Like' button. The
owners of the Page can then post informational content, which consequently will appear in
the news feed of their fans. The fans can choose to react to the posts in few different ways
such as liking, commenting or sharing it with their own networks. In other words,
2
www.facebook.com
20
Facebook is a medium that can give any brand a voice and allows it to establish an active
conversation with Facebook users. It has therefore been largely employed by various
brands and used as a tool for customer engagement. According to the statistics, some of the
largest Facebook brands in terms of number of fans belong to food and drinks product
category (FanPageList, 2012), which has also been chosen to be the focus of this study.
Even though one recent paper about Facebook has showed that the degree of fan
engagement with brands from any given category is highly similar (Nelson-Field & Taylor,
2012), it is still important to narrow it down to a single category as to assure that the
antecedents and the consequences of engaging with the brands are more or less
homogeneous.
The data collection procedure comprised two stages - a pilot study to pretest the survey
instrument and a full-scale field study. During the pretest a self-administrated online
questionnaire was created on an online survey tool Qualtrics3 and distributed to a number of
selected web forums. A total of 57 responses were collected. The results of the pilot test
have been used for reviewing and refining the questions. The full-scale questionnaire has
also been launched online and distributed using various web tools such as email, social
networking platforms (e.g. Facebook) as well as various international forums. The
questionnaire comprised a few basic parts. It started out with an introduction to the survey
and a screening question, to make sure that only those, who have a Facebook account,
participate in the survey. Further questions were related to the usage of and perceptions
about Facebook, such as involvement, participation, ease of use and telepresence as well as
three control variables (customer goals, resources and perceived cost/benefit). In order to
get to the next part of the questionnaire the participants had to state whether they are fans of
any of food or drink brands on Facebook, which then allowed to divide the total sample (N)
into two major groups 1) respondents who are fans of at least one food or drink brands on
Facebook (N1); 2) respondents who are not fans of any food or drink brand on Facebook
(N2). The respondents in the first group were then asked about their engagement with a
certain brand of their choice on Facebook as well as the ongoing relationship with that
brand and future intentions related to loyalty and recommending the brand to others. A list
www.qualtrics.com
21
of 15 most popular food and drink product brands at that very moment was provided to the
respondents to choose from. The brand popularity ranks were retrieved from a social media
counter application previously called Famecount4 (see Appendix 2).The respondents also
had an option to enter a brand name of their own liking, in case it was not provided on the
list. Since the respondents in the second group were not fans of any food or drink brands,
they were simply asked to pick a brand from the same list that they liked most (also with
the option of entering a brand name of their own). They were then directed straight to the
questions relating to the customer brand relationship and its outcomes. The final part of the
questionnaire included socio-demographic questions, such as age, gender, country of origin
as well as the usage of other online social media platforms.
The final survey sample (N) contained a total of 419 internet users from all over the
world, who also had an account on Facebook. Almost 27% of those Facebook users have
identified themselves as fans of at least a single Facebook page dedicated to a brand in the
food and drinks product category, meaning that N1=112 and N2=307. The total sample
included respondents from various age groups ranging from teenagers to seniors, with the
largest group consisting of 20-29 year olds (70% of all respondents). A chi-square test was
performed on age and other demographic variables to investigate whether there are
statistically significant differences between the two groups of N1 and N2. Table 1 below
reports demographic characteristics of the two sub-samples along with the results of the
chi-square test.
The findings of the test suggest that there were no statistically significant differences
between the two groups of fans and non-fans with respect to age and gender of the
respondents. However, the use of other online social media platforms and the time spent on
it per day were found to be related to the group of the respondent. In particular, those
respondents who indicated themselves as fans of at least one food and drink brand on
Facebook showed a tendency of using a higher number of various online social media
platforms and were to spend more time on Facebook and other platforms per day.
From the 1st of May 2012 Famecount has changed its name to Starcount. For more information visit
http://www.starcount.com/pages/starcount.
22
Age
Younger than 20
20 24
25 29
30 34
35 39
40 44
45 49
50 and older
Gender
Male
Female
Use of other online social media platforms
No other
1-2 others
3-5 others
6-9 others
10 and more others
Time spent on online social media platforms per day
Less than 30 mins
30 mins 1 hour
1 hour 2 hours
2 hours 3 hours
More than 3 hours
Time spent on Facebook per day
Less than 30 mins
30 mins 1 hour
1 hour 2 hours
2 hours 3 hours
More than 3 hours
Fans
Non-fans
(N1) %
(N2) %
2
X (7) = 4.43, = 0.729
7
4
37
33
33
38
11
15
4
4
4
2
2
1
3
3
2
X (1) = 0.70, = 0.401
58
53
42
47
2
X (4) = 10.51, = 0.033
8
9
32
43
35
35
23
12
2
1
2
X (4) = 15.92, = 0.003
12
26
22
25
33
26
16
14
17
8
2
X (4) = 13.97, = 0.007
19
36
29
27
27
23
14
9
11
6
supplemented with seven other self-constructed items referring to the frequency of the
different forms of behavioral engagement. Nelson-Field & Taylor (2012) suggest that in
social media, and particularly on Facebook, engagement takes the form of all kinds of
direct interaction with the fan page. The inclusion of seven additional Facebook specific
items was also based on this premise. Thus, the self-constructed items refer to the
frequency of various interactions with a particular fan page, such as visiting the page,
noticing, reading, liking, commenting and sharing its contents as well as creating and
posting contents on the fan page yourself. The response format chosen for these seven
items has been a seven point frequency scale (1=Never, 2=Almost never, 3=Rarely,
4=Sometimes, 5=Often, 6=Almost all the time, 7=All the time). The response
format used for the rest of the items in the questionnaire was a seven point Likert scale
anchored by 1=Strongly disagree, 7=Strongly agree.
The consequences of customer brand engagement on online social media platforms
have been measured in terms of behavioral brand loyalty and word-of-mouth. The scale for
behavioral brand loyalty contained two items relating to future purchase intentions
(Chaudhuri & Holbrook, 2001). Word-of-mouth, which can also be defined as the intention
to recommend the brand to others, has been measured with three items suggested by
Zeithaml, Berry & Parasuraman (1996).
In addition to the 62 mentioned items, there were also three control variables included
in the questionnaire and measured by two self-constructed items each. These were goals,
resources, and the perceived cost/benefit of interacting with the brand pages on Facebook
specifically. These control variables have been included in the survey as the literature
suggests that they can also be expected to influence how customers engage with brands
(van Doorn et al., 2010). The two specific goals accounted for in the questionnaire were: 1)
maximizing the consumption benefits (e.g. interacting with the brand on Facebook out of
interest); 2) maximizing the relational benefits (e.g. becoming a member of a brand
community). The resource items referred to the time available for browsing on Facebook
fan pages and the effort that it takes. Finally, the perceived cost/benefit items were focusing
on the respondents perceived levels of enjoyment while browsing on Facebook fan pages
and its value in comparison to the time and effort spent on it. A summary of all the
25
mentioned questionnaire items including the sources of reference and the resulting
Cronbachs alpha for each scales are displayed in Table 2 below.
Table 2: Construct measurement items, sources and scale reliabilities
Measure/Source
Items
Antecedents
Customer brand relationship quality related
Involvement 1.In general I have a strong interest in [BN]5
(Beatty & Talpade, 2.[BN] is very important to me
1994) 3.[BN] matters a lot to me
4.I get bored when other people talk to me about [BN]*6
5.[BN] is relevant to me
Satisfaction 6.Overall I am satisfied with [BN]
(Gustafsson et al., 7.[BN] exceeds my expectations
2005) 8.The performance of [BN] is very close to the ideal brand
in the product category
Commitment 9.I am very loyal to [BN]
(Aaker, Fournier, 10.I am willing to make small sacrifices in order to keep
& Brasel, 2008) using the products of [BN]
11.I would be willing to postpone my purchase if the
products of [BN] were temporarily unavailable
12.I would stick with [BN] even if it would let me down
once or twice
13.I am so happy with [BN] that I no longer feel the need
to watch out for other alternatives
14.I am likely to be using [BN] one year from now
Trust 15.I trust [BN]
(Chaudhuri & 16.I rely on [BN]
Holbrook, 17.[BN] is an honest brand
2001) 18.[BN] is safe to use
Online social media platform related
Involvement 19.In general, I have a strong interest in Facebook
(Beatty & Talpade, 20.Facebook is very important to me
1994) 21.Facebook matters a lot to me
22.I get bored when other people talk to me about
Facebook*
23.Facebook is relevant to me
Participation 24.I consider myself an active user of Facebook
(Self-constructed) 25.I log on to Facebook everyday
26.I spend long periods of time on Facebook
Reliability
0.80
0.75
0.84
0.81
0.83
0.82
The abbreviation BN stands for brand name, as different respondents have answered the questions with
a different brand name in mind.
6
The items marked with * were reverse scored.
26
27
0.90
0.89
0.89
0.89
0.90
Consequences
Behavioral brand
loyalty
(Chaudhuri &
Holbrook, 2001)
Word-of-mouth
(Zeithaml, Berry,
& Parasuraman,
1996)
Control variables
Goals
(Self-constructed)
Resources
(Self-constructed)
Perceived
cost/benefit
(Self-constructed)
0.61
0.89
0.59
0.53
0.56
The coefficient reliability analysis revealed that all the scales consisting of more than
two items exceeded the recommended Cronbachs alpha benchmark of 0.70 (Nunnally,
1978). However, the construct of behavioral brand loyalty measured by two items only has
performed an internal consistency of 0.61, which is considered to be questionable (George
& Mallery, 1999). In addition, the same happened to be the case with the three control
variables that were also operationalized by two items each and did not meet the 0.70
benchmark. However, the nature of the Cronbachs alpha dictates that its value is
determined not only by the mean of inter-item correlations, but also depends on the number
of the items in the scale, which implies that the scales with fewer items will generally be
expected to yield lower reliability coefficients. Therefore, the four underperforming two
item scales were not eliminated and used further in the analysis.
29
There are two types of statistical techniques for estimating structural equation models
covariance-based (e.g. LISREL) and variance-based (e.g. PLS) (Henseler et al., 2009). The
method used in this study is PLS (partial least squares) path modeling, which can be viewed
as a combination of principal component and multiple regression analysis. The main
reasons behind choosing PLS relate to the highly favorable features of this technique
(Henseler et al., 2009). PLS allows analyzing highly complex models without making the
estimation problematic even when both formative and reflective measurement models are
employed. Moreover, it can be used with a relatively small sample size and there are no
distributional requirements. In this study only the data collected from respondents who
belonged to the group of fans of at least one food or drink brand on Facebook could be used
for testing the full model, which implies that the sample size equaled N1=112. Given that
the model contained a total of 15 latent constructs (13 first-order constructs and 2 secondorder constructs), a sample of N1=112 was considered to be relatively small. In addition,
some of the observations turned out to be skewed. Therefore, PLS was the more appropriate
technique to apply in this study. Because of its flexible nature PLS path modeling is also
generally suggested to be more adequate for causal modeling applications with no prior
theoretical background. Thus, it goes well with the purposes of this study developing and
testing a conceptual model of customer brand engagement on online social media
platforms. All data analysis were performed using a predictive analysis software SPSS and
a path modeling software application SmartPLS (Ringle, Wende, & Will, 2005).
4. Results
4.1 Descriptive analysis
Means and standard deviations were calculated for each of the constructs in order to
compare the differences between the two groups of fans and non-fans (Table 3). The
independent samples t-test revealed that the two groups showed significant differences in
several aspects. With regards to customer brand relationship related antecedents, the group
of fans was found to be more involved and expressed more trust in the brands than the nonfans. On the other hand, no significant differences were discovered in the levels of
30
satisfaction or commitment to the brands. Furthermore, the fans also showed a higher
tendency of involvement, participation and telepresence in Facebook than the non-fans.
Table 3: Means, standard deviations and results of t-test for equality of means
(N1=112, N2=307)
Fans (N1)
Non-fans (N2)
Construct
Mean
SD
Mean
SD
Dimensions of customer brand relationship related antecedents
1. Involvement
3.85
1.38
3.09
1.48
2. Satisfaction
4.76
1.41
4.75
1.24
3. Commitment
4.09
1.42
3.94
1.36
4. Trust
4.66
1.31
4.20
1.35
Dimensions of online social media platform related antecedents
5. Involvement
4.05
1.38
3.71
1.48
6. Participation
5.31
1.41
4.98
1.62
7. Ease of use
5.49
1.04
5.33
1.07
8. Telepresence
3.06
1.50
2.67
1.42
9. Emotional engagement
2.90
1.30
n.a.
n.a.
10. Behavioral engagement
2.54
1.04
n.a.
n.a.
11. Cognitive engagement
2.33
1.19
n.a.
n.a.
12. Behavioral brand loyalty
4.50
1.31
4.22
1.47
13. Word-of-mouth
4.63
1.55
3.92
1.67
t-value
4.77***
0.09
1.02
3.11**
2.11*
2.06*
1.37
2.48*
n.a.
n.a.
n.a.
1.75
3.93***
Note: n.a. = not applicable; SD = standard deviation; t-values were obtained by performing the
independent samples t-test; *significant at <0.05 level, **significant at <0.01 level, ***significant at
<0.001 level.
Behavioral engagement
Emotional engagement
Cognitive engagement
31
As the non-fans did not have to answer the questions about emotional, behavioral or
cognitive engagement, the data is only available for the group of fans. Inspecting the means
of fan engagement showed that it is mainly concentrated in the lower part of the scale and
ranges from 2.33 to 2.90 on average, with emotional engagement scoring the highest.
Hence, the engagement level with the brand pages on Facebook can be considered
relatively low. Figure 3 illustrates fan distribution in percentage based on engagement level
(on a scale from 1 to 7) for all three dimensions.
Finally, the group of fans showed a significantly higher intention to recommend their
favorite brand than the non-fans. On the other hand, the observed levels of behavioral brand
loyalty were found to be similar for both groups, which would suggest that the fans are no
more likely to be loyal to their brands than the non-fans. However, it must also be taken
into consideration that low levels of engagement will also influence the levels of customer
outcomes to be lower. Hence, the sample of 112 fans was split at the median (2.47) into two
equal sub-groups of low engaged and highly engaged fans, and another t-test was
performed in order to determine whether the two types of fans differ in their behavioral
loyalty to the brand. The test results portrayed in Table 4 reveal that the highly engaged
fans show a significantly higher level of behavioral brand loyalty than those of low
engagement. Thus, it can be concluded that a certain level of engagement has to be
achieved before the level of behavioral brand loyalty increases notably.
Table 4: Means, standard deviations and results of t-test for equality of means in
behavioral brand loyalty of high and low engaged fans (N1a=56, N1b=56)
Construct
Behavioral brand
loyalty
Low engaged
(N1a)
Mean
SD
4.23
Highly engaged
(N1b)
Mean
SD
1.43
4.78
1.13
t-value
2.232*
Note: SD = standard deviation; t-values were obtained by performing the independent samples ttest; *significant at <0.05 level.
32
33
AVE
0.71
0.74
0.63
0.66
0.77
0.74
0.68
0.74
0.66
0.57
0.69
0.71
0.81
Item loadings were inspected next. Literature suggests that item loadings on their
respective latent variables should be at least 0.6 and ideally above 0.7 (Chin, 1998a), which
implies that the construct should share more variance with the item than the error term. The
analysis revealed that most of item loadings exceeded the more stringent threshold of 0.7.
One of the items measuring involvement in both online social media and the brand (INV4
I get bored when other people talk to me about Facebook/[brand name], reverse scored) had
a construct loading of 0.30 and -0.30 respectively. As the loading values were way below
the accepted threshold and expressed low item reliability, INV4 has been eliminated from
each of the involvement constructs. Other two constructs (commitment and behavioral
engagement) each had an item loading just below 0.6 (COMM6 I am likely to be using
[brand name] one year from now; BEH2 I devote a lot of energy to the Facebook fan page
of [brand name]). Even though the loading values of these two items (COMM6 0.59 and
BEH2 0.58) were rather close to passing the threshold, they were still discarded as it
consequently helped increase the reliability and validity of the two respective constructs.
Table 5 reports the reliability and validity measures after removing the four items.
In order to assess the validity of the constructs two measures were used. Average
variance extracted (AVE) is usually employed as the criterion for convergent validity,
which signifies that a block of indicators is unidimentional and represents the exact same
construct. The requirements for the convergent validity of the constructs were met as all
AVE values exceeded the suggested cut-off threshold of 0.5 (Henseler et al., 2009). The
discriminant validity was inspected by using the Fornell-Larcker criterion (Fornell &
Larcker, 1981), which requires the AVE of each latent construct to be higher than its
highest squared correlation with any other latent construct. This means, that a latent
construct should share more variance with its own measurement indicators, than any other
latent construct. Table 6 below illustrates that the squares of absolute correlation
coefficients between constructs are mostly higher than the respective AVEs. However, the
construct of emotional engagement seems to share slightly more variance with the construct
of cognitive engagement than its own set of indicators. It is therefore inherent that the same
tendency also appears when assessing the discriminant validity on the indicator level, i.e.
inspecting the cross-loadings. However, as the difference between the AVE of emotional
34
engagement and its squared correlation with cognitive engagement is only 0.02 (see Table
6), the discriminant validity of the construct was still deemed acceptable.
Validation of the second-order constructs should follow the exact same assessment
process (Chin, 1998a). The first second-order construct CBRR (customer brand relationship
related antecedents) is modeled in the reflective mode. Therefore, both the reliability and
the validity of the construct have to be evaluated. The construct of CBRR was deemed
satisfactory by the previously discussed conventional standards. Table 7 below reports the
parameters of the composite reliability and AVE as well as the loadings of the first-order
latent constructs on the CBRR construct. The thresholds for reliability and validity are met
as the composite reliability is equal to 0.95, the AVE exceeds 0.5 and the component
loadings range from 0.85 to 0.93 (all significant).
Table 6: Average variance extracted and squared correlations between first-order
latent constructs (N1=112)
1
2
3
4
5
6
7
Dimensions of customer brand relationship related antecedents
1. Involvement
0,71
2. Satisfaction
0,39 0,74
3. Commitment 0,58 0,54 0,63
0,39 0,70 0,46 0,66
4. Trust
Dimensions of online social media platform related antecedents
0,00 0,01 0,01 0,01 0,77
5. Involvement
6. Participation
0,00 0,00 0,00 0,00 0,33 0,74
7. Ease of use
0,00 0,03 0,00 0,01 0,17 0,32 0,67
8. Telepresence 0,02 0,00 0,01 0,00 0,14 0,03 0,02
9. Emotional
engagement
10. Behavioral
engagement
11. Cognitive
engagement
12. Behavioral
brand loyalty
13. Word-ofmouth
10
11
12
0,74
0,20
0,11
0,12
0,15
0,05
0,01
0,00
0,06
0,66
0,17
0,08
0,13
0,07
0,06
0,02
0,01
0,07
0,58
0,57
0,17
0,04
0,12
0,06
0,05
0,00
0,00
0,09
0,68
0,53
0,69
0,42
0,26
0,44
0,28
0,00
0,00
0,01
0,01
0,08
0,13
0,06
0,71
0,45
0,49
0,48
0,37
0,01
0,02
0,02
0,00
0,06
0,04
0,02
0,28
35
13
0,81
Indicators
Commitment
Involvement
Satisfaction
Trust
Indicator
loadings
0.90*
0.86*
0.87*
0.88*
Composite
reliability
0.95
AVE
0.52
When assessing the formative measurement models, the same concepts of validation no
longer apply as the assumption of error-free measures eliminates the issue of reliability all
together (Henseler et al., 2009). The criteria used for the formative indicators are therefore
focused on validity (Diamantopoulos, Riefler, & Roth, 2008).
At the indicator level, each of the four OSMPR (online social media platform related
antecedents) dimensions was checked for the weight and significance of the delivered
contributions to the formative index. Table 8 presents the indicator weights and their
significance estimated by means of bootstrapping. All indicators were found to have a
significant impact on the OSMPR construct.
Table 8: Estimated weights and variance inflation factors for formative dimensions of
second-order latent construct of online social media platform related antecedents
(N1=112)
Construct
Online social media platform
related antecedents
Indicators
Weight
t-value
VIF range
Ease of use
Involvement
Participation
Telepresence
0.43**
0.41**
0.25**
0.26*
4.76
8.13
7.95
2.56
1.11 1.57
1.02 1.45
1.11 1.28
1.45 1.98
Note: *Significant at <0.05 level, **significant at <0.001 level. The VIF values were calculated by
regressing each of the indicators on the other three.
The next step in validating the formative indicators is to assess the degree of
multicollinearity by calculating the variance inflation factor (VIF) (Henseler et al., 2009). A
rule of thumb is that any VIF greater than one shows a presence of multicollinearity.
However, only a VIF value above ten indicates a critical level of multicollinearity, which is
already harmful. In the case of OSMPR indicators all VIF values ranged from 1.02 to 1.98
36
(Table 8), meaning that the information of the indicators was not redundant and, therefore,
each of them contributes to the formative index. Finally, as the outer model was assessed to
be valid and reliable, the estimation of the inner path model was performed next.
Cognitive
engagement
Emotional
engagement
Behavioral
brand loyalty
Word-ofmouth
Predictors
Customer brand relationship
related antecedents
Online social media
platform related antecedents
Customer brand relationship
related antecedents
Online social media
platform related antecedents
Customer brand relationship
related antecedents
Online social media
platform related antecedents
Behavioral engagement
Cognitive engagement
Emotional engagement
Behavioral engagement
Cognitive engagement
Emotional engagement
Path
0.38
t-value
4.81***
f2
0.17
0.21
1.83
0.05
0.35
4.11***
0.14
0.11
0.77
0.01
0.43
5.76***
0.23
0.15
1.40
0.04
0.40
-0.08
0.04
0.12
-0.26
0.38
3.20**
0.38
0.20
0.59
1.41
2.24*
0.07
0.00
0.00
0.01
0.03
0.04
R2
0.20
0.14
0.22
0.14
0.09
Note: *Significant at <0.05 level, **significant at <0.01 level, ***significant at <0.001 level; the effect size f2
is calculated as the relationship of the determination coefficients when including or excluding each of the
predictors from the structural model, i.e. f2= (R2included-R2excluded)/(1-R2included).
The relationships between the latent exogenous and endogenous variables were
assessed first. The t-values and the significance of the structural coefficients were
computed for each path by means of a bootstrapping procedure using 500 subsamples as
recommended by Chin (1998b). Inspection of the paths revealed that not all the
relationships in the inner model turned out statistically significant (see Table 9). Online
37
social media platform related antecedents have shown no significant direct effect on either
of the three engagement dimensions. However, customer brand relationship related
antecedents were found to have a strong effect on each of the three dimensions behavioral
engagement (0.38), cognitive engagement (0.35) and emotional engagement (0.43). Yet,
only two of the paths connecting the behavioral, cognitive and emotional engagement with
their expected outcomes turned out to be significant. That is, a strong and positive
relationship was found between the behavioral engagement and the behavioral brand
loyalty (0.40), and between the emotional engagement and word-of-mouth (0.38).
The size of the predictor effect (f2) was also assessed for each of the paths. The effect
size determines the relevance of each predictor in a latent endogenous variable. The f2
values of 0.02, 0.15 and 0.35 can be classified as weak, medium and large, respectively
(Cohen, 1988). The values provided in Table 9 above show that all of the insignificant
predictors were found to have a weak effect on their latent endogenous variables. Customer
brand relationship related antecedents turned out to have a medium influence on the
behavioral and cognitive engagement. However, it had a more prominent effect on
emotional engagement. On the other hand, the significant predictor effects on behavioral
brand loyalty and word-of-mouth were found to be relatively weak.
Table 9 also provides the R2 values for endogenous latent variables, which determine
the explanatory power of the underlying models. The suggested classification for the R2
values of 0.67, 0.33 and 0.19 is substantial, moderate and weak, respectively (Chin, 1998b).
When referring to the endogenous latent variables in this model, the low R2 values ranging
from 0.09 to 0.22 would seem to suggest that the model is relatively weak in explaining the
constructs. However, given the early stage of research in this field, where little is known
about the variables observed, this result provides some useful insights and is, therefore,
considered acceptable.
variables relating to the usage of and attitudinal perceptions about Facebook fan pages:
goals, resources and perceived cost/benefit. According to van Doorn et al. (2010), these
three factors are potential antecedents of customer engagement behaviors. However, as no
further support was found in alternative sources of academic literature, the three factors
were not included in the main model. Yet, it is possible, that the goals of browsing on
Facebook fan pages along with the time available, the effort that it takes, and the perceived
cost and benefit could moderate the effect of the online social media platform related
antecedents. That is, the effect of telepresence, involvement, ease of use and participation
in the online social media platform on customer engagement may vary with the level of
perceived cost/benefit, existing resources or goals. Therefore, the tests for the potential
moderating effects between the online social media platform related antecedents and the
three control variables were performed on each of the engagement dimensions.
Since the construct of online social media platform related antecedents was formative,
a two-stage PLS procedure recommended by Henseler et al. (2009) for estimating
moderating effects was applied. In the first stage, the main effects PLS model including a
predictor, a moderator and a latent endogenous variable was run in order to obtain the
estimates for latent variable scores. The latent variables scores were then saved and
subsequently used in the second stage. In the second stage an interaction term was created
between the predictor and the moderator using the latent variable scores, and used in a
linear multiple regression as the independent variable together with the latent variable
scores of the predictor and the moderator alone on the endogenous latent variable scores as
the dependent variable. The existence of a moderation effect is determined by a significant
path coefficient (or regression coefficient in this case) of the interaction term regardless of
the values of path coefficients between the predictor or the moderator and the dependent
variable. After identifying the significant moderation effects, the next step in the analysis is
to assess their strength (Henseler & Fassott, 2010).
A total of nine moderation effects were tested between each of the three control
variables and the online social media platform related antecedents on behavioral, cognitive
and emotional engagement dimensions. Out of nine potential moderation effects, five
interaction terms turned out to be significant (see Table 10). The perceived cost/benefit
39
(AVE = 0.68, composite reliability = 0.80, item loadings of 0.96 and 0.66) was found to
moderate the effect of online social media platform related antecedents on cognitive
engagement. The goals (AVE = 0.69, composite reliability = 0.82, item loadings of 0.79
and 0.87) turned out to have a moderating effect regarding the behavioral engagement.
And, finally, resources (AVE = 0.67, composite reliability = 0.80, item loadings of 0.77 and
0.86) were found to moderate the effect of online social media platform related antecedents
on all three engagement dimensions.
Table 10: Results of the two-stage PLS approach for estimating moderating effects
(N1=112)
DV
Behavioral
engagement
Cognitive
engagement
Emotional
engagement
IVs
OSMPR
BEN
OSMPR x BEN
OSMPR
GOAL
OSMPR x GOAL
OSMPR
RES
OSMPR x RES
OSMPR
BEN
OSMPR x BEN
OSMPR
GOAL
OSMPR x GOAL
OSMPR
RES
OSMPR x RES
OSMPR
BEN
OSMPR x BEN
OSMPR
GOAL
OSMPR x GOAL
OSMPR
t-value
R2
0.14
0.31
0.14
0.14
0.33
0.17
0.20
0.21
0.28
0.03
0.35
0.21
0.03
0.30
0.11
0.08
0.19
0.23
0.07
0.34
0.15
0.04
0.38
-0.01
0.14
1.41
3.21**
1.60
1.47
3.49**
2.02*
2.21*
2.38*
3.53**
0.29
3.58**
2.30*
0.31
3.06**
1.33
0.86
1.98*
2.67**
0.76
3.54**
1.68
0.47
4.00***
-0.01
1.53
0.15
f2 of the interaction
term
0.02
0.19
0.05
0.19
0.12
0.16
0.06
0.11
0.01
0.10
0.07
0.15
0.02
0.16
0.00
0.17
RES
0.27
3.01**
0.08
OSMPR x RES
0.23
2.92**
Note: DV = dependent variable, IV = independent variable, OSMPR = Online social media platform related
antecedents, BEN = Perceived cost/benefit, GOAL = Goals, RES = Resources; *significant at <0.05 level,
**significant at <0.01 level, ***significant at <0.001 level.
40
These results imply that the positive effect of online social media platform related
antecedents on behavioral, cognitive and emotional engagement increases as the level of
resources available to browse on Facebook fan pages increases. Moreover, an increase in
perceived benefit of browsing on Facebook fan pages will increase the effect of online
social media platform related antecedents on cognitive engagement. Whereas an increase in
levels of interest in the brand or the desire to become a part of a brand community when
browsing on Facebook fan pages will result in increased effect of the online social media
platform related antecedents on behavioral engagement. Yet, the inspection of R2 values
and, especially, the effect size of the interaction terms on the engagement dimensions
reveals that most of the moderations are weak in explaining the latent endogenous variable
and have a rather small effect size, except for the interaction term between online social
media platform related antecedents and resources on behavioral engagement, which is
closer to being classified as a medium effect.
Hence, even though the online social media platform related antecedents did not have a
significant direct effect on customer brand engagement in this particular context, their
effect was found to be moderated by attitudinal customer perceptions towards the Facebook
fan pages in terms of goals and benefits of using them as well as availability of time and
necessary effort.
satisfaction and trust) were all together found to have significant direct effects on all three
engagement dimensions, which imply that they all are valid predictors of customer brand
engagement on online social media platforms. However, the results given in Table 9
suggest that the larger portion of variance in the engagement levels will remain unexplained
if measured by customer brand relationship related constructs only. This outcome can be
explained by the highly contextual nature of customer engagement (Vibert & Shields,
2003), which implies that the context specific factors will influence the engagement itself.
Thus, it is merely inherent that online social media platform related antecedents would play
an important role in the formation of customer brand engagement in this specific context.
Even though the suggested online social media platform related antecedents such as ease of
use, involvement, telepresence and participation were not significant in affecting customer
brand engagement directly, their effect was found to be moderated by three contextual
factors concerning the goals, resources and perceived cost/benefit of browsing on Facebook
fan pages specifically. As a result, variation in available resources such as time and effort to
engage with brands on Facebook influences the effect of online social media platform
related antecedents on all three facets of engagement, whereas the level of perceived
cost/benefit and the prevailing goals moderate the effects on cognitive and behavioral
engagement, respectively.
Finally, the above mentioned findings helped refining and validating the
multidimensional concept of customer brand engagement in the context of online social
media platforms. Both customer brand relationship related and online social media platform
related factors were found to influence all three dimensions of customer engagement.
However, only two of them - behavioral and emotional engagement - turned out to be
critical in order to achieve the desired customer outcomes such as behavioral brand loyalty
and intention to recommend.
performance (Nelson-Field & Taylor, 2012). However, the low levels of engagement
observed in this study show that businesses still lack the knowledge and skill to achieve a
substantial level of customer engagement. The conceptualization of customer brand
engagement on online social media platforms presented in this paper provides the managers
with a better understanding of the newly emerged concept and delivers empirical evidence
of the potential returns.
First of all, the findings of this research allow drawing a line and defining the main
differences between the two groups of customers - fans and non-fans. The knowledge of the
fan base on social media platforms will allow marketers setting more realistic goals and
targeting the communication better. Facebook users who engage with fan pages dedicated
to brands are more trusting and involved in the relationship with a brand. They are also
more involved, telepresent and participate on Facebook and other online social media
platforms more actively. Second, even if they are heavy users of online social media, the
final decision to engage with Facebook fan pages will depend on the perceived level of
benefit, available resources and goals. Thus, the managers need to realize that a Facebook
user who decides to become a fan of a brand is driven by certain goals and expectations.
The task of the marketers is therefore to fulfill these expectations and respond accordingly.
Based on the findings of this research, businesses should especially focus on engaging the
customer emotionally and behaviorally, which means that the communication transmitted
through online social media platforms should excel in emotional appeal and encourage
various forms of interaction with the brand. Yet, it will be more effective if the
communication can be perceived purposeful, valuable and not too complicated to respond
to. Even though this study was focused on Facebook fan pages, the group of fans was
found to use and spend time on other online social media platforms as well. The managers
should therefore consider integrating their social media effort on different platforms as it
will provide the brand with increased exposure and, therefore, even more ways to interact
with and engage the customer.
Finally, literature suggests that engaged customers can lead businesses to their ultimate
objective increased sales (Kumar et al., 2010). The rationale behind this assertion is that
engaged customers are highly important for successful viral marketing as they are more
44
likely to influence other existing and prospect customers by providing referrals and
recommendations, which in turn will help businesses to acquire new and retain existing
customers. The results of this study support this statement and present empirical evidence
that customer engagement will lead to an enhanced business performance. Even if it may
not be visible at first, increasing the levels of customer engagement will also gradually lead
to a significant increase in behavioral brand loyalty and the intention to recommend the
brand.
45
6. Conclusion
This study was an attempt to introduce and investigate the newly emerged concept of
customer brand engagement in the context of online social media platforms. With the
diminishing role of traditional media and the evolution of Internet technologies the rules of
the marketing game have changed. As a result, customer engagement was brought to the
attention of the marketers as a way to improve customer brand relationships and therefore
gain competitive advantage in the new era of social media.
The concept and its roots were introduced by reviewing the existing academic literature.
While the notion of engagement was not new in other disciplines, it has only emerged in
the field of marketing in the past few years. Building on various conceptualizations adapted
from other academic disciplines, it has been concluded that the concept of customer brand
engagement on online social media platforms is characterized by interactive customer
experiences with the brand. It is a process of dynamic and iterative nature, which stems
from S-D logic and the relationship marketing domain, which imply that creating superior
value in cooperation with the customer is becoming the source of competitive advantage
and it is therefore important for businesses to put their focus on building and maintaining
long-term interactive value-driven relationships with their customers. Customer brand
engagement on online social media platforms is the central element embedded in a broader
network of other relational constructs serving as the antecedents and the consequences. The
concept of engagement is multidimensional and comprises the expressions of emotional,
behavioral and cognitive engagement specific to this particular context.
Furthermore, the conceptual model of customer brand engagement on online social
media platforms was established by identifying the potential drivers and outcomes, and
consequently tested in a quantitative online consumer study. Two groups of antecedents
were found to influence the overall level of customer engagement: customer brand
relationship related factors such as commitment, involvement, satisfaction and trust, and
online social media platform related factors such as ease of use, involvement, participation
and telepresence. While the brand relationship related factors had a direct effect on
customer brand engagement, the effect of online social media platform related factors was
46
moderated by the perceived level of cost/benefit, available resources and goals when
interacting with the brand. The concepts of behavioral brand loyalty and word-of-mouth
were identified to be the consequences of engagement, driven by the dimensions of
behavioral engagement and emotional engagement, respectively.
In conclusion, the findings of this study have important implications for both academic
marketing literature and practice. As the scholarly inquiries into the notion customer
engagement have mostly remained conceptual to date, this research is one of the first few
attempts to test the concept in an empirical setting. On the other hand, the managers will
also find some useful implications that are relevant and can be applied in designing the
strategies for engaging the customers. Yet, further testing and refinement of the model is
necessary in order to fully leverage the potential of customer brand engagement in the
context of online social media platforms.
47
References
Aaker, J., Fournier, S., & Brasel, S. A. (2008). When good brands do bad. Working Paper
Series. Center for Responsible Business. UC Berkeley.
Baker, S. (2009). Beware Social Media Snake Oil. Retrieved from BusinessWeek website:
http://www.businessweek.com/magazine/content/09_50/b4159048693735.htm
Beatty, S., E. , & Talpade, S. (1994). Adolescent Influence in Family Decision Making: A
Replication with Extension. Journal of Consumer Research, 21(2), 332-341.
Bowden, J. L.-H. (2009). The Process of Customer Engagement: a Conceptual Framework.
[Article]. Journal of Marketing Theory & Practice, 17(1), 63-74. doi:
10.2753/mtp1069-6679170105
Brodie, R. J., Hollebeek, L. D., Juri, B., & Ili, A. (2011a). Customer Engagement.
Journal of Service Research, 14(3), 252-271. doi: 10.1177/1094670511411703
Brodie, R. J., Ilic, A., Juric, B., & Hollebeek, L. (2011b). Consumer engagement in a
virtual brand community: An exploratory analysis. Journal of Business Research(0).
doi: 10.1016/j.jbusres.2011.07.029
Chaudhuri, A., & Holbrook, M. B. (2001). The Chain of Effects from Brand Trust and
Brand Affect to Brand Performance: The Role of Brand Loyalty. [Article]. Journal
of Marketing, 65(2), 81-93.
Cheung, C., Lee, M., & Jin, X. (2011). Customer Engagement in an Online Social
Platform: A Conceptual Model and Scale Development. ICIS 2011 Proceedings. ,
Paper 8. .
Chin, W. W. (1998a). Commentary: Issues and opinion on structural equation modeling.
MIS Quarterly, 22(1), vii-xvi.
Chin, W. W. (1998b). The partial least squares approach to structural equation modeling.
In: Marcoulides, George A. (Ed), Modern Methods for Business Research (pp. 295358). Mahwah, NJ: Lawrence Erlbaum Associates.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale,
NJ: Lawrence Erlbaum Associates.
Constantinides, E., & Fountain, S. J. (2008). Web 2.0: Conceptual foundations and
marketing issues. Journal of Direct, Data and Digital Marketing Practice, 9(3),
231-244.
Creamer, M. (2012). Study: Only 1% of Facebook 'Fans' Engage With Brands. Retrieved
from AdAge.com website: http://adage.com/article/digital/study-1-facebook-fansengage-brands/232351/
Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of
information technology. MIS Quarterly, 319-340.
Deighton, J., & Kornfeld, L. (2009). Interactivity's unanticipated consequences for
marketers and marketing. Journal of Interactive Marketing, 23(1), 4-10.
48
Diamantopoulos, A., Riefler, P., & Roth, K. P. (2008). Advancing formative measurement
models. Journal of Business Research, 61(12), 1203-1218. doi:
10.1016/j.jbusres.2008.01.009
Facebook.
(2012).
http://newsroom.fb.com/content/default.aspx?NewsAreaId=22
Retrieved 6th June 2012
FanPageList. (2012). http://fanpagelist.com/category/products/ Retrieved 7th June 2012
Fornell, C., & Larcker, D. F. (1981). Structural equation models with unobservable
variables and measurement error: Algebra and statistics. Journal of Marketing
Research, 18(3), 382-388.
George, D., & Mallery, P. (1999). SPSS for Windows Step by Step: A simple guide and
reference: Needham Heights, MA: Allyn & Bacon.
Gustafsson, A., Johnson, M. D., & Roos, I. (2005). The Effects of Customer Satisfaction,
Relationship Commitment Dimensions, and Triggers on Customer Retention.
[Article]. Journal of Marketing, 69(4), 210-218. doi: 10.1509/jmkg.2005.69.4.210
Hennig-Thurau, T., Malthouse, E. C., Friege, C., Gensler, S., Lobschat, L., Rangaswamy,
A., & Skiera, B. (2010). The Impact of New Media on Customer Relationships.
Journal of Service Research, 13(3), 311-330. doi: 10.1177/1094670510375460
Henseler, J., & Fassott, G. (2010). Testing Moderating Effects in PLS Path Models: An
Illustration of Available Procedures. In V. Esposito Vinzi, W. W. Chin, J. Henseler
& H. Wang (Eds.), Handbook of Partial Least Squares: Concepts, Methods and
Applications (pp. 713-735). Berlin et al.: Springer.
Henseler, J., Ringle, C. M., & Sinkovics, R. R. (2009). The Use of Partial Least Squares
Path Modeling in International Marketing. In R. R. Sinkovics & P. N. Ghauri (Eds.),
Advances in International Marketing (Vol. 20, pp. 277-320). Bingley: Emerald
Higgins, E. T., & Scholer, A. A. (2009). Engaging the Consumer: The Science and Art of
the Value Creation Process. Journal of Consumer Psychology, Vol. 19, No. 2, pp.
100-114, 2009.
Hirschman, A. O. (1970). Exit, voice, and loyalty: Responses to decline in firms,
organizations, and states. Cambridge, MA: Harvard University Press.
Hollebeek, L. (2011a). Exploring customer brand engagement: definition and themes.
Journal
of
Strategic
Marketing,
19(7),
555-573.
doi:
10.1080/0965254x.2011.599493
Hollebeek, L. D. (2011b). Demystifying customer brand engagement: Exploring the loyalty
nexus. Journal of Marketing Management, 27(7-8), 785-807. doi:
10.1080/0267257x.2010.500132
Jarvis, C. B., MacKenzie, S. B., & Podsakoff, P. M. (2003). A critical review of construct
indicators and measurement model misspecification in marketing and consumer
research. Journal of Consumer Research, 30(2), 199-218.
Karpen, I. O., Bove, L. L., & Lukas, B. A. (2012). Linking Service-Dominant Logic and
Strategic Business Practice. Journal of Service Research, 15(1), 21-38. doi:
10.1177/1094670511425697
49
Kim, T., & Biocca, F. (1997). Telepresence via Television: Two Dimensions of
Telepresence May Have Different Connections to Memory and Persuasion.[1].
Journal of ComputerMediated Communication, 3(2).
Kumar, V., Aksoy, L., Donkers, B., Venkatesan, R., Wiesel, T., & Tillmanns, S. (2010).
Undervalued or Overvalued Customers: Capturing Total Customer Engagement
Value.
Journal
of
Service
Research,
13(3),
297-310.
doi:
10.1177/1094670510375602
Malthouse, E. C., & Calder, B. J. (2011). Comment: Engagement and Experiences:
Comment on Brodie, Hollenbeek, Juric, and Ilic (2011). Journal of Service
Research, 14(3), 277-279. doi: 10.1177/1094670511414584
Mollen, A., & Wilson, H. (2010). Engagement, telepresence and interactivity in online
consumer experience: Reconciling scholastic and managerial perspectives. Journal
of Business Research, 63(910), 919-925. doi: 10.1016/j.jbusres.2009.05.014
Morgan, R. M., & Hunt, S. D. (1994). The Commitment-Trust Theory of Relationship
Marketing. [Article]. Journal of Marketing, 58(3), 20.
MSI. (2010). 2010-2012 Research Priorities
Retrieved August 15, 2012, from
http://www.msi.org/research/index.cfm?id=271
Nelson-Field, K., & Taylor, J. (2012). Facebook fans: A fan for life? Admap, 25-27.
Nunnally, J. (1978). Psychometric theory: New York: McGraw-Hill.
Ringle, C. M., Wende, S., & Will, A. (2005). SmartPLS 2.0: www.smartpls.de.
Sawhney, M., Verona, G., & Prandelli, E. (2005). Collaborating to create: The Internet as a
platform for customer engagement in product innovation. Journal of Interactive
Marketing, 19(4), 4-17. doi: 10.1002/dir.20046
Sprott, D., Czellar, S., & Spangenberg, E. (2009). The Importance of a General Measure of
Brand Engagement on Market Behavior: Development and Validation of a Scale.
Journal of Marketing Research, 46(1), 92-104. doi: 10.1509/jmkr.46.1.92
van Doorn, J., Lemon, K. N., Mittal, V., Nass, S., Pick, D., Pirner, P., & Verhoef, P. C.
(2010). Customer Engagement Behavior: Theoretical Foundations and Research
Directions.
Journal
of
Service
Research,
13(3),
253-266.
doi:
10.1177/1094670510375599
Vargo, S., & Lusch, R. (2008). Service-dominant logic: continuing the evolution. Journal
of the Academy of Marketing Science, 36(1), 1-10. doi: 10.1007/s11747-007-0069-6
Verhoef, P. C., Reinartz, W. J., & Krafft, M. (2010). Customer Engagement as a New
Perspective in Customer Management. Journal of Service Research, 13(3), 247-252.
doi: 10.1177/1094670510375461
Vibert, A. B., & Shields, C. (2003). Approaches to student engagement: Does ideology
matter? McGill Journal of Education, 38(2), 221-240.
WARC. (2012a). Social media gains ground. Retrieved 13 August, 2012, from
http://www.warc.com/LatestNews/News/EmailNews.news?ID=29421&Origin=WA
RCNewsEmail
50
WARC. (2012b). Brands need Facebook "stickiness". Retrieved August 13, 2012, from
http://www.warc.com/LatestNews/News/EmailNews.news?ID=29986&Origin=WA
RCNewsEmail
WARC. (2012c). Brands seeking ROI metrics for Facebook. Retrieved 17th August, 2012,
from
http://www.warc.com/LatestNews/News/EmailNews.news?ID=30250&Origin=WA
RCNewsEmail
Wilkins, J. (2007). Web 2.0. Infonomics, 21(4), 10.
Zeithaml, V. A., Berry, L. L., & Parasuraman, A. (1996). The behavioral consequences of
service quality. The Journal of Marketing, 31-46.
51
52
Strongly
Agree
53
Strongly
Disagree
I browse on Facebook fan pages
because I am interested in being a
part of a brand community
I have enough time to browse on
Facebook fan pages
I think that browsing on Facebook
fan pages is not worth the time and
effort
Browsing on Facebook fan pages
does not take too much effort
I enjoy browsing on Facebook fan
pages
I browse on Facebook fan pages
because I am interested in the
brands they are dedicated to
Strongly
Agree
54
5. Which of the following food and drink brands are you a fan of on Facebook, if any?
You may select more than one answers.
Coca
Red Bull
Oreo
Skittles
Pringles
Monster Energy
Ferrero Rocher
Nutella
Dr Pepper
Starburst
Reeses
Starbucks Frappuccino
Sprite
Pepsi
Mountain Dew
Other food or drink brand (please name one only): _____
None
Coca
Red Bull
Oreo
Skittles
Pringles
Monster Energy
Ferrero Rocher
Nutella
Dr Pepper
Starburst
Reeses
Starbucks Frappuccino
55
Sprite
Pepsi
Mountain Dew
[Other food or drink brand, if selected and entered in the previous question]
The following questions concern your engagement with the Facebook fan page of
[selected brand].
7. How often do you..
Never
visit the Facebook fan page
of [selected brand]
notice posts by [selected
brand] in your news feed?
read posts by [selected
brand]?
like posts by [selected
brand]?
comment on Facebook wall
posts by [selected brand]?
share posts by [selected
brand] with your friends?
post on the Facebook fan
page of [selected brand]
yourself?
Almost
Rarely Sometimes Often
never
Almost
all the
time
All the
time
56
Strongly
Agree
Jump to question 10 and continue answering the remaining questions with the same brand in
mind.
FP = Fan page
57
Coca
Red Bull
Oreo
Skittles
Pringles
Monster Energy
Ferrero Rocher
Nutella
Dr Pepper
Starburst
Reeses
Starbucks Frappuccino
Sprite
Pepsi
Mountain Dew
Other food or drink brand (please name one only):______
58
Strongly
Agree
59
Consequences
11. How much do you agree with the following statements?
Strongly
Disagree
I will buy [selected brand] the next
time I buy food/drinks
I intend to keep purchasing
[selected brand]
I say positive things about
[selected brand] to other people
I often recommend [selected
brand] to others
I encourage friends to buy
[selected brand]
Strongly
Agree
Socio-demographic
12. What other online social media platforms do you have an account on, if any? You
may select more than one answer.
Twitter
LinkedIn
MySpace
Google+
Bebo
Badoo
Tagged
Orkut
Friendster
hi5
Netlog
YouTube
Instagram
Flickr
Pinterest
Foursquare
Tumblr
Other (please specify): _______
None
60
13. How much time do you spend on online social media platforms on average every day?
61
62