Original Article

Examining the drivers and brand

performance implications of
customer engagement with brands
in the social media environment
Received (in revised form): 21st May 2014

Natalie Jane De Vries

is a research analyst in Marketing at the Faculty of Business and Law, the University of Newcastle, Australia. She holds a
Bachelor of Business (Marketing) with Honours I. Her research interests include branding, customer engagement and social
media marketing.

Jamie Carlson
is a Senior Lecturer in Marketing at the Faculty of Business and Law, the University of Newcastle, Australia. He holds a
Bachelor of Business with Honours I and a PhD in Marketing. His research interests include consumer behaviour, service
performance and branding. His work has appeared in numerous journals including the Journal of Marketing Management, Journal
of Consumer Marketing, Journal of Services Marketing, Internet Research, Electronic Markets, Managing Service Quality.

ABSTRACT A key issue for marketers resulting from the dramatic rise of social media is
how brand pages can be leveraged to engage customers and enhance relationships with
brands. The article examines the role of gratifications consumers derive from brand
pages together with customer-brand relationship characteristics influencing customer
engagement (CE) with Facebook brand pages. Data was gathered via a survey of 404
consumers of brand pages and analysed using structural equation modelling. The find-
ings show that co-creation value, social value, usage intensity and brand strength influ-
ence CE with brand pages. CE was also found to influence brand performance
outcomes of CE behaviours directed at the brand page and brand loyalty. The findings
are of value to brand managers of social media sites and focus on how managing critical
user gratifications together with customer-brand relationship variables acts as a
mechanism for unlocking CE with brand pages. In addition, the study examines CE
effects on both behaviours central to the brand page and brand loyalty outcomes in the
research framework.
Journal of Brand Management (2014) 21, 495–515. doi:10.1057/bm.2014.18;
published online 27 June 2014

Correspondence:
Jamie Carlson
Keywords: social media; customer engagement; consumer behaviour; brands; motivation
Newcastle Business School,
University of Newcastle, Level 3,
brand relationships
University House, Corner King
and Auckland Streets, Newcastle,
Callaghan 2300, Australia
E-mail: jamie.carlson@newcastle.
edu.au

© 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515

www.palgrave-journals.com/bm/
 De Vries and Carlson

INTRODUCTION ‘comment’ over 3.2 billion times a day

In recent years, greater focus has been (Facebook, 2013), and recent forecasts
placed upon customer engagement (CE) in project that interactive marketing expendi-
branding and relationship marketing, which ture by 2016 will reach up to US $77 billion
has been argued to act as a vehicle for in the United States alone (Forrester
enhancing consumer relationships, profit- Research, 2011).
ability and growth (Brodie et al, 2011; Scholars have begun to argue that by
Hollebeek, 2011a; Vivek et al, 2012). CE leveraging social media platforms, organiza-
refers to a psychological state reflecting cus- tions can create online brand communities.
tomers’ interactive, co-creative experiences Perhaps the best-known example of this is
with a firm, which highlights the active role Brand Facebook Pages (BFP) found on the
of the consumer (Brodie et al, 2011; Verleye Facebook platform. In practice, users follow
et al, 2014). CE is distinct relative to tradi- a BFP by pressing the ‘like button’, which
tional concepts such as consumer involve- indicates to their social network that they
ment, which reflects a consumer’s level of like the brand. This then enables new con-
interest in, and personal relevance of, a tent to be automatically sent from the brand
brand, whereas CE seeks to explain or pre- to be posted to their personal Facebook
dict the dynamics characterizing focal inter- newsfeed, where customers can interact
active consumer/brand relationships more with a brand via anecdotes, photos, videos
explicitly such as in social media (Brodie or other brand-related material that users
et al, 2011; Hollebeek et al, 2014). can like, share or comment on with other
Interest in CE has also attracted attention followers of the brand or their own friends
in the social media environment, and there (De Vries et al, 2012; Jahn and Kunz, 2012;
is interest in how brands can leverage new Labrecque et al, 2013). Interestingly for
media and online communities on platforms marketers, social media technologies enable
such as Facebook, Twitter or YouTube to brands to co-create brand stories with, and
engage and collaborate with customers due among, active networked consumers
to social media’s multi-dimensional, two- through higher levels of customer-brand
way, peer-to-peer communication proper- interaction (Gensler et al, 2013). Such
ties (Yan, 2011; Jahn and Kunz, 2012; interactive communications can act as
Brodie et al, 2013; Hutter et al, 2013). mechanisms for value co-creation and
Specifically, there has been significant extraction possibilities such as improved
growth in the adoption and use of brand brand meaning (Hatch and Schultz, 2010;
pages found on the Facebook social net- Gensler et al, 2013) and collaborative pro-
working platform by consumers to com- duct innovation opportunities (Prahalad and
municate with their favourite brands and Ramaswamy, 2004; Sawhney et al, 2005;
with one another. In this sense, consumers Kozinets et al, 2010), thereby enhancing
are becoming pivotal authors of brand stor- consumer perceptions of CE with BFP,
ies through the easy sharing of brand leading to favourable consumer behaviours
experiences due to the triad of commu- towards the brand.
nication arising from new dynamic net-
works among consumers and brands formed
through social media (Gensler et al, 2013). PROBLEM STATEMENT AND
These developments are of significant PURPOSE
interest to brand managers, considering the Emerging CE research in the social media
fact that on the Facebook platform more context of BFP has shown that consumers
than 1.11 billion users worldwide ‘like’ and who become more psychologically engaged

496 © 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515
 Customer engagement, brands and social media

with these BFP tend to be more committed deliver co-creation activities which they
and loyal to the brand (Jahn and Kunz, consider of value to them, we posit that the
2012), to visit the physical retail store more concept of co-creation value may serve as
and to generate more positive word of an antecedent driver of CE with BFP in
mouth (Hutter et al, 2013), and are more addition to other variables that have
emotionally attached to the brand than hitherto been empirically considered in the
non-brand fans (Dholakia and Durham, context of BFP. Further to the inclusion of
2010). Despite deployment of BFP quickly value co-creation into the research frame-
increasing by many consumer brands glob- work, little is known about the impact of
ally to better reach, interact with and serve consumers’ perceived strength of the custo-
customers, the rate of theorizing to under- mer-brand relationship and its effect on the
stand how to better facilitate and optimize development of CE with BFP, in addition
CE with BFP and its effects on brand loyalty to sought after needs and gratifications as
outcomes has not kept pace with this investigated by Jahn and Kunz (2012). The
growth in social media marketing. The issue CE and branding literature has conceptually
is of critical importance because little is argued that a consumer’s relationship with a
known about how the interactions that take brand impacts the intensity of CE towards
place in social media environments con- the brand (for example, Loureiro et al, 2012;
tribute to effective brand-building efforts. Vivek et al, 2012) including firm-based
In response to these concerns and gaps in activities that personify the brand including
the literature, we aim to offer a more in- websites or other computer-mediated enti-
depth understanding of CE in BFP by ties (Mollen and Wilson, 2010), and brand
focusing on three important issues. performance in social media (Gensler et al,
First, we develop a theoretical framework 2013). As such, incorporating these theore-
that places greater emphasis on the identifi- tical insights we empirically examine the
cation of relevant antecedents to CE with a role of consumer’s perceived strength with
BFP than previously studied. Among the the brand to assess its effect on the devel-
first works to enter the literature, Jahn and opment of CE with computer-mediated
Kunz (2012) empirically found that the entities such as the BFP in the social media
intensity of prior usage, seeking social environment.
interactions and brand interactions influ- Second, we place greater emphasis on the
ence CE levels with a BFP on the Facebook identification of relevant consequences of
platform. However, further theorizing and CE with a BFP than do previous studies.
investigation is needed on the notion of co- Although Jahn and Kunz (2012) found that
creation activity which has been argued to the psychological state of CE with the BFP
take place between the brand’s BFP and the influences brand loyalty outcomes includ-
customer in social media environments as ing brand commitment, positive word of
well as examine its contribution towards the mouth and future purchase intentions, we
formation of CE with the BFP. Past social expand upon this view by drawing on
media research has suggested the existence work by van Doorn et al (2010) to incor-
of value co-creation as a form of benefit porate relevant CE behaviours (CEB)
derived by the consumer as a result from the towards the BFP. According to van Doorn
BFP consumption experience (Hennig- et al (2010), CEBs are behavioural mani-
Thurau et al, 2010; Sashi, 2012), but expli- festations of CE towards a firm after and
cit research on this phenomenon has beyond purchase, which can contribute to
remained underexposed. In order to assess firm performance in two ways: (1) CEBs in
whether consumers perceive a BFP to interactions with the firm and their

© 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515 497
 De Vries and Carlson

Gratifications

Functional
Value
H1

Hedonic
Value H2 Usage Intensity
H7a
H7b CEB with BFP
H3a
Social
H6
Value
H3b

H4a H8a
Brand Loyalty
Co-Creation H4b H8b
Value CE with BFP

H5a
H5b

Brand Strength

Brand Relationship
Characteristics

Figure 1: Proposed conceptual model.

employees – such as giving suggestions for in four key sections. The first section presents
service improvement and (2) CEBs in a brief review of the literature with specific
interactions with other customers such as attention given to CE in the context of social
spreading word of mouth and/or writing media. The second section introduces the
reviews that influence other customer’ atti- conceptual model and system of relation-
tudes and behaviours. In the social media ships. Third, there is a discussion of the
context, the inclusion of CEBs towards the methodology and subsequent analysis of
BFP (for example, sharing, liking and posting findings. Finally, limitations and future
brand-related content online) has yet to be research directions are presented.
investigated within CE frameworks in the
social media marketing literature. Research-
ing this topic and adding to this literature THEORETICAL BACKGROUND
provides a deeper understanding of the pro- This section presents the proposed model
cess and practices that encourage successful of this study (Figure 1), illustrating the
CEB management practices. antecedent factors that influence CE with
Third, to test the robustness of our model, BFP and its consequences on CEB with
we apply it to a broad range of consumer BFP and brand loyalty. Underpinned by
brands in our empirical study. Furthermore, uses and gratifications theory, branding
to account for heterogeneity in the sample, theory and consumer engagement theory,
we differentiate between product and service the antecedent drivers are argued to be
brand BFP to detect whether the research related to a customer’s usage intensity and
framework differs by product versus service CE with a BFP. Finally, the constructs of
brands. To this end, this article is structured usage intensity and CE with a BFP are

498 © 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515
 Customer engagement, brands and social media

proposed to influence brand relationship Hypothesis 1: Functional value of the

outcomes including brand loyalty and BFP positively influences usage inten-
CEBs towards the BFP. The framework sity of the BFP.
explains that if the BFP satisfies the parti-
cular needs of a user, operating together
with the influence of favourable percep- The effect of hedonic value
tions of the brand relationship, this should Prior empirical research into CE in the
lead to higher usage levels of the BFP, context of BFP has examined seeking
which should in turn lead to higher brand hedonic value, or satisfying hedonic gratifi-
relationship outcomes. cations such as fun and enjoyment. For
instance, Madupu and Cooley (2010) found
hedonic value to be a major driver of online
The effect of functional value brand community participation. In the
Consumer motivations for adopting and context of BFPs, a customer satisfying a
using a wide range of media can be hedonic gratification need requires access to
explained using uses and gratifications fun, entertaining and exciting content (Jahn
theory. One of the main reasons for and Kunz, 2012). For instance, a customer
adopting a new media type has been func- using a BFP may do this in their spare time
tional or ‘information’ gratification (Luo knowing that this particular brand uploads
et al, 2011; De Vries et al, 2012). In the interesting and entertaining content in rela-
context of BFP, a customer satisfying a tion to that brand, and/or topics related to
functional gratification means to be able to what the brand personifies and symbolises.
have access to helpful, functional, practical This high perceived hedonic value of the
and useful content (Jahn and Kunz, 2012). BFP as derived by the customer then leads
For instance, a customer using the BFP of a to that customer using the brand page more
service brand may want to know how the frequently (Jahn and Kunz, 2012). In a
service works, what is involved and other similar fashion, Cvijikj and Michahelles
practical information. Consequently, this (2013) also found that entertaining content
study proposes that if a customer satisfies was one of the main drivers of BFP partici-
these needs for useful and practical content, pation and usage. Thus,
and in doing so satisfies their functional
gratification, they are more likely to use Hypothesis 2: Hedonic value of the BFP
that brand page more intensely. Empirical positively influences usage intensity of
support for this assertion can be drawn the BFP.
from Jahn and Kunz (2012), who found a
significant, positive influence from func-
tional value to usage intensity of the BFP. The effect of social-interaction value
Furthermore, Cvijikj and Michahelles Social value has also been argued to drive
(2013) found that informational content the adoption and usage levels of new media
was one of the main drivers that led active (Hennig-Thurau et al, 2010). Social net-
users to participate in BFP behaviours such working sites such as BFPs provide greater
as liking, commenting and sharing the opportunities for social interactions facili-
brand’s posts, as well as affecting duration tated via the Facebook platform where
of BFP interaction. On this basis, con- consumers can derive social value from
sumers with higher perceptions of func- computer-mediated interactions with one
tional value are more likely to influence another. Furthermore, it has been estab-
higher usage intensity with the BFP. Thus: lished that customers seek ‘linking value’

© 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515 499
 De Vries and Carlson

based on peer-to-peer bonds (Libai, 2010), brand story, which can contribute to build-
which act as a motivation for customer-to- ing awareness, comprehension, empathy,
customer interactions to take place. In the recognition and recall, and enhance the
context of new social media, for a customer meaning of the brand (Hatch and Schultz,
to satisfy a social-interaction gratification, 2010; Singh and Sonnenburg, 2012). In the
they need to be able to interact and com- social media context, consumers can dyna-
municate with other customers, perceive mically interact with brands quickly, on a
other customers as similar to themselves, and real-time basis, which means that they can
feel as if they have the opportunity to inter- co-create value for themselves by commu-
act, meet and communicate with people nicating, providing and receiving feedback,
similar to themselves (De Choudhury et al, share and interact with the brand, and
2010; Jahn and Kunz, 2012). Therefore, a therefore actually cooperate with the brand
higher perception of social-interaction value through the social networking platform
of the BFP derived by the customer may (Sashi, 2012; Gensler et al, 2013). Conse-
then lead to that customer using that brand quently, this study advances the notion that
page more frequently and/or becoming if a customer derives co-creation value from
more highly engaged with the BFP. Empiri- co-creation enabling interactions with the
cal evidence for this assertion can be drawn BFP, that customer is more likely to use that
from Jahn and Kunz (2012), who found a brand page more intensely and experience
significant, positive influence from social- higher levels of CE with the BFP. Drawing
interaction value on BFP on the Facebook on the arguments found in the consumer
platform. As such, it is argued in this study behaviour and co-creation of value litera-
that consumers with higher perceptions of ture, it appears that customers seek co-
the social-interaction value are more likely to creation value with brands on BFPs. As
influence higher usage intensity and higher such, it is argued in this study that con-
levels of CE on the BFP. Thus, sumers with higher perceptions of co-crea-
tion value are more likely to influence
Hypothesis 3a: Social interaction value of
higher usage intensity and higher levels of
the BFP positively influences usage
CE with the BFP. Thus,
intensity of the BFP.
Hypothesis 4a: Co-creation of value on
Hypothesis 3b: Social interaction value of
the BFP positively influences usage
the BFP positively influences CE with
intensity of the BFP.
the BFP.
Hypothesis 4b: Co-creation of value on
the BFP positively influences CE with
the BFP.
The effect of co-creation value
The notion of co-creation of value leading to
higher levels of CE in an online context has The effect of brand strength on
been introduced by Sawhney et al (2005), customer engagement
where the customer is able to interact, com- Based on the analysis of the literature, it can
municate, and in certain cases cooperate to be concluded that variables that reflect the
achieve experiences, services and offerings consumer’s strength of relationship with
that serve the customer better. From a the brand impact on attachment to and
branding perspective, such brand-related engagement with the brand (Loureiro
conversations enable consumers to integrate et al, 2012; Vivek et al, 2012), including
their own experiences and thoughts into the firm-based activities that personify the

500 © 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515
 Customer engagement, brands and social media

brand such as branded websites and social literature (cf. Dolich, 1969; Sirgy, 1982;
media sites (Mollen and Wilson, 2010; Sirgy et al, 2000). Self-brand congruency
Gensler et al, 2013). In this sense, we theo- refers to the match between a consumer’s
rize that when the relationship between the self-concept and the ‘personality’ of a brand
customer and the brand is perceived to be perceived by the customer, and when the
strong and the brand has high perceived match is high this will lead to increased
brand strength in the customer’s mind, the brand preference and loyalty (Jamal and
greater the levels of CE with the BFP. Goode, 2001; Kressmann et al, 2006), as
Therefore, to explore the role of the brand well as pre-consumption evaluations such
acting as a mechanism to facilitate CE with as attitude, preference and intention (for
BFPs, we focus here on developing our example, Aaker, 1999). Moreover, the
term perceived brand strength, which may saliency of self-image congruity was
be viewed through the lens of involvement demonstrated in the adoption process of
and self-brand congruency theory, andthese information systems such as mobile ser-
important brand variables converge into a vices (Kleijnen et al, 2005) and social net-
multi-dimensional construct, namely, brand working sites (Kang et al, 2009). Building
strength. on this logic that self-image congruence
Within consumer behaviour, involve- between an individual and the brand is an
ment is an important concept and has been important factor in understanding brand
identified as being at the heart of the per- preference, we anticipate the same effect
son-object relationship, the relational vari- with BFPs where it acts as a driver that
able predictive of consumers’ behaviour influences the propensity for usage inten-
(Evrard and Aurier, 1996), and largely based sity and CE with the BFP.
on the interaction of consumers with Based on the above discussion, it is the-
objects or stimuli including e-commerce orized in this study that these important
websites (Carlson and O’Cass, 2012). In the brand variables converge into a multi-
internet environment, it has been found dimensional construct of brand strength
that involved consumers are more interested that has a positive impact on CE. This is
in a website if the information and other because if a customer has a higher level of
attributes delivered via the website are rela- perceived brand strength (that is, strength
ted to the object of involvement (that is, the of relationship with a particular brand),
brand), are motivated to gain and process they are more likely to experience higher
more information about the brand on the levels of CE and become a more integrated
website, possess higher exploratory beha- and participating member of the BFP.
viour related to the website, exhibit more Therefore, we argue that involvement and
extensive information search habits, and self-brand congruency are two critical
conduct online purchases (Balabanis and components of brand strength. Impor-
Reynolds, 2001; Richard and Chandra, tantly, as consumers participate in con-
2005). Building on this view, we anticipate templative assessments, some will assess a
the same effect to be applicable to social brand as being significant to their life and
media sites such as BFP where higher brand congruent with their self-concept. On that
involvement will influence higher usage basis, brand strength is likely to, in part,
intensity and CE levels. contribute to the formation of a favourable
Regarding a consumer’s attitude towards, assessment of CE in collateral aspects of the
and mental image of, a brand, the con- brand’s activity such as the BFP, as well as
gruence between self-image and brand behavioural activity towards the BFP such
image has been extensively examined in the as increased usage levels. As such, it is

© 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515 501
 De Vries and Carlson

argued in this study that consumers with brand, the brand relationship and increased
higher perceptions of a brand’s brand likelihood of brand loyalty has previously
strength are more likely to influence higher been made in involvement theory (see, for
levels of usage intensity and higher levels instance, Olsen, 2007). Following the
of CE with the BFP Thus, direction of Jahn and Kunz (2012) on the
link between usage intensity and CE and
Hypothesis 5a: Brand strength positively
brand loyalty with the BFP, we extend this
influences usage intensity.
link to also incorporate CEBs directed at the
Hypothesis 5b: Brand strength positively BFP and theorize this construct as a con-
influences CE on the BFP. sequence of usage intensity. On this basis,
we hypothesize the collective effects of
usage intensity on CE, CEBs directed at
BFPs and brand loyalty. Based on this rea-
The effect of usage intensity on CE with soning, we expect:
BFP Hypothesis 6: Usage intensity positively
As CE has been defined as a psychological
influences CE with the BFP.
state reflecting customers’ interactive and
integrative participation with a firm (that is, Hypothesis 7a: Usage intensity positively
a behavioural orientation) (van Doorn et al, influences brand loyalty.
2010; Jahn and Kunz, 2012), it is expected Hypothesis 7b: Usage intensity positively
that there is a correlation between usage influences CEB with the BFP.
intensity and CE. That is, a logical con-
sequence of a customer using a BFP more
frequently is heightened engagement levels
with that brand. However, as Jahn and The effect of CE with BFP on brand
Kunz (2012) state, it is possible for a con- performance outcomes
sumer to be a frequent user of a particular Favourable performance responses towards
BFP without becoming an integrated and the brand as a consequence of CE have been
participating member. It is therefore fruitful proposed by multiple authors (Hollebeek,
to investigate this proposed relationship 2011b; Vivek et al, 2012; Brodie et al, 2013).
further. As such, consistent with Jahn and This link has since been empirically sup-
Kunz (2012), we argue that where a positive ported in the social media context where
relationship exists between a customer’s CE has influenced offline loyalty behaviours
usage of the BFP and their engagement, it towards the brand (including purchase
would be expected to influence CE with a intentions, positive offline word of mouth
BFP when the customer uses the BFP more and willingness to pay more) (Jahn and
frequently, often and on a regular basis. Kunz, 2012; Hutter et al, 2013). However,
Furthermore, it has been empirically the link from CE to CEBs on the BFP has
found that users who are in regular contact yet to be investigated. Although not inves-
with the brand and thus have high levels of tigating CE specifically, Cvijikj and
usage intensity would positively impact on Michahelles (2013) argue that active users
their brand relationship and therefore the of the BFP are more likely to participate in
likelihood of their engaging in word of CEB like responses on the BFP such as
mouth, repurchase or other loyalty and intentions to ‘like’, ‘share’ and post/share
brand-related behaviours (Jahn and Kunz, comments. CEBs directed at the BFP are
2012). As Jahn and Kunz (2012) explain, the particularly pertinent in the social media
link between regular contact with the context, given the potential in the

502 © 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515
 Customer engagement, brands and social media

magnitude of social contagion effects of their demographic and behavioural char-

customer-to-customer interactions among acteristics (Wyllie et al, 2014).
social networks of users (Hennig-Thurau Items to measure each construct were
et al, 2010; Libai, 2010; Labrecque et al, drawn from the literature and are shown in
2013). Based on the above, we argue that as Table 1. For instance, items to measure
CE with the BFP increases, loyalty to the the constructs of functional value, hedonic
brand will also increase. We further argue value, social-interaction value, usage inten-
that such favourable responses to the brand sity and CE were drawn from Jahn and
will also include future CEBs directed at the Kunz (2012). A six-item measure of
BFP such as intentions to ‘like’, ‘share’ and co-creation value was adapted from O’Cass
post/share comments on the BFP, and and Ngo (2011). A six-item measure of
continued usage of the brand’s BFP. This brand involvement was drawn from Carlson
reasoning leads to the following hypotheses: and O’Cass (2012). A five-item measure
of self-brand congruency was adapted
Hypothesis 8a: CE with the BFP posi-
from Hohenstein et al (2007). Measures
tively influences brand loyalty.
used to assess brand loyalty were adapted
Hypothesis 8b: CE with the BFP in the from Zeithaml et al (1996) and three
offline channel CEB with the BFP. items to measure CEBs directed at
the BFP were adapted from Cvijikj and
Michahelles (2013). Each item was mea-
sured on a seven-point Likert scale
METHODOLOGY anchored with (1) ‘strongly disagree’ to (7)
‘strongly agree’.
Study design
Data was collected via a self-administered
paper-and-pencil intercept survey with stu- Sample profile
dents at a large public university in Australia. In total, 452 completed surveys were col-
The advantage of conducting an intercept lected with 404 useable surveys remaining
survey is that the researcher (or interviewer) after screening. Of this sample, 42.3 per cent
comes to the respondents, making it more were males and 57.7 per cent females. The
efficient and less difficult for the respondent age ranged from 17 to 49 with an average
to participate, and respondents can be age of 21.44. Owing to the sample being a
screened by the researcher for eligibility to student sample, the large majority was in the
participate (Malhotra, 2010). Such a sample 18–24 years age bracket (90.4 per cent) fol-
is considered highly suitable because the lowed by the 25–30 years bracket (6.4 per
majority of Facebook users are young adults cent) and the 30+ years bracket (2.7 per cent).
between the ages of 18 and 30 years (Chu When asked to indicate their usage of their
and Kim, 2011). This makes it appropriate nominated BFP, the majority responded once
to draw inferences from a student-based a month or more (43.5 per cent) followed by
sample on the population of Facebook once a week or more (37.5 per cent) and daily
users, as this type of sample has the appro- (19 per cent). The majority of respondents
priate demographic and technology usage also indicated that they had been a user of the
characteristics to yield useful results (Taylor BFP they indicated for six months or more
et al, 2012). Thus, a student sample qualifies (60.3 per cent), followed by 3-5 months (21.9
as a relevant sample for the theory-building per cent), 1-2 months (11.3 per cent) and less
purposes of this study due to the homo- than one month (6.3 per cent). Finally, the
genous nature of students in regard to both respondents were also asked to identify what

© 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515 503
 De Vries and Carlson

device they usually used to access the first evaluated the measurement model and
Facebook brand page, with the majority second estimated the structural model.
answering mobile smartphone (54.1 per cent). Although the measurement model results
This was followed by those using a laptop are akin to those of principal component
(30.5 per cent), a home desktop or PC (9.4 analyses, the path coefficients calculated in
per cent), a mobile tablet (4.1 per cent) and the structural model can be interpreted in a
finally a work PC (1.95 per cent). On the similar fashion to that of beta (path) coef-
whole, the sample of this study comprises ficients in an ordinary least squares regres-
experienced users of BFP, with most accessing sion (Vock et al, 2013).
these pages from smartphone and laptop
devices.
Measurement model evaluation
In order to assess the psychometric prop-
erties of the multiple item scales, we esti-
EMPIRICAL RESULTS AND mated the measurement model by
FINDINGS calculating individual indicator reli-
Structural equation modelling (SEM) using abilities, composite reliability (CR), con-
Partial Least Squares (PLS) was employed for vergent validity and discriminant validity
analysis of the survey data. PLS is ideal for (for example, Hair et al, 2014). To do so,
studies with smaller sample sizes (for exam- we ran the PLS algorithm (path weighting
ple, less than 500) (Hair et al, 2014) and is scheme) and bootstrapping procedure (500
more suitable to the investigation of rela- samples). To assess the reliability of indi-
tionships in a predictive rather than con- vidual items, we checked whether the
firmatory fashion (Fornell and Bookstein, loading of each item on its respective
1982) such as predicting the relationships latent construct was higher than the
between CE and BFP and its consequences. recommended threshold of 0.70. As
Furthermore, PLS allows for simultaneous shown in Table 1, all loadings exceeded
analysis for reflective and formative con- this threshold, ranging from 0.73 to 0.96,
structed models (Hair et al, 2014). This is and were significant (t-statistic > 1.96) and
particularly the case where a second-order hence retained. To assess the internal
formative latent construct (in our case per- consistency, we evaluated the CR scores
ceived brand strength) is formed by several for each latent construct. They all excee-
first-order reflective latent constructs (that is, ded the recommended threshold of 0.70,
brand involvement, self-brand image con- which is considered satisfactory (Table 2).
gruency). We modelled the second-order CR scores are interpreted in a similar way
construct by using the hierarchical compo- to Cronbach’s α and are calculated in PLS.
nent model, where the indicators of the first- AVE scores (average variance extracted)
order reflective constructs are repeated to indicate the level of convergent validity as
measure the second-order formative con- they specify the amount of shared variance
struct. SEM using the PLS method is also between reflective measurement items and
consistent with previous branding and inter- their respective latent construct. All AVE
net-based studies examining consumer scores in Table 1 are higher than the
behaviour (for example, Loureiro et al, 2012; recommended benchmark of 0.50. In
O’Cass and Carlson, 2012). addition, it is recommended to assess any
We conducted our analyses with statis- formative measurement constructs on the
tical software SmartPLS v2.0 (Ringle et al, basis of convergent validity and the sig-
2005) using a two-step procedure that nificance and relevance of outer weights

504 © 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515
 Customer engagement, brands and social media

Table 1: PLS measurement model results

Components and manifest variables AVE CR Loading t-value

Functional value 0.80 0.94 — —

The content of the Facebook brand page is helpful for me 0.91 64.25
The content of the Facebook brand page is useful for me 0.92 61.50
The content of the Facebook brand page is functional for me 0.90 54.23
The content of the Facebook brand page is practical 0.84 33.06

Hedonic value 0.77 0.93 — —

The content of the Facebook brand page is fun 0.89 49.43
The content of the Facebook brand page is exciting 0.91 53.35
The content of the Facebook brand page is pleasant 0.87 34.27
The content of the Facebook brand page is entertaining 0.85 28.22

Co-creation value 0.72 0.94 — —

The Facebook brand page interacts with me to serve me better 0.81 28.08
The Facebook brand page works together with me to produce offerings that better suit me 0.88 50.61
The Facebook brand page interacts with me to design offerings that meet my needs 0.85 41.85
The Facebook brand page provides services in conjunction with me 0.85 47.90
The Facebook brand page allows my involvement in providing services to me to get the 0.85 39.18
experience that I want
The Facebook brand page provides me with services that I help create 0.79 26.73

Social value 0.86 0.96 — —

I can meet people like me on this Facebook brand page 0.93 85.58
I can meet new people like me on Facebook brand page 0.94 77.22
I can find out about people like me on this Facebook brand page 0.92 68.82
I can interact with people like me on this Facebook brand page 0.91 49.20

Brand involvement 0.70 0.93 — —

This brand means a lot to me 0.87 59.13
This brand is significant to me 0.91 81.42
I consider this brand to be a relevant part of my life 0.88 55.32
For me personally, this brand is important 0.91 81.09
I am interested in this brand 0.72 21.18
I am involved in/with this brand 0.70 19.48

Self-brand image congruency 0.77 0.94 — —

This brand is a lot like me 0.86 39.72
This brand reflects what I am 0.91 80.36
This brand is exactly how I see myself 0.91 70.57
This brand image corresponds to my self-image in many respects 0.90 72.85
Through this brand, I can express what I find important in life 0.80 30.80

Usage intensity 0.87 0.95 — —

I frequently use the Facebook brand page 0.94 16.12
I often use the Facebook brand page 0.93 61.80
I regularly use the Facebook brand page 0.93 68.72

Customer engagement with the BFP 0.81 0.95 — —

I am an integrated member of this Facebook brand page community 0.88 55.80
I am an engaged member of this Facebook brand page community 0.91 57.56
I am an active member of this Facebook brand page community 0.93 77.52
I am a participating member of this Facebook brand page community 0.88 38.55
I am an interacting member of this Facebook brand page community 0.91 61.90

© 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515 505
 De Vries and Carlson

Table 1: Continued

Components and manifest variables AVE CR Loading t-value

Brand loyalty 0.68 0.93 — —

I recommend this brand to other people 0.73 18.46
I introduce this brand to other people 0.78 24.30
I say positive things about this brand to other people 0.84 33.58
I intend to remain loyal to this brand in the future 0.87 54.39
I will not stop supporting this brand 0.85 43.44
I think of myself as a loyal customer/supporter of this brand 0.86 37.42

CEB on the BFP 0.71 0.88 — —

I will share this brand’s Facebook page content in the future 0.86 36.96
I intend to ‘like’ this brand’s Facebook page content 0.81 30.68
I will comment on this brand’s Facebook content in the future 0.85 24.62

(Formative constructs) Weights t-value

Brand strength (Formative type II outer model) weights — —
Brand involvement 0.58 39.59
Self-brand image congruency 0.52 35.31

Abbreviations: AVE: average variance extracted; BFP: Brand Facebook Pages; CEB: customer engagement behaviours; CR:
composite reliability.

Table 2: Latent variable correlations

Construct Mean SD 1 2 3 4 5 6 7 8 9 10

1. Co-creation value 4.12 1.31 (0.94)

0.85
2. CE BFP 3.70 1.55 0.49 (0.95)
0.90
3. BFP CEB 4.93 1.37 0.38 0.50 (0.88)
0.84
4. Functional value 5.08 1.13 0.34 0.38 0.36 (0.94)
0.89
5. Hedonic value 5.33 1.16 0.24 0.33 0.36 0.32 (0.93)
0.88
6. Brand involvement 4.74 1.31 0.28 0.42 0.31 0.52 0.32 (0.93)
0.84
7. Brand loyalty 5.32 1.91 0.35 0.43 0.69 0.43 0.42 0.55 (0.93)
0.82
8. Self-brand image congruency 3.96 1.42 0.32 0.43 0.30 0.46 0.31 0.65 0.45 (0.94)
0.88
9. Social value 3.86 1.67 0.34 0.43 0.34 0.32 0.26 0.39 0.31 0.44 (0.96)
0.93
10. Usage intensity 4.41 1.54 0.24 0.44 0.38 0.37 0.30 0.36 0.32 0.33 0.20 (0.95)
0.93

Abbreviations: BFP: Brand Facebook Pages; CE: customer engagement; CEB: customer engagement behaviours.

(Hair et al, 2014). The formative outer Next we assessed discriminant validity,
model weights for brand strength are which is described as the extent to which a
included in Table 1, indicating satisfactory construct is truly distinct from other con-
loading and t-value scores. structs by empirical standards. As Hair et al

506 © 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515
 Customer engagement, brands and social media

Table 3: Structural model results

Hypotheses Predictor variables Predicted variables β R2 Critical ratio

H1 Functional value — 0.20 — 3.65*

H2 Hedonic value — 0.18 — 3.59*
H3a Social value — 0.09 — 0.63**
H4a Co-creation value — 0.07 — 1.44**
H5a Brand strength Usage intensity 0.20 0.23 3.09*
H3b Social Value — 0.22 — 3.50*
H4b Co-creation value — 0.33 — 5.98*
H5b Brand strength — 0.12 — 3.28*
H6 Usage intensity CE w/BFP 0.28 0.47 4.83*
H7a Usage intensity — 0.17 — 2.90*
H8a CE w/BFP CEB w/BFP 0.43 0.28 8.38*
H7b Usage intensity — 0.19 — 3.03*
H8b CE w/BFP Brand loyalty 0.32 0.20 6.32*
— — — AVA 0.30 —

N.B. AVA = Average Variance Accounted for; *Significant 1.96; **Not significant < 1.65.
Abbreviations: BFP: Brand Facebook Pages; CE: customer engagement; CEB: customer engagement behaviours.

(2014) explain, the Fornell-Larcker criter- explained by the predictor variables (Götz
ion (Fornell and Larcker, 1981) is a et al, 2010). Some scholars suggest that the
widely accepted approach to assessing dis- recommended R2 benchmark should
criminant validity where each latent con- exceed 0.10 (cf. Falk and Miller, 1992);
struct should share more variance with its however, a value of 0.20 is considered
own block of indicators than with any high in consumer behaviour studies (Vock
other latent variables from the model. et al, 2013).
Therefore, the AVE score of a latent vari- Our results are shown in Table 3. In line
able should be higher than the construct’s with Hypotheses 1 and 2, functional value
squared correlation with any other latent and hedonic value positively impact usage
variable, which is confirmed for our model intensity. However, the estimation of our
shown in Table 2. structural model did not confirm Hypotheses
3a and 4a, which suggested that social value
and co-creation value do not have a sig-
RESULTS nificant, positive impact on usage intensity.
The adequacy of the hypotheses that are We hence reject Hypotheses 3a and 4a. We
represented in the research model in also expected a positive effect of social value
Figure 1 was examined by running the (hypothesis 3b), co-creation value (hypoth-
PLS algorithm along with the boot- esis 4b), usage intensity (hypothesis 5) and
strapping procedure to obtain path coeffi- brand strength (hypothesis 6b) on CE with
cients, t-values and R2 coefficients of the BFP. Our results confirmed our hypotheses,
endogenous constructs. The quality of with the explained variance in CE with BFP
PLS models is assessed in a similar manner accounted for by social value, co-creation
to multiple regression models, which are value, usage intensity and brand strength
evaluated based on the direction and sig- was R2 = 0.47 which is considerably higher
nificance of path coefficients and the than for usage intensity. In line with
magnitude of R2, which indicate the Hypotheses 7a and 8a, CE with BFP and
amount of variance in a construct that is usage intensity positively impact brand

© 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515 507
 De Vries and Carlson

loyalty. The explained variance in brand respondents selected include popular and
loyalty accounted for by CE with BFP and well-known brands such as Apple, Nike,
usage intensity was R2 = 0.20. Our results Red Bull and Toyota. Some examples of
also confirmed our hypotheses for 7b and the wide variety of service brands selected
8b where CE with BFP and usage intensity by respondents include the online fashion
positively impact BFP CE behaviours with retail aggregator ASOS.com, McDonalds,
the explained variance R2 = 0.28. Subway, Jetstar Airlines and NAB Bank.
The predictive relevance of the structural To examine the differences between the
model was assessed via the Average Variance two groups, product brands and service
Accounted (AVA). The AVA is simply the brands were compared to determine whether
mean of each dependent construct R2 within differences existed in the direction and
the model and it represents the predictive strength of the paths. Comparisons were
power of the structural model without regard conducted visually by inspecting the path
to the measurement model (Fornell and coefficients (representing H1 to H8b) across
Bookstein, 1982). The results revealed that the two models. The results of this examina-
the AVA value is at a satisfactory level for the tion are displayed in Table 4 where paths
inner-structural model at 0.30. Given that the relating to H1, H2, H3a, H4b and H5b were
indices for predictive relevance of the struc- stronger for product brands than for service
tural model are higher than the recommended brands. Paths relating to H3b, H5a, H6, H7a,
0.10 benchmark, the theoretical soundness of H7b, H8a and H8b were stronger for service
the conceptual model is supported. brands than for product brands. Furthermore,
H4a was not supported for either brand type,
which is consistent with the finding for the
Post-hoc analysis total sample as displayed in Figure 2.
The examination of paths within the
Differences between product brands and research framework reveals differences in
service brands their importance in the context of brand
Following the hypothesis testing, post-hoc type. For instance, both the effects of func-
analysis was conducted to explore the tional value and hedonic value of the BFP
influence of brand type on the research content to the customer’s usage intensity,
model. In this study, the respondents were brand strength to CE with BFP, and co-
asked to identify a BFP they previously creation value to CE with BFP were stron-
‘liked’ or engaged with in an open question ger for product brands than for service
at the start of the survey. The sample has brands. Conversely, the effects of co-crea-
been manually coded and these brands have tion value to usage intensity, brand strength
been categorized as either being a ‘product’ to usage intensity, usage intensity to CE
brand or ‘service’ brand. As previous with BFP, usage intensity to CEB with BFP
empirical studies investigating CE with BFP and brand loyalty, and CE with BFP to
(for example, Jahn and Kunz, 2012) did not CEB with BFP were found to be stronger
account for differences in brand type, we for service brands than for product brands.
split the sample according to product brands
and service Brands. After separation of the
total data set into two subsets, the subset DISCUSSION
including all product brands contained 103 Following the model proposed by Jahn and
participants and the subset including all ser- Kunz (2012), the aim of this study was to
vice brands contained 301 participants. expand our understanding of the drivers of
Some examples of product brands that CE in the social media environment, and

508 © 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515
 Customer engagement, brands and social media

Table 4: Comparison of product brands and service brands

Hypothesis Products sample Services sample Result

Paths t-value Paths t-value

H1 0.36 7.21** 0.11 1.85* + and stronger for products

H2 0.21 5.62** 0.18 3.05** + and stronger for products
H3a −0.09 1.77* −0.04 0.73*** − and stronger for products
H3b 0.11 2.86** 0.23 4.37** + and stronger for services
H4a 0.05 0.84*** 0.06 1.24*** Not supported for either
H4b 0.35 8.16** 0.27 6.46** + and stronger for products
H5a −0.09 1.51** 0.32 5.19** + and stronger for services
H5b 0.32 8.00** 0.13 2.26** + and stronger for products
H6 0.16 3.68** 0.30 6.30** + and stronger for services
H7a 0.11 2.14** 0.23 4.23** + and stronger for services
H7b 0.09 1.76* 0.22 3.45** + and stronger for services
H8a 0.40 9.23** 0.41 8.31** + and stronger for services
H8b 0.28 5.82** 0.35 6.78** + and stronger for services

NB: *Significant > 1.65; **Significant > 1.96; ***Not significant < 1.65.

Gratifications

Functional
Value

Hedonic
Value Usage Intensity

Social CEB with BFP

Value

Co-Creation
Value Brand Loyalty
CE with BFP

Brand Strength

Brand Relationship
Characteristics
Figure 2: Final model.
Notes: …………New hypotheses, prior research has not considered these relationships. - - - - - - - Replication hypotheses tested in the
literature.
NB: All paths exceed t > 1.96; non-significant paths not shown.

investigate how CE translates to improving the BFP and brand loyalty. This study con-
key brand performance outcomes. We find tributes to the academic literature in several
that CE is formed by co-creation and social ways.
value together with usage intensity and First, it adds insights to the current lit-
brand strength. Moreover, our findings erature on CE for online social networks
show that CE influences CEB directed at (Dholakia and Durham, 2010; Jahn and

© 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515 509
 De Vries and Carlson

Kunz, 2012) by introducing co-creation framework, which has not been studied in
value and brand strength as mechanisms to prior work. For instance, in an examination
explain CE with the BFP. In terms of co- of the strength of paths between BFPs of
creation value, we empirically validate the- product and service brands, the results
oretical arguments by Hennig-Thurau et al revealed that functional and hedonic value
(2010), Sashi (2012) and others that co- acted as more important drivers of usage
creation activity between the brand and the intensity, while co-creation value and brand
consumer is an important benefit derived by strength acted as more important drivers of
the consumer as a result of the BFP con- CE with the brand page than for service
sumption experience. With regard to brand brands. On the other hand, the links
strength and its effects, we extend existing between co-creation value and usage
research that has been conducted primarily intensity, and brand strength and CE with a
in the website and social networking con- BFP, and the collective linkages between
text (for example, Kang et al, 2009; Carlson usage intensity, CE with a BFP, CEB and
and O’Cass, 2012), with our study demon- brand loyalty, were stronger for service
strating the viability of the construct influ- brands. As such, our study thus provides a
encing CE with BFPs. first step with preliminary evidence that
Second, this study adds insights to the brand type of the BFP seems to be guided
influence that CE has on two important by different relationship principles and
brand performance outcomes: CEBs with hence consumers seek out different sources
the BFP and brand loyalty. Although past of customer value.
researchers have found that CE with a BFP From a managerial perspective, the study
(Jahn and Kunz, 2012) and CE-like con- has several interesting implications. Pri-
structs such as brand page commitment marily, it provides a greater understanding
(Hutter et al, 2013) influence brand loyalty, and further clarification of mechanisms that
studies have yet to address non-purchase influence the formation of CE with a brand
engagement behaviours that are central to as personified through a BFP. The impor-
the BFP that provide benefits to the firm. tance of developing CE with a BFP was
These benefits are derived in the form of (1) further reinforced in this study given the
customer-firm interactions on the BFP to strong effects found on CEBs with BFP and
improve the customer relationship such as brand loyalty. This has important implica-
posting comments and brand-related com- tions for brand page moderators since the
munication dialogue and (2) customer-to- mechanisms by which to facilitate CE are
customer interactions that provide positive largely under the control of the marketing
brand advocacy and brand exposure (for manager. As the findings show, customers
example, awareness, observational learning) derive several gratifications from online
opportunities to social networks through content; therefore, the functional and
the liking and sharing of brand content. We hedonic content provided by a BFP impacts
extend present insights in the branding and their customers’ participation and the usage
marketing literature by examining the intensity of the BFP, which in turn influ-
impact of CE with the BFP, along with ence CE with BFP. Thus, the content a
usage intensity, as a direct predictor of brand BFP uploads is entirely moderated by the
loyalty and CEBs with the BFP in a unified brand, which needs to be practical, helpful
framework. and useful, as well as entertaining, fun and
Finally, our results underscore the pleasant. These aspects are critical, given
importance of considering the type of that the information and the subsequent
brand in the application of the research experience it enables on the BFP inform the

510 © 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515
 Customer engagement, brands and social media

vision behind the brand and build differ- In addition to brand involvement, the
entiations for it. It was further found that aspect of self-image–brand-image con-
the social-interaction and co-creation values gruency also needs to be considered by mar-
exhibited a strong positive influence on CE keting managers. As such, when designing
formation, indicating that it is important to social media strategies, brand managers
provide opportunities and processes for could consider that consumers might prefer
value co-creation and facilitate open, BFP that have content and images compa-
prompt communication with/among cus- tible with their perceptions of self, which
tomers and the brand to allow for greater may reflect the symbolic properties of the
levels of interaction, collective cooperation brand. Thus, brand managers can sig-
and collaboration. This may present chal- nificantly improve the effectiveness of their
lenges for firms as such processes and sub- brand positioning strategy of the BFP by
sequent moderation is demanding, as it measuring the image of their brand and self-
requires balancing the stimulation of CE images of their target audience. Further-
while steering that engagement in a direc- more, brand managers could also consider
tion that is consistent with the company’s segmenting the audience in terms of self-
objectives and brand image. This is further image–brand-image congruency. Analysing
intensified by the constantly changing the brand’s customer base may disclose dif-
interfaces and functionalities of large social ferent clusters of images with various brand-
networking platforms (for example, Face- related content (that is, imagery, brand
book, Instagram, Pinterest and Google+ information, reward campaigns and so on)
and so on) to create and manage content and interactivity requirements (for example,
with community members. volume of content, intensity of interactivity
This study also provides managerial and participation) developed to meet the
guidance by incorporating the notion of needs of these segments.
brand strength into the CE framework. As detailed in the research model in
For instance, it is important to recognise Figure 2, the brand loyalty outcomes of CE
the components that form brand strength with BFP provide brand managers with
in this study, namely, brand involvement reasons to be interested in the concept of
and self-brand image congruency. Brand CE. The loyalty outcomes include con-
managers should consider these concepts tinued support for the brand, positive word
when developing their social media stra- of mouth, and introducing and recom-
tegies. Specifically, it has been found that mending the brand to other people. Addi-
when a consumer is more highly involved tionally, this study provides practitioners
with a brand, they are more likely to pro- with a new finding linking CE to specific
cess information about the brand on CEBs with BFP, namely, higher levels of
websites and demonstrate exploratory liking, sharing and commenting on content
behaviours (for example, Carlson and in the future. Such CEBs are important in
O’Cass, 2012). Thus, increasing involve- the social media environment due to the
ment levels with consumers via the BFP potential for social contagion effects, as they
and appealing to mid-to-high-involve- encourage the users of BFP to engage in
ment consumers to adopt the BFP may dialogue and promote/evangilize the brand
yield higher levels of participation and with their social network to those who the
CE with the BFP, greater learning, and brand has not yet reached. The findings of
comprehension of the brand values this study shed more light on this beha-
and associations to ultimately achieve vioural aspect of CEB arising from engaged
brand performance outcomes. customers with BFPs, and provide brand

© 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515 511
 De Vries and Carlson

managers deeper insights into how to better industry compared with fashion retail or res-
elicit CE on their BFP and how this leads to taurants and the hospitality sector, would
favourable CEBs. provide the literature with a much greater
understanding of the drivers of CE specific to
a particular product or industry category.
LIMITATIONS AND FURTHER The findings from such research may also
RESEARCH reveal different drivers or different levels of
It is important that the limitations of the importance of the drivers of CE, and this is
study are detailed to ensure clarity for read- therefore a rich area of research.
ers and future researchers. First, the Face- Second, as stated in the limitations of this
book platform has a large global user base study, a non-random sample of university
and the external validity of the findings to students was used for this study. Future
global segments of users should be con- research should consider a larger random
sidered. Relating to this, a non-probability sample of respondents in order to reflect the
convenience sample made up of students wide variety of BFP users globally. This
was used in this study, which may limit the could include different demographic, indi-
external validity and generalizability. vidual consumer characteristics (for exam-
Therefore, the results should be interpreted ple, opinion leaders, market mavens,
with caution. A second limitation of this personality traits) and cultural groups to
study arises from the surveys being self- further enrich the findings. In addition to
administered questionnaires. This approach this, cross-cultural or cross-country studies
may produce data collection errors, and should also be considered in future research
non-response errors tend to be more fre- to prove the universality of the research
quent than in interview-administered for- framework. This is particularly relevant to
mat (Wright et al, 1998). Third, the sample informing how managers can ensure con-
of this study comprised many different sistent brand meaning on one BFP when it
brands that were assessed on the basis of is used by consumers around the world who
product brands versus service brands. How- might have completely different interpreta-
ever, no analysis of more detailed categories tions of brand meaning.
has been conducted in this study in order to Third, there are opportunities to enrich
provide more insights into the antecedent this research using a mixed-methods
drivers of online CE typical of a particular approach. For example, rather than relying
industry or product. Nevertheless, future solely on self-reporting via surveys, future
researchers should consider comparing studies could employ qualitative content
multiple brand categories with greater sam- analysis techniques such as netnography
ple sizes for more conclusive results across (Brodie et al, 2011) to analyse actual activity
industries and product categories. by customers online such as ‘likes’, ‘shares’
Despite these limitations, a number of and comments on the brand page and its
opportunities for future research have posts. Such analysis was conducted by
emerged from this study. First, no specific Cvijikj and Michahelles (2013), and future
product or industry categories were exam- research could incorporate such methods
ined in this study, which leaves many into a study examining the behavioural
opportunities for future research. Comparing aspects of CE with BFPs. Future research
the same model across different industries or addressing this topic could include analysis
product categories of different natures, for of positive and negative comments about
instance a comparison across different indus- the brand to ascertain the direction and
tries such as the consumer technology goods valence of these sentiments, which may

512 © 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515
 Customer engagement, brands and social media

influence future customer behaviour, as Brodie, R.J., Ilic, A., Juric, B. and Hollebeek, L. (2013)
Consumer engagement in a virtual brand com-
well as the impact on broader customer-to- munity: An exploratory analysis. Journal of Business
customer interactions. Research 66(1): 105–114.
Fourth, in order to capture the real-time Carlson, J. and O’Cass, A. (2012) Optimizing the online
channel in professional sport to create trusting and
nature of customers’ online behaviours, we loyal consumers: The role of the professional sports
suggest that future researchers adopt differ- team brand and service quality. Journal of Sport
ent measurements of a customer’s usage Management 26(6): 463–478.
intensity. For instance, measurement items Chu, S.C. and Kim, Y. (2011) Determinants of
consumer engagement in electronic word-of-mouth
measuring the number of times a day, week (eWOM) in social networking sites. International
or month a customer accessed, used and Journal of Advertising 30(1): 47–75.
interacted with the BFP would provide Cvijikj, I.P. and Michahelles, F. (2013) Online
engagement factors on Facebook brand pages. Social
more detailed insight into that customer’s Network Analysis and Mining 3(4): 843–861.
actual usage. An additional avenue for fur- De Choudhury, M., Sundaram, H., John, A., Seligmann,
ther research could be to examine the D. and Kelliher, A. (2010) Birds of a feather: Does
user homophily impact information diffusion in
moderating effect of usage intensity on the social media? arXiv preprint, arXiv:1006.1702,
research model to determine whether dif- http://arxiv.org/abs/1006.1702.
ferences exist in the research model under De Vries, L., Gensler, S. and Leeflang, P.S. (2012)
these behavioural conditions. Popularity of brand posts on brand fan pages: An
investigation of the effects of social media
Finally, given that the number of Amer- marketing. Journal of Interactive Marketing 26(2):
ican social media users accessing social net- 83–91.
working sites via mobile is expected to Dholakia, U.M. and Durham, E. (2010) One café
chain’s facebook experiment. Harvard Business Review
exceed 200 million by 2018 (Forrester 88(3): 26.
Research, 2011), future studies should also Dolich, I.J. (1969) Congruence relationships between
explore the role of mobile devices in how self images and product brands. Journal of Marketing
Research 6(1): 80–84.
consumers engage with BFP and the impli- Evrard, Y. and Aurier, P. (1996) Identification and
cations this has for brand management, parti- validation of the components of the person-object
cularly the possibility of real-time brand relationship. Journal of Business Research 37(2):
interactions taking place in the physical 127–134.
Facebook (2013) Facebook, https://www.facebook.
environment (service experiences, in-store com/facebook.
product trials in retail stores) influencing par- Falk, R.F. and Miller, N.B. (1992) A Primer for Soft
ticipation and CE with BFP. Understanding Modelling. Akron, OH: University of Akron Press.
Fornell, C. and Bookstein, F.L. (1982) Two structural
of this issue is becoming of strategic impor- equation models: LISREL and PLS applied to
tance since the share of time spent on mobile consumer exit-voice theory. Journal of Marketing
social networking continues to increase. Research. 19(4): 440–452.
Fornell, C. and Larcker, D. (1981) Evaluating structural
equation models with unobservable variables and
measurement error. Journal of Marketing Research
REFERENCES 18(1): 39–50.
Aaker, J.L. (1999) The malleable self: The role of self- Forrester Research (2011) U.S. Interactive Marketing
expression in persuasion. Journal of Marketing Research Forecast, 2011 to 2016, http://www.forrester.com/
1(36): 45–57. US+Interactive+Marketing+Forecast+2011+To
Balabanis, G. and Reynolds, N. (2001) The impact of +2016/fulltext/-/E-RES59379, accessed 8 October
involvement, brand attitude, internet knowledge, 2013.
and visit duration on consumers’ attitudes towards Gensler, S., Völckner, F., Liu-Thompkins, Y. and
electronic shopping web sites. The Journal of Business Wiertz, C. (2013) Managing brands in the social
Strategy 18(2): 1–22. media environment. Journal of Interactive Marketing
Brodie, R.J., Hollebeek, L.D., Juric, B. and Ilic, A. 27(4): 242–256.
(2011) Customer engagement: Conceptual Götz, O., Liehr-Gobbers, K. and Krafft, M. (2010)
domain, fundamental propositions, and impli- Evaluation of structural equation models using the
cations for research. Journal of Service Research partial least squares (pls) approach. In: V. Esposito
14(3): 252–271. Vinzi, W.W. Chin, J. Hensler and H. Wang (eds.)

© 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515 513
 De Vries and Carlson

Handbook of Partial Least Squares, 1st edn. Berlin and Evolution in the digital age. Journal of Interactive
Heidelberg, Germany: Springer-Verlag. Marketing 4(27): 257–269.
Hair, J.F., Hult, G.T.M., Ringle, C.M. and Sarstedt, M. Libai, B. (2010) Customer-to-customer interactions:
(2014) A Primer on Partial Least Squares Structural Broadening the scope of word of mouth research.
Equation Modeling (PLS-SEM). Los Angeles, CA: Journal of Service Research 13(3): 267–282.
Sage Publications. Loureiro, S., Ruediger, K. and Demetris, V. (2012)
Hatch, M. and Schultz, M. (2010) Toward a theory Brand emotional connection and loyalty. Journal of
of brand co-creation with implications for brand Brand Management 20(1): 13–27.
governance. Journal of Brand Management 17(8): Luo, M.M., Chea, S. and Chen, J.S. (2011) Web-based
590–604. information service adoption: A comparison of the
Hennig-Thurau, T. et al. (2010) The impact of new motivational model and the uses and gratifications
media on customer relationships. Journal of Service theory. Decision Support Systems 51(1): 21–30.
Research 13(3): 311–330. Madupu, V. and Cooley, D.O. (2010) Antecedents and
Hohenstein, N., Sirgy, M.J., Herrmann, A. and consequences of online brand community
Heitmann, M. (2007) Self-congruity: Antecedents participation: A conceptual framework. Journal of
and consequences. In: D. Merunka (ed.) Proceedings of Internet Commerce 9(2): 127–147.
the 34th La Londe International Research Conference in Malhotra, N. (2010) Marketing Research: An Applied
Marketing Communications and Consumer Behaviour. Orientation, 6th edn. Upper Saddle River, NJ:
France: France University Paul Cezanne Aix en Pearson Education New Jersey.
Province, pp. 118-130. Mollen, A. and Wilson, H. (2010) Engagement,
Hollebeek, L. (2011a) Exploring customer brand telepresence and interactivity in online consumer
engagement: Definitions and themes. Journal of experience: Reconciling scholastic and managerial
Strategic Marketing 19(7): 555–573. perspectives. Journal of Business Research 63(9–10):
Hollebeek, L. (2011b) Demystifying customer brand 919–925.
engagement: Exploring the loyalty nexus. Journal of O’Cass, A. and Carlson, J. (2012) An empirical assessment
Marketing Management 27(7–8): 785–807. of consumers’ evaluations of web site service quality:
Hollebeek, L., Glynn, M. and Brodie, R. (2014) Conceptualizing and testing a formative model.
Consumer brand engagement in social media: Journal of Services Marketing 26(6): 419–434.
Conceptualization, scale development and O’Cass, A. and Ngo, L. (2011) Examining the firm’s
validation. Journal of Interactive Marketing, Article (in value creation process: A managerial perspec-
press). tive of the firm’s value offering strategy and
Hutter, K., Hautz, J., Dennhardt, S. and Fuller, J. (2013) performance. British Journal of Management 22(4):
The impact of user interactions in social media on 646–671.
brand awareness and purchase intention: The case of Olsen, S.O. (2007) Repurchase loyalty: The role of
MINI on Facebook. Journal of Product & Brand involvement and satisfaction. Psychology & Marketing
Management 5/6(22): 342–351. 24(4): 315–341.
Jahn, B. and Kunz, W. (2012) How to transform Prahalad, C.K. and Ramaswamy, V. (2004) Co-creation
consumers into fans of your brand. Journal of Service experiences: The next practice in value creation.
Management 23(3): 344–361. Journal of Interactive Marketing 18(3): 5–14.
Jamal, A. and Goode, M.M.H. (2001) Consumers and Richard, M. and Chandra, R. (2005) A model of
brands: A study of the impact of self-image consumer web navigational behavior: Conceptual
congruence on brand preference and satisfaction. development and application. Journal of Business
Marketing Intelligence & Planning 19(7): 482–492. Research 55(8): 1019–1029.
Kang, Y.S., Hong, S. and Lee, H. (2009) Exploring Ringle, C.M., Wende, S. and Will, A. (2005) SmartPLS
continued online service usage behavior: The roles of (Version 2.0 (beta)). Hamburg, Germany: University
self-image congruity and regret. Computers in Human of Hamburg.
Behavior 1(25): 111–122. Sashi, C.M. (2012) Customer engagement: Buyer seller
Kleijnen, M., De Ruyter, K. and Andreassen, T.W. relationships, and social media. Management Decision
(2005) Image congruence and the adoption of 50(2): 253–272.
service innovations. Journal of Service Research 7(4): Sawhney, M., Verona, G. and Prandelli, E. (2005)
343–359. Collaborating to create: The internet as a platform
Kozinets, R.V., De Valck, K., Wojnicki, A.C. and for customer engagement in product innovation.
Wilner, S.J. (2010) Networked narratives: Journal of Interactive Marketing 19(4): 4–17.
Understanding word-of-mouth marketing in online Singh, S. and Sonnenburg, S. (2012) Brand performances
communities. Journal of Marketing 74(2): 71–89. in social media. Journal of Interactive Marketing 26(4):
Kressmann, F., Sirgy, M.J., Herrmann, A., Huber, F., 189–197.
Huber, S. and Lee, D.J. (2006) Direct and indirect Sirgy, M.J. (1982) Self-concept in consumer behavior:
effects of self-image congruence on brand loyalty. A critical review. Journal of consumer research 9(3):
Journal of Business Research 59(9): 955–964. 287–300.
Labrecque, I., vor dem Esche, J., Mathwick, C., Novak, Sirgy, M.J., Grewal, D. and Mangleburg, T. (2000)
T.P. and Hofacker, C.F. (2013) Consumer power: Retail environment, self-congruity, and retail

514 © 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515
 Customer engagement, brands and social media

patronage: An integrative model and a research Vock, M., van Dolen, W. and de Ruyter, K. (2013)
agenda. Journal of Business Research 49(2): 127–138. Understanding willingness to pay for social network
Taylor, D., Strutton, D. and Thompson, K. (2012) sites. Journal of Service Research 16(3): 311–325.
Self-enhancement as a motivation for sharing Wright, D., Aquilino, W.S. and Supple, A.J. (1998) A
online advertising. Journal of Interactive Advertising comparison of computer-assisted and paper-and-
12(2): 13–28. pencil self-administered questionnaires in a survey
van Doorn, J. et al. (2010) Customer engagement on smoking, alcohol, and drug use. Public Opinion
behavior: Theoretical foundations and research Quarterly 62(3): 331–353.
directions. Journal of Service Research 13(3): 253–266. Wyllie, J., Carlson, J. and Rosenberger III P.J. (2014)
Verleye, K., Gemmel, P. and Rangarajan, D. (2014) Examining the influence of different levels of
Managing engagement behaviors in a network of sexual-stimuli intensity by gender on advertising
customers and stakeholders evidence from the effectiveness. Journal of Marketing Management
nursing home sector. Journal of Service Research 17(1): 30(7–8): 1–22.
68–84. Yan, J. (2011) Social media in branding: Fulfilling a need.
Vivek, S.D., Beatty, S.E. and Morgan, R.M. (2012) Journal of Brand Management 18(9): 688–696.
Customer engagement: Exploring customer relation- Zeithaml, V., Berry, L. and Parasuraman, A. (1996) The
ships beyond purchase. Journal of Marketing Theory behavioral consequences of service quality. Journal of
and Practice 20(2): 127–145. Marketing 60(2): 2.

© 2014 Macmillan Publishers Ltd. 1350-231X Journal of Brand Management Vol. 21, 6, 495–515 515