Embed Size (px)
Citation preview
lable at ScienceDirect
Tourism Management 59 (2017) 597e609
Contents lists avai
Tourism Management
journal homepage: www.elsevier .com/locate/ tourman
Customer engagement with tourism social media brands
Paul Harrigan a, Uwana Evers a, Morgan Miles b, *, Timothy Daly c
a The University of Western Australia, UWA Business School M263, 35 Stirling Highway, Crawley, WA 6009, Australiab University of Canterbury, Department of Management, Marketing and Entrepreneurship, Private Bag 4800, Christchurch, New Zealandc United Arab Emirates University, Business Administration Department, PO BOX 15551, Al Ain, Saudi Arabia
a r t i c l e i n f o
Article history:Received 29 April 2016Received in revised form18 September 2016Accepted 19 September 2016
Keywords:Customer engagementSocial mediaTourismBrand loyaltyCustomer engagement scale
* Corresponding author.E-mail addresses: [email protected] (P. H
edu.au (U. Evers), [email protected] (M(T. Daly).
http://dx.doi.org/10.1016/j.tourman.2016.09.0150261-5177/© 2016 Elsevier Ltd. All rights reserved.
a b s t r a c t
In tourism, customer engagement has been found to boost loyalty, trust and brand evaluations. Customerengagement is facilitated by social media, but neither of these phenomena is well-researched in tourism.This research contributes in two ways. First, we validate the Customer Engagement with Tourism Brands(CETB) 25-item scale proposed by So, King & Sparks (2014) in a social media context, and offer analternative three-factor 11-item version of the scale. Second, we replicate their proposed structuralmodel, and test our alternative model, to predict the behavioural intention of loyalty from engagement,and to test customer involvement as an antecedent to engagement. Ultimately, we propose a customerengagement scale and a nomological framework for customer engagement, both of which can be appliedin both tourism and non-tourism contexts. Managers of tourism brands on social to better assess thenature of customer engagement with the parsimonious 11-item scale.
© 2016 Elsevier Ltd. All rights reserved.
1. Introduction
Customer engagement has gained much attention in the recentliterature. This is due to engagement being linked with numerousimportant brand performance indicators including sales growth,customer involvement in product development, customer feed-back, and referrals (Bijmolt et al., 2010; Bowden, 2009; Kumar et al.,2010; Nambisan & Baron, 2007; Sawhney, Verona, & Prandelli,2005; Van Doorn et al., 2010). Not surprisingly, much of thisbrand engagement occurs online through social media (Malthouse& Hofacker, 2010). Customers engaged with brand communitiesonline feel more connected to their brands, trust their preferredbrands more, are more committed to their chosen brands, havehigher brand satisfaction, and are more brand loyal (Brodie, Ili�c,Juri�c, & Hollebeek, 2013; Jahn & Kunz, 2012).
In the tourism context, customer engagement has been found toboost loyalty, trust and brand evaluations (So, King& Sparks, 2014).
arrigan), uwana.evers@uwa.. Miles), [email protected]
Social media facilitate customer engagement, but neither of thesephenomena are well researched in the tourism context. This hasresulted in a need for practical social media recommendations fortourism organizations (Cabiddu, Carlo, & Piccoli, 2014; Hudson,Roth, Madden, & Hudson, 2015; Mistilis & Gretzel, 2013). Socialmedia use is high among tourism organizations, particularly Face-book and Twitter (Leung, Bai, & Stahura, 2015); Instagram andother social media like TripAdvisor, Airbnb and Booking.com aregrowing in popularity and influence (Cabiddu et al., 2014; Filieri,2014; Munar & Jacobsen, 2014). TripAdvisor is the world's largesttravel review company and turned over $1.246 billion in 2014, up32 percent from the previous year (Forbes, 2015).
The goal of this research is to investigate the nature of customerengagement with tourism social media brands. We contribute tothe tourism literature in two ways. First, we test the CustomerEngagement with Tourism Brands (CETB) scale proposed by So et al.(2014) in a social media context. Further, we offer a psychometri-cally sound, concise eleven-item version of the scale. The socialmedia context is very different to the offline hospitality brands(hotels and airlines) context in which the CETB scale was originallydeveloped. Social media are driving fundamental business change,where they enable interactive, two-way communications betweencustomers and organizations (Dijkmans, Kerkhof, & Beukeboom,
P. Harrigan et al. / Tourism Management 59 (2017) 597e609598
2015; Hollebeek, Glynn, & Brodie, 2014; Lee & Choi, 2014; Vivek,Beatty, & Morgan, 2012). Social media allow customers tocomment, review, create and share content across online networks.They also allow customers direct access to organizations, brandsand marketers (Chau & Xu, 2012). This creates challenges and op-portunities for marketers, where they must engage with customersin real-time and manage the significant amounts of incomingcustomer data (Cui, Lui, & Guo, 2012; Hennig-Thurau et al., 2010;Pagani&Mirabello, 2012). The disruptive nature of the social mediacontext means that it is valuable to test if a scale developed tomeasure engagement in an offline context performs similarly onsocial media.
Second, we empirically replicate So et al.'s (2014) proposedconceptual model and test the nomological framework that in-corporates customer involvement as an antecedent of customerengagement and behavioural intention of loyalty as an outcome ofcustomer engagement. The next sections in this paper discuss theconceptualisation of customer engagement, its dimensions, andpossible antecedents and consequences. The method, a survey ofU.S. consumers, is outlined before the presentation of the scalevalidation and model testing results. Finally, there is a discussionaround the implications for theory and practice.
2. Literature
Customer engagement is characterised by repeated interactionsbetween a customer and an organization that strengthen theemotional, psychological or physical investment a customer has inthe brand and the organization (Hollebeek et al., 2014; Phang,Zhang, & Sutanto, 2013). Social exchange theory underpins thisnotion of investment, which holds that individuals evaluate thetangible and intangible costs and benefits of engaging in relation-ships (Thibaut & Kelley, 1959). For customer-brand engagement topersist, customersmust at least achieve a balance in these costs andbenefits over time (Brodie, Hollebeek, Juri�c, & Ili�c et al., 2011;Hollebeek, 2011). For example, consumers may invest enthusiasmand attention in engaging with a brand to receive benefits such asproduct news, offers, through to a sense of belonging (Blau, 1964;Foa & Foa, 1980).
Social media are the dominant enablers of customer engage-ment, and these technologies are very different from previousmarketer-customer technology platforms. They are owned by thecustomer but are transparent, and facilitate two-way interactionsbetween customers and organizations (e.g. Deighton & Kornfeld,2009; Dwyer, 2007; Hennig-Thurau et al., 2010; Vivek et al.,2012). Goh, Heng, and Lin's (2013) finding that engaged customers'messages were 22 times more valuable than those of marketersunderlines the importance of understanding customer engage-ment. Social media are defined as the ‘group of Internet-based ap-plications that build on the ideological and technological foundationsof Web 2.0 and that allow the creation and exchange of User-Generated Content’ (Kaplan & Haenlein, 2010, p. 61). This defini-tion means that tourism sites like TripAdvisor, Booking.com,Airbnb, and Lonely Planet are considered as social media (Cabidduet al., 2014; Munar & Jacobsen, 2014). They allow customers tocomment, review, spread and even create content online that noweven appears in search engine results. The importance of socialmedia as a means for customer engagement within the tourismindustry cannot be ignored (Cabiddu et al., 2014; Cheng& Edwards,2015; Dijkmans et al., 2015; Hudson et al., 2015; Munar& Jacobsen,2014).
Customer engagement has been conceptualised in differentways (see Table 1). The majority of customer engagement research
has been based on a multidimensional conceptualisation, encom-passing some form of cognitive, emotional, and behavioural com-ponents (Bowden, 2009; Brodie et al. 2013; Cheung, Lee,& Jin, 2011,pp. 1e8; Dessart, Veloutsou, & Morgan-Thomas, 2015; Dwivedi,2015; Hollebeek, 2011; Hollebeek et al., 2014; Patterson, Yu, & DeRuyter, 2006; So et al., 2014). The broader conceptualisation ofcustomer engagement behaviours proposed by Van Doorn et al.(2010) encompasses valence, form, scope, impact of engagement,and the customers' goals. All of these conceptualisations assert thatcustomer engagement is discriminately different from involve-ment, a construct with which it is frequently compared. As So et al.(2014) state, involvement tends to be limited to a cognitivecomponent, whereas engagement incorporates cognitive,emotional, and behavioural components (Hollebeek, 2011; Mollen& Wilson, 2010; Vivek et al., 2012).
A recent analysis of customer engagement dimensionalityconcluded that customer engagement is a multi-dimensionalconstruct consisting of three dimensions: cognitive (customerfocus and interest in a particular brand), emotional (feelings ofinspiration or pride caused by a particular brand) and behavioural(customer effort and energy necessary for interaction with aparticular brand) (Kuvykait _e & Tarute, 2015). Conceptualisations ofengagement that do not explicitly refer to underlying cognitive,affective, and behavioural components are still likely to encompassthese dimensions. The proposed dimensions of customer engage-ment with tourism brands (So et al., 2014) and Dwivedi's (2015)customer brand engagement conceptualisation can be mapped,largely, onto the customer brand engagement dimensions offeredby Hollebeek et al. (2014) (Table 2). Interaction is similar to Acti-vation and Vigor, representing the behavioural component ofcustomer engagement; Identification relates to Affection andDedication as the emotional component of customer brandengagement, while Attention and Absorption, the cognitivecomponent. The definitions of the Absorption, Enthusiasm andAttention dimensions (So et al., 2014) have both affective andcognitive elements.
2.1. Dimensions of customer engagement
This research builds on So et al.'s (2014) conceptualisation ofcustomer engagement, which incorporates five dimensions, iden-tification, enthusiasm, attention, absorption, interaction, andidentification. As So et al. (2014) undertake a comprehensive dis-cussion around these dimensions, the purpose of this paper is bestserved by briefly introducing each dimension.
2.1.1. EnthusiasmEnthusiasm represents an individual's “strong level of excite-
ment or zeal” and interest in a brand (Vivek, 2009, p. 60). So et al.(2014, p. 308) note that the dimension of enthusiasm “representsan individual's strong level of excitement and interest regarding thefocus of engagement … and differentiate the construct of engage-ment from other similar constructs such as satisfaction.”
2.1.2. AttentionAttention refers to a customer's level of focus, consciously or
sub-consciously, on the brand. Persistent attention towards a brandis likely to lead to higher levels of engagement (Lin, Gregor, &Ewing, 2008; Scholer & Higgins, 2009).
2.1.3. AbsorptionAbsorption goes further than attention, where it refers to a
customer's high level of concentration and engrossment in a brand
Table 1Examples of conceptualisations of customer engagement.
Construct Dimensions of engagement Author(s)
Customer brand engagement (1) Identification (2) Enthusiasm (3) Attention (4) Absorption (5) Interaction So, Sparks & King (2014)Consumer brand engagement (1) Cognitive (2) Emotional (3) Behavioural Hollebeek (2011)Consumer brand engagement (1) Cognitive processing (cognitive) (2) Affection (emotional) (3) Activation (behavioural) Hollebeek et al. (2014)Consumer brand engagement (1) Cognitive (2) Emotional (3) Behavioural Dessart et al. (2015)Consumer brand engagement (1) Vigor (2) Dedication (3) Absorption Dwivedi (2015)Consumer engagement (1) Cognitive (2) Emotional (3) Behavioural Brodie et al. (2013)Customer engagement (1) Vigor (2) Dedication (3) Absorption (4) Interaction Patterson et al. (2006)Customer engagement (1) Vigor (2) Dedication (3) Absorption Cheung et al. (2011)Customer engagement (1) Trust (2) Dedication (3) Reputation Enginkaya and Esen (2014)Customer engagement (process of) (1) Involvement (behavioural) (2) Commitment (cognitive and affective) Bowden (2009)
Table 2Merging conceptualisations of customer engagement.
Customer engagement with tourism brandsSo et al. (2014)
Consumer brand engagementDwivedi (2015)
Consumer brand engagementHollebeek et al. (2014)
Dimension Definition Dimension Dimension Definition
Behavioural Interaction “Various participation (bothonline and offline) that acustomer has with a brandorganization or othercustomers outside ofpurchase” (p.311)
Vigor “Vigor denotes high levelsof energy and mentalresilience when interactingwith a brand, and theconsumer willingness andthe ability to invest effort insuch interactions” (p. 100)
Activation “A consumer's level ofenergy, effort and timespent on a brand in aparticular consumer/brandinteraction” (p. 154)
Emotional Identification “The degree of a consumer'sperceived oneness with orbelongingness to thebrand” (p.311)
Cognitive Absorption “A pleasant state whichdescribes the customer asbeing fully concentrated,happy and deeplyengrossed while playingthe role as a consumer ofthe brand” (p. 311)
Dedication “In the context ofconsumer-brandrelationships … dedicationrefers to a sense ofsignificance, enthusiasm,inspiration, pride andchallenge” (p. 100)
Affection “A consumer's degree ofpositive brand-relatedaffect in a particularconsumer/brandinteraction” (p.154)
Enthusiasm “The degree of excitementand interest that aconsumer has in the brand”(p. 311)
Attention “The degree ofattentiveness, focus andconnection that a consumerhas with a brand” (p. 311)
Absorption “Absorption refers to thesense of being fullyconcentrated and happilyengrossed in brandinteractions and in whichtime passes quickly” (p.101)
Cognitive Processing “A consumer's level ofbrand-related thoughtprocessing and elaborationin a particular consumer-brand interaction” (p. 154)
P. Harrigan et al. / Tourism Management 59 (2017) 597e609 599
(Csikszentmihalyi, 1990; Schaufeli, Salanova, Gonzalez-Roma, &Bakker, 2002). Absorption is a positive trait, where customers willbe contently absorbed in or with the brand, most likely unaware ofhow much time they are devoting to the brand (Patterson et al.,2006; Scholer & Higgins, 2009).
2.1.4. InteractionInteraction is fundamental to customer engagement, and in-
volves sharing and exchanging ideas, thoughts, and feelings aboutexperiences with the brand and other customers of the brand(Vivek, 2009). Interaction between customers of the brand is sup-ported by the brand community literature (e.g. Muniz & O'Guinn,2001). This interaction, as well as direct brand interaction, is abehavioural element of customer engagement.
2.1.5. IdentificationCustomers will identify more with certain brands over others,
particularly with those that match their self-image (Bagozzi &
Dholakia, 2006). This notion draws on social identity theory,where individuals have both a personal identity and a socialidentity. The groups one is a member of, in this context thebrands with which one engages, are a manifestation of thebrand's social identity function (Mael & Ashforth, 1992; Tajfel &Turner, 1985).
These five dimensions of customer engagement are readilyapplicable to tourism brands on social media. Where social mediaare generally a powerful enabler of customer engagement, it fol-lows that tourism social media brands like TripAdvisor, Booking.com, Airbnb, and Lonely Planet will seek to inspire customerengagement in each of the five dimensions. Following So et al.(2014), we treat customer engagement as a second-order, reflec-tive construct. Covin andWales (2012, p. 682) note “in the reflectivemeasurement model the latent construct is modeled as producingits measures.” This means that the five dimensions above are likelyto be caused by customer engagement, and to be inter-correlated(Hair, Ringle, & Sarstedt, 2011).
Fig. 1. Conceptual model of customer engagement. Note. The latent factor labels represent the following: INV ¼ customer involvement with a tourism social media brand;CE ¼ customer engagement with a tourism social media brand; BIL ¼ behavioural intention of loyalty toward a tourism social media brand.
P. Harrigan et al. / Tourism Management 59 (2017) 597e609600
2.2. Customer engagement antecedents
As noted earlier in the paper, engagement is conceptuallydifferent from involvement. Based on this premise, it may be thatinvolvement is an antecedent to customer engagement (Hollebeeket al., 2014). This relationship was proposed by So et al. (2014) intheir conceptual model, that we test (Fig. 1).
Zaichkowsky (1994) defines customer involvement as a cus-tomer's perceived relevance of an object (brand) based on theirneeds, values, and interests. Zaichkowsky (1985, p. 342) states thatinvolvement is “independent of the behavior that results frominvolvement” and reflects an object's relevance in meeting a cus-tomer's value based needs. There are certain tourism brands onsocial media that may elicit higher involvement than others may,for different types of customers. For example, Airbnb's #mankindcampaign seeks to involve the consumer at a cognitive levelthrough focusing on positive news stories and the trust and kind-ness aspect of opening one's home to strangers. Other sites likeBooking.com or TripAdvisor, may elicit involvement throughdifferent strategies, such as ease of use, soliciting reviews, trans-parency or social influence.
2.3. Customer engagement consequences
Building on So et al. (2014), we test the relationship proposedbetween customer engagement and behavioural intention of loy-alty (BIL), which is a widely used outcome variable. BIL, as oper-ationalised by Zeithaml, Berry, and Parasuraman (1996), measures acustomer's intention to say positive things about a brand, torecommend a brand generally and to friends, and to purchase thisbrand in the near future. As illustrated in Fig. 1 customer involve-ment with a tourism social media brand is an antecedent ofcustomer engagement; while customer's behavioural intention ofloyalty is a consequence of customer engagement.
It is important, theoretically and managerially, that customerengagement is not treated as an outcome but rather a process thatleads to more measurable outcomes such as customer satisfactionor loyalty. For example, Hudson et al. (2015) considered word-of-mouth as an outcome in their research on social media interac-tion. For engagement, there is evidence to support that it is a pre-dictor of loyalty (Bowden, 2009; Hollebeek, 2011; Patterson et al.,2006). For tourism social media brands, such as Lonely Planet,Travelocity or Expedia, the relationship between engagement andloyalty is essential. The fragmentation of the tourism market,particularly online, has led to hyper-competition. In turn, this hasled to these brands using social media to try to increase customerengagement, at cognitive, affective and behavioural levels, with theprincipal aim of encouraging higher customer retention.
This paper contributes to tourism research, first, by validating acustomer engagement scale previously developed by So et al.(2014). This scale consists of five dimensions, identification,enthusiasm, attention, absorption, and interaction. We take these
dimensions, and the scale as whole, and test it in a social mediacontext among U.S. customers. Second, we develop a concise 11-item version of the So et al. (2014) scale. Third, we examine thebehavioural loyalty intention as a consequence of customerengagement (Bowden, 2009; Brodie, Hollebeek, Juri�c, & Ili�c, 2011;De Villiers, 2015; De Vries & Carlson, 2014; Dwivedi, 2015;Hollebeek, 2011; So et al., 2014). Fourth, we also examinecustomer involvement as an antecedent to customer engagement(Hollebeek et al., 2014; So et al., 2014).
3. Method
This study extends recent work by Evers, Harrigan, and Daly(2015) and uses the dataset developed for that study. The datawere gathered from Amazon Mechanical Turk (MTurk) market-place during the first half of 2015 using an online survey aboutsocial media use in tourism-related decisions.
There were three constructs of interest. First, consumerengagement with tourism brands was measured using the original25- item So et al. (2014) CETB scale. This scale included five sub-scales, namely (1) identification; (2) enthusiasm; (3) attention;(4) absorption; and (5) interaction (items in Table 5). Second,behavioural intention of loyalty was measured using the 4-itemZeithaml et al. (1996) behavioural intention of loyalty (BIL) scale(items reported in Table 3). Third, customer involvement wasmeasured using Zaichkowsky's (1994) 10-item scale (items re-ported in Table 3). Finally, basic demographics were collected.
MTurk is Amazon's crowdsourcing employment website whereanonymous participants find and complete tasks (HITs) posted byemployers. Studies by Buhrmester, Kwang, and Gosling (2011) andPaolacci, Chandler, and Ipeirotis (2010) found that MTurk userswere as representative of the population as online panels. In orderto reduce potential fatigue effects, the current survey was split intotwo 15-min surveys. Using the method outlined by Daly andNataraajan (2015) the 300 respondents who completed the firstsurvey were invited to complete the second approximately twoweeks later. In total 195 respondents completed both surveys,resulting in an attrition rate of 35%. Respondents were compen-sated $1.50 per completed survey. A HIT0 was posted on MTurk toinvite individuals registered on the U.S.-based site to participate intwo surveys over twoweeks. The surveywas separated into two 15-min phases due to length (Daly & Nataraajan, 2015). Respondentswere matched across both phases with 195 respondents completedboth survey phases (see Daly & Nataraajan, 2015). Three hundredcompleted the first phase but not the second phase, resulting in anattrition rate of 35%. Respondents were paid a monetary induce-ment of $1.50 USD. Only respondents located in the United Stateswere eligible to complete the surveys (this was verified by IP geo-location). Factor analysis of the customer engagement with tourismbrands scale and descriptive statistics were carried out using SPSS(22.0) and the structural equation modeling in AMOS (22.0).
Table 3Items used to measure behavioural intention of loyalty and customer involvement.
Items for consumer involvement (INV)(Adapted from Zaichkowsky, 1994)10 items presented on a 7-point semantic differential scale
Thinking about your favourite tourism site, please indicate your attitudes toward the site from the descriptive words below:INV1 Important e UnimportantINV2 Boring e InterestingINV3 Relevant e IrrelevantINV4 Exciting - UnexcitingINV5 Means nothing - Means a lot to meINV6 Appealing - UnappealingINV7 Fascinating - MundaneINV8 Worthless e ValuableINV9 Involving e UninvolvingINV10 Not needed e Needed
Items for Behavioural Intention of Loyalty (BIL)(adapted from Zeithaml et al., 1996)4 items presented on a 7-point Likert scale (strongly disagree to strongly agree)
Thinking of your favourite tourism site, indicate the extent to which you agree or disagree with the following statements:BIL1 I would say positive things about this tourism site to other people.BIL2 I would recommend this tourism site to someone who seeks my advice.BIL3 I would encourage friends and relatives to do business with this tourism site.BIL4 I would do more business with this tourism site in the next few years.
Table 4The regional dispersion of the sample compared with 2015 U.S. Census estimates.
Region % of sample in region % of U.S. households in regiona
Northeast 21.9% 17.5%Midwest 19.9% 21.1%West 15% 23.7%South 43.4% 37.7%
a www.census.gov/popclock/data_tables.php?component¼growth.
P. Harrigan et al. / Tourism Management 59 (2017) 597e609 601
4. Results
It is not possible to estimate a response rate from a samplingframe based on MTurk (Evers et al., 2015). The sample was splitevenly across the genders (50.8% male), and the average age of therespondents was 36 years (SD ¼ 11). The majority of respondentshad completed university education (67.8%) and currently workedoutside of the home (70.3%); the average annual household incomebracket was $50,000-$59,999 (Evers et al., 2015). In addition, thesample was also broadly representative of the regional geographicdispersion of the U.S. population (Table 4). Importantly, 42 out ofthe 50 U.S. states were represented in the final data set. Overallthese results suggest that the sampling should have limited expo-sure to geo-demographic based biases.
Questions for involvement, engagement, and behaviouralintention of loyalty were answered in relation to respondents'favourite tourism social media site; the top five favourite tourismbrand sites specified by respondents were TripAdvisor (29%,n ¼ 56), Expedia (19%, n ¼ 37), Priceline (14%, n ¼ 27), Kayak (9%,n ¼ 18), and Orbitz (9%, n ¼ 18). A one-way ANOVA confirmed thatthere were no differences in the level of brand engagement acrossthe nominated tourism brands (F (13, 181) ¼ 1.370, p ¼ 0.178).
4.1. Study 1. the customer engagement with tourism brands (CETB)factor structure
To identify the underlying factor structure of So et al.'s (2014)25-item customer engagement scale, data collected from 195 re-spondents were subjected to exploratory factor analysis (maximumlikelihood estimation); oblique rotation (promax) was chosen overorthogonal (varimax) because the factors were highly correlated
with one another (Tabachnick & Fidell, 2007), and it can betteridentify the ‘simple structure’ of factors (Finch, 2006). The datasatisfied the factor analysis assumptions; the Kaiser-Meyer-Olkinmeasure of sampling adequacy was ideal at 0.951, and Bartlett'stest of sphericity was significant (c2 (300) ¼ 6048.58, p < 0.001).Although So et al. (2014) proposed a five-factor structure, the cur-rent analysis only supported four-factors (based on eigenvaluesgreater than 1). Our subsequent analyses proceeded in three steps:first, we present the analysis with five-factors as per So et al. (2014),second we test a four-factor structure to better fit the data, andthird, following further analysis, we propose a three-factor 11-itemscale that best fits the data.
4.2. The five-factor CETB scale
We initially extracted five factors from the data, though the lastfactor had an eigenvalue of only 0.701. The interaction factor aloneaccounted for 61.6% of variance, while the remaining four factorsaccounted for an additional 22% of variance. In total, the five factorsaccounted for 83.8% of variation in the scale data.
The scale items loaded onto separate factors as indicated by theoriginal sub-scales. The loadings onto each factor ranged as follows:identification (0.755-0.861), enthusiasm (0.544e1.015), attention(0.695-0.935), absorption (0.505-0.869), and interaction (0.874-0.981).
The analysis also supported the reliability and validity of theoriginal CETB scale. The reliability of all factor scales was examinedby internal consistency analyses; the Cronbach's alpha for inter-action (0.977), absorption (0.934), enthusiasm (0.951), identifica-tion (0.910), attention (0.951), and overall customer engagementwith brands (0.974) all indicated high internal consistency.Maximum shared variance (MSV) and average shared squaredvariance (ASV) were both lower than the average variance extrac-ted (AVE) for all factors demonstrating discriminant validity of thescale (Table 5).
4.3. The four-factor CETB scale
As a result of our initial analysis, we re-ran the exploratoryfactor analysis and extracted four factors from the data, each withan eigenvalue greater than 1. In this four-factor model, all
Table 5Confirmatory factor analysis of the five-factor customer engagement with tourism brands scale (So et al., 2014).
Factor and item description Model and item indices
SL CR SMC AVE MSV ASV
Identification 0.913 0.725 0.546 0.451ID1. When someone criticizes this tourism site, it feels like a personal insult. 0.804 0.647ID2. When I talk about this tourism site, I usually say ‘we’ rather than ‘they’. 0.826 0.683ID3. This tourism site's successes are my successes. 0.895 0.801ID4. When someone praises this tourism site, it feels like a personal compliment. 0.877 0.769
Enthusiasm 0.948 0.785 0.743 0.561EN1. I am heavily into this tourism site. 0.907 0.822EN2. I am passionate about this tourism site. 0.955 0.912EN3. I am enthusiastic about this tourism site. 0.881 0.777EN4. I feel excited about this tourism site. 0.913 0.833EN5. I love this tourism site. 0.763 0.582
Attention 0.948 0.786 0.743 0.563AT1. I like to learn more about this tourism site. 0.841 0.707AT2. I pay a lot of attention to anything about this tourism site. 0.929 0.863AT3. Anything related to this tourism site grabs my attention. 0.924 0.855AT4. I concentrate a lot on this tourism site. 0.890 0.793AT5. I like learning more about this tourism site. 0.844 0.712
Absorption 0.937 0.712 0.643 0.527AB1. When I am interacting with the tourism site, I forget everything else around me. 0.802 0.644AB2. Time flies when I am interacting with the tourism site. 0.896 0.804AB3. When I am interacting with the tourism site, I get carried away. 0.866 0.750AB4. When interacting with the tourism site, it is difficult to detach myself. 0.844 0.712AB5. In my interaction with the tourism site, I am immersed. 0.872 0.760AB6. When interacting with the tourism site intensely, I feel happy. 0.775 0.601
Interaction 0.977 0.893 0.487 0.436INT1. In general, I like to get involved in the tourism site community discussions. 0.957 0.916INT2. I am someone who enjoys interacting with like-minded others in the tourism site community. 0.948 0.899INT3. I am someone who likes actively participating in the tourism site community discussions. 0.974 0.948INT4. In general, I thoroughly enjoy exchanging ideas with other people in the tourism site community. 0.965 0.932INT5. I often participate in activities of the tourism site community. 0.879 0.772
Note. c2 ¼ 407.410 (p < 0.001, df ¼ 252) c2/df ¼ 1.617; goodness-of-fit index (GFI) ¼ 0.857; adjusted GFI ¼ 0.816; comparative fit index (CFI) ¼ 0.974; normed fit index(NFI) ¼ 0.936; root mean square error of approximation (RMSEA) ¼ 0.056; PCLOSE ¼ 0.147; SL ¼ standardised loadings; CR ¼ composite reliability; AVE ¼ average varianceextracted; SMC ¼ squared multiple correlation.
Table 6Confirmatory factor analysis of the proposed four-factor 20-item customer engagement with tourism brands scale.
Factor and item description Model and item indices
SL CR SMC AVE MSV ASV
Identification 0.922 0.748 0.452 0.415ID1. When someone criticizes this tourism site, it feels like a personal insult. 0.851 0.725ID2. When I talk about this tourism site, I usually say ‘we’ rather than ‘they’. 0.821 0.674ID3. This tourism site's successes are my successes. 0.922 0.850ID4. When someone praises this tourism site, it feels like a personal compliment. 0.862 0.744
Attraction 0.961 0.805 0.570 0.487EN2. I am passionate about this tourism site. 0.928 0.861EN3. I am enthusiastic about this tourism site. 0.896 0.803EN4. I feel excited about this tourism site. 0.935 0.875AT1. I like to learn more about this tourism site. 0.868 0.753AT2. I pay a lot of attention to anything about this tourism site. 0.906 0.820AT3. Anything related to this tourism site grabs my attention. 0.848 0.719
Absorption 0.932 0.734 0.570 0.487AB1. When I am interacting with the tourism site, I forget everything else around me. 0.813 0.661AB2. Time flies when I am interacting with the tourism site. 0.900 0.811AB3. When I am interacting with the tourism site, I get carried away. 0.870 0.756AB4. When interacting with the tourism site, it is difficult to detach myself. 0.836 0.700AB5. In my interaction with the tourism site, I am immersed. 0.863 0.745
Interaction 0.977 0.893 0.480 0.433INT1. In general, I like to get involved in the tourism site community discussions. 0.957 0.916INT2. I am someone who enjoys interacting with like-minded others in the tourism site community. 0.948 0.900INT3. I am someone who likes actively participating in the tourism site community discussions. 0.974 0.948INT4. In general, I thoroughly enjoy exchanging ideas with other people in the tourism site community. 0.965 0.932INT5. I often participate in activities of the tourism site community. 0.879 0.772
Note. c2 ¼ 194.834 (p ¼ 0.003, df ¼ 143) c2/df ¼ 1.362; goodness-of-fit index (GFI) ¼ 0.908; adjusted GFI ¼ 0.866; comparative fit index (CFI) ¼ 0.989; normed fit index(NFI) ¼ 0.960; root mean square error of approximation (RMSEA) ¼ 0.043; PCLOSE ¼ 0.765; SL ¼ standardised loadings; CR ¼ composite reliability; AVE ¼ average varianceextracted; SMC ¼ squared multiple correlation.
P. Harrigan et al. / Tourism Management 59 (2017) 597e609602
P. Harrigan et al. / Tourism Management 59 (2017) 597e609 603
Enthusiasm and Attention items, plus one Absorption item (AB6),loaded onto the same factor. This result suggested that the Enthu-siasm and Attention items measure the same underlying construct.This factor was renamed ‘Attraction’ to better represent the mea-surement items. The factor loadings for Attraction ranged from0.504 to 1.02. The other five Absorption items (0.792-0.867) loadedonto one factor as expected. All five Interaction items (0.879-0.991)and all four Identification items (0.745-0.878) loaded onto separatefactors.
Five items (EN1, EN5, AT4, AT5, AB6) were removed from themodel based upon inspection of the standardised loadings andmodel fit. The four factors of the 20-item scale accounted for acumulative 83.4% of variation in the data. Therefore, in terms ofexplained variance, the 20-item CETB scale performs equally as wellas the original 25-item scale. The reduced 20-item CETB scale alsodemonstrates good reliability and validity. The reliability of the fourfactors individually and the scale as a whole were examined byinternal consistency analyses; the Cronbach's alpha for interaction(0.977), absorption (0.931), identification (0.910), attraction(0.952), and overall customer engagement with brands (0.967) allindicated high internal consistency. A confirmatory factor analysiswas conducted in AMOS to examine the validity of the scale.Convergent validity was demonstrated by the average varianceextracted (AVE) exceeding 0.5 for all constructs (Fornell & Larcker,1981; Hair, Black, Babin, & Anderson, 2010). Further, the maximumshared variance (MSV) and average shared squared variance (ASV)were both lower than the AVE for all factors demonstratingdiscriminant validity of the scale (Gaskin, 2016; Hair et al., 2010)(Table 6).
4.4. The three-factor CETB scale
High inter-item correlations in the 20-item scale suggestedpotential item redundancy, and the factor loadings pointed to anoverlap, whereby some items were loading onto multiple factors(see Streiner, 2003). We therefore removed the highly correlateditems (ID3, EN3, AT1, AT2, AB2, AB3, AB4, INT3, INT4) to reduceredundancy and move toward a more parsimonious version of theCETB scale. We re-ran the factor analysis on the remaining 11 itemsand discovered that the items were loading onto three factors.Identification and Interaction remained independent factors, whilethe items from the CETB's three factors Absorption, Enthusiasm,and Attention all collapsed into one factor measuring a singleconstruct - Absorption. The three factors of the 11-item scale
Table 7Confirmatory factor analysis of the proposed three-factor 11-item customer engagement
Factor and item description
IdentificationID1. When someone criticizes this tourism site, it feels like a personal insult.ID2. When I talk about this tourism site, I usually say ‘we’ rather than ‘they’.ID4. When someone praises this tourism site, it feels like a personal compliment.
AbsorptionEN2. I am passionate about this tourism site.EN4. I feel excited about this tourism site.AT3. Anything related to this tourism site grabs my attention.AB1. When I am interacting with the tourism site, I forget everything else around mAB5. In my interaction with the tourism site, I am immersed.
InteractionINT1. In general, I like to get involved in the tourism site community discussions.INT2. I am someone who enjoys interacting with like-minded others in the tourismINT5. I often participate in activities of the tourism site community.
Note. c2 ¼ 172.101 (p ¼ 0.000, df ¼ 82) c2/df ¼ 2.10; goodness-of-fit index (GFI) ¼ 0.(NFI) ¼ 0.935; root mean square error of approximation (RMSEA) ¼ 0.075; PCLOSE ¼ 0.0extracted; SMC ¼ squared multiple correlation.
accounted for a cumulative 80.5% of variation in the data,explaining almost as much variance in the data as the original 25-item scale. The 11-item CETB scale also demonstrated reliability andvalidity. The Cronbach's alpha for interaction (0.948), absorption(0.906), identification (0.869), and overall customer engagementwith brands (0.936) all indicated high internal consistency. Lastly, aconfirmatory factor analysis examined the validity of the scale.Convergent validity was demonstrated by the average varianceextracted (AVE) exceeding 0.5 for all constructs (Fornell & Larcker,1981; Hair et al., 2010) and discriminant validity was shown as themaximum shared variance (MSV) and average shared squaredvariance (ASV) were both lower than the AVE for all factors (Gaskin,2016; Hair et al., 2010) (Table 7).
4.5. Study 2: an antecedent and consequence of customerengagement
We tested the predictive validity of the five-factor CE scale witha structural model, placing a path from CE to BIL (Fig. 2). The fitindices suggested that the model fit the data fairly well; (c2
(346) ¼ 559.375, p < 0.001), c2/df ¼ 1.617, GFI ¼ 0.829;AGFI ¼ 0.785; CFI ¼ 0.968; NFI ¼ 0.922; RMSEA ¼ 0.056;PCLOSE¼ 0.111. The results suggest that original 25-item CETB scaleis a significant predictor of BIL (b ¼ 0.648, CR ¼ 7.668, p < 0.001)and explained 42% of the variance in BIL.
We tested the predictive validity of the 20-item CETB scale witha structural model, placing a path from CE to BIL (Fig. 3). The fitindices suggested that the model fit the data well; (c2
(224) ¼ 323.499, p < 0.001), c2/df ¼ 1.444, GFI ¼ 0.874;AGFI ¼ 0.832; CFI ¼ 0.982; NFI ¼ 0.943; RMSEA ¼ 0.048;PCLOSE¼ 0.611. The results suggest that the four-factor CE scale is asignificant predictor of BIL (b ¼ 0.625, CR ¼ 7.493, p < 0.001) andexplained 39% of the variance in BIL.
The final examinationwas of the predictive ability of the 11-itemthree-factor CETB scale. As with the previous two structuralmodels, we tested the three-factor model by placing a path from CEto BIL (Fig. 4). The fit indices suggested that the model fit the datamoderately well; (c2 (82) ¼ 172.101, p < 0.001), c2/df ¼ 2.10,GFI ¼ 0.890; AGFI ¼ 0.839; CFI ¼ 0.965; NFI ¼ 0.935;RMSEA ¼ 0.075; PCLOSE ¼ 0.005. The results suggest that thethree-factor CE scale is a significant predictor of BIL (b ¼ 0.635,CR ¼ 7.364, p < 0.001) and explained 40% of the variance in BIL.
with tourism brands scale.
Model and item indices
SL CR SMC AVE MSV ASV
0.874 0.697 0.569 0.5010.832 0.6960.809 0.6400.878 0.757
0.906 0.663 0.569 0.5300.920 0.8500.930 0.8430.858 0.712
e. 0.642 0.4360.675 0.472
0.948 0.860 0.491 0.4620.949 0.904
site community. 0.956 0.9090.874 0.767
890; adjusted GFI ¼ 0.839; comparative fit index (CFI) ¼ 0.965; normed fit index05; SL ¼ standardised loadings; CR ¼ composite reliability; AVE ¼ average variance
Fig. 2. The five-factor model of CE structural model with behavioural intention of loyalty. Note. The latent factor labels represent the following: INT ¼ interaction; AB ¼ absorption;EN ¼ enthusiasm; ID ¼ identification; AT ¼ attention; CE ¼ customer engagement; BIL ¼ behavioural intention of loyalty.
P. Harrigan et al. / Tourism Management 59 (2017) 597e609604
4.6. A test of the conceptual model
Finally, we examined the initial conceptual model, wherecustomer involvement predicts customer engagement, which inturn, predicts behavioural intentions of loyalty. We tested thepredictive validity of the proposed three-factor 11-item CETB scalewith a structural model, placing paths from INV to CE and CE to BIL(Fig. 4). The fit indices suggested that the model fit the datamoderately well; (c2 (248) ¼ 390.581, p < 0.001), c2/df ¼ 1.575,GFI ¼ 0.859; AGFI ¼ 0.815; CFI ¼ 0.962; NFI ¼ 0.903;RMSEA¼ 0.054; PCLOSE¼ 0.234. The results suggest that customerinvolvement can predict customer engagement (b ¼ 0.694,CR ¼ 6.781, p < 0.001), accounting for 48% of the variance in CE.Importantly, the three-factor CE scale remained a significant pre-dictor of BIL (b¼ 0.663, CR¼ 7.505, p < 0.001) and explained 44% of
the variance in BIL when testing the relationship between customerinvolvement, customer engagement and loyalty proposed in Soet al.'s (2014) nomological framework (see Fig. 5).
5. Discussion
The first objective of our research was to examine and validatethe CETB scale proposed by So et al. (2014). The initial analysesfound that the original scale has four instead of five underlyingfactors. In this phase of our study, the original items for the factorsEnthusiasm and Attention loaded together onto the same factor.Therefore, for So et al.'s (2014) original CETB scale it is proposedthese factors be merged into one factor named Attraction. Theresults demonstrate that the proposed four-factor, twenty-itemscale has better structural model fit than the original five-factor
Fig. 3. The four-factor model of CE structural model with behavioural intention of loyalty, Note. The latent factor labels represent the following: ID ¼ identification; ATT ¼ attraction;AB ¼ absorption; INT ¼ interaction; CE ¼ customer engagement; BIL ¼ behavioural intention of loyalty.
P. Harrigan et al. / Tourism Management 59 (2017) 597e609 605
twenty-five-item scale (So et al., 2014). In addition, the 20-itemfour-factor scale demonstrated a similar ability to predict thebehavioural intention of loyalty. However, additional analysisrevealed high inter-item correlations in the four-factor, twenty-item scale that suggested item redundancy (Streiner, 2003).Highly correlated items were removed from the twenty-itemscale. This resulted in a psychometrically sound three-factor,eleven-item scale that is more parsimonious and better fits thedata. This finding has implications for future research intocustomer engagement with tourism brands.
Empirically, the results strongly support the 11-item reducedscale. Conceptually, the collapsing of Enthusiasm and Attention intoAbsorption is supported by considering Hollebeek's et al.’s (2014, p.160) “attitudinal CBE factors”. These three factors map againstHollebeek's et al.'s (2014) factors of cognitive processing andaffection, which they label attitudinal. Looking beyond the defini-tions into the items, and thus what the constructs are measuring,we can see that there is also conceptual argument for constructvalidity. Absorption is described as “(a) pleasant state with whichdescribes the customer as being fully concentrated, happy, anddeeply engrossed while playing the role as a consumer of thebrand,” (So et al. 2014, p. 311) which involves customer passion
with, excitement about, and attraction to the tourism site. In thisway, absorption encompasses enthusiasm and attention.
This leads to the second objective of this research, which was toinvestigate the nature of customer engagement within So et al.'s(2014) proposed nomological framework. Specifically, withcustomer involvement as an antecedent and behavioural intentionof loyalty as a consequence. Providing empirical support for Soet al.'s (2014) nomological framework, the current research findsthat customer engagement is a predictor of brand loyalty using the11-item CETB scale. This is a relationship proposed by many otherresearchers (e.g. De Villiers, 2015; De Vries & Carlson, 2014;Dwivedi, 2015; Hollebeek, 2011; Hollebeek et al., 2014; Viveket al., 2012). The findings also build on Hudson et al.'s (2015) par-allel work on the effects of social media interaction on word-of-mouth, which did not explicitly consider the role of customerengagement.
The finding that involvement is a predictor of engagement withtourism brands on social media is important. Brands must use so-cial media, among other channels, to elicit involvement with theirbrand if they seek to engage with consumers effectively. Involve-ment is characterised, for example, by the brand's level of appeal,meaning, and value to customers (Zaichkowsky, 1994). By placing
Fig. 4. The proposed three-factor model of CE structural model with behavioural intention of loyalty. Note. The latent factor labels represent the following: ID ¼ identification;AB ¼ absorption; INT ¼ interaction; CE ¼ customer engagement; BIL ¼ behavioural intention of loyalty.
Note. The latent factor labels represent the following: INV = customer involvement; CE = customer engagement; BIL = behavioural intention of loyalty; the 11-item scale was used to measure CE.
r2=.44 r2=.48 .66.69
CE INV BIL
Fig. 5. The relationship between customer involvement, engagement and loyalty. Note. The latent factor labels represent the following: INV ¼ customer involvement;CE ¼ customer engagement; BIL ¼ behavioural intention of loyalty; the 11-item scale was used to measure CE.
P. Harrigan et al. / Tourism Management 59 (2017) 597e609606
and testing customer engagement as part of So et al.'s (2014)nomological framework, we emphasise its interdependence onexisting constructs.
6. Conclusion and implications
This research contributes to tourism research, first, by validatinga customer engagement scale previously developed by So et al.(2014). The original scale consists of five dimensions, identifica-tion, enthusiasm, attention, absorption, and interaction. We takethese dimensions, and the scale as whole, and test it in a socialmedia context among U.S. customers. Second, we offer a psycho-metrically sound, parsimonious 11-item version of the CETB. Third,we examine the behavioural loyalty intention as a consequence ofcustomer engagement (Bowden, 2009; Brodie et al., 2011; DeVilliers, 2015; De Vries & Carlson, 2014; Dwivedi, 2015;Hollebeek, 2011; So et al. 2014). Fourth, we examine customerinvolvement as an antecedent to customer engagement (Hollebeeket al., 2014; So et al., 2014).
Customer engagement is an area of significant theoretical andpractical relevance. How brands can utilise social media to increaseengagement among their customers is a key question in the hyper-competitive tourism sector. This research extends the excellent
work by Cabiddu et al. (2014) on how organizations engage withcustomers by articulating how customers engage with the organi-zation's brands. Another key question is around the outcomes ofcustomer engagement. This paper has addressed these questionswith two specific contributions. First, we validate and revise acustomer engagement scale developed by So et al. (2014) specif-ically for the tourism sector. We applied this scale to tourism brandson social media. Second, we place customer engagement within ameaningful nomological framework for researchers and practi-tioners, where customer involvement leads to customer engage-ment, which in turn leads to brand loyalty.
The multi-dimensionality of customer engagement is confirmedby our analysis. We find there to be three dimensions of customerengagement in the 11-item scale that can be used for futureresearch within and beyond the tourism sector. These dimensionsare Absorption, Identification and Interaction. Previous research,notably by Dwivedi (2015) and Hollebeek et al. (2014), has similarlyconceptualised and operationalised dimensions of customerengagement. This research builds on So et al. (2014) and provides aconceptualisation of customer engagement, operationalised in thetourism sector.
This study's theoretical contributions include: (1) the develop-ment of a parsimonious 11-item customer engagement scale that
P. Harrigan et al. / Tourism Management 59 (2017) 597e609 607
can be taken and tested in other tourism or non-tourism contexts;and (2) an empirically tested nomological framework that placescustomer engagement as a consequence of involvement and as anantecedent of behavioural intention of loyalty that can be applied toand assessed in other tourism or non-tourism contexts. The focuson tourism brands on social media is an important contributionthat adds theoretical weight to the social media marketing litera-ture. Although the tourism sector is unique, findings may beapplied to other sectors. We believe social exchange theory is anappropriate theoretical underpinning for customer engagementresearch, no matter the sector. It is clear that exchange betweenconsumer and marketer is essential for consumers to identify with,absorb themselves in, and interact with brands.
Managerially, tourism brand managers using social media willbe able to better understand and shape the nature of customerengagement and the nuances of its dimensions using a concise 11-item CETB scale. Social media is the ideal channel through which toinspire customers' absorption, identification and interaction with abrand. However, these are complex cognitive, affective andbehavioural components. Brands must understand how to effec-tively use various functions of social media, such as pictures, videos,polls, reviews, comments, blogs, all of which can be both marketer-and user-generated, to foster these three different dimensions ofengagement with their brand over another. For example, brandscan provide entertaining or educational content through blogs toabsorb customers. Through these activities and others, brands candevelop a unique image on social media that can enable customersto identify with their brand over others. A final example would bebrands that provide honest and transparent responses to customerreviews can experience positive interaction with their customers.
Managers must also understand the ecosystem within whichcustomer engagement exists and functions. We have illustratedthat involvement is an antecedent to customer engagement, whichmeans that brand managers are particularly responsible for
Consumer engagement with tourism brand items N Mean SD Skewness Kurtosis
ID1. CBE - Identification -When someone criticizes this tourism site, it feels like a personal insult. 195 1.72 1.225 2.108 4.519ID2. CBE - Identification -When I talk about this tourism site, I usually say ‘we’ rather than ‘they’. 195 1.63 1.166 2.463 6.450ID3. CBE - Identification -This tourism site's successes are my successes. 195 1.79 1.308 1.769 2.412ID4. CBE - Identification -When someone praises this tourism site, it feels like a personal compliment. 195 2.02 1.489 1.452 1.220EN1. CBE - Enthusiasm -I am heavily into this tourism site. 195 2.73 1.717 0.612 -0.785EN2. CBE - Enthusiasm -I am passionate about this tourism site. 195 2.79 1.709 0.608 -0.680EN3. CBE - Enthusiasm -I am enthusiastic about this tourism site. 195 3.36 1.938 0.160 �1.326EN4. CBE - Enthusiasm -I feel excited about this tourism site. 195 3.31 1.934 0.248 �1.213EN5. CBE - Enthusiasm -I love this tourism site. 195 4.09 1.747 -0.254 -0.720AT1. CBE - Attention -I like to learn more about this tourism site. 195 3.96 1.870 -0.210 �1.066AT2. CBE - Attention -I pay a lot of attention to anything about this tourism site. 195 3.55 1.845 0.111 �1.144AT3. CBE - Attention -Anything related to this tourism site grabs my attention. 195 3.31 1.827 0.264 �1.094AT4. CBE - Attention -I concentrate a lot on this tourism site. 195 3.32 1.765 0.143 �1.027AT5. CBE - Attention -I like learning more about this tourism site. 195 3.77 1.816 -0.155 -0.999AB1. CBE - Absorption -When I am interacting with the tourism site, I forget everything else around me. 195 2.89 1.755 0.573 -0.790AB2. CBE - Absorption -Time flies when I am interacting with the tourism site. 195 3.51 1.898 0.155 �1.149AB3. CBE - Absorption -When I am interacting with the tourism site, I get carried away. 195 3.00 1.830 0.418 �1.171AB4. CBE - Absorption -When interacting with the tourism site, it is difficult to detach myself. 195 2.45 1.557 0.892 -0.138AB5. CBE - Absorption -In my interaction with the tourism site, I am immersed. 195 3.24 1.790 0.122 �1.234AB6. CBE - Absorption -When interacting with the tourism site intensely, I feel happy. 195 3.73 1.720 -0.267 -0.983INT1. CBE - Interaction -In general, I like to get involved in the tourism site community discussions. 195 2.58 1.775 0.851 -0.451INT2. CBE - Interaction -I am someone who enjoys interacting with like-minded others in the tourism site community. 195 2.81 1.856 0.584 �1.026INT3. CBE - Interaction -I am someone who likes actively participating in the tourism site community discussions. 195 2.64 1.851 0.868 -0.484INT4. CBE - Interaction -In general, I thoroughly enjoy exchanging ideas with other people in the tourism site community. 195 2.76 1.852 0.732 -0.680INT5. CBE - Interaction -I often participate in activities of the tourism site community. 195 2.46 1.750 1.046 -0.070
developing a brand that inspires involvement. Such high levelbranding decisions will influence customer engagement on socialmedia. We have also illustrated that loyalty is an outcome ofcustomer engagement, which emphasises to managers the
importance of, one, being on social media, and two, developingstrategies for customer engagement on social media.
There are limitations to the research, which can be mitigated byfuture research. For example, this study is based on one-country(the U.S.) non-random sample and therefore the findings cannotbe generalized beyond the sample. Future research should validatethe customer engagement scale and model using random samplesin countries with varying cultures. A second, related point is thatthe United States itself is extremely large and diverse both from atourism accessibility perspective and overall with state andregional cultural differences. It is potentially important to investi-gate if the results reported in this research vary depending on stateor regional comparisons. As discussed in the results, the sample ofthis study is broadly representative of the major U.S. regions e
however the sample size is not large enough to do this fine-grainedintra-national investigation. Third, we looked at the most populartourism brands on social media. It would be useful for futureresearch to assess the scale and model on other brands' socialmedia, such as tourism organizations (e.g. Tourism Australia,Discover America), major attractions, and small and large hotels.
Funding
This research was funded by a BHP Billiton DistinguishedResearch Award.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.tourman.2016.09.015.
Appendix
References
Bagozzi, R. P., & Dholakia, U. M. (2006). Antecedents and purchase consequences ofcustomer participation in small group brand communities. International Journalof Research in Marketing, 23, 45e61.
P. Harrigan et al. / Tourism Management 59 (2017) 597e609608
Bijmolt, T. H. A., Leeflang, P. S. H., Block, F., Eisenbeiss, M., Hardie, B. G. S.,Lemmens, A., et al. (2010). Analytics for customer engagement. Journal of ServiceResearch, 13(3), 341e356.
Blau, P. M. (1964). Exchange and power in social life. New York: John Wiley & Sons.Bowden, J. L. (2009). The process of customer engagement: A conceptual frame-
work. Journal of Marketing Theory and Practice, 17(1), 63e74.Brodie, R. J., Hollebeek, L. D., Juri�c, B., & Ili�c, A. (2011). Customer engagement:
Conceptual domain, fundamental propositions, and implications for research.Journal of Service Research, 14(3), 252e271.
Brodie, R. J., Ili�c, A., Juri�c, B., & Hollebeek, L. D. (2013). Consumer engagement in avirtual brand community: An exploratory analysis. Journal of Business Research,66, 105e114.
Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazon's Mechanical Turk a newsource of inexpensive, yet high-quality, data? Perspectives on Psychological Sci-ence, 6(1), 3e5.
Cabiddu, F., Carlo, M. D., & Piccoli, G. (2014). Social media affordances: Enablingcustomer engagement. Annals of Tourism Research, 48, 175e192.
Chau, M., & Xu, J. (2012). Business intelligence in Blogs: Understanding consumerinteractions and communities. MIS Quarterly, 36(4), 1189e1216.
Cheng, M., & Edwards, D. (2015). Social media in tourism: A visual analyticapproach. Current Issues in Tourism, 18, 1080e1087.
Cheung, C. M. K., Lee, M. K. O., & Jin, X.-L. (2011). Customer engagement in an onlineplatform: A conceptual model and scale development. In The proceedings of thethirty-second international conference on information systems (ICIS), Shanghai, China.
Covin, J. G., & Wales, W. J. (2012). The measurement of entrepreneurial orientation.Entrepreneurship Theory and Practice, 36(4), 677e702.
Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience. New York,NY: Harper & Row.
Cui, G., Lui, H.-K., & Guo, X. (2012). The effect of online consumer reviews on newproduct sales. International Journal of Electronic Commerce, 17(1), 39e57.
Daly, T., & Nataraajan, R. (2015). Swapping bricks for clicks: Crowdsourcing longi-tudinal data on Amazon Turk. Journal of Business Research, 68(12), 2603e2609.
De Villiers, R. (2015). Consumer brand enmeshment: Typography and complexitymodeling of consumer brand engagement and brand loyalty enactments.Journal of Business Research, 68(9), 1953e1963.
De Vries, N. J., & Carlson, J. (2014). Examining the drivers and brand performanceimplications of customer engagement with brands in the social media envi-ronment. Journal of Brand Management, 21, 495e415.
Deighton, J., & Kornfeld, L. (2009). Interactivity's unanticipated consequences formarketers and marketing. Journal of Interactive Marketing, 23(1), 4e10.
Dessart, L., Veloutsou, C., & Morgan-Thomas, A. (2015). Consumer engagement inonline brand communities. The Journal of Product and Brand Management, 24(1),28e42.
Dijkmans, C., Kerkhof, P., & Beukeboom, C. J. (2015). A stage to engage: Social mediause and corporate reputation. Tourism Management, 47, 58e67.
Dwivedi, A. (2015). A higher-order model of consumer brand engagement and itsimpact on loyalty intentions. Journal of Retailing and Consumer Services, 24,100e109.
Dwyer, P. (2007). Measuring the value of electronic word of mouth and its impact inconsumer communities. Journal of Interactive Marketing, 21(2), 63e79.
Enginkaya, E., & Esen, E. (2014). Dimensions of online customer engagement.Journal of Business, Economics & Finance, 3(1), 106e114.
Evers, U., Harrigan, P., & Daly, T. (2015). Brand engagement with tourism organi-sations on social media. In Innovation, and growth strategies in marketing con-ference proceedings of the Australian and New Zealand marketing associationconference, Sydney, Australia, 2e4 December (pp. 898e904).
Filieri, R. (2014). Why do travellers trust TripAdvisor? Antecedents of trust towardsconsumer-generated media and its influence on recommendation adoption andword of mouth. Tourism Management, 51, 174e185.
Finch, H. (2006). Comparison of the performance of Varimax and Promax rotations:Factor structure recovery for dichotomous items. Journal of Educational Mea-surement, 43(1), 39e52.
Foa, E. B., & Foa, U. G. (1980). Resource theory: Inter-personal behavior as socialexchange. In K. J. Gergen, M. S. Greenberg, & R. H. Willis (Eds.), Social Exchange:Advances in theory and research. New York: Plenum Press.
Forbes. (2015). TripAdvisor's growth plans for 2015 and beyond. Forbes.com. Availableat: http://www.forbes.com/sites/greatspeculations/2015/03/06/tripadvisors-growth-plans-for-2015-and-beyond/ (Accessed 04.08.15).
Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with un-observable variables and measurement error. Journal of Marketing Research,18(1), 39e50.
Gaskin, J. (2016). ValidityMaster, stats tools package. http://statwiki.kolobkreations.com.
Goh, K. Y., Heng, C. S., & Lin, Z. (2013). Social media brand community and consumerbehavior: Quantifying the relative impact of user-and marketer-generatedcontent. Information Systems Research, 24(1), 88e107.
Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2010). Multivariate dataanalysis: A global perspective. Upper Saddle River, NJ: Pearson.
Hair, J. F., Ringle, C. M., & Sarstedt, M. (2011). PLS-SEM: Indeed a silver bullet. Journalof Marketing Theory and Practice, 19(2), 139e151.
Hennig-Thurau, T., Malthouse, E. C., Friege, C., Gensler, S., Lobschat, L.,Rangaswamy, A., et al. (2010). The impact of new media on customer re-lationships. Journal of Service Research, 13(3), 311e330.
Hollebeek, L. D. (2011). Demystifying customer brand engagement: Exploring theloyalty nexus. Journal of Marketing Management, 27(7e8), 785e807.
Hollebeek, L. D., Glynn, M. S., & Brodie, R. J. (2014). Consumer brand engagement insocial media: Conceptualization, scale development and validation. Journal ofInteractive Marketing, 28(2), 149e165.
Hudson, S., Roth, M. S., Madden, T. J., & Hudson, R. (2015). The effects of social mediaon emotions, brand relationship quality, and word of mouth: An empiricalstudy of music festival attendees. Tourism Management, 47, 68e76.
Jahn, B., & Kunz, W. (2012). How to transform consumers into fans of your brand.Journal of Service Management, 23(3), 344e361.
Kaplan, A. M., & Haenlein, M. (2010). Users of the world, unite! The challenges andopportunities of social media. Business Horizons, 53(1), 59e68.
Kumar, V., Aksoy, L., Donkers, B., Venkatesan, R., Wiesel, T., & Tillmanns, S. (2010).Undervalued or overvalued customers: Capturing total customer engagementvalue. Journal of Service Research, 13(3), 297e310.
Kuvykait _e, R., & Tarute, A. (2015). A critical analysis of consumer engagementdimensionality [abstract]. Kaunas, Lithuania: International Scientific ConferenceEconomics and Management (ICEM).
Lee, J.-N., & Choi, B. (2014). Strategic role of IT and its impact on organizations.Information & Management, 51, 881e882.
Leung, X. Y., Bai, B., & Stahura, K. A. (2015). The marketing effectiveness of socialmedia in the hotel industry a comparison of Facebook and Twitter. Journal ofHospitality & Tourism Research, 39(2), 147e169.
Lin, A., Gregor, S., & Ewing, M. (2008). Developing a scale to measure the enjoymentof Web experiences. Journal of Interactive Marketing, 22(4), 40e57.
Mael, F., & Ashforth, B. E. (1992). Alumni and their alma mater: A partial test of thereformulated model of organizational identification. Journal of OrganizationalBehavior, 13, 103e123.
Malthouse, E., & Hofacker, C. (2010). Looking back and looking forward withinteractive marketing. Journal of Interactive Marketing, 24(3), 181e184.
Mistilis, N., & Gretzel, U. (2013). Tourism operators' digital uptake benchmark survey2013. Research report http://www.tra.gov.au/documents/tourism_operators_surveypdf (Accessed February 2015).
Mollen, A., & Wilson, H. (2010). Engagement, telepresence and interactivity in on-line consumer experience: Reconciling scholastic and managerial perspectives.Journal of Business Research, 63(9e10), 919e925.
Munar, A. M., & Jacobsen, J. K. S. (2014). Motivations for sharing tourism experiencesthrough social media. Tourism Management, 43, 46e54.
Muniz, J. A. M., & O'Guinn, T. C. (2001). Brand community. Journal of ConsumerResearch, 27(4), 412e432.
Nambisan, S., & Baron, R. A. (2007). Interactions in virtual customer environments:Implications for product support and customer relationship management.Journal of Interactive Marketing, 21(2), 42e62.
Pagani, M., & Mirabello, A. (2012). The influence of personal and social-interactiveengagement in social TV web sites. International Journal of Electronic Com-merce, 16(2), 41e67.
Paolacci, G., Chandler, J., & Ipeirotis, P. G. (2010). Running experiments on amazonmechanical turk. Judgment and Decision Making, 5(5), 411e419.
Patterson, P., Yu, T., & De Ruyter, K. (2006). Understanding customer engagement inservices. In Advancing theory, maintaining relevance: Proceedings of ANZMAC2006 conference, Brisbane, Australia (pp. 4e6).
Phang, C. W., Zhang, C., & Sutanto, J. (2013). The influence of user interaction andparticipation in social media on the consumption intention of niche products.Information & Management, 50, 661e672.
Sawhney, M., Verona, G., & Prandelli, E. (2005). Collaborating to create: The Internetas a platform for customer engagement in product innovation. Journal ofInteractive Marketing, 19(4), 4e17.
Schaufeli, W. B., Salanova, M., Gonzalez-Roma, V., & Bakker, A. B. (2002). Themeasurement of engagement and burnout: A two sample confirmatory factoranalytic approach. Journal of Happiness Studies, 3(1), 71e92.
Scholer, A. A., & Higgins, E. T. (2009). Exploring the complexities of value creation: Therole of engagement strength. Journal of Consumer Psychology, 19(2), 137e143.
So, K. K. F., King, C., & Sparks, B. (2014). Customer engagement with tourism Brands:Scale development and validation. Journal of Hospitality & Tourism Research,38(3), 304e329.
Streiner, D. L. (2003). Starting at the beginning: An introduction to coefficient alphaand internal consistency. Journal of Personality Assessment, 80(1), 99e103.
Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5th ed). UpperSaddle River, NJ: Pearson Allyn & Bacon.
Tajfel, H., & Turner, J. C. (1985). The social identity theory of intergroup behaviour. InS. Worchel, & W. G. Austin (Eds.), Psychology of intergroup relations (pp. 7e24).Chicago, IL: Nelson-Hall.
Thibaut, J. W., & Kelley, H. H. (1959). The social psychology of groups. New York: JohnWiley & Sons.
Van Doorn, J., Lemon, K. E., Mittal, V., Nab, S., Pick, D., Pirner, P., et al. (2010).Customer engagement behavior: Theoretical foundations and research di-rections. Journal of Service Research, 13(3), 253e266.
Vivek, S. D. (2009). A scale of consumer engagement, doctor of philosophy dissertation,department of management & marketing. The University of Alabama. http://libcontent1.lib.ua.edu/content/u0015/0000001/0000096/u0015_0000001_0000096.pdf.
Vivek, S. D., Beatty, S. E., & Morgan, R. M. (2012). Customer engagement: Exploringcustomer relationships beyond purchase. Journal of Marketing Theory andPractice, 20(2), 127e145.
Zaichkowsky, J. L. (1985). Measuring the involvement construct. Journal of ConsumerResearch, 12(3), 341e352.
Zaichkowsky, J. L. (1994). The personal involvement inventory: Reduction, revision,
P. Harrigan et al. / Tourism Management 59 (2017) 597e609 609
and application to advertising. Journal of Advertising, 23(4), 59e70.Zeithaml, V., Berry, L., & Parasuraman, A. (1996). The behavioral consequences of
service quality. Journal of Marketing, 60(2), 31e46.
Dr. Paul Harrigan has been at UWA since July 2012.Before this, he was a Lecturer in Marketing at the Uni-versity of Southampton in the UK from 2008 to 2012.PhD, from the University of Ulster in 2008, looked atcustomer relationship management (CRM) in SMEs. Hiscurrent research interests lie in CRM, spanning themarketing and information systems literature. Socialmedia is my specific expertise, with current projectslooking at social CRM (i.e. the impact of social media onCRM) from business and consumer perspectives. He haspublished in journals such as the International Journal ofElectronic Commerce, the Journal of Marketing Manage-ment, the Journal of Strategic Marketing and the Journal ofMarketing Education and has close engagement with in-
dustry and marketing practitioners.
Uwana Evers is a Research Fellow and Lecturer inMarketingat the UWA Business School, University ofWestern Australia.She is a BPS Chartered Psychologist, has a PhD in Psychologyfrom the University of Wollongong and has worked at theUniversity College London. Uwana currently teaches under-graduate Marketing Research, and has published and pre-sented research in social marketing, behaviour change,personal values, and cross-cultural psychology.
Professor Morgan P. Miles is Miles is Professor ofEntrepreneurship and Innovation at the University ofCanterbury, Previously he had been the Tom HendrixChair of Excellence at the University of Tennessee Mar-tin, Professor of Enterprise Development at the Univer-sity of Tasmania, and Professor of Marketing at GeorgiaSouthern University. He has been a visiting scholar atGeorgia Tech, Cambridge University, University ofStockholm, the University of Otago, University of Auck-land, and an Erskine Fellow at the University of Canter-bury. He holds a D.B.A. in marketing from MississippiState University.
Dr. Timothy Daly is Assistant Professor at United ArabEmirates University. Dr Daly graduated in 2010 from theUniversity of Western Australia with a PhD in Marketing.Since then he has worked as an Assistant Professor at theUniversity of Akron, Ohio (USA), the University ofWestern Australian and United Arab Emirates University.He currently teaches Consumer Behavior.
