THE MARKETING EFFECTIVENESS OF

SOCIAL MEDIA IN THE HOTEL

INDUSTRY: A COMPARISON OF
FACEBOOK AND TWITTER

Xi Y. Leung
University of Mississippi
Billy Bai
University of Nevada, Las Vegas
Kurt A. Stahura
University of Nevada, Las Vegas–Singapore Campus

The unprecedented popularity of social media outlets have forced scholars to inquire about
their marketing effectiveness, especially in the hotel industry. This study attempted to explore
the marketing effectiveness of two different social media sites (Facebook and Twitter) in the
hotel industry. Integrating the attitude-toward-the-ad (Aad) model with the concepts of
attitude-toward-social-media-page, the study proposed a theoretical model of hotel social
media marketing effectiveness. Based on the data collected from an online survey, the goodness
of fit of the model implied that the Aad model provides an appropriate theoretical framework
to explain the marketing effectiveness of social media in the hotel industry. The results
revealed that hotel customers’ social media experiences influence their attitudes-toward-
social-media-site, which in turn influences their attitudes-toward-hotel-brand, and that hotel
customers’ attitudes-toward-hotel-brand affects their hotel booking intentions and, in turn,
intentions to spread electronic word of mouth. The study also indicated that different social
media sites demonstrate the same marketing effectiveness, suggesting that hotel managers
use the same marketing tactics for Facebook and Twitter marketing.

KEYWORDS: attitude-toward-the-ad; attitude-toward-social-media-page; eWOM;

marketing effectiveness; social media experience

INTRODUCTION

In the past few years, the development of new Internet applications has changed
the characteristics of websites toward more participatory, interactive, and user-
centric social media. The Merriam-Webster dictionary defines social media as
“forms of electronic communication through which users create online communities

Authors’ Note: This article was partially supported by the Conrad N. Hilton Foundation. We are
also thankful for the partial support by the Caesars Foundation for this research.
Journal of Hospitality & Tourism Research, Vol. XX, No. X, Month XXXX, xx-xx
DOI: 10.1177/1096348012471381
© 2013 International Council on Hotel, Restaurant and Institutional Education
Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
1
2   JOURNAL OF HOSPITALITY & TOURISM RESEARCH

to share information, ideas, personal messages, and other content” (Merriam-

Webster, 2012). These newer versions of social media (Web 2.0 technologies)
have become extremely popular on the Internet and are predicted to be future
marketing media (Cooke & Buckley, 2008). A recent study by Madden and Zickuhr
(2011) reported that 65% of U.S. adult Internet users (50% of all U.S. adults)
now use social networking sites, more than double the percentage reported in
2008 (29%) and that 43% of online adults use social networking sites daily. Social
media includes applications such as social networking sites, blogs, content
communities, forums/bulletin boards, and other interactive applications, which
allow users to create and share information (Alarcó-del-Amo, Lorenzo-Romero,
& Gómez-Borja, 2011; Constantinides & Fountain, 2008). Among various social
media sites, Facebook and Twitter were ranked as the top two most popular sites
in terms of website’s traffic (“Top 15,” 2011).
The explosive growth of social media channels has transformed the way
many consumers interact with each other and with businesses of all kinds.
Ultimately this is changing the way we do business and how businesses attract
and retain consumers. A study showed that social media channels were commonly
used by U.S. businesses to connect with their customers or prospects than
Google (The eMarketer, 2011). Seventy percent (70%) of businesses use
Facebook, followed by 46% using Twitter, 37% using LinkedIn, and 25% using
YouTube. In an annual study conducted by Stelzner (2010, 2011), Facebook
and Twitter were the top two social media tools used by marketers for the past
2 years. In the hotel industry, about 75% of hotels have used social media for
marketing purposes (Hotelmarketing.com, 2011). Facebook and Twitter were
also found to be the two most useful social media channels for marketing
purposes in the hospitality industry (Friebe & Campbell, 2010; Hotelmarketing.
com, 2011).
As more hotels are incorporating social media into their integrated marketing
communications, the attention on its effectiveness is rising. Although social
media is claimed to be effective in improving marketing practices, there has
been limited quantitative support to reinforce these claims. According to Stelzner’s
(2009, 2010, 2011) annual industry report, how to measure the marketing
effectiveness of social media is one of the most important questions marketers
need to continue to investigate. However, very few studies have examined the
effectiveness of social media promotions in the academic world (Mabry & Porter,
2010). In advertising research, to understand the process through which advertising
influences consumer behavior has been a long-standing focus and several major
theories have been proposed to study advertising effectiveness (MacKenzie,
Lutz, & Belch, 1986). Therefore, the purpose of this study was to explore the
marketing effectiveness of social media using advertising theory, namely, attitude
toward the ad model and attitude toward the website model. Specifically, this
study intended to achieve the following objectives: (a) to develop a hypothesized
model of hotel social media marketing effectiveness based on the Attitude-
toward-the-ad and the Attitude-toward-the-website models; (b) to test the
hypothesized model in the context of the two most popular social media sites:
Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
 Leung et al. / MARKETING EFFECTIVENESS OF SOCIAL MEDIA  3

Facebook and Twitter; (c) to compare the marketing effectiveness of two hotel
social media pages; and (d) to provide suggestions for the hotel industry in
leveraging social media marketing.

LITERATURE REVIEW

Attitude-Toward-the-Ad Model
The studies on the effects of persuasive advertising on attitude formation and
change have led to a very important concept in marketing and advertising research:
Attitude-toward-the-ad (Aad; Edell & Burke, 1984). The concept of Aad was first
introduced by Mitchell and Olson (1981) and Shimp (1981). They found it to be
an affective construct referring to individuals’ favorable/unfavorable feelings
toward a particular advertisement after ad exposure.
Numerous studies have tested that Aad has a mediating influence on brand
attitudes and purchase intentions (Lutz, Mackenzie, & Belch, 1983; MacKenzie
et al., 1986; Mitchell & Olson, 1981; Shimp, 1981). The Aad model describes
possible sequences of exposure to a persuasive advertisement and generally posits
that a recipient of an advertising message develop an Aad, which, in turn, exerts
an influence on subsequent measures of advertising effectiveness such as brand
attitude and purchase intentions (Lutz et al., 1983). Studies on Aad have proposed
four competing Aad models representing different mediating roles of Aad (Lutz
et al., 1983; MacKenzie et al., 1986). The four Aad models are based on four
alternative hypotheses: affect transfer hypothesis (ATH), dual mediation hypothesis
(DMH), reciprocal mediation hypothesis (RMH), and independent influences
hypothesis (IIH; Figure 1).
In all four Aad models, ad cognitions and brand cognitions have direct impacts
on Aad and attitude toward the brand, respectively. However, different hypotheses
posit that Aad has different impacts on attitude toward the brand and purchase
intention. The ATH model postulates a direct one-way influence of Aad and brand
cognitions on attitude toward the brand (Mitchell & Olson, 1981; Shimp, 1981).
The DMH model posits both a direct effect of Aad on attitude toward the brand
and an indirect effect through the mediation of brand cognitions (Holbrook, 1978;
Lutz & Swasy, 1977). The RMH model is based on the balance theory (Heider,
1946) and asserts a reciprocal causal flow between Aad and attitude toward the
brand, that is, these two constructs affect each other. The basic premise of balance
theory is that a person has a preference to maintain balance among a set of cognitive
components in a system (Heider, 1946). In advertising, the consumer, the ad, and
the brand constitute the system (Edell & Burke, 1984). Thus, the ad–brand
relationship is always a balanced state, suggesting Aad and attitude toward the
brand are highly positively correlated (Edell & Burke, 1984). Finally, the IIH
model assumes no causal relationship between Aad and attitude toward the brand
while both have direct impacts on purchase intentions (Howard, 1977). Both
Mackenzie et al.’s (1986) and Homer’s (1990) studies compared the four competing
Aad models using experiment data and demonstrated that the DMH model provides
Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
4   JOURNAL OF HOSPITALITY & TOURISM RESEARCH

Figure 1
Four Alternative Attitude-Toward-the-Ad Models

A. Affect transfer hypothesis B. Dual mediation hypothesis

Cad Aad Cad Aad
Ad Ad
Exposure Exposure
Cb Ab PI Cb Ab PI

C. Reciprocal mediation hypothesis D. Independent influences hypothesis

Cad Aad Cad Aad
Ad Ad
Exposure Exposure
Cb Ab PI Cb Ab PI

Cad: ad cognitions
Cb: brand cognitions
Aad: attitude toward the ad
Ab: attitude toward the brand
PI: purchase intention

Note: Adapted from MacKenzie et al. (1986).

the best fit to the data. That is, Aad has both direct and indirect effects through
brand cognitions on attitude toward the brand.
The Aad model has been rarely applied in the hospitality field. Only a study
by Miller and Stoica (2004) compared the effects of photograph print ad versus
artistic rendition ad for a fictional Caribbean resort island based on the Aad model.
The study found that the photograph ad was better than the artistic rendition ad
in generating favorable attitude toward the ad, attitude toward the brand, and visit
intention. Also, in hospitality research, brand cognition has often been studied as
brand equity or brand awareness. Brand equity as customer’s cognitive awareness
or mindset of a brand has been demonstrated to influence customers’ purchase
intentions and brand choices in the hospitality industry (Cobb-Walgren, Ruble,
& Donthu, 1995). Kim, Jin-Sun, and Kim (2008) again stated that brand awareness
has a positive effect on midpriced hotel guests’ revisit intention. In this study, the
measures of brand awareness were used to measure brand cognition.

Attitude-Toward-the-Website Model
The Aad model has also been extended to explain web advertising effects.
Under the web environment, a new construct, attitude-toward-the-website (Aws),
is added to be as important as Aad in evaluating advertising effectiveness (Chen
& Wells, 1999). Similar to Aad, Aws is defined as web users’ “predispositions to
respond favorably or unfavorably to web content in natural exposure situations”
(Chen & Wells, 1999, p. 28). The rationale for adding this new construct is that
customers’ reactions to the context where an advertisement is presented (the
Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
 Leung et al. / MARKETING EFFECTIVENESS OF SOCIAL MEDIA  5

website) are proposed to affect how consumers react to the ad (Bruner & Kumar,
2000). Chen and Wells (1999) developed a reliable and valid scale that measures
Aws and concluded that entertainment, informativeness, and organization are
three dimensions of Aws. Stevenson, Bruner, and Kumar (2000) proposed that
Aws plays an important role in the traditional Aad model, especially the ATH
model. Bruner and Kumar (2000) further tested their new model of web advertising
effectiveness, which proposed that one’s web experience play an important role
along with webpage complexity and interestingness on Aws, which in turn affects
Aad, attitude-toward-the-brand, and finally purchase intention. Poh and Adam
(2002) incorporated the three dimensions of Aws (Chen & Wells, 1999) with the
web advertising effectiveness model proposed by Bruner and Kumar (2000) and
developed an integrative Aws model.
In the hospitality field, only a few studies have focused on the concept of Aws.
McMillan, Hwang, and Lee (2003) explored determining factors of Aws of hotel
websites. Two structural variables (number of features and creative strategy) and
two perceptual variables (involvement and perceived interactivity) were tested
in terms of their effects on Aws. They found that perceptual variables have greater
impacts on Aws than structural variables. Jeong and Choi (2004) examined the
potential effects of different picture presentations on hotel websites on Aws and
customers’ behavioral intentions. The findings indicated that the picture content
and picture realism significantly influence Aws of hotel websites and Aws is a
strong predictor of behavioral intentions.

WOM and eWOM

Word-of-mouth (WOM) has been an important concept in the marketing field
for decades. Since the early 1950s, researchers have noticed the importance of
personal conversation and informal exchange of information among acquaintances
in marketing (Arndt, 1967; Whyte, 1954). Arndt (1967) was one of the earliest
researchers who defined WOM as oral, person-to-person communication between
a communicator and receiver that is perceived as noncommercial message. In the
late 1990s, the WOM concept was expanded to be applied in the Internet-based
communications and the power of WOM has become stronger using Web 2.0
technologies (Hennig-Thurau & Walsh, 2004). The so-called electronic word-of-
mouth (eWOM) is defined as “all informal communications directed at consumers
through Internet-based technology related to the usage or characteristics of
particular goods and services, or their sellers” (Litvin, Goldsmith, & Pan, 2008,
p. 461).
WOM and eWOM can serve as both marketing tools and marketing outcomes
(Hennig-Thurau, Gwinner, & Gremler, 2002; Litvin et al., 2008). As marketing
tools, WOM and eWOM are powerful in influencing consumers’ attitudes, product
evaluation, customer loyalty, and purchase intentions and decisions (Arndt, 1967;
Chan & Ngai, 201; Litvin et al., 2008). Research generally demonstrates that
WOM and eWOM are more influential than conventional marketing tools (Day,
1971; Godes & Mayzlin, 2004; Herr, Kardes, & Kim, 1991). On the other hand,
WOM and eWOM are also considered as one of the key marketing outcomes,
Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
6   JOURNAL OF HOSPITALITY & TOURISM RESEARCH

which are as important as customer loyalty (Casaló, Flavián, & Guinalíu, 2010;
Hennig-Thurau et al., 2002). Numerous studies have been conducted in an attempt
to gain insights into the antecedents and motives of WOM and eWOM (Dichter,
1966; Hennig-Thurau, Gwinner, Walsh, & Gremler, 2004; Sundaram, Mitra, &
Webster, 1998).
In the hospitality field, the drivers of WOM have also been studied in several
different contexts and attitude has been identified as an important attribute of
WOM communication (Cheng, Lam, & Hsu, 2006; Leach, Liu, & Winsor, 2008).
Leach et al. (2008) proposed a model of conference attendance that included both
intention to attend and intention to recommend (WOM). The findings suggested
that attitude toward conference is the most important factor affecting both intention
to attend a future conference and intention to recommend the conference to others.
Thus, we also incorporated intention to spread WOM with hotel booking intention
in the hypothesized model.

HYPOTHESIZED MODEL

This study posited that the affect transfer hypothesized (ATH) Aad model can
be applied as the theoretical framework to explain the marketing effectiveness of
social media for hotels. The ATH model postulates that Aad and brand cognitions
both have direct impact on attitude toward the brand, which in turn affects purchase
intention (here referred to as hotel booking intention). Since social media are
types of websites, the study incorporated the concept of Aws into the ATH model,
in which Aws is actually attitude-toward-social-media-page. In addition, this study
focused on hotel social media pages as a whole rather than a specific advertising
message on the social media page. As a result, the Aad concept in the Aad model
was left out in the proposed model. Thus, in the model proposed by the study,
hotel customers’ attitudes toward social media pages are proposed to have direct
impacts on their attitudes toward hotel brand, which then influences their hotel
booking intention. Hotel customers’ social media experience and brand cognition
are postulated to affect their attitudes toward social media sites and attitudes
toward hotel brand, respectively. Besides, the study added intention of eWOM
(in this study, eWOM refers to positive eWOM) as another major marketing
outcome of social media that is expected to be determined by both attitudes toward
hotel brand and hotel booking intention. Specifically, six directional hypotheses
were proposed:

Hypothesis 1: The more positive a customer’s social media experience, the more
positive the customer’s attitude toward the social media site.
Hypothesis 2: The more positive a customer’s attitude toward the social media site,
the more positive the customer’s attitude toward the hotel brand.
Hypothesis 3: The higher a customer’s cognition of a hotel brand, the more positive
the customer’s attitude toward the hotel brand.
Hypothesis 4: The more positive a customer’s attitude toward a hotel brand, the more
likely the customer will book this hotel brand.

Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015

 Leung et al. / MARKETING EFFECTIVENESS OF SOCIAL MEDIA  7

Hypothesis5: The more positive a customer’s attitude toward a hotel brand, the more
likely the customer will spread positive word-of-mouth about this hotel brand online.
Hypothesis 6: The more likely a customer to book a hotel brand, the more likely the
customer to spread positive word-of-mouth about this hotel brand online.

In addition, the study also posited that different social media sites may have
different marketing effectiveness:

Hypothesis 7a: The relationships between a customer’s social media experience and
attitude-toward-social-media-page are different when the customer uses Facebook
or Twitter.
Hypothesis 7b: The relationships between a customer’s attitude-toward-social-media-
page and attitude-toward-hotel-brand are different when the customer uses Facebook
or Twitter.
Hypothesis 7c: The relationships between a customer’s brand cognition and attitude-
toward-hotel-brand are different when the customer uses Facebook or Twitter.
Hypothesis 7d: The relationships between a customer’s attitude-toward-hotel-brand
and hotel booking intention are different when the customer uses Facebook or Twitter.
Hypothesis 7e: The relationships between a customer’s attitude-toward-hotel-brand
and intention of eWOM are different when the customer uses Facebook or Twitter.
Hypothesis 7f: The relationships between a customer’s hotel booking intention and
intention of eWOM are different when the customer uses Facebook or Twitter.

To examine these hypothesized relationships, a path model was proposed (see

Figure 2). In the hypothesized model, all aforementioned hypothesized structural
relationships were proposed and tested simultaneously. All hypotheses were tested
in the context of two social media sites: Facebook and Twitter. These two social
media channels were selected because they are the most useful and most used
social media tools for marketing purposes (Friebe & Campbell, 2010; Stelzner,
2010, 2011). All factor loadings and effect sizes of the relationships were tested
for invariance across the two social media site groups.

METHODOLOGY

An online survey was conducted by Qualtrics in November 2011. A convenient

sample was selected using a database provided by the Utah-based online research
company Qualtrics. Established in 1997, Qualtics panel members now include
nearly 4 million individuals within the United States who have already agreed to
be contacted for survey participation. An email was sent to all the panel members
in search of people who were Facebook users or Twitter users. Qualified participants
were invited to take the survey via a link contained in the email. A total number
of 408 participants completed the survey within a week of data collection. All
participants were assigned to two groups (Facebook or Twitter) randomly, resulting
in 204 respondents in each group. The Facebook page or the Twitter page of one
major hotel brand was chosen as the stimulus for each group. The conditions for
participating in this study were that the respondents had to have a Facebook account

Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015

8   JOURNAL OF HOSPITALITY & TOURISM RESEARCH

Figure 2
Hypothesized Model of Social Media Marketing Effectiveness

Brand
cognition
Intention to make
H3
hotel booking
H4 decision
Attitude toward
Social media H1 H2 Attitude toward
social media H6
experience hotel brand
website
H5
Intention of
eWOM

to be in the Facebook group or to have a Twitter account to be in the Twitter group.

The survey comprised three steps. First, all participants took a presurvey, which
consisted of questions on demographics and brand cognition. Second, after finishing
the presurvey, the two groups were led to different social media sites. One group
was asked to read the Facebook page of the hotel brand. The other group was
required to read the Twitter page of the same hotel brand. The participants had as
much time to look at the webpages as they needed. Finally, after reading the
webpages, all participants were requested to complete a postsurvey, which consisted
of questions on social media experience, attitude-toward-social-media-page,
attitude-toward-hotel-brand, booking intention, and intention of eWOM.

Instrument
The survey comprised two questionnaires: a presurvey questionnaire and a
postsurvey questionnaire. The presurvey questionnaire consisted of two sections.
The first section included several demographic questions regarding gender, age,
ethnicity, education level, and Internet usage. The second section included questions
regarding brand cognition. The respondent’s cognition of the selected hotel brand
was measured using multi-item scales. All items were measured by 5-point Likert-
type scales anchored by strongly disagree (1) and strongly agree (5).
The postsurvey questionnaire consisted of five sections. The first section asked
the respondent’s experience on the hotel social media page he/she just visited. All
14 items were measured by 5-point semantic differential scales anchored by not
at all (1) and very much (5). The second and third sections measured the respondent’s
attitude toward social media site he/she just visited and attitude toward the hotel
brand. Both of them were measured using 5-point semantic differential scales. The
fourth and fifth sections asked about the participant’s intentions to book the hotel
and to recommend the hotel online in the future. Both of the intentions were
measured by 3 items using 5-point Likert-type scales. Table 1 lists all the scale
items used to measure each construct and their reference sources.
Before data analysis, measurement validity and reliability were first evaluated.
Measurement validity of the instrument was evaluated by conducting principal
axis factor analysis with oblimin rotation on all of the items. After performing
Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
 Leung et al. / MARKETING EFFECTIVENESS OF SOCIAL MEDIA  9

Table 1
Measurement Items of the Constructs
Constructs Measurement Items Reference

Brand I have heard of Hilton hotel brand. Buil, de Chernatony,

cognition I have been staying in Hilton hotel. and Martinez (2008);
I know what Hilton hotel looks like. So and King (2010)
I can recognize Hilton hotel
When I book a hotel, Hilton hotel comes
into my mind.
Experience Important: unimportant Choi (2000); Moital,
Interesting: boring Vaughan, Edwards,
Relevant: irrelevant and Peres (2009)
Exciting: unexciting
Means a lot to me: means nothing
Appealing: unappealing
Fascinating: mundane
Valuable: worthless
Involving: uninvolving
Needed: not needed
Beneficial: prejudicial
Desirable: undesirable
Useful: useless
Essential: dispensable
Attitude- Good: bad Choi (2000)
toward- Likable: dislikable
social- Favorable: unfavorable
media-page Enjoyable: unenjoyable
Pleasant: unpleasant
Entertaining: not entertaining
Interesting: uninteresting
Nice: awful
Attitude- Positive: negative Chiang and Jang
toward- Satisfactory: unsatisfactory (2006); Choi (2000)
hotel-brand Favorable: unfavorable
Good: bad
Likable: dislikable
Nice: awful
Booking My willingness to book Hilton hotels is very Chiang and Jang
intention high. (2006)
The probability that I would consider
booking Hilton hotels is very high.
The likelihood of booking Hilton hotels is
very high.
Intention of My willingness to recommend Hilton hotels to Chiang and Jang
eWOM others on Facebook/Twitter is very high. (2006); Gruen,
The probability that I would consider Osmonbekov, and
recommending Hilton hotels to others on Czaplewski (2006)
Facebook/Twitter is very high.
The likelihood of recommending Hilton
hotels to others on Facebook/Twitter is
very high.

Note: eWOM = electronic word-of-mouth.

Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
10   JOURNAL OF HOSPITALITY & TOURISM RESEARCH

Table 2
Internal Consistency Estimates (α) of Constructs
Construct Facebook Group Twitter Group

Brand cognition .86 .83

Experience .96 .97
Attitude-toward-social-media-page .96 .96
Attitude-toward-hotel-brand .97 .97
Booking intention .96 .96
Intention of eWOM .98 .97

Note: eWOM = electronic word-of-mouth.

validity analyses, two items in the measures for attitude-toward-social-media-page

were deleted to maintain both convergent and discriminant validities. The
remaining items were tested for internal consistency by Cronbach’s alpha. The
results demonstrated good internal consistency for all measures in both groups
(.83-.98; see Table 2). The lowest internal consistency (reliability) is from the
brand cognition construct in both groups.

Data Analysis Procedure

Data were analyzed using multigroup structural equation modeling (SEM).
Measurement models were tested first in each group (Facebook or Twitter),
followed by invariance tests on factor loadings across two sample groups. The
full structural model was tested, followed by invariance tests on structural
coefficients between Facebook and Twitter groups. The models were tested using
EQS 6.1 software.
The evaluation of model adequacy was based on comparative fit index (CFI),
nonnormed fit index (NNFI), standardized root mean square residual (SRMR),
and root mean square error of approximation (RMSEA) and its lower and upper
confidence interval boundaries. In addition, standardized residuals and the results
of Lagrange multiplier tests and Wald tests were inspected along with the theoretical
literature of the research area.
Each construct in the model had three indicators. For brand cognition, social
media experience, attitude-toward-social-media-page, attitude-toward-hotel-brand
constructs, item parceling was conducted to group more than three items into
three indicators. Item parceling was used because these constructs (latent variable)
had too many items. Using all individual items in the measurement model may
harm the overall model fit. Item parceling can lead to a better fitting solution and
less bias in estimates of structural parameters when the items are unidimensional
(Bandalos, 2002).
Item parceling in this study used the following procedure: First, items belonging
to each construct (latent variable) were subjected to exploratory factor analysis
to determine if the unidimensional assumptions were met. After finding that the
unidimensional assumptions were met, items were grouped into three indicators
Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
 Leung et al. / MARKETING EFFECTIVENESS OF SOCIAL MEDIA  11

for each factor based on factor loadings in the exploratory factor analysis conducted
in the validity check. Finally, the average score of the items was calculated to
represent the subscale scores of each indicator. This procedure was applied for
each group.
Linearity, multicollinearity, and singularity assumptions for the SEM analyses
were met. The multivariate kurtosis indicated that the data distributions were less
than optimal (normalized estimates were 33.40 and 31.97 for the Facebook and
the Twitter groups, respectively). However, the data distributions and outlier
analysis suggested there was no outlier. Using both the maximum likelihood
estimation and the robust methods estimation, multigroup SEMs were run. As
the results from both methods were very similar, the results of the maximum
likelihood estimation were reported.

RESULTS

Demographic Profile of Respondents

The final sample was composed of 50% (102) males and 50% (102) females
in both groups. The demographics of Facebook and Twitter users indicated that
there were approximately the same numbers of males and females (Infographic,
2010). The age of the respondents ranged from 18 to 78 years in the Facebook
group and from 18 to 72 years in the Twitter group. The average age of the
respondents was 43 years in the Facebook group and 41 years in the Twitter group.
The biggest age group in Facebook was 18 to 25 years, which is consistent with
the Facebook users’ age distribution (Infographic, 2010). The biggest age groups
in Twitter were 26 to 34 years and 35 to 44 years, which is also consistent with
the Twitter users’ age distribution (Infographic, 2010).
More than three fourths of the respondents were White/Caucasian in both the
Facebook group (77.9%) and the Twitter group (75.9%). These numbers were
similar to the results of a survey conducted by Pew Research Center that 78% of
Facebook users and 71% of Twitter users were White (Hampton & Goulet, 2011).
The education level of the respondents was high. The demographics revealed that
79.9% of the respondents in the Facebook group and 82.8% in the Twitter group
had at least some college education. In addition, 39.2% of the respondents in the
Facebook group and 47.5% in the Twitter group had at least a bachelor’s degree.
These numbers were also similar to the education distribution of Facebook and
Twitter users (Hampton & Goulet, 2011).
The respondents reported a long history of Internet usage and social media
usage. All the respondents in the Facebook group had used the Internet for at
least 6 months, and all the respondents in Twitter group had used the Internet
for at least 1 year. Finally, 95.6% of the respondents in both groups had used the
Internet for at least 4 years. The data revealed that 84.8% of the respondents in
the Facebook group and 94.6% in the Twitter group had used social media
websites for at least 1 year.

Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015

12   JOURNAL OF HOSPITALITY & TOURISM RESEARCH

Table 3
Standardized Factor Loadings and Variances (R2), After Equality Constraints Were
Tested
Attitude- Attitude-
Brand Toward-Social- Toward- Booking Intention
Indicator Cognition Experience Media-Page Hotel-Brand Intention of eWOM R2

BC1 .86 (.86) .74 (.74)

BC2 .79 (.73) .62 (.53)
BC3 .96 (.94) .92 (.89)
EP1 .95 (.96) .90 (.93)
EP2 .88 (.89) .77 (.80)
EP3 .96 (.96) .92 (.92)
AM1 .95 (.96) .90 (.91)
AM2 .95 (.96) .90 (.92)
AM3 .93 (.93) .87 (.87)
AB1 .95 (.97) .90 (.93)
AB2 .98 (.96) .95 (.92)
AB3 .96 (.96) .93 (.92)
BI1 .94 (.92) .87 (.84)
BI2 .97 (.97) .94 (.94)
BI3 .95 (.94) .89 (.88)
WOM1 .94 (.93) .89 (.86)
WOM2 .98 (.99) .96 (.97)
WOM3 .97 (.95) .94 (.91)

Note: eWOM = electronic word-of-mouth. Values in parentheses are loadings for Facebook group.
Values not in parentheses are loadings for Twitter group. All ps < .001. All indicators are indicators after
item parceling of the measurements.

Measurement Model
The measurement model specified six factors: brand cognition, social media
experience, attitude-toward-social-media-page, attitude-toward-hotel-brand,
booking intention, and intention of eWOM. To test the model in each group
(Facebook and Twitter), indicators were constrained to load only on the factor it
was designated to measure. The residual terms for all indicators were fixed to be
uncorrelated. No equality constraints on the factor loadings were imposed and
the factor covariances were free to be estimated.
Goodness-of-fit indices indicated that the measurement model fit the data well
in both groups: χ2(120, N = 204) = 237.72, p < .001, CFI = .98, NNFI = .97,
SRMR = .03, RMSEA = .07 (CI = 0.06, 0.08) for the Facebook group, and χ2(120,
N = 204) = 193.36, p < .001, CFI = .99, NNFI = .98, SRMR = .03, RMSEA = .05
(CI = 0.04, 0.07) for the Twitter group. All factor loadings of the indicators were
statistically significant, ps < .001, ranging from .80 to .98 for the Facebook group
and from .72 to .99 for the Twitter group.
Next, whether all factor loadings were invariant between the two groups was
tested. The parameters to be tested were constrained to be equal across the two
social media groups. Results of Lagrange multiplier tests of multivariate statistics
and univariate increment indicated that all factor loadings were invariant.
Standardized factor loadings are presented for the Facebook and Twitter groups
in Table 3.
Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
 Leung et al. / MARKETING EFFECTIVENESS OF SOCIAL MEDIA  13

Table 4
Correlation Between Constructs
Construct (Factor) 1 2 3 4 5 6

1.  Brand cognition — .48* .36* .37* .68* .57*

2. Experience .56* — .85* .75* .58* .54*
3. Attitude-toward-social-media-page .50* .83* — .92* .51* .56*
4. Attitude-toward-hotel-brand .50* .80* .94* — .54* .59*
5.  Booking intention .61* .64* .56* .56* — .87*
6.  Intention of eWOM .53* .67* .60* .61* .84* —

Note: Values below the diagonal are correlation coefficients for Facebook group. Values
above the diagonal are correlation coefficients for Twitter group.
*p < .001.

Variances (R2) of the indicators were accounted for by their corresponding

constructs, which ranged from .62 to .96 for the Facebook group and from .53 to
.97 for the Twitter group. The two smallest explained variances were both from
the indicators of brand cognition for both groups (.62 and .74 for the Facebook
group; .53 and .74 for the Twitter group).
The correlations among factors in the measurement model are presented in
Table 4. In the Facebook group, correlation coefficients ranged from .50 to .94,
all ps < .001. The highest correlation was the relationship between attitude-toward-
social-media-page and attitude-toward-hotel-brand. In the Twitter group,
correlation coefficients ranged from .36 to .92, all ps < .001. The highest correlation
was also found to be the relationship between attitude-toward-social-media-page
and attitude-toward-hotel-brand.

Structural Model
To examine the goodness-of-fit of the hypothesized model, the measurement
model was respecified by imposing the structure of the model. The equality
constraints on factor loadings across grade that were tested plausible remained
fixed. Goodness-of-fit indices indicated that the hypothesized model only
represented a marginal fit to the data in both groups: χ2(127, N = 204) = 299.52,
p < .001, CFI = .97, NNFI = .96, SRMR = .1, RMSEA = .08 (CI = 0.07, 0.09)
for the Facebook group, and χ2(127, N = 204) = 288.99, p < .001, CFI = .97,
NNFI = .96, SRMR = .1, RMSEA = .08 (CI = .07, .09) for the Twitter group. For
both the Facebook and the Twitter groups, the LM test suggested that one parameter
that was not specified in the hypothesized model contributed a lot to model fit;
that is, brand cognition had a direct effect on hotel booking intention. Given the
findings of marginal fit of hypothesized models to both groups, the model was
respecified to both groups based on the LM statistics.
After being respecified, the two new models represented good fits to data in two
groups, respectively: χ2(126, N = 204) = 257.89, p < .001, CFI = .98, NNFI = .97,
SRMR = .05, RMSEA = .07 (CI = .06, .08) for the Facebook group, and χ2(126, N
= 204) = 210.24, p < .001, CFI = .98, NNFI = .98, SRMR = .03, RMSEA = .06
(CI = .04, .07) for the Twitter group.
Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
14   JOURNAL OF HOSPITALITY & TOURISM RESEARCH

Figure 3
The Structural Model of Social Media Experience With Standardized Path
Coefficients

0.84* (0.84)* Attitude toward 0.92* (0.90)* Attitude toward

Social media
social media site hotel brand
experience R2 = 0.89 (0.83)
R2 = 0.71 (0.70)

ns (ns) 0.17* (0.20)*

0.57* (0.48)* 0.30* (0.36)*

Intention to Intention of
Brand
book hotel e-WOM
cognition *
0.51 (0.50) * R2 = 0.50 (0.52) R2 = 0.75 (0.77)
0.76* (0.75)*
*p < 0.001

Note: Values not in parentheses are parameter estimates for Facebook group. Values in
parentheses are parameter estimates for Twitter group.

We then examined the invariance of structural coefficients between two groups.

All structural coefficients were tested to be invariant between groups, ps > .05.
We also tested the invariance of factor covariance (brand cognition and social
media experience) between the two groups. The factor covariance between brand
cognition and social media experience was also found to be invariant across
groups, p > .05.
Combining two data sets, the final configural model represented a good model
fit. The goodness-of-fit indices for the final model were as follows: χ2(270,
N = 408) = 480.69, p < .001, CFI = .98, NNFI = .98, SRMR = .05, RMSEA =
.04 (CI = .04, .05). Figure 3 presents the structural part of the hypothesized model.
The dashed line represents the nonsignificant path.
The results showed that customers’ experiences on a hotel social media page
had a significant effect on their attitudes toward a social media site (β = .84 for
both the Facebook and Twitter groups, ps < .001). Customers’ attitudes toward a
social media site had a significant effect on their attitudes toward hotel brand
(β = .92 and β = .90 for the Facebook and Twitter groups, respectively, ps < .001).
Customers’ brand cognition did not have a significant effect on their attitudes
toward the hotel brand, p > .05. However, customers’ brand cognition had a
significant effect on their intentions to make hotel booking decisions (β = .51 and
β = .50 for the Facebook and Twitter groups, respectively, ps < .001). Customers’
attitudes toward hotel brand had significant effects on both their intentions to
make hotel booking decisions (β = .30 and β = .36 for the Facebook and Twitter
groups, respectively, ps < .001) and their intentions to spread positive word-of-
mouth online (β = .17 and β = .20 for the Facebook and Twitter groups, respectively,
ps < .001). Finally, customers’ intentions to make hotel booking decisions had a
significant effect on their intentions to spread positive word-of-mouth online
(β = .76 and β = .75 for the Facebook and Twitter groups, respectively, ps < .001).
The results also revealed a number of significant indirect effects among
constructs. Social media experience had a significant indirect effect on
Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
 Leung et al. / MARKETING EFFECTIVENESS OF SOCIAL MEDIA  15

attitude-toward-hotel-brand mediated by attitude-toward-social-media-page (β =

.78 and β = .75 for the Facebook and Twitter groups, respectively, ps < .001). It
also had a significant indirect effect on booking intention mediated by attitude-
toward-social-media-page and attitude-toward-hotel-brand (β =.24 and β = .27
for the Facebook and Twitter groups, respectively, ps < .001) and on intention of
eWOM mediated by attitude-toward-social-media-page, attitude-toward-hotel-
brand, and booking intention (β = .31 and β = .35 for the Facebook and Twitter
groups, respectively, ps < .001). Attitude-toward-social-media-page had a
significant indirect effect on booking intention mediated by attitude-toward-hotel-
brand (β = .28 and β = .32 for the Facebook and Twitter groups, respectively, ps
< .001) and on intention of eWOM mediated by attitude-toward-hotel-brand and
booking intention (β = .37 and β = .42 for the Facebook and Twitter groups,
respectively, ps < .001). Brand cognition had a significant indirect effect on
intention of eWOM mediated by booking intention (β = .40 and β = .39 for the
Facebook and Twitter groups, respectively, ps < .001). Attitude-toward-hotel-
brand had a significant indirect effect on intention of eWOM mediated by booking
intention (β = .23 and β = .27 for the Facebook and Twitter groups, respectively,
ps < .001).
Variances (R2) in attitude-toward-social-media-page accounted for by social
media experience were .71 for the Facebook group and .70 for the Twitter group.
Variances (R2) in attitude-toward-hotel-brand explained by attitude-toward-social-
media-page and brand cognition were .92 for the Facebook group and .90 for the
Twitter group. Variances (R2) in booking intention accounted for by attitude-
toward-hotel-brand, social media experience, and brand cognition were .50 for
the Facebook group and .52 for the Twitter group. Variances (R2) in intention of
eWOM attributed by attitude-toward-hotel-brand and booking intention were .75
for the Facebook group and .77 for the Twitter group. The correlations between
brand cognition and social media experience were significant for both the Facbook
group (r = .57) and the Twitter group (r = .48), ps < .001 (Figure 2).

DISCUSSION AND IMPLICATIONS

This study attempted to combine the Aad model with the concepts of attitude-
toward-social-media-page and eWOM to measure the marketing effectiveness of
social media in the hotel industry. Based on the Aad model, the study proposed
an integrative model to relate hotel customers’ social media experiences and brand
cognitions with their attitudes-toward-social-media-site, attitudes-toward-hotel-
brand, booking intentions, and intentions of eWOM. An online survey was
conducted by Qualtrics to test the hypothesized model in two different social
media contexts and the model was tested to fit the data in both the Facebook and
Twitter groups. Hypotheses 1, 2, 4, 5, and 6 were supported by the SEM results.
That is, a positive hotel social media experience will lead to positive attitude-
toward-social-media-page and positive attitude-toward-hotel-brand; thus, a hotel
customer’s intention to book this hotel room and to recommend this hotel brand
online will increase. However, Hypothesis 3 was refuted, suggesting that a hotel
customer’s brand Downloaded
cognition does not affect his/her attitude-toward-hotel-brand.
from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
16   JOURNAL OF HOSPITALITY & TOURISM RESEARCH

Hypotheses 7a, 7b, 7c, 7d, 7e, and 7f were all refuted by the invariance test of
path coefficients, suggesting that all hypothesized relationships in the model were
not statistically different between Facebook and Twitter. Besides, there was one
new significant relationship revealed by the SEM results, which was not proposed
in the model. It was the relationship between brand cognition and hotel booking
intention. The relationship was not statistically different between the Facebook
and Twitter groups.
The support of the hypothesis regarding the relationship between social media
experience and attitude-toward-social-media-page suggests that if a hotel customer
enjoys the hotel social media page, the customer’s attitude toward the hotel social
media page is more positive. This finding is consistent with Bruner and Kumar’s
(2000) notion that an individual’s web experience has a positive effect on Aws.
Their study also suggested that complexity and interestingness are two important
factors affecting individual’s web experience (Bruner & Kumar, 2000). Thus, if
a hotel wants to achieve favorable attitudes toward a social media page from its
customers, it should try to enhance its customers’ experiences with its social
media page. Specifically, a hotel should make its social media page more
interesting, appealing, informative, interactive, and consumer centric so the
customers can enjoy their social media experiences. Besides, when customers
feel the hotel social media page is useful, valuable, and beneficial, their experiences
are more likely to be positive.
The support of the hypothesis regarding the relationship between attitude-
toward-social-media-page and attitude-toward-hotel-brand suggests that if a hotel
customer has a favorable attitude toward the hotel social media page, the customer’s
attitude toward the hotel brand is more positive. This finding is consistent with
all studies involving the Aws model, which tests that Aws is positively correlated
with attitude-toward-the-brand (Bruner & Kumar, 2000; Poh & Adam, 2002;
Stevenson et al., 2000). It demonstrates the importance of social media marketing
since hotel customers’ attitudes toward hotel social media pages positively affect
their attitudes toward the hotel brand. Thus, it is very important to leverage social
media pages to impress customers and change their attitudes in becoming more
favorable to the brand.
The disapproval of the hypothesis regarding the relationship between brand
cognition and attitude-toward-hotel-brand indicates that no matter what a customer
knows about a hotel brand, the customer’s attitude toward the hotel brand does
not change. This finding is interesting because it is contrary to the proposed
relationship in the Aad model that brand cognition is positively related to attitude-
toward-the-brand (Mitchell & Olson, 1981; Shimp, 1981). However, in the study,
brand cognition is tested to directly affect customers’ hotel booking intention
instead of affecting booking intention indirectly through the mediation of attitude-
toward-the-brand. This finding indicates that a hotel’s potential customers will
book the hotel as long as they are familiar with the hotel brand. Therefore, it
implies that hotels need to market their brand name heavily so that the potential
customers can strongly identify with the hotel brand. Thus, when they need to
book a hotel room, the preconceived hotel brand will influence their booking
decision. This finding alsofromsuggests
Downloaded another
jht.sagepub.com alternative
at Selcuk Universitesi hypothesis
on February 10, 2015 of Aad model
 Leung et al. / MARKETING EFFECTIVENESS OF SOCIAL MEDIA  17

besides the four Aad models proposed in the literature. In this alternative Aad
model, brand cognition directly affects purchase intention, whereas attitude toward
the ad indirectly affects purchase intention through the mediation of attitude
toward the brand.
The support of the hypothesis regarding the relationship between attitude-
toward-hotel-brand and hotel booking intention suggests that if a hotel customer
has a favorable attitude toward the hotel brand, the customer is more willing to
book this hotel in the future. This finding supports the Aad model and the Aws
model, both of which proposed that attitude-toward-the-brand has a positive
impact on purchase intention (Bruner & Kumar, 2000; Mitchell & Olson, 1981).
Thus, hotels should try to use their social media pages to build a positive and
impressive hotel brand image to build their potential customers’ affection for their
hotel brand. This suggests that social media marketing should not only be used
for promotions but also for brand building purposes.
The support of the hypothesis regarding the relationship between attitude-
toward-hotel-brand and intention of (positive) eWOM implies that if a hotel
customer has a favorable attitude toward the hotel brand, the customer is more
willing to recommend this hotel online to others. Studies on electronic word-of-
mouth have identified consumers’ affective elements as factors motivating them
to share experiences and recommend products to others (Dichter, 1966;
Neelamegham & Jain, 1999; Nyer, 1997). This study extends this finding by the
notion that electronic word-of-mouth can also be triggered by customers’ favorable
attitude toward the hotel brand. The positive eWOM is another important outcome
of social media marketing. Previous studies demonstrate that word-of-mouth and
eWOM can change consumers’ attitudes toward the brand and even influence
their purchase intentions (Arndt, 1967; Day, 1971; Jones, Aiken, & Boush, 2009).
Thus, attitude toward hotel brand, booking intention, and eWOM constitute a
positive feedback loop. Therefore, to build and promote hotel brand online is of
significance for hotels to stimulate and manage eWOM on social media sites.
The support of the hypothesis regarding the relationship between booking
intention and intention of eWOM suggests that if a hotel customer has a stronger
intention to book the hotel, the customer is more willing to recommend this hotel
online to others. The study also found that the relationship between booking
intention and intention of eWOM is much stronger than the relationships between
attitude-toward-hotel-brand and booking intention and intention of eWOM. This
implies that the two marketing outcomes of social media marketing, booking
intention and intention of eWOM, are closely related to or even tied to each other.
One positive outcome leads to the others positive outcome, and vice versa. Since
in the literature WOM as marketing outcome fell short, this finding supports the
statement that both purchase intention and WOM are important marketing
outcomes.
There were a lot of significant indirect effects revealed in the study. First,
social media experience has an indirect effect on booking intention and the
intention of eWOM, strengthening the importance of social media experience in
social media marketing. When customers enjoy their social media experience,
they tend to have Downloadedpositive attitudes-toward-social-media-site
from jht.sagepub.com at Selcuk Universitesi on February 10, 2015 and
18   JOURNAL OF HOSPITALITY & TOURISM RESEARCH

attitudes-toward-hotel-brand, which, in turn, increases their booking intention

and intention to spread positive WOM on social media sites. Second, there are
the indirect effects of attitude-toward-social-media-page on booking intention
and the intention of eWOM. There also exist indirect effects of brand cognition
and attitude-toward-hotel-brand on intention of eWOM. All these indirect effects
demonstrate that attitude-toward-social-media-page and attitude-toward-hotel-
brand are important mediators of social media marketing effects, which is
consistent with Lutz et al.’s (1983) and Bruner and Kumar’s (2000) findings that
Aad and Aws are significant mediators of advertising effects.
The invariance test of path coefficients showed that all the relationships among
the constructs are invariant between the Facebook and the Twitter groups. This
finding implied that different social media sites share the same marketing
mechanism. Thus, hotels can use the same marketing strategy in marketing on
different social media sites. The important factors that drive customer’s booking
intention and eWOM are not what social media site the hotel uses, but how
customers feel about the social media experience and what their attitudes are.
Therefore, no matter whether the hotel use Facebook or Twitter as marketing tool,
the managers should focus on the social media experience they create for customers
and the brand building on social media. This finding makes hotel social media
marketing simpler since hotels can employ the same strategy and tactics in using
different social media sites. Finally, the correlation coefficients between brand
cognition and social media experience were significant in both social media
groups. There was no difference in correlation coefficients across the two groups.
This finding indicates that hotel customers’ prebeliefs of the hotel brand are
positively related to their experiences on hotel social media pages. Assuming a
customer has higher brand cognition toward a hotel, the customer’s social media
experience is more likely to be positive.

CONCLUSION AND LIMITATIONS

The purpose of this study was to propose a theoretical framework to understand

the marketing effectiveness of different social media sites in the hotel industry.
The Aad model and the concepts of Aws and eWOM have been introduced into
the social media context to form the hypothesized model. Results of the study find
that hotel customers’ social media experiences influence their attitudes-toward-
social-media-site, which in turn influences their attitudes-toward-hotel-brand. Then
their attitudes-toward-hotel-brand affect their hotel booking intentions and eWOM,
which are two major marketing outcomes of hotel social media marketing. Contrary
to the Aad model, hotel customers’ brand cognition does not affect their attitudes-
toward-hotel-brand. The goodness of fit of the model to the data further suggests
that the Aad model serves as an appropriate theoretical framework to analyze the
marketing effectiveness of social media. The other important finding of the study
is that different social media sites show the same marketing mechanism in the
marketing effectiveness model. Thus, hotel can employ the same marketing strategy
and tactics in using different social media sites.
Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
 Leung et al. / MARKETING EFFECTIVENESS OF SOCIAL MEDIA  19

This study makes valuable contributions to both the academia and industry.
From a theoretical perspective, the research on the marketing effectiveness of
social media is very limited. This study is one of the first attempts to apply the
Aad model, a commonly used model in advertising area, into social media context
to explore the marketing effectiveness of social media. The study provides
empirical evidence to support the use of the Aad model combined with the Aws
concept in the social media field. That is, a customer’s attitude toward hotel social
media page affects his/her attitude toward hotel brand, which in turn influence
his/her hotel booking intention. Besides, the study also suggests another alternative
hypothesis of Aad model besides the four commonly used ones in the literature,
in which brand cognition directly affects purchase intention. In terms of
methodology, this study attempts to identify the marketing effectiveness of
different social media sites using multigroup SEM. Since multigroup SEM is
rarely used in the hospitality and tourism field, this study extends the literature
by introducing a new theory and applying a new method for future research.
In a practical sense, hotel managers can use the findings of this study to better
their social media marketing plans. The study identified social media experience
as a very important construct in social media marketing, indicating that hotels
need to focus on how to create a positive and satisfactory experience for their
social media users. As a result, hotels should focus on social media page design
to enhance customers’ experiences. Specifically, a hotel should make its social
media pages more interesting, appealing, informative, interactive, and consumer
centric so the customers can enjoy their social media experiences. The hotel social
media pages should also be useful, valuable, and beneficial to customers. Another
important factor leading to hotel booking intention is brand cognition. Thus, a
hotel should try to enroll more existing customers or even loyal customers as fans
on its Facebook page or followers on its Twitter page. Since existing customers
are found to have higher brand cognition toward the hotel, they are more likely
to be influenced by social media marketing and are more likely to book hotel
rooms in the future. Besides, attitude-toward-the-hotel-brand was found as an
important mediator for social media marketing to achieve effective outcomes.
Thus, hotels should emphasize brand building on their social media pages. Hotel
managers should remember that the quintessential function of social media page
is not to exchange information, but to enhance brand reputation. The same
marketing mechanism across different social media sites let hotel managers
simplify their social media marketing strategy and tactics. Managers can employ
the same marketing tactics focusing on the major factors in the marketing
effectiveness model when using different social media sites. That is, hotels should
focus on customers’ social media experience, brand cognition, and attitudes of
customers regardless what social media site they are using as the marketing tool.

Limitations and Future Research

As is always the case with any research there are several limitations of this
examination that should be noted. One major limitation of this study is the inclusion
of data only from two social media sites using one hotel brand as sample, rendering
Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
20   JOURNAL OF HOSPITALITY & TOURISM RESEARCH

its inability to generalize research findings. Since this study is an exploratory

study in the hotel social media marketing area, further work using more hotel
brands and more social media sites to test the model proposed in the study would
certainly be beneficial to hospitality researchers from better generalizing the
results. Also, the measurement of the brand cognition construct should be further
refined because of the low internal consistency (reliability) and factor loadings.
Besides, the time the participants spent on the specific Facebook page or Twitter
page was not controlled. This may have also affected their social media experience
and the final results. Another notion of the study was the large range of ages in
the study sample. In this study, the impacts of the sample demographics on their
attitudes and behavior intentions were not considered in the hypotheses and the
model, which might be an area of future study. Finally, the study compared the
marketing effectiveness of Facebook and Twitter because these two are the most
commonly used social media channels. However, as we all know, Facebook and
Twitter have different functions. Facebook marketing focuses on the website,
whereas Twitter marketing uses mobile devices. These different functions may
affect the comparison results. Future research should address the limitations of
this study and explore the application of the Aad model and the Aws model in
other contexts in the hospitality and tourism field.

REFERENCES

Alarcó-del-Amo, M., Lorenzo-Romero, C., & Gómez-Borja, M. (2011). Classifying and

profiling social networking site users: A latent segmentation approach.
Cyberpsychology, Behavior, and Social Networking, 14, 547-553.
Arndt, J. (1967). Role of product-related conversations in the diffusion of a new product.
Journal of Marketing Research, 4, 291-295.
Bandalos, D. L. (2002). The effects of item parceling on goodness-of-fit and parameter
estimate bias in structural equation modeling. Structural Equation Modeling, 9, 78-102.
Bruner, G. C., II, & Kumar, A. (2000). Web commercials and advertising hierarchy-of-
effects. Journal of Advertising Research, 40(1/2), 35-42.
Buil, I., de Chernatony, L., & Martinez, E. (2008). A cross-national validation of the
customer-based brand equity scale. Journal of Product & Brand Management, 17(6),
384-92.
Casaló, L. V., Flavián, C., & Guinalíu, M. (2010). Determinants of the intention to
participate in firm-hosted online travel communities and effects on consumer behavioral
intentions. Tourism Management, 31, 898-911.
Chan, Y. Y., & Ngai, E. W. T. (2011). Conceptualising electronic word of mouth activity:
An input-process-output perspective. Marketing Intelligence & Planning, 29,
488-516.
Chen, Q., & Wells, W. D. (1999). Attitude toward the site. Journal of Advertising Research,
39(5), 27-37.
Cheng, S., Lam, T., & Hsu, C. H. C. (2006). Negative word-of-mouth communication
intention: An application of the theory of planned behavior. Journal of Hospitality &
Tourism Research, 30, 95-116.
Chiang, C., & Jang, S. (2006). The effects of perceived price and brand image on value
and purchase intention: Leisure travelers’ attitudes toward online hotel booking.
Journal of Hospitality & Leisure Marketing, 15(3), 49-69.
Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
 Leung et al. / MARKETING EFFECTIVENESS OF SOCIAL MEDIA  21

Choi, Y. K. (2000). Effects of presence on the effectiveness of web site advertising

(Doctoral dissertation). Available from ProQuest Dissertations and Theses database.
(UMI No. 9985366)
Cobb-Walgren, C. J., Ruble, C. A., & Donthu, N. (1995). Brand equity, brand preference,
and purchase intent. Journal of Advertising, 24(3), 25-40.
Constantinides, E., & Fountain, S. J. (2008). Web 2.0: Conceptual foundations and
marketing issues. Journal of Direct, Data, and Digital Marketing Practice, 9,
231-244.
Cooke, M., & Buckley, N. (2008). Web 2.0, social networks and the future of market
research. International Journal of Market Research, 50, 267-292.
Day, G. S. (1971). Attitude change, media and word of mouth. Journal of Advertising
Research, 11(6), 31-40.
Dichter, E. (1966). How word-of-mouth advertising works. Harvard Business Review,
44(6), 147-166.
Edell, J. A., & Burke, M. C. (1984). The moderating effect of attitude toward an ad on ad
effectiveness under different processing conditions. In T. C. Kinnear (Ed.), Advances
in consumer research (Vol. 11, pp. 644-649). Provo, UT: Association for Consumer
Research.
The eMarketer. (2011, November 22). Facebook still top social marketing venue for SMBs.
Retrieved from http://www.emarketer.com/Article.aspx?R=1008703
Friebe, K., & Campbell, D. (2010). Social media in the hospitality and leisure industry. DLA
Piper. Retrieved from http://www.dlapiper.com/files/Publication/157721b8-d5ee-45ce-
8911-11451f0738df/Presentation/PublicationAttachment/a9f81f84-7cbb-49c1-865d-
19c9283af4d3/DLA_Piper_UK_Social_Media_in_Hospitality_and_Leisure_Sector.pdf
Godes, D., & Mayzlin, D. (2004). Using online conversations to study word-of-mouth
communication. Marketing Science, 23, 545-560.
Gruen, T. W., Osmonbekov, T., & Czaplewski, A. J. (2006). eWOM: The impact of
customer-to-customer online know-how exchange on customer value and loyalty.
Journal of Business Research, 59, 449-456.
Hampton, K., & Goulet, L. S. (2011). Social networking sites and our lives (Pew Internet &
American Life Project). Retrieved from http://pewinternet.org/~/media//Files/
Reports/2011/PIP%20-%20Social%20networking%20sites%20and%20our%20lives.pdf
Heider, F. (1946). Attitudes and cognitive organization. Journal of Psychology, 21,
107-112.
Hennig-Thurau, T., Gwinner, K. P., & Gremler, D. D. (2002). Understanding relationship
marketing outcomes: An integration of relational benefits and relationship quality.
Journal of Service Research, 4, 230-247.
Hennig-Thurau, T., Gwinner, K. P., Walsh, G., & Gremler, D. D. (2004). Electronic word-
of-mouth via consumer-opinion platforms: What motivates consumers to articulate
themselves on the internet? Journal of Interactive Marketing, 18(1), 38-52.
Hennig-Thurau, T., & Walsh, G. (2003). Electronic word-of-mouth: Motives for and
consequences of reading customer articulations on the Internet. International Journal
of Electronic Commerce, 8(2), 51-74.
Herr, P. M., Kardes, F. R., & Kim, J. (1991). Effects of word-of-mouth and product
attribute information on persuasion: An accessibility-diagnosticity perspective. Journal
of Consumer Research, 17, 454-462.
Holbrook, M. P. (1978). Beyond attitude structure: Toward the informational determinants
of attitude. Journal of Marketing Research, 15, 545-556.
Homer, P. M. (1990). The mediating role of attitude toward the Ad: Some additional
evidence. Journal of Marketing Research, 27(1), 78-86.
Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
22   JOURNAL OF HOSPITALITY & TOURISM RESEARCH

Hotelmarketing.com. (2011). One quarter of hotels not using social media. Retrieved from
http://hotelmarketing.com/index.php/content/article/one_quarter_of_hotels_
not_using_social_media
Howard, J. A. (1977). Consumer behavior: Application of theory. New York, NY:
McGraw-Hill.
Infographic: Facebook vs Twitter demographics. (2010, December 21). Retrieved from
http://www.digitalbuzzblog.com/infographic-facebook-vs-twitter-demographics-
2010-2011/
Jeong, M., & Choi, J. (2004). Effects of picture presentations on customers’ behavioral
intentions on the web. Journal of Travel & Tourism Marketing, 17, 193-204.
Jones, S. A., Aiken, K. D., & Boush, D. M. (2009). Integrating experience, advertising,
and electronic word of mouth. Journal of Internet Commerce, 8, 246-267.
Kim, W. G., Jin-Sun, B., & Kim, H. J. (2008). Multidimensional customer-based brand
equity and its consequences in midpriced hotels. Journal of Hospitality & Tourism
Research, 32, 235-254.
Leach, M. P., Liu, A. H., & Winsor, R. D. (2008). The impact of attitudes, word-of-mouth,
and value congruence on conference participation: A comparison of attending and
non-attending organizational members. Journal of Hospitality & Leisure Marketing,
16, 249-269.
Litvin, S. W., Goldsmith, R. E., & Pan, B. (2008). Electronic word-of-mouth in hospitality
and tourism management. Tourism Management, 29, 458-468.
Lutz, R. J., & Swasy, J. (1977). Integrating cognitive structure and cognitive response
approaches to monitoring communication effects. In W. Perreault Jr. (Ed.), Advances
in consumer research (Vol. 4, pp. 363-371). Atlanta, GA: Association for Consumer
Research.
Lutz, R. J., MacKenzie, S. B., & Belch, G. E. (1983). Attitude toward the Ad as a mediator
of advertising effectiveness: Determinants and consequences. In R. P. Bagozzi &
A. M. Tybout (Eds.), Advances in consumer research (Vol. 10, pp. 532-539). Ann
Arbor, MI: Association for Consumer Research.
Mabry, E., & Porter, L. (2010). Movies and MySpace: The effectiveness of official web
sites versus online promotional contests. Journal of Interactive Advertising, 10(2),
1‐15.
MacKenzie, S. B., Lutz, R. J., & Belch, G. E. (1986). The role of attitude toward the Ad
as a mediator of advertising effectiveness: A test of competing explanations. Journal
of Marketing Research, 23, 130-143.
Madden, M., & Zickuhr, K. (2011, August 26). 65% of online adults use social networking
sites. Retrieved from http://pewinternet.org/~/media//Files/Reports/2011/PIP-SNS-
Update-2011.pdf
Merriam-Webster. (2012). Social media. Retrieved from http://www.merriam-webster.
com/dictionary/social%20media
McMillan, S. J., Hwang, J., & Lee, G. (2003). Effects of structural and perceptual factors
on attitudes toward the website. Journal of Advertising Research, 43, 400-409.
Miller, D. W., & Stoica, M. (2004). Comparing the effects of a photograph versus artistic
renditions of a beach scene in a direct-response print ad for a Caribbean resort island:
A mental imagery perspective. Journal of Vacation Marketing, 10(1), 11-21.
Mitchell, A. A., & Olson, J. C. (1981). Are product attribute beliefs the only mediator of
advertising effects on brand attitude? Journal of Marketing Research, 18, 318-332.
Moital, M., Vaughan, R., Edwards, J., & Peres, R. (2009). Determinants of intention to
purchase over the Internet. Anatolia: An International Journal of Tourism and
Hospitality Research, 20, 345-358.
Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015
 Leung et al. / MARKETING EFFECTIVENESS OF SOCIAL MEDIA  23

Neelamegham, R., & Jain, D. (1999). Consumer choice process for experience goods: An
econometric model and analysis. Journal of Marketing Research, 36, 373-386.
Nyer, P. U. (1997). A study of the relationships between cognitive appraisals and
consumption emotions. Journal of the Academy of Marketing Science, 25, 296-304.
Poh, D. M. H., & Adam, S. (2002). An exploratory investigation of attitude toward the
website and the advertising hierarchy of effects. In A. Treloar & A. Ellis (Eds.),
Proceedings of AusWeb 02, the 8th Australian World Wide Web Conference (pp. 620-631).
Queensland, Australia: Southern Cross University. Retrieved from http://ausweb.scu.edu.
au/aw02/papers/refereed/poh/paper.html
Shimp, T. A. (1981). Attitude toward the Ad as a mediator of consumer brand choice.
Journal of Advertising, 10(2), 9-14.
So, K. K. F., & King, C. (2010). “When experience matters”: Building and measuring
hotel brand equity: The customers’ perspective. International Journal of Contemporary
Hospitality Management, 22, 589-608.
Stelzner, M. A. (2009). Social media marketing industry report: How marketers are using
social media to grow their businesses. Retrieved from http://marketingwhitepapers.
s3.amazonaws.com/smss09/SocialMediaMarketingIndustryReport.pdf
Stelzner, M. A. (2010). 2010 social media marketing industry report: How marketers are
using social media to grow their businesses. Retrieved from http://marketingwhitepapers.
s3.amazonaws.com/SocialMediaMarketingReport2010.pdf
Stelzner, M. A. (2011). 2011 social media marketing industry report: How marketers are
using social media to grow their businesses. Retrieved from http://www
.socialmediaexaminer.com/SocialMediaMarketingReport2011.pdf
Stevenson, J. S., Bruner, G. C., II, & Kumar, A. (2000). Webpage background and viewer
attitudes. Journal of Advertising Research, 40(1), 29-34.
Sundaram, D. S., Mitra, K., & Webster, C. (1998). Word-of-mouth communications: A
motivational analysis. Advances in Consumer Research, 25, 527-531.
Top 15 most popular social networking sites. (2011, November 21). Retrieved from http://
www.ebizmba.com/articles/social-networking-websites
Whyte, W. H., Jr. (1954). The web of word of mouth. Fortune, 50, 140-212.

Submitted March 7, 2012

Accepted September 5, 2012
Refereed Anonymously

Xi Y. Leung (e-mail: [email protected]) is an assistant professor in the Department

of Nutrition and Hospitality Management at the University of Mississippi. Billy Bai (e-mail:
[email protected]) is an associate professor in the William F. Harrah College of Hotel
Administration at University of Nevada, Las Vegas. Kurt A. Stahura (e-mail: kurt.
[email protected]) is a professor and director of student affairs at University of Nevada,
Las Vegas, Singapore campus.

Downloaded from jht.sagepub.com at Selcuk Universitesi on February 10, 2015