Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=rjmm20

Download by: [ James Madison University] Date: 04 October 2016, At: 20:01

Journal of Marketing Management

ISSN: 0267-257X (Print) 1472-1376 (Online) Journal homepage: http://www.tandfonline.com/loi/rjmm20

Brand engagement on social media: will firms’social media efforts influence search engineadvertising effectiveness?

Shuai Yang, Shan Lin, Jeffrey R. Carlson & William T. Ross Jr.

To cite this article: Shuai Yang, Shan Lin, Jeffrey R. Carlson & William T. Ross Jr. (2016)Brand engagement on social media: will firms’ social media efforts influence searchengine advertising effectiveness?, Journal of Marketing Management, 32:5-6, 526-557, DOI:10.1080/0267257X.2016.1143863

To link to this article: http://dx.doi.org/10.1080/0267257X.2016.1143863

Published online: 23 Feb 2016.

Submit your article to this journal

Article views: 1079

View related articles

View Crossmark data

Citing articles: 1 View citing articles

Brand engagement on social media: will firms’ social mediaefforts influence search engine advertising effectiveness?Shuai Yanga*, Shan Linb*, Jeffrey R. Carlsonc and William T. Ross Jr.d

aSchool of Business and Management, Donghua University, Shanghai, China; bCollege of Business, LewisUniversity, Romeoville, IL, USA; cRobins School of Business, University of Richmond, Richmond, VA, USA;dSchool of Business, University of Connecticut, Storrs, CT, USA

ABSTRACTAlthough firms leverage social media platforms such asFacebook to engage with customers, they often treat socialmedia elements and other online marketing activities such assearch engine advertising as stand-alone, rather than part of anintegrated online activities system. Hence, it is important tosystematically understand how specific elements of socialmedia, signifying and representing behavioural manifestationsof brand engagement, relate to other online activities. We studyhow three types of brand engagement on social media – affilia-tion, conversation and responsiveness – influence search engineadvertising effectiveness, including click-through rate andconversion rate. Specifically, we find that affiliation,conversation and responsiveness increase click-through rateand conversion rate. Moreover, brand engagement on socialmedia strengthens the relationships between advertisementrank and search engine advertising effectiveness.

ARTICLE HISTORYReceived 20 June 2015Accepted 15 December 2015

KEYWORDSBrand engagement on socialmedia; search engineadvertising; social media;advertising effectiveness;advertisement rank

Introduction

Marketing executives and brand managers have a long-standing interest in engagingwith consumers through social media (De Vries, Gensler, & Leeflang, 2012). Such effortsare defined as building brand engagement on social media, which refers to theinteractive behaviours among consumers and brands. Consequently, according to arecent US survey of 351 marketing executives, spending on social media platformsnow represents 9% of marketing budgets1 (Tadena, 2014). From a consumer’s point ofview and compared to traditional media, social media provides a platform for two-waydialogue between consumers and brands; it also provides a means to seek and discovernew brands, compare alternative brands and read comments and reviews from otherconsumers.

Firms’ efforts in engaging with customers through social media have gainedincreasing attention from academic researchers. Some research has demonstratedthe critical role of engagement with consumers through social media (Habibi,

CONTACT Shan Lin linsh@lewisu.edu

*Shuai Yang and Shan Lin have contributed equally to this article.

JOURNAL OF MARKETING MANAGEMENT, 2016VOL. 32, NOS. 5–6, 526–557http://dx.doi.org/10.1080/0267257X.2016.1143863

© 2016 Westburn Publishers Ltd.

Laroche, & Richard, 2014; Hollebeek, Glynn, & Brodie, 2014; Lin, Ross, & Liu, 2015). Forexample, Lin et al. (2015) find that brand engagement on social media is positivelyassociated with a firm’s financial performance. Hollebeek et al. (2014) show thatconsumer engagement in social media can increase self-brand connection andbrand usage intent. However, there is scant research investigating relationshipsbetween brand engagement on social media and online advertising. Althoughresearch has linked some types of engagement to online banner advertising (e.g.Calder, Malthouse, & Schaedel, 2009), it has not explored how engagement via onlinesocial media platforms influences online advertising effectiveness, specifically searchengine advertising. Search engine advertising is the predominant form of onlineadvertising (Rutz & Bucklin, 2011). It is a service offered by Internet search enginesthrough which advertisers select specific keywords and create textual advertisingbased on consumers’ search entries.

It should be noted that marketers recognise that they face a similar problem: in thesame survey of marketing executives, only 15% of marketers claimed that they can showthe impact of social media on their businesses (Tadena, 2014). There is evidence that, asfirms continue to leverage online media to better reach and engage with customers,they often treat social media platforms (e.g. Facebook) and other online marketingstrategies (e.g. search engine advertising) as stand-alone elements, rather than part ofan integrated system (Hanna, Rohm, & Crittenden, 2011). Therefore, we ask: how doesbrand engagement on social media influence the effectiveness of search engineadvertising?

To explore the relationship between brand engagement on social media and searchengine advertising effectiveness, we collected industry data from two different sources:brand engagement data from Facebook and search engine advertisement data fromGoogle. We examined two outcomes for search engine advertisements: (1) click-throughrate, a measure of the degree to which consumers’ attention was attracted by the ad;and (2) conversion rate, a measure of the sales return from the advertisement. Resultsshow that brand engagement on social media positively influences the effectiveness ofsearch engine advertising (i.e. click-through rate and conversion rate). We also find thatbrand engagement on social media amplifies the positive effect of a top advertisementrank on search engine advertising effectiveness.

Our study contributes to academic research in several ways. First, we contribute toresearch on engagement. Concepts related to ‘engagement’ are gaining popularity inthe marketing literature, reflected by a wide range of concepts: consumer engagement(e.g. Brodie, Hollebeek, Jurić, & Ilić, 2011; Brodie, Ilic, Juric, & Hollebeek, 2013); consumerbrand engagement (e.g. Hollebeek, 2011a, 2011b; Hollebeek et al., 2014); brandengagement in self-concept (Sprott, Czellar, & Spangenberg, 2009); onlineengagement (Calder et al., 2009); and brand community engagement (e.g.Algesheimer, Dholakia, & Herrmann, 2005; Baldus, Voorhees, & Calantone, 2015; Habibiet al., 2014). Despite significant interest in engagement, there is limited researchexamining consumer–brand relationships in social media. We contribute to this streamof literature by providing a lens through which to view and measure the relationshipbetween brand engagement on social media and search engine advertising. Specifically,we identify three dimensions of brand engagement on social media – affiliation,conversation and responsiveness – and demonstrate that both consumer-initiated

JOURNAL OF MARKETING MANAGEMENT 527

social media engagement efforts (i.e. affiliation and conversation) and firm-initiatedengagement efforts (i.e. responsiveness) are important in influencing search engineadvertising effectiveness.

Second, we contribute to the literature on search engine advertisement rank. Agrowing body of research has examined the effect of search engine advertisementrank on advertisement performance. These studies indicate that the lower (better)the rank of the advertisement in the search list, the higher is the advertisement’sclick-through rate and/or conversion rate, and that the effect of rank can be influencedby: keyword characteristics (Ghose & Yang, 2009; Rutz, Trusov, & Bucklin, 2011; Yang &Ghose, 2010); the advertisers’ positioning strategy (Animesh, Viswanathan, & Agarwal,2011); and consumers’ knowledge and beliefs about firm qualities (Jerath, Ma, Park, &Srinivasan, 2011). We extend this line of work by examining the moderating role ofbrand engagement on social media on the relationship between search engineadvertisement rank and advertisement performance.

Finally, our findings offer practical managerial advice. Overall, we provide evidencethat brand engagement on social media and online advertising such as search engineadvertising should not be considered independent marketing tools, but part of anintegrated online marketing strategy (Hanna et al., 2011). Specifically, we provide firmsand managers with three different ways – affiliation, conversation and responsiveness– to understand, and measure their level of brand engagement on social media. In turn,this provides a basis that helps managers better understand how their social mediaefforts relate to search engine advertising. For instance, if a firm has a limited advertisingbudget and cannot pay for a higher rank in search engine advertising, they could investin social media such as Facebook to facilitate interaction between their brand(s) andconsumers. In turn, these efforts can improve search engine advertising effectiveness.Overall, the contributions of our study are threefold: (1) we contribute to the strong andgrowing literature related to brand and consumer engagement (e.g. Hollebeek, 2011a,2011b; Hollebeek et al., 2014); (2) we contribute to the literate related to search engineadvertising and (3) we provide practical managerial advice on how to leverage socialmedia to boost the effectiveness of search engine advertising.

In the next section, we provide an overview of brand engagement on social media,introduce search engine advertising and discuss the theoretical background about howbrand engagement on social media should influence search engine advertisingeffectiveness. After presenting the data, methodology and results, we conclude with adiscussion of managerial implications and suggestions for future research.

Theoretical background

Marketing literature has demonstrated that the Internet has fundamentally changed theway consumers generate and obtain consumption-related information. In particular,consumers rely on various Internet-based information sources to make purchasedecisions (Anderson & Palma, 2012; Ghose, Ipeirotis, & Li., 2014), including using socialmedia platforms’ brand-related content with others (Xiang & Gretzel, 2010), and usingsearch engines to obtain brand-related information (Xiang, Wöber, & Fesenmaier, 2008).Consequently, firms have increasingly invested in multiple online marketing channels,including social media and search engine advertising. Research also suggests that the

528 S. YANG ET AL.

relationship between social media and search engines is evolving in how one influencesthe other (Blackshaw & Nazzaro, 2006; Gretzel, 2006; Xiang & Gretzel, 2010). Thus, themajor goal of this study is to explore how brand engagement on social media impactsthe effectiveness of search engine advertising.

Next, we discuss brand engagement on social media, followed by a discussion onhow three dimensions of brand engagement on social media (i.e. affiliation,conversation and responsiveness) are associated with search engine advertisementperformance.

Brand engagement on social media

Brand engagement on social media stems from a multidisciplinary theoreticalperspective – including marketing (Lin et al., 2015), sociology (Bourdieu, 1986),psychology (Achterberg et al., 2003) and information systems (Millen & Patterson,2002) – and from marketing practice (Weinberg & Pehlivan, 2011; Zinnbauer & Honer,2011). The practical foundation of brand engagement on social media draws from aconcept known as ‘social currency’, the extent to which people share brand-relatedinformation with others as part of their everyday social lives (Zinnbauer & Honer, 2011).The primary theoretical foundation of brand engagement on social media originatesfrom Bourdieu’s social capital theory, which suggests that social networks have a rangeof value, and are dependent upon network size, and interactivity (Bourdieu, 1986;Zinnbauer & Honer, 2011).

Complementing this line of work, and underscoring the importance of interactivity,research in psychology and information systems recognises the importance ofstimulating engagement in social contexts (i.e. involvement with social activities andinteractions) in contexts such as online community networks (Achterberg et al., 2003;Millen & Patterson, 2002; Zhang, Jiang, & Carroll, 2011). For example, Zhang et al. (2011)find that social engagement, defined as ‘the commitment of a member to stay in thegroup and interact with other members’ reinforces social identities and social capitalwith the affordances of Facebook. In addition, Millen and Patterson (2002) suggest thatcreating online mechanisms that facilitate conversations also stimulates interactivity.Consistent with this work, brand engagement on social media extends the practicalapplication of social currency by capturing how consumers interact with otherconsumers and brands in discussing a brand in online social platforms. Formally, andparalleling previous engagement research (Gummerus, Liljander, Weman, & Pihlström,2012; Van Doorn et al., 2010), we define brand engagement as the customers’behavioural manifestation towards a brand – beyond purchase – resulting frommotivational drivers, which is captured through the interactive behaviours betweenconsumers and brands. While we recognise that this definition applies to both onlineand offline contexts, we draw specific attention to brand engagement on social mediawhich captures the extent to which people use social media to engage with brands andinteract with other consumers as part of their everyday lives.

Brand engagement on social media parallels a wide range of work related toengagement (Brodie et al., 2013; Hollebeek et al., 2014). According to Hollebeek et al.(2014), there are many engagement-based concepts, reflecting the nascentdevelopmental state of engagement research in marketing. Broadly, engagement

JOURNAL OF MARKETING MANAGEMENT 529

concepts such as consumer brand engagement, represent a two-way interactionbetween a customer and an engagement object (e.g. brand, firm) and arecharacterised by specific levels of cognitive, emotional and behavioural activities(Hollebeek, 2011a). In this research, and consistent with previous works (Gummeruset al., 2012; Van Doorn et al., 2010), we focus on brand engagement in a specific context,social media, and focus on the behavioural manifestations of brand engagement.According to Gummerus et al. (2012), this type of customer engagement is directlyrelated to the emergence of new media, which recognises that consumers carry out anumber of brand-related behaviours that did not exist a decade ago, such as onlinediscussions, and participating in online brand communities.

As depicted in Table 1, few studies have focused on engagement in an onlinecontext, such as online websites (Calder et al., 2009), social media (Hollebeek et al.,2014) and online brand communities (Gummerus et al., 2012). Moreover, and with

Table 1. An overview of selected engagement research in marketing.

Author(s) ConceptEngagement perspective (firm- or consumer-

initiated vs. both brand- and consumer-initiated)Focal research

context

This article Brand engagementon social media

Firm- and consumer-initiated Social media

Avnet and Higgins (2006a) Engagement Consumer-initiated –Gummerus et al. (2012) Customer

engagementConsumer-initiated Online brand

communityHiggins (2006) Engagement Consumer-initiated –Higgins and Scholer (2009) Engagement Consumer-initiated –Abdul-Ghani, Hyde, andMarshall (2011)

Engagement Consumer-initiated Website

Vivek, Beatty, and Morgan(2010)

Consumerengagement

Consumer-initiated –

Mollen and Wilson (2010) Consumerengagement

Consumer-initiated Website

Brodie et al. (2013) Consumerengagement

Consumer-initiated Online brandcommunity

Bowden (2009) Customerengagement

Consumer-initiated –

Brodie et al. (2011) Customerengagement

Consumer-initiated –

Gambetti, Graffigna, andBiraghi (2012)

Consumer brandengagement

Firm-initiated –

Hollebeek et al. (2014) Consumer brandengagement

Consumer-initiated Social media

Van Doorn et al. (2010) Customerengagementbehaviour

Firm- and consumer-initiated –

Hollebeek (2011a, 2011b) Customer brandengagement

Consumer-initiated –

Sprott et al. (2009) Brand engagement inself-concept

Consumer-initiated –

Calder et al. (2009) Online engagement Consumer-initiated A websitePhillips and McQuarrie (2010) Advertising

engagementConsumer-initiated Fashion

magazineAlgesheimer et al. (2005) Community

engagementConsumer-initiated –

Baldus et al. (2015) Online brandcommunityengagement

Consumer-initiated Social media

Note: Focal research contexts marked with ‘–’denotes that no specific research context existed.

530 S. YANG ET AL.

one exception (Van Doorn et al., 2010), previous research has not considered brandengagement from both a consumer-initiated and a firm-initiated perspectives. In atheoretical overview of customer brand engagement, Van Doorn et al. (2010)emphasise the notion that firms can engage customers by utilising online socialplatforms to encourage behavioural expressions that reflect customer engagement.According to Van Doorn et al. (2010), firms can influence consumers’ engagingbehaviours by developing online platforms that entice engagement. Indeed, severalstudies demonstrate how social media, such as Facebook, can be leveraged tofoster customer engagement behaviours (Ashley & Tuten, 2015; Kabadayi & Price,2014). Our study is the first to specifically address this idea by considering howcustomers engage with brand-related content that is produced and initiated by thefirm. In addition, while brand engagement on social media does not capture thethoughts and feelings a brand may evoke in individuals (Hollebeek et al., 2014), itdoes capture brand-related behavioural interactivity from an individual perspectiveand brand perspective. Overall, brand engagement on social media signifiesspecific, interactive behaviours among consumers and brands in a social mediacontext.

Prior research focuses on consumers’ perceptions of engagement and relies primarilyon survey research (Hollebeek et al., 2014). In particular, Hollebeek et al. (2014) developa scale that measures three dimensions of consumer brand engagement on socialmedia, including cognitive processing (e.g. interest in using social media), affection(e.g. feeling about using social media) and activation (e.g. behaviours related to socialmedia use). Our study complements this work by showing how industry data can beleveraged to measure different types of consumer-initiated and firm-initiated brandengagement behaviours on social media. Our work is also consistent with more recentwork that demonstrates the use of industry data to measure types of brand andconsumer engagement (e.g. Ashley & Tuten, 2015; Brodie et al., 2013; Malhotra,Malhotra, & See, 2013; Miller & Tucker, 2014). Next, we discuss the details of eachdimension of brand engagement on social media.

Derived from the concept of brand community (McAlexander, Schouten, & Koenig,2002; Muniz & O’Guinn, 2001), affiliation, a consumer-initiated action, is defined as thebrand-related connections among consumers. Today, consumers are becomingincreasingly interested in the social status that derives from brand affiliations (Fournier& Lee, 2009). In online social platforms, every time a consumer becomes a fan of a brand(e.g. a Facebook fan page), the brand gets additional exposure to and affiliation withthat consumer and other consumers.

There are several concepts that although they are distinct relative to affiliation, theyhave conceptual similarities. For example, there are similarities to Hollebeek et al.(2014)’s construct named affection, which captures the emotional interaction betweena consumer and a brand. Because of its focus on overt behaviours, affiliation focusesmore on how a consumer uses a brand as an instrument to interact with otherconsumers. Thus, affection may represent an antecedent variable of affiliation. Anothersimilar construct with affiliation is identification, which refers to an employee’s sense ofcloseness to an institution (Ashforth & Mael, 1989). Identification serves as a salientidentity that may form part of the identity-based motivation of employee engagement(He, Zhu, & Zheng, 2014; Saks, 2006). Therefore, identification may also represent a key

JOURNAL OF MARKETING MANAGEMENT 531

emotional antecedent condition that enables individuals to become behaviourallyaffiliated with an organisation (Ashforth & Mael, 1989).

Based on social capital theory (Bourdieu, 1986; Zinnbauer & Honer, 2011),conversation, also a consumer-initiated action, refers to brand-related talk andbrand-supporting actions among consumers. Overall, conversation capturesinformation being shared within a consumer group (Brown, Barry, Dacin, & Gunst,2005; Keller, 2010). Consumers can voice their opinions and experiences related tothe firm with other social media users, creating social interactions on social media(Agarwal, Animesh, & Prasad, 2009). A similar construct with conversation iselectronic word-of-mouth (e-WOM). e-WOM refers to ‘any positive or negativestatement made by potential, actual or former customers about a product orcompany, which is made available to a multitude of people and institutions via theInternet’ (Hennig-Thurau, Gwinner, Walsh, & Gremler, 2004, p. 39). But whileconversation occurs only on social media sites, eWOM can take place throughother online channels, such as emails, forums, instant messaging, blogs andproduct review sites (Vilpponen, Winter, & Sundqvist, 2006). Moreover, aconversation initiated by consumers could be positive, negative or neutral, withoutany purposeful influencing activity (Kabadayi & Price, 2014; Naylor, Lamberton, &West, 2012); whereas consumers’ eWOM tends to influence others’ attitudes andbehaviours (Chu & Kim, 2011).

Finally, responsiveness, a firm-initiated consumer engagement, captures howengaged consumers respond to firm-initiated and firm-generated content on socialmedia. When a consumer responds to firm-initiated content on a brand’s social mediapage, anyone who views the brand’s social media page can see the responsive activities,even though the viewers are not the consumer’s personal friend on social media. Byresponding to firm-initiated and firm-generated content on social media, a consumercan strengthen their personal relationship with the brand (Rishika, Kumar, Janakiraman,& Bezawada, 2013). For instance, responsiveness provides a platform to view otherconsumers’ opinions and experiences, which in turn affects consumers’ relationshipintensity with the brand (McAlexander et al., 2002).

Overall, prior research demonstrates that brand engagement is related to severalonline contexts and behaviours, including for example: a brand community (Trusov,Bucklin, & Pauwels, 2009); conversation (Miller & Tucker, 2014) and consumer interactionwith a firm brand page (Rishika et al., 2013). However, previous research has notsystematically examined the differential effects of brand engagement on social mediaon paid search advertising effectiveness. As such, and consistent with previous works(Lin et al., 2015; Miller & Tucker, 2014; Rishika et al., 2013; Trusov et al., 2009;), weconceptualise three dimensions – affiliation, conversation and responsiveness – thatcapture both consumer-initiated and firm-initiated brand engagement on social mediaefforts and study how these types of brand engagement influence search engineadvertising effectiveness.

Search engine advertising

With the prominence of search engines, such as Google and Yahoo!, search engineadvertising is a dominant form of online advertising. Among various online advertising

532 S. YANG ET AL.

formats, search engine advertising accounts for nearly 40% of online advertisingspending (Agarwal, Hosanagar, & Smith, 2011). Overall, search engine advertising isconsidered to be an inexpensive and scalable form of internet marketing, and isconsidered to be one of the best ways to draw visitors to an advertiser’s website(Karjaluoto & Leinonen, 2009; Rangaswamy, Giles, & Seres, 2009).

The process of data generation for search engine advertising is different from that fortraditional offline advertising. In search engine advertising, advertisers bid on keywordsin the auction process. Paid-search results are rank-based, dependent on whether andhow much a firm bids for the keywords relevant to the search query. The higher theprice an advertiser bids for a keyword, the better (lower) the rank2 for the search engineadvertisement. Once the advertisers get a rank for the keywords, the advertisements aredisplayed on the search pages in response to consumers’ queries.

The serving of the advertisement in response to a query for a keyword is called animpression. If the consumer clicks on the ad, which is denoted as a click, he/she will beled to the landing page of the advertiser’s website. If the consumer purchases productsor services from the advertiser’s website, a conversion is recorded. In general, there aretwo indicators of effectiveness for search engine advertisements: click-through rate, andconversion rate (Ghose & Yang, 2009). Following previous literature (e.g. Ghose & Yang,2009), conversion rate is defined as the ratio of the number of conversions to thenumber of clicks, and click-through rate is the ratio of the number of clicks for aspecific advertisement to the number of impressions of the advertisement. Figure 1summarises the process of paid search advertising.

Generally, click-through rate and conversion rate are low (Rutz & Trusov, 2011). This isdue, in part, to an overload of online information, and the observation that consumersspend very limited time (e.g. only a few seconds) actually looking at content a websitebefore clicking through to the next site; overall, this notion refers to thin-slice judgmentswhich reflect consumers’ automatic and quick responses to Internet content (Peracchio& Luna, 2006). Thus, click-through rate and conversion rate are distinct from brandengagement on social media in that they represent consumers’ thin-slice judgment onadvertising in a context of search engine advertising. In contrast, brand engagement onsocial media represents consumers’ time investment in interacting with other consumersand brands (Rothbard, 2001).

Because of the popularity of search engine advertising, researchers have a long-standing interest in studying what drives search engine advertising effectiveness(Agarwal et al., 2011; Ghose & Yang, 2009). To our knowledge, however, research hasnot investigated how brand engagement on social media influences the effectiveness of

Search queries that match the keyword

Search engine advertisement

Advertisement rank

Advertisement effectiveness

Click-through

Conversion

Keywords

Figure 1. The process of search engine advertising.

JOURNAL OF MARKETING MANAGEMENT 533

search engine advertising or how brand engagement on social media interacts withadvertisement rank to impact click-through rate and conversion rate. Next, we discussthese relationships.

The link between brand engagement on social media and search engineadvertising effectiveness

Previous research demonstrates that some types of engagement are positivelyassociated with advertising effectiveness (Calder et al., 2009; Wang, 2006). Consistentwith previous work (Calder et al., 2009), we draw from categorisation theory (Cohen &Basu, 1987) to explain why brand engagement on social media influences consumers’attitude and behaviour towards the brand’s search engine advertisement. Categorisationis a fundamental cognitive activity which occurs in response to different stimulussituations (Mervis & Rosch, 1981). According to categorisation theory (Cohen & Basu,1987), consumers’ knowledge about brands forms in structures in their memory. Appliedto engagement and advertising effectiveness, the main premise is that contexts withinteractive engagements, versus contexts that are not engaging, positively influenceadvertising effectiveness (Calder et al., 2009; Wang, 2006). For example, Calder et al.(2009) find that consumers’ engagement with a website results in a greater positiveattitude towards banner advertisements and more intention to click on theadvertisement. According to Dahlén (2005), such a context could serve as a cognitiveprime that activates a semantic network of related material that facilitates interpretationof the advertisement. Applied to the context of search engines, if a search engineadvertisement evokes knowledge about a brand that includes interactiveengagements, there could be a positive relationship between brand engagement onsocial media and search engine advertising effectiveness.

Relevant to our study is the way in which consumers’ affect or emotional response isassociated with the category, such as a specific brand. Fishbein and Ajzen (1975) viewpeople’s attitude towards an object or concept as the location of the object or concepton a dimension of affect. If people identify a new instance as belonging to a categorydefined previously, their attitudes associated with the previously defined categorywould be transferred to the new instance (Boush & Loken, 1991; Dahlén, 2005). In thisarticle, we view a brand as a category and the brand’s search engine advertising as anew instance of that brand. A high level of brand engagement on social media ignitesconsumers’ involvement, attention and elaboration about the brand (Wang, 2006),which would be transferred to the brand’s search engine advertising when consumerssee it. The transferred affect associated with the brand’s search engine advertising islikely to increase consumers’ intention to click on the search engine advertisements andmake a subsequent purchase.

As a dimension of brand engagement on social media, affiliation refers to the totalnumber of consumers connected to a brand. McAlexander et al. (2002) suggest thatbrand communities strengthen brand affiliation by providing a specific context andplatform for consumers to connect with a specific brand. A brand community isdefined as ‘a specialized, non-geographically bound community, based on a structuredset of social relationships among admirers of a brand’ (Muñiz & O’Guinn, 2001, p. 4).Research suggests that online brand communities are important because they provide

534 S. YANG ET AL.

platforms that facilitate concepts closely related to engagement (Algesheimer et al.,2005). For example, Algesheimer et al. (2005) demonstrate that brand communities arerelated to engagement; they use a construct of engagement conceptually similar tobrand engagement called community engagement, which they define as the consumer’sintrinsic motivation to interact and cooperate with community members. In short, theyfind that a brand community can foster engagement (Algesheimer et al., 2005). Inaddition, Gummerus et al. (2012) stress that a brand community, such as a brand’sFacebook page, is an important platform that fosters and captures customers’engagement behaviours. Gummerus et al. (2012) specifically demonstrate thatengagement behaviours, such as liking and creating brand-related posts positivelyimpact brand-related outcomes, including satisfaction and loyalty.

Overall, brand communities provide structure to the relationship between marketersand consumers and facilitate additional marketing functions, such as sharinginformation, providing assistance and fostering engagement among consumers(Algesheimer et al., 2005; Gummerus et al., 2012; Muñiz & O’Guinn, 2001). Research onbrand community suggests that because brand communities increase affiliation, aspecific type of brand engagement on social media, it results in consumers having agreater positive affect associated with the community, and as a result, the brand(Algesheimer et al., 2005; Bagozzi & Dholakia, 2006; Gummerus et al., 2012; Martin,Schouten, & McAlexander, 2006; McAlexander et al., 2002; Muñiz & O’Guinn, 2001).This greater positive affect in turn should increase consumers’ likelihood to click onthe brand’s search engine advertisements and make a purchase post-click. Thus, wepropose that:

Hypothesis 1: Affiliation positively influences search engine advertising effectiveness,such that: (H1a) affiliation increases the click-through rate for a brand’s search engineadvertisement; and (H1b) affiliation increases the conversion rate for a brand’s searchengine advertisement.

Conversation, a consumer-initiated effort, refers to when a consumer ‘talks’ or discussesa brand on social media platforms. Conversation may represent information aboutproducts, services, stores and companies (Brown et al., 2005). Research suggests thatconversation enhances consumers’ affect, attitudes and behaviours towards a brand(Hansen, 1969; Harvey, 1960; Moschis, 1976; Stafford, 1966; Witt, 1969; Witt & Bruce,1970). Thus, if consumers are highly engaged in online conversations about a brand,they are more likely to have positive attitudes and behaviours towards the brand’ssearch engine advertisement. Formally, we propose that:

Hypothesis 2: Conversation positively influences search engine advertising effective-ness, such that: (H2a) conversation increases the click-through rate for a brand’s searchengine advertisement; and (H2b) conversation increases the conversion rate for abrand’s search engine advertisement.

Responsiveness captures how consumers respond to a brand’s online activities and howthey interact with brands on the brand’s social media outlets. Compared to conversationamong consumers, responsiveness refers to the interaction between consumers and

JOURNAL OF MARKETING MANAGEMENT 535

firms. Previous research shows that responsiveness enhances brand communication, createsbrand identity and increases brand equity (Lin et al., 2015). In addition, Cobb-Walgren, Ruble,and Donthu (1995) show that improvement in a firm’s brand equity is positivelyassociated with an increase of consumer preferences, affect and purchase intentions.Conceptually consistent with the link between brand equity and consumer affect, wecontend that responsiveness can enhance search engine advertising effectiveness.Therefore, we propose:

Hypothesis 3: Responsiveness positively influences search engine advertising effec-tiveness, such that: (H3a) responsiveness increases the click-through rate for a brand’ssearch engine advertisement; and (H3b), responsiveness increases the conversion ratefor a brand’s search engine advertisement.

We also suggest that affiliation, conversation and responsiveness moderate the impactof advertisement rank on click-through rate and conversion rate. Previous researchshows that online search result ranks influence advertising effectiveness (Pan et al.,2007; Westerwick, 2013). For example, Westerwick (2013) finds that a low (good)search engine ranking increases perceptions of sponsor credibility. In addition,Agarwal et al. (2011) demonstrate that click-through rate decreases as the rank of theadvertisement increases (gets worse). Recent studies also show that the degree ofattractiveness of the product with the best rank influences consumers, and thatadvertisement rank influences brand recall (Agarwal et al., 2011). We contend that ifconsumers are highly engaged in and involved with a brand, a low (good) search enginerank is more likely to influence consumers’ affect and perceptions of sponsor credibilityand brand recall. As a result, their likelihood to click on the advertisement and make apurchase will increase. Specifically, a brand with high brand engagement on socialmedia should boost teffect of result rank on search engine advertising effectiveness.Formally, we hypothesise that:

Hypothesis 4: Affiliation moderates the relationship between advertisement rank andsearch engine advertising effectiveness, such that: (H4a) higher (lower) affiliationincreases (reduces) the positive effect of a top advertisement rank on the click-throughrate; and (H4b) higher (lower) affiliation increases (reduces) the positive effect of a topadvertisement rank on the conversion rate.

Hypothesis 5: Conversation moderates the relationship between advertisement rankand search engine advertising effectiveness, such that: (H5a) higher (lower) conversationincreases (reduces) the positive effect of a top advertisement rank on the click-throughrate; and (H5b) higher (lower) conversation increases (reduces) the positive effect of atop advertisement rank on the conversion rate.

Hypothesis 6: Responsiveness moderates the relationship between advertisement rankand search engine advertising effectiveness, such that: (H6a) higher (lower) responsivenessincreases (reduces) the positive effect of a top advertisement rank on the click-throughrate; and (H6b) higher (lower) responsiveness increases (reduces) the positive effect of a topadvertisement rank on the conversion rate.

536 S. YANG ET AL.

Figure 2 depicts our conceptual model, proposing that (1) three dimensions of brandengagement on social media (i.e. affiliation, conversation and responsiveness), (2) areassociated positively with search engine advertisement performance and that (3) theeffects of advertisement rank are moderated by brand engagement on social media.

Empirical study

Data

To examine the impact of affiliation, conversation and responsiveness on search engineadvertising effectiveness, we collected a comprehensive industry dataset merged fromdifferent industry sources, including brand engagement data from a top social networksite, Facebook, and search engine advertisement data from the world’s most popularsearch engine, Google. There are several reasons to utilise industry data, such hasdemonstrating the use of ‘real-time’ data in making decisions (Winer, 1999), andmoreover, compared to laboratory or survey data, industry data are collected in lessobtrusive manners (Houston, 2004).

Industry publications collect firms’ real data on their social media activities tomeasure engagement. For example in 2009, a leading digital media consultancyAltimeter Group and a social content hosting firm Wetpaint evaluated and scored thedepth of brands’ engagement in social media for the top 100 valued brands in theworld. The study measured brands’ engagement based on the data of their actualactivities in multiple social media channels, such as publishing social media content,building networks among consumers and updating brand profiles. For instance,Starbucks received the highest engagement score among the top 100 brands withnearly 200,000 fans on the official Facebook page. Because of the published contenton Facebook and consumer sharing of information, Starbucks ‘has built that fan base to

Conceptual model

Search engine advertisement variables

Advertisement rank

Dependent variables Advertisement performance

Click-through rate

Brand engagement on social media variables

Conversation

Conversion rate

Responsiveness

H1a; H2a; H3a

H4a; H5b; H6a

H1b; H2b; H3b

H4b; H5b; H6b

Affiliation

Figure 2. Conceptual model.

JOURNAL OF MARKETING MANAGEMENT 537

nearly 2.5 million members’ (Elowitz & Charlene, 2009). According to Lee Odden, CEO ofTop Rank Online Marketing, consumers’ engagement in brands is typically measured byconsumers’ ‘linking, bookmarking, blogging, referring, clicking, friending, connecting,subscribing, submitting inquiry forms and buying’ behaviours (Falls, 2010).

Some academic studies also measure engagement with industry data. Accordingto Gummerus et al. (2012), customer engagement behaviours can be collectedunobtrusively by analysing social networking sites, brand communities and othersources. For example, Ashley and Tuten (2015) measured firms’ social mediaengagement by calculating the number of firms’ Twitter followers and number ofFacebook fans, as well as the Klout score from www.klout.com and the engagementscore from ENGAGEMENTdb. Brodie et al. (2013) collected consumers’ engagementdata by observing users’ real blog activities in an exploratory analysis, such as thelength of users’ blog posts and the intervals between their activities. Malhotra et al.(2013) suggest that engagement through Facebook is increasingly critical inmarketing, including Facebook fan ‘likes’, comments and shares. Given thesetrends, we also leveraged industry data to operationalise brand engagement onsocial media.

We collected brand engagement data from Facebook for a recognised globalapparel brand from 1 April 2012 to 1 April 2013. The data were provided bySocialbakers.3 Facebook is an online social networking site where, among otherthings, any firm can set up a fan page that enables users to communicate withboth the firm and other users of the brand. We gathered data from Facebook tomeasure three dimensions of brand engagement on social media: affiliation,conversation and responsiveness. Table 2 shows the descriptive statistics for thedimensions. The average daily affiliation is 5218.3, the average daily conversation is6472.8 and the average responsiveness is 3969.3. Figure 3 displays a representativetime-series snapshot of daily levels of affiliation, conversation and responsiveness. Asdepicted in Figure 3, there is substantial daily variation in the value of affiliation,conversation and responsiveness over time. The greatest activity for both affiliationand conversion was in January 2013, but the greatest activity for responsiveness wasin December 2012. In addition, there is day-of-the-week fluctuation in the dailyactivities of brand engagement on social media for the brand that must beaccounted for.

Consistent with previous literature (e.g. Ghose & Yang, 2009; Rutz & Bucklin, 2011; Rutz &Trusov, 2011), we collected search engine effectiveness data for the apparel brandadvertisements from Google Adwords from 1 April 2012 to 1 April 2013.4 Impressionscapture how many times an advertisement is seen, that is, how many times theadvertisement appears as a result of the search. If a consumer clicks on the advertisement,this denotes a click, resulting in the appearance of a brand landing page for a given website.Further, if the consumer purchases products from the website, a conversion is recorded. We

Table 2. Descriptive statistics of brand engagement on social media data.Variable Mean SD Min. Max.

Affiliation 5218.3 20,115.5 −324 304,010Conversation 6472.8 5,792.9 1698 38,261Responsiveness 3969.3 1,939.8 1708 10,097

538 S. YANG ET AL.

collected data for advertisement effectiveness in daily intervals, including click-through rate,conversion rate and advertisement rank. The data for search engine advertising include166,344 observations, with 949 unique keywords. Table 3 shows the descriptive statistics forthe search engine advertisement data. The average click-through rate is 3.4% and theaverage conversion rate is 2.3%. The average advertisement rank is 4.2.

Affiliation

Conversation

Responsiveness

Responsiveness

Conversation

Affiliation14000

100008000600040002000

0 1-Apr

21-May

10-Jul

29-Aug

18-Oct

7-Dec26-Jan

17-Mar

1-Apr

21-May

10-Jul

29-Aug

18-Oct

7-Dec26-Jan

17-Mar

1-Apr

21-May

10-Jul

29-Aug

18-Oct

7-Dec26-Jan

17-Mar

12000

150001300011000

90007000500030001000

11000

9000

7000

5000

3000

1000

Figure 3. Daily activities for brand engagement on social media data across time.

Table 3. Descriptive statistics of search engine data (N = 166,344).Variable Mean SD Min. Max.

Click-through rate 3.4% 13.4% 0% 100%Conversion rate 2.3% 13.6% 0% 100%Advertisement rank 4.2 2.7 0 11

JOURNAL OF MARKETING MANAGEMENT 539

Measures

Measures of brand engagement on social media and search engine advertisingperformanceOur measures of affiliation, conversation and responsiveness are consistent withprevious research (Lin et al., 2015). Affiliation is measured as the daily increase inthe number of fans for an official brand page on Facebook. Conversation ismeasured as the number of individual Facebook users who initially talk about theapparel brand on their own Facebook page each day, such as posting brand-relatedcontent, sharing brand-related Facebook posts and tagging brand-related photoson their individual Facebook page. All of these conversations are initiated by theindividual Facebook users and are shown on each individual’s own Facebook page.Responsiveness is measured as the total number of likes, comments and shares byconsumers of the content posted by the brand on the brand’s Facebook page andinitiated by the brand administrator. Overall, the key difference betweenconversation and responsiveness is that conversations are shown on eachindividual’s own Facebook page, whereas responses are initiated by a brandadministrator and displayed on a brand’s Facebook brand page.5

Our measures of advertisement effectiveness and advertisement rank are basedon previous search engine advertising work (Ghose & Yang, 2009; Rutz & Bucklin,2011; Rutz & Trusov, 2011). Recall that advertisement effectiveness refers to theclick-through rate and the conversion rate for the advertisement. The click-throughrate (CTR) is measured as the ratio of the number of clicks on an advertisement tothe number of times the advertisement is shown to consumers (CTR = clicks/impressions). The conversion rate (CONV) is measured as the ratio of the numberof purchases that consumers make from the advertisement as a percentage of thenumber of clicks on the advertisement (CONV = conversions/clicks). Advertisementrank represents how a firm’s advertisement ranks against competitors’advertisements, measured as the order in which advertisements appear on thepage after a consumer’s search query, with lower ranks being better (Table 4).

Table 4. Measurements of variables.Variable Measurement Source

Brand engagement on social mediaAffiliation A daily measure of the increase in the number of daily fans for the official brand

page on Facebook each dayFacebook

Conversation The number of individual Facebook users who initially talk about the apparel brand ontheir own Facebook page each day by posting brand-related content, sharing brand-related Facebook posts and tagging brand-related photos on their own Facebook page

Facebook

Responsiveness The total number of likes, comments and shares of the content posted by the brand,which are initiated by the brand administrator and displayed on the brand’sFacebook brand page

Facebook

Search engine advertisingClick-through rate The ratio of the number of clicks on an advertisement to the number of impressions

the advertisement is shown (CTR = clicks/impressions)Google

Conversion rate The ratio of the number of conversions (purchases) that consumers make from theadvertisement to the number of clicks on the advertisement (CONV = conversions/clicks)

Google

Advertisement rank The order in which advertisements appear on the page after a consumer’s search query Google

540 S. YANG ET AL.

Further examination of measures of brand engagement on social mediaIn this study, we employ industry data to measure marketing constructs. Industry dataproxies have become increasing popular and are strongly preferred over self-report scalemeasure in disciplines, such as finance, economics, information systems and health caremanagement (Day & Montgomery, 1999; Srivastava, Shervani, & Fahey, 1999). Sincethere are no other direct industry measures of brand engagement, we followed Linet al. (2015) and used industry data to measure three dimensions of brand engagementon social media. Aforementioned, each dimension was measured by a single item fromindustry data.

Despite the advantages of utilising industry data, there are possible drawbacksrelated to reliability and validity (Houston & Johnson, 2000). Therefore, we followedHouston’s (2004) suggestion on how to assess reliability and validity of our industry data.However, since one-item data proxies for affiliation, conversation and responsiveness,traditional methods were used, it is not appropriate to assess reliability using traditionalmethods, such as the use of Cronbach’s alpha, and the use of a confirmatory factoranalysis (Houston, 2004). Specifically, we followed Houston (2004)’s suggestions, andcollected additional survey data to complement industry data and test the reliability andvalidity of our measurements. A questionnaire comprising items applied to the apparelfirm was administered to a sample of 108 university undergraduate students inSeptember 2015 (46.3% male). Each student was invented to participate in the surveywith a university bookstore gift card. All participants were familiar with the apparel firm,and the survey completion time was approximately 5 minutes. Consistent with previouswork (Churchill, 1979), two weeks after the initial survey completion, participants wereasked to complete the same survey as a post-measure. Although there were twoincomplete responses which were omitted from our analysis, all questionnaires werereturned to us.

We measured affiliation, conversation and responsiveness using a single item asindustry data proxies did. Table 5 contains the item for each construct. To check the

Table 5. Construct measurements in survey.Variable Measurement

Brand engagement on social mediaAffiliation I like to become a fan of [the apparel brand] for the official brand page on FacebookConversation I like to initially talk about [the apparel brand] on my own Facebook page, such as posting content

related to [the apparel brand], sharing Facebook posts related to [the apparel brand] and taggingphotos related to [the apparel brand] on my own Facebook page

Responsiveness I like to respond to the content posted by [the apparel brand] on the brand’s Facebook brandpage, such as liking, commenting and sharing the content

Consumer brand engagement (Hollebeek et al., 2014)Cognitiveprocessing

Using [the apparel brand] gets me to think about [the apparel brand]

I think about [the apparel brand] a lot when I’m using itUsing [the apparel brand] stimulates my interest to learn more about [the apparel brand]

Affection I feel very positive when I use [the apparel brand]Using [the apparel brand] makes me happyI feel good when I use [the apparel brand]I’m proud to use [the apparel brand]

Activation I spend a lot of time using [the apparel brand], compared to other [category] brandsWhenever I’m using [category], I usually use [the apparel brand][The apparel brand] is one of the brands I usually use when I use [category]

JOURNAL OF MARKETING MANAGEMENT 541

validity of our measures, we also included items that measured consumer brandengagement (Hollebeek et al., 2014). This included three items for cognitiveprocessing, four items for affection and three items for activation (Table 5). All theseitems were rated on 7-point Likert scales (1 = strongly disagree, 7 = strongly agree).

Table 6 reports reliability and validity statistics for each construct. Because measuresof internal consistency reliability, such as Cronbach’s coefficient alpha, cannot be appliedto the single-item scales of brand engagement on social media, we employed test–retestreliability for the single-item scales (Heise, 1969). To investigate test–retest reliability, weexamined the t-tests of the paired difference between each brand engagement on socialmedia items. The reliability estimate for each construct is: affiliation, 0.94; conversation,0.91; and, responsiveness, 0.92 (all ps < 0.001), providing evidence for the reliability ofour measures of affiliation, conversation and responsiveness.

Second, we assessed nomological validity by calculating correlation coefficientsbetween other relevant constructs (Houston, 2004). Affiliation correlated positivelywith cognitive processing (r = 0.82, p < 0.01), affection (r = 0.86, p < 0.01) andactivation (r = 0.81, p < 0.01); conversation correlated positively with cognitiveprocessing (r = 0.77, p < 0.01), affection (r = 0.83, p < 0.01) and activation (r = 0.75,p < 0.01); and responsiveness correlated positively with cognitive processing (r = 0.79,p < 0.01), affection (r = 0.87, p < 0.01) and activation (r = 0.77, p < 0.01). These resultsprovide support for nomological validity (Stanton, Sinar, Balzer, & Smith, 2002).

Third, we used the traditional multitrait-multimethod (MTMM) approach to assessconvergent and discriminant validity (Campbell & Fiske, 1959; Houston, 2004). Table 7

Table 6. Reliability and validity test results of the constructs using survey data.

Reliability

Inter-construct correlations

BSE CBE

1 2 3 4 5 6

BSE Affiliation (1) 0.94 /Conversation (1) 0.91 0.87 /Responsiveness (1) 0.92 0.8 0.8 /

CBE Cognitive processing (3) 0.98 0.82 0.77 0.79 0.98Affection (4) 0.96 0.86 0.83 0.87 0.78 0.94Activation (3) 0.97 0.81 0.75 0.77 0.950 0.77 0.97

Note: Reliability of BSE is test–retest reliability and reliability of CBE is Cronbach’s coefficient alpha, all of which aresignificant at p < 0.01; the diagonal elements in the inter-construct correlations sub-table are the square roots of AVEfor the respective construct.

Table 7. A multitrait-multimethod (MTMM) approach to assess convergent and discriminant validity.Industry data Survey data

A C R A C R

Industry data Affiliation ()Conversation 0.08 ()Responsiveness 0.78 0.38 ()

Survey data Affiliation 0.29 0.05 0.07 (0.94)Conversation 0.06 0.19 0.08 0.87 (0.91)Responsiveness 0.03 0.02 0.10 0.80 0.80 (0.92)

Note: The validity diagonals are the italicised values. The reliability diagonals are the values in parentheses. Emptyparentheses report no appropriate reliability estimates.

542 S. YANG ET AL.

illustrates three dimensions of brand engagement, each measured by two methods,generating for separate variables. The entries in the validity diagonal (i.e. the italicisedvalues) are significantly different from zero. Moreover, the entries in the validity diagonalare greater than those lying in corresponding columns and rows in the heterotrait-heteromethod triangles. Further, the entries in the heterotrait-monomethod triangle aregreater than those in the heterotrait-heteromethod triangles. These results providesupport for convergent and discriminant validity.

Empirical analysesTypically, linear regression models use an ordinary least squares (OLS) approach, whichassumes a normal data distribution (Y ~ N(μ, φ)). However, in search engine advertisingdata, a normal distribution is inappropriate (Agarwal et al., 2011; Ghose & Yang, 2009). Thisis because the utility of the individual choice in clicking on an advertisement or convertingthe advertisement to a sale follows an independent and identically distributed (i.i.d.)extreme value distribution (Agarwal et al., 2011). In other words, two of our dependentvariables, click-through rate and conversion rate, are bounded between 0 and 1.Considering this, we use a logit model to capture the click-through probability orconversion probability of a search engine advertisement for a keyword as follows:

Pr ¼ expðUÞ1þ expðUÞ ; (1)

where U is the latent utility of clicking on an advertisement or converting, whichdepends on brand engagement on social media, advertisement rank and theirinteractions. We also include indicator variables for day of the week and time trend tocontrol for their effects. The latent utility can be expressed as follows:

UCTR¼αCTR0 þβCTR1 �AffiliationþβCTR2 �ConversationþβCTR3 �ResponsivenessþβCTR4 �Affiliation�RankþβCTR5 �Conversation�RankþβCTR6 Responsiveness�Rank

þβCTR7 �RankþX6

d¼1

γCTR1d �DaydþγCTR2 �TimetþεCTR;

(2)

UCONV¼αCONV0 þβCONV1 �AffiliationþβCONV2 �ConversationþβCONV3 �ResponsivenessþβCONV4 �Affiliation�RankþβCONV5 �Conversation�RankþβCONV6 Responsiveness�Rank

þβCONV7 �RankþX6

d¼1

γCONV1d �DaydþγCONV2 �TimetþεCTR;

(3)

where CTR, the click-through rate; CONV, the conversion rate; Rank, advertisement rankin the result listing. Top rank is coded as 1; Affiliation, the affiliation dimension of brandengagement on social media; Conversation, the conversation dimension of brandengagement on social media; Responsiveness, the responsiveness dimension of brandengagement on social media; Day, day of week; Time, time trend.

JOURNAL OF MARKETING MANAGEMENT 543

As mentioned previously, click-through rate (CTR) is measured by tabulating the totalnumber of clicks on an advertisement and dividing by the number of impressions for theadvertisement; conversion rate (CONV) is measured by taking the total number ofconversions/purchases and dividing by the total number of clicks. Advertisement rankrepresents the daily average rank at which the advertisement is displayed in the searchresult listing. A lower value for advertisement rank denotes an advertisement that isdisplayed closer to the top of the search result listings. For example, an advertisementrank of 1 means that the advertisement would be the first listing displayed.

Estimation results

Effects of brand engagement on social media on advertisement effectiveness(hypotheses 1–3)Table 8 shows the estimates for the effects of affiliation, conversation andresponsiveness on search engine advertising effectiveness, as well as their moderatingeffects. To test the explanatory power of our model, we compared a full model withthree base models. Base model (1) contains only the three dimensions of brandengagement on social media as direct effects and includes day-of-the-week and timetrend as our standard control variables. Base model (2) contains only advertisement rankand the control variables. Base model (3) includes the three brand engagement on socialmedia constructs and advertisement rank, as well as the control variables. In the fullmodel, we added the interaction terms between brand engagement on social mediaand advertisement rank to base model (3). We used the likelihood ratio test to comparethe base models against each other and the full model against the base models(Koschate-Fischer, Cramer, & Hoyer, 2014), where the test statistic (D) is twice thedifference in log-likelihoods (D = (−2 ln(likelihood for the null model) – (−2 ln(likelihood for the alternative model)))) and follows a chi-squared distribution. Whetherthe null model is better than the alternative model is determined by the p-value of D.

Compared with base models (1) and (2), base model (3), which includes both thedirect effects of both brand engagement on social media and advertisement rank, leadsto a statistically significantly better fit; the likelihood ratio test is statistically significant

Table 8. Regression results for effects of brand engagement on social media.Base model (1) Base model (2) Base model (3) Full model

Variables CTR CONV CTR CONV CTR CONV CTR CONV

Affiliation −3.8E-7*** 3.3E-6*** 2.3E-7*** 4.0E-6*** 1.4E-10*** 6.4E-6***Conversation 3.4E-6*** 1.7E-5*** 6.7E-7*** 2.3E-5*** 4.4E-6** 6.2E-5***Responsiveness 1.3E-4*** 6.8E-5*** 7.5E-5*** 4.0E-5*** 2.6E-4*** 1.6E-4***Affiliation * rank −2.9E-7*** −2.2E-6a

Conversation * rank −4.4E-6*** −4.0E-5***Responsiveness * rank −1.4E-4*** −1.0E-4***Advertisement rank −0.82*** −0.77*** −0.81*** −0.77*** −0.47*** −0.42***Day of week effects Yes Yes Yes Yes Yes Yes Yes YesTime effect 0.004*** 0.002*** 0.001*** −0.001*** 0.001*** −0.001*** 0.001*** −0.001***Constant −4.31*** −1.20*** −1.35*** 0.52*** −1.49*** 0.44*** −2.03*** −0.01No. of observations 166,344 166,344 166,344 166,344 166,344 166,344 166,344 166,344−2 log likelihood 597,400 128,212 375,319 108,631 373,530 107,845 369,642 107,356AIC 597,422 128,234 375,337 108,649 373,554 107,869 369,672 107,386

Notes: ***p < 0.001, **p < 0.01, ap < 0.1.

544 S. YANG ET AL.

(ps < 0.001). This indicates that adding the main effects for affiliation, conversation andresponsiveness or advertisement rank improves the model fit for search engineadvertising effectiveness. Further, the full model, which includes the interaction terms,provides a statistically significant improvement over base model (3) which includes onlythe direct effects; the likelihood ratio test is significant (ps < 0.001), showing that addingthe interaction terms improves the model fit. Moreover, coefficients for the majorvariables retain their sign and significance across specifications, indicating therobustness of our model.

We find that affiliation exerts a positive and significant effect on both click-throughrate (β = 1.4E-10, p < 0.001) and conversion rate (β = 6.4E-6, p < 0.001), suggesting thatconsumers’ affiliation engagement for the brand enhances both click-through rate andconversion rate. We find that conversation has a positive and significant effect on click-through rate (β = 4.4E-6, p < 0.01) and conversion rate (β = 6.2E-5, p < 0.001), whichimplies that consumers’ conversation engagement on the brand improves its searchengine advertisements’ click-through rate and conversion rate. We also find that theeffects of responsiveness on click-through rate (β = 2.6E-4, p < 0.001) and conversionrate (β = 1.6E-4, p < 0.001) are positive and significant, suggesting that consumers’responsiveness to the brand also improve its advertisements’ effectiveness. Together,these results support H1a, H1b, H2a, H2b, H3a and H3b.

Moderating effects of brand engagement on social media (hypotheses 4–6)We tested the moderating effects of brand engagement on social media by examiningthe significance of the interaction terms and assessing the effect sizes for theinteractions. The results in Table 8 show that the coefficient for the interactionbetween advertisement rank and affiliation on click-through rate (β = −2.9E-7,p < 0.001) is negative and statistically significant, which implies that higher affiliationstrengthens the positive effect of a top advertisement rank on click-through rate. Thecoefficient for the interaction between advertisement rank and affiliation for conversionrate (β = −2.2E-6, p < 0.1) is also negative and marginally statistically significant. Thisindicates that higher affiliation helps the effect of a top advertisement rank onconversion rate. Thus, H4a and H4b are supported.

With respect to conversation, the interaction between advertisement rank andconversation has a negative and statistically significant effect for both click-throughrate (β = −4.4E-6, p < 0.001) and conversion rate (β = −4.0E-5, p < 0.001). In support ofH5a and H5b, this suggests that higher conversation strengthens the positive effect of atop advertisement rank on click-through rate and conversion rate. The coefficients forthe interaction between advertisement rank and responsiveness for both click-throughrate (β = −1.4E-4, p < 0.001) and conversion rate (β = −1.0E-4, p < 0.001) are negativeand statistically significant. This implies that higher responsiveness strengthens thepositive impact of a top advertisement rank on advertisement effectiveness. Thus, H6aand H6b are supported.

After testing the significance of the interaction terms, we assessed their effect size (f2)to determine the size of the moderating effects. We use the Cohen effect size formula(Cohen, 1988) to compare the squared multiple correlation (R2) for the full model with R2

for the basic model without interaction terms:

JOURNAL OF MARKETING MANAGEMENT 545

f 2 ¼ R2ðFullÞ � R2ðBasicÞ1� R2ðBasicÞ : (4)

The results show that the full model, including interactions terms betweenadvertisement rank and affiliation, conversation and responsiveness has significantlybetter explanatory power than the base model without interaction terms (p < 0.001).The effect size f2 for the moderating effects is 0.03 for click-through rate and 0.04 forconversion rate. Interaction effect sizes are deemed large if it is 0.35, medium if 0.15 andsmall if 0.02 (Cohen, 1988). Thus, the moderating effects of affiliation, conversation andresponsiveness are significant, with effect sizes that are deemed somewhat larger thansmall.

Control variablesAs expected, advertisement rank has a negative and significant effect on click-throughrate (β = −0.47, p < 0.001) and conversion rate (β = −0.42, p < 0.001), which supports thenotion that advertisements displayed in higher positions in the search results positivelyincrease both click-through rate and conversion rate. Also as expected, the results showthat the effects of day-of-the-week on click-through rate and conversion rate aresignificant. This suggests that the likelihood of consumers clicking on anadvertisement and that of the following purchase depends to some extent on the dayof the week on which they are searching. Further, the effects of time on click-throughrate (β = 0.001, p < 0.001) and conversion rate (β = −0.001, p < 0.001) are significant,showing that the probabilities of consumers clicking on an advertisement and making apurchase vary across time.

Robustness checksWith respect to robustness of the regression results, our main concern was a possiblelack of independence between a consumer’s decision to click on a search engineadvertisement and their decision to purchase. In other words, are click-through rateand conversion rate correlated? Based on previous works (Agarwal et al., 2011; Ghose &Yang, 2009; Rutz & Trusov, 2011), we correlated the error terms of Equation (2) and (3) torepresent their correlations. The following distribution was used:

εCTRit

εCONVit

24

35,N½0; V�; where V ¼

V11 V12

V21 V22

24

35: (5)

The results of Table 9 provide evidence from estimates of this more structured model.The general insights of the generalised linear model (GLM) results with correlated errorterms are similar to that of GLM results without correlated error terms, suggesting thatthe correlated nature of the dependent variables does not drive our results.

Tests for alternative explanationsPrevious research questions the quality of social media sites’ data. For instance, Nelson-Field, Riebe, and Sharp (2012) show that the Facebook fan base for two fast movingconsumer goods companies is skewed towards the brands’ heavy buyers, but that theactual purchase base is in an opposite pattern for the two brands. This raises an

546 S. YANG ET AL.

important question: what supports our findings that brand engagement works incombination with search engine advertising?

One possible explanation is that the user base of search engines is also skewedtowards the brands’ heavy buyers. To find evidence with respect to this notion, weseparated users of search engines into two categories: users who search for genericproducts, and users who search specifically for the apparel brand. Rutz and Bucklin(2011) suggest that consumers who search using a generic keyword are not aware thatthe brand name is relevant to the search, while those who search using a specificbranded keyword are aware that the brand is relevant to their current search. Thus,users who search for a generic product are more likely to be non-buyers or light buyersof the brand, whereas users who search for the specific brand are more likely to bemoderate to heavy buyers of the brand. To test this, we compared the impressions ofnon-branded searches with branded searches. The results show a similar negativebinomial distribution (NBD)-distributed user base, with high numbers of non-brandedsearches and much fewer branded searches (see Figure 4). This is consistent withprevious literature (Rutz & Bucklin, 2011). Thus, the user base of search engines is

Table 9. Regression results with correlated error terms.Variables CTR CONV

Affiliation 1.3E-10*** 6.4E-6***Conversation 4.4E-6** 6.2E-5***Responsiveness 2.5E-4*** 1.7E-4***Affiliation * rank −2.8E-7*** −2.1E-6a

Conversation * rank −4.4E-6*** −4.0E-5***Responsiveness * rank −1.4E-4*** −1.0E-4***Advertisement rank −0.47*** −0.42***Day of week effects Yes YesTime effect 0.001*** −0.001***Constant −2.03*** −0.01No. of observations 166,344 166,344

Notes: ***p < 0.001, **p < 0.01, ap < 0.1.

6000000

5000000

4000000

3000000

2000000

1000000

0Non-Branded Searches Branded Searches

Buyer Categories

Ave

rage

Pro

duct

Sale

s

Figure 4. Search concentration across the search engine searches.

JOURNAL OF MARKETING MANAGEMENT 547

skewed towards non-buyers of the brands, which undermines the plausibility of thispossible explanation.

A second explanation is that leveraging social media platforms to engage with heavybuyers of brands also works for non-buyers who start to search the brands in searchengines. Therefore, we estimated the effects for how brand engagement on social mediaimpacts search engine advertising effectiveness for two user categories. Table 10 showsthat brand engagement on social media has a significant and positive effect on searchengine advertising effectiveness for both non-branded and branded searches. Theresults offer some support for the second alternative explanation that engaging withheavy buyers of the brands in social media platforms also works for their non-buyers insearch engines.

Conclusions, implications and limitations

Conclusions and discussion

The results of this study are twofold (see Table 11). First, we show that there is a positiveassociation between brand engagement on social media – affiliation, conversation andresponsiveness – and search engine advertising effectiveness. Second, the study revealsthat brand engagement on social media moderates the relationship betweenadvertisement rank and search engine advertising effectiveness. In particular,affiliation, conversation and responsiveness amplify the positive effect of a topadvertisement rank on search engine advertising effectiveness.

There are three primary implications for the results. The first is that brandengagement on social media can be measured using at least three sub-dimensions,two of which, affiliation and conversation, are consumer-initiated, either purely personalor provided by the brand, and one of which is in response to the brand’s social mediaactivities and take place on the brand’s social media sites. The second is that brandengagement on social media matters to firms’ other online marketing activities. All threeaspects of brand engagement on social media have a significantly positive effect on theeffectiveness of search advertising and, moreover, positively moderate the effect of atop search advertising rank on the effectiveness of search advertising. Third, these

Table 10. Effects of brand engagement on social media for two search categories.Non-branded searches Branded searches

Variables CTR CONV CTR CONV

Affiliation 1.3E-7*** 6.8E-6* 1.8E-8* 2.0E-6*Conversation 4.1E-6* 1.8E-5* 2.0E-5*** 3.5E-4***Responsiveness 1.5E-4*** 1.3E-4*** 9.0E-4*** 3.0E-4***Affiliation * rank −1.7E-7*** −1.0E-5** −2.04E-6a −2.4E-5**Conversation * rank −9.7E-7 −4.6E-5** −1.7E-5*** −2.3E-4***Responsiveness * rank −5.0E-5*** −9.9E-4*** −1.3E-4*** −1.6E-5*Advertisement rank −0.49*** −0.34*** −3.58*** −6.51***Day of week effects Yes Yes Yes YesTime effect −0.0003*** 0.003*** −0.001*** −0.003***Constant −2.41*** −0.15** 2.44*** 7.04***−2 log likelihood 225,523 41,101 114,342 51,747AIC 225,553 41,131 114,372 51,777

Notes: ***p < 0.001, **p < 0.01, *p < 0.05, ap < 0.1.

548 S. YANG ET AL.

results are a first demonstration that the combination of online marketing efforts is moreuseful than focusing on one to the exclusion of others and that social media is, in fact, avaluable brand management tool.

The results lead us to an additional speculation with respect to types of searchengine advertising. In response to a particular keyword search, search enginesdisplay two sets of results: a paid search engine advertising list and an unpaid,organic list. While firms can increase the rank of their search engine advertising bybidding more, the rankings of the organic results are based on a complex algorithmdevised by the search engines that include the website’s quality and ‘relativeimportance’ compared to other links (Yang & Ghose, 2010). We might speculatethat positive relationships should exist between the level of brand engagement onsocial media and organic search results for the brand for two different reasons.First, paid search engine advertising and unpaid organic listings are positivelycorrelated with each other’s click-through rate (Yang & Ghose, 2010). We havedemonstrated that firms’ efforts on brand engagement on social media areassociated with higher click-through and conversion rates for search engineadvertising, which may lead to higher click-through and conversion rates oforganic listings. Second, and more generally, increased brand engagement onsocial media and improved search engine advertising’s performance could raisethe site’s general popularity or inherent value (Katona & Sarvary, 2010). This wouldhelp search engines’ judgment on the site’s quality and importance, thus improvingthe rank of organic search listings. The click-through rate and the conversion rateof organic search listings would therefore be increased. Overall, these concernmerits future consideration.

Managerial implications

In today’s brand management environment, brands engaging with customersthrough online social platforms have become commonplace. However, what is

Table 11. Summary of the hypotheses.(a) Hypotheses 1, 2, 3

Hypothesis Independent variable Dependent variable Expected effect Result

H1a Affiliation CTR Positive SupportedH1b Affiliation CONV Positive SupportedH2a Conversation CTR Positive SupportedH2b Conversation CONV Positive SupportedH3a Responsiveness CTR Positive SupportedH3b Responsiveness CONV Positive Supported

(b) Hypotheses 4, 5, 6

Hypothesis Independent variable Dependent variable Hypothesised moderatorExpected

moderating effect Result

H4a Ad rank CTR Affiliation Positive SupportedH4b Ad rank CONV Affiliation Positive SupportedH5a Ad rank CTR Conversation Positive SupportedH5b Ad rank CONV Conversation Positive SupportedH6a Ad rank CTR Responsiveness Positive SupportedH6b Ad rank CONV Responsiveness Positive Supported

JOURNAL OF MARKETING MANAGEMENT 549

the return on this marketing investment? To better measure a return oninvestment, marketers need evidence linking social media spending to consumerengagement and, ultimately, to brand sales. In addition, search engine advertisingis an increasing trend in online advertising spending, not only because ofconsumers’ increasing usage of search engines, such as Google and Yahoo!, butalso because search engine advertising is less costly than traditional mediaadvertising. However, for these two major online activities, little has been knownabout how social media elements and investments impact search engineadvertising effectiveness. Our study addresses this gap by showing thatcollectively, the dimensions of brand engagement on social media have greatereffect on advertising effectiveness compared to each individual dimension.

This study generates several managerial implications. First, firms are increasinglyusing ‘engagement metrics’ to track their social media activities, but themeasurements remain a challenge (Moorman, Ross, & Gorman, 2015). Following Linet al. (2015) and Gummerus et al. (2012), this study demonstrates that brandengagement behaviours can be collected unobtrusively by analysing social networkingsites. We provide firms with three different ways – affiliation, conversation andresponsiveness – to understand, and measure their level of brand engagement onsocial media. This, for example, not only includes how consumers talk about thebrand to other consumers, but also how consumers respond to the brand contentposted by the company.

Second, Hanna et al. (2011) underscore the notion that social media efforts arepart of an online ecosystem – these elements not only relate to, but also ofteninfluence, one another. However, results from The CMO Survey referencedpreviously reveal that firms still manage social media efforts as activities that areseparate and distinct from other marketing strategy activities (Moorman, Ross, &Gorman, 2014). In fact, the level of social media integration with marketing strategyhas remained both relatively moderate and relatively stable, from 3.8 in 2011 to 3.9in 2014 (where 1 is ‘not integrated’ and 7 is ‘very integrated’). Our study providesevidence that seemingly unrelated social media investments are, in fact,interdependent activities. Firms’ interactions with customers by way of socialmedia can improve the effectiveness of their online search advertising. In fact,search engines, such as Bing Social Search and TripAdvisor, have begun tointegrate richer social information from social media platforms into their rankingmechanism design to improve consumers’ search experience (Ghose, Ipeirotis, & Li,2012). Our findings offer strong support for firms’ investment in, and perhapseven greater investment in, social media marketing activities (Goh, Heng, & Lin,2013).

Finally, our findings suggest an alternative means for firms to boost onlineadvertising effectiveness with a limited advertisement budget. Firms with limitedbudgets that cannot afford to pay for a top advertisement rank search resultsshould consider investing in social media. In turn, these efforts may improvesearch engine effectiveness. For large firms with considerable resources, boostingbrand engagement on social media can amplify the positive impact of bidding for atop advertisement rank on search engine advertising effectiveness. Our findings

550 S. YANG ET AL.

offer practical managerial advice for firms that use search engine advertising; thesefirms should consider hiring employees devoted to attracting and responding toconsumers via social media platforms (Tadena, 2014) to increase the performanceof their search advertising. In short, firms should continue to utilise social media toprovide content and a platform that entices brand engagement among consumersand between the brand and consumers.

Moreover, because responsiveness is a firm-initiated social media action, it can beinduced more easily by brand managers compared to consumer-initiated engagement.Thus, firms should invest more in increasing consumers’ engagement using the firms’social media brand site. For instance, firms can post more news or blogs on their socialmedia platforms and monitor consumers’ responses. Alternatively, firms can generatecampaigns (e.g. free products for top answers) to encourage consumer–brandinteraction and engagement.

Limitations and future research

There are limitations to this study and, consequently, future-research opportunities.First, we relied upon industry data to measure brand engagement on social media.Thus, the nature of our data limits the generalisability of the results of our study.Similarly, it can be argued that the timing of our data, from 2012 to 2013 may limitthe generalisability of our results. While this is true for the magnitude of theparticular coefficients for the main effects and interactions, it does not apply tothe overall results that brand engagement on social media directly affects and alsointeracts with search advertising rank to affect search advertising effectiveness.There is no reason to suggest that the increasing importance of social media willreduce or eliminate these effects. However, there are additional future researchopportunities that may focus on validating our results with additional data sources,perhaps including surveys or experiments. Second, future work should considerapplying our framework to the individual level. Third, we relied upon the data ofonly one brand and one type of social medium (i.e. Facebook) to test ourhypotheses. This limits our ability to generalise our results to other brandsand social media contexts. For example, how may the use of different socialmedia platforms affect our findings? In addition, what types of engagementbehaviour would Twitter foster among consumers? Fourth, in this study weexamined only the total numbers for conversation and responsiveness; we didnot consider the nature and valence (positive versus negative) of their content.The valence of the content of conversation and responsiveness (i.e. positive versusnegative) might have an impact on search engine advertising effectiveness. Futureinvestigation of how the valences of conversation and responsiveness influencesearch engine advertising effectiveness and other online investments would beuseful. Finally, whereas this study provided evidence that firms’ efforts on socialmedia platforms affect their search engine advertising effectiveness, it did notprovide a behavioural mechanism to demonstrate why, which also creates anopportunity for future work.

JOURNAL OF MARKETING MANAGEMENT 551

Notes

1. The CMO Survey was compiled by Christine Moorman of Duke University, McKinsey & Co.and the American Marketing Association (AMA). The sample consisted of 351 top marketersfrom leading organisations in the United States reached through Duke University’s alumninetwork and the AMA’s membership.

2. Note: In this article, the lower the rank (1–10), the higher the position in the search results.For example, an advertisement rank of 1 means that the advertisement is displayed at thetop of the search listing.

3. Socialbakers (http://www.socialbakers.com) is one of Facebook’s preferred marketingdevelopers. It has been awarded three Preferred Marketing Developer badges (Pages,Apps and Insights) from Facebook.

4. One-year historical Google data is sufficient to examine the search engine advertisingperformance. For example, Ghose and Yang (2009)’s search engine advertising datacontain advertising on Google from a large nationwide retail chain during a period of sixmonths in 2007. Rutz and Bucklin (2011)’s data include search engine advertising for amajor lodging chain from both Google (from 1 March 2004 to 20 December 2004) andYahoo! (from 6 May 2004 to 31 August 2004). Rutz and Trusov (2011) use a search engineadvertising dataset from the ringtone industry over 20 days in 2007.

5. Technically, conversation data were obtained from individual users’ own Facebook pagewhose posts or sharing related to the apparel brand, while responsiveness data wereobtained from the apparel brand’s Facebook brand page. Thus, each data point was usedfor a single element.

Disclosure statement

No potential conflict of interest was reported by the authors.

Funding

This work was supported by National Natural Science Foundation of China (NSFC) [grant number71172036].

References

Abdul-Ghani, E., Hyde, K., & Marshall, R. (2011). Emic and etic interpretations of engagement with aconsumer-to-consumer online auction site. Journal of Business Research, 64(10), 1060–1066.doi:10.1016/j.jbusres.2010.10.009

Achterberg, W., Pot, A. M., Kerkstra, A., Ooms, M., Muller, M., & Ribbe, M. (2003). The effect ofdepression on social engagement in newly admitted Dutch nursing home residents. TheGerontologist, 43(2), 213–218. doi:10.1093/geront/43.2.213

Agarwal, A., Hosanagar, K., & Smith, M. D. (2011). Location, location, location: An analysis ofprofitability of position in online advertising markets. Journal of Marketing Research, 48(6),1057–1073. doi:10.1509/jmr.08.0468

Agarwal, R., Animesh, A., & Prasad, K. (2009). Research note—Social interactions and the “digitaldivide”: Explaining variations in internet use. Information Systems Research, 20(2), 277–294.doi:10.1287/isre.1080.0194

Algesheimer, R., Dholakia, U. M., & Herrmann, A. (2005). The social influence of brand community:Evidence from European car clubs. Journal of Marketing, 69(3), 19–34. doi:10.1509/jmkg.69.3.19.66363

Anderson, S. P., & Palma, A. (2012). Competition for attention in the information (overload) age.The RAND Journal of Economics, 43(1), 1–25. doi:10.1111/j.1756-2171.2011.00155.x

552 S. YANG ET AL.

Animesh, A., Viswanathan, S., & Agarwal, R. (2011). Competing “creatively” in sponsored searchmarkets: The effect of rank, differentiation strategy, and competition on performance.Information Systems Research, 22(1), 153–169. doi:10.1287/isre.1090.0254

Ashforth, B. E., & Mael, F. (1989). Social identity theory and the organization. Academy ofManagement Review, 14, 20–39. doi:10.5465/AMR.1989.4278999

Ashley, C., & Tuten, T. (2015). Creative strategies in social media marketing: An exploratory study ofbranded social content and consumer engagement. Psychology & Marketing, 32(1), 15–27.doi:10.1002/mar.20761

Avnet, T., & Higgins, E. T. (2006a). How regulatory fit affects value in consumer choices andopinions. Journal of Marketing Research, 43(1), 1–10. doi:10.1509/jmkr.43.1.1

Bagozzi, R. P., & Dholakia, U. M. (2006). Antecedents and purchase consequences of customerparticipation in small group brand communities. International Journal of Research in Marketing,23(1), 45–61. doi:10.1016/j.ijresmar.2006.01.005

Baldus, B. J., Voorhees, C., & Calantone, R. (2015). Online brand community engagement: Scaledevelopment and validation. Journal of Business Research, 68(5), 978–985. doi:10.1016/j.jbusres.2014.09.035

Blackshaw, P., & Nazzaro, M. (2006). Consumer-generated media (CGM) 101: Word-of-mouth in theage of the web-fortified consumer. New York, NY: Nielsen BuzzMetrics. Retrieved from http://www.nielsen-online.com/downloads/us/buzz/nbzm_wp_CGM101.pdf

Bourdieu, P. (1986). The forms of capital. In J. G. Richardson (Ed.), Handbook of theory and researchfor the sociology of education. New York, NY: Greenwood Press.

Boush, D. M., & Loken, B. (1991). A process-tracing study of brand extension evaluation. Journal ofMarketing Research, 28(1), 16–28. doi:10.2307/3172723

Bowden, J. H. (2009). The process of customer engagement: A conceptual framework. The Journalof Marketing Theory and Practice, 17(1), 63–74. doi:10.2753/mtp1069-6679170105

Brodie, R. J., Hollebeek, L. D., Jurić, B., & Ilić, A. (2011). Customer engagement: Conceptual domain,fundamental propositions, and implications for research. Journal of Service Research, 14(3), 252–271. doi:10.1177/1094670511411703

Brodie, R. J., Ilic, A., Juric, B., & Hollebeek, L. (2013). Consumer engagement in a virtual brandcommunity: An exploratory analysis. Journal of Business Research, 66(1), 105–114. doi:10.1016/j.jbusres.2011.07.029

Brown, T. J., Barry, T. E., Dacin, P. A., & Gunst, R. F. (2005). Spreading the word: Investigatingantecedents of consumers positive word-of-mouth intentions and behaviors in a retailingcontext. Journal of the Academy of Marketing Science, 33(2), 123–138. doi:10.1177/0092070304268417

Calder, B. J., Malthouse, E. C., & Schaedel, U. (2009). An experimental study of the relationshipbetween online engagement and advertising effectiveness. Journal of Interactive Marketing, 23(4), 321–331. doi:10.1016/j.intmar.2009.07.002

Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait-multimethod matrix. Psychological Bulletin, 56(2), 81–105. doi:10.1037/h0046016

Chu, S.-C., & Kim, Y. (2011). Determinants of consumer engagement in electronic word-of-mouth(ewom) in social networking sites. International Journal of Advertising, 30(1), 47–75. doi:10.2501/ija-30-1-047-075

Churchill, G. A., Jr. (1979). A paradigm for developing better measures of marketing constructs.Journal of Marketing Research, 16(1), 64–73. doi:10.2307/3150876

Cobb-Walgren, C. J., Ruble, C. A., & Donthu, N. (1995). Brand equity, brand preference, andpurchase intent. Journal of Advertising, 24(3), 25–40. doi:10.1080/00913367.1995.10673481

Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale, NJ: LawrenceErlbaum Associates.

Cohen, J. B., & Basu, K. (1987). Alternative models of categorization: Toward a contingentprocessing framework. Journal of Consumer Research, 13(4), 455–472. doi:10.1086/209081

Dahlén, M. (2005). The medium as a contextual cue: Effects of creative media choice. Journal ofAdvertising, 34(3), 89–98. doi:10.1080/00913367.2005.10639197

JOURNAL OF MARKETING MANAGEMENT 553

Day, G. S., & Montgomery, D. B. (1999). Charting new directions for marketing. Journal of Marketing,63, 3. doi:10.2307/1252096

De Vries, L., Gensler, S., & Leeflang, P. S. (2012). Popularity of brand posts on brand fan pages: Aninvestigation of the effects of social media marketing. Journal of Interactive Marketing, 26(2), 83–91. doi:10.1016/j.intmar.2012.01.003

Elowitz, B., & Charlene, L. (2009). The world’s most valuable brands. Who’s most engaged?ENGAGEMENTdb: Ranking the top 100 global brands . Retrieved from http://www.ENGAGEMENTdb.com

Falls, J. (2010). What is engagement and how do we measure it? Social media explorer. Retrievedfrom https://www.socialmediaexplorer.com/social-media-marketing/what-is-engagement-and-how-to-we-measure-it/

Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior. Reading, MA: Addison-WesleyPublishing Company.

Fournier, S., & Lee, L. (2009). Getting brand communities right. Harvard Business Review, 87(4), 105–111. Retrieved from http://www.wsuakpsi.com/wp-content/uploads/2015/02/Brand-community-3.pdf

Gambetti, R. C., Graffigna, G., & Biraghi, S. (2012). The grounded theory approach to consumer–brand engagement: The practitioner’s standpoint. International Journal of Market Research, 54(5), 659–687. doi:10.2501/ijmr-54-5-659-687

Ghose, A., Ipeirotis, P. G., & Li, B. (2012). Designing ranking systems for hotels on travel searchengines by mining user-generated and crowdsourced content. Marketing Science, 31(3), 493–520. doi:10.1287/mksc.1110.0700

Ghose, A., Ipeirotis, P. G., & Li., B. (2014). Examining the impact of ranking on consumer behaviorand search engine revenue. Management Science, 60(7), 1632–1654. doi:10.1287/mnsc.2013.1828

Ghose, A., & Yang, S. (2009). An empirical analysis of search engine advertising: Sponsoredsearch in electronic markets. Management Science, 55(10), 1605–1622. doi:10.1287/mnsc.1090.1054

Goh, K.-Y., Heng, C.-S., & Lin, Z. (2013). Social media brand community and consumer behavior:Quantifying the relative impact of user- and marketer-generated content. Information SystemsResearch, 24(1), 88–107. doi:10.1287/isre.1120.0469

Gretzel, U. (2006). Consumer generated content – Trends and implications for branding. E-Reviewof Tourism Research, 4(3), 9–11. Retrieved from http://3ws1wk1wkqsk36zmd6ocne81.wpengine.netdna-cdn.com/files/2012/09/199_c-4-3-1.pdf

Gummerus, J., Liljander, V., Weman, E., & Pihlström, M. (2012). Customer engagement in aFacebook brand community. Management Research Review, 35(9), 857–877. doi:10.1108/01409171211256578

Habibi, M. R., Laroche, M., & Richard, M.-O. (2014). The roles of brand community and communityengagement in building brand trust on social media. Computers in Human Behavior, 37(37),152–161. doi:10.1016/j.chb.2014.04.016

Hanna, R., Rohm, A., & Crittenden, V. L. (2011). We’re all connected: The power of the social mediaecosystem. Business Horizons, 54(3), 265–273. doi:10.1016/j.bushor.2011.01.007

Hansen, F. (1969). Primary group influence and consumer conformity. In P. R. McDonald (Eds.),Proceedings of the American Marketing Association’s educators conference (pp. 300–305). Chicago,IL: American Marketing Association.

Harvey, O. J. (1960). Reciprocal influence of the group and three types of leaders in anunstructured situation. Sociometry, 23(1), 57–71. doi:10.2307/2786138

He, H., Zhu, W., & Zheng, X. (2014). Procedural justice and employee engagement: Roles oforganizational identification and moral identity centrality. Journal of Business Ethics, 122(4),681–695. doi:10.1007/s10551-013-1774-3

Heise, D. R. (1969). Separating reliability and stability in test-retest correlation. AmericanSociological Review, 34(1), 93–101. doi:10.2307/2092790

554 S. YANG ET AL.

Hennig-Thurau, T., Gwinner, K. P., Walsh, G., & Gremler, D. D. (2004). Electronic word-of-mouth viaconsumer-opinion platforms: What motivates consumers to articulate themselves on theinternet? Journal of Interactive Marketing, 18(1), 38–52. doi:10.1002/dir.10073

Higgins, E. T. (2006). Value from hedonic experience and engagement. Psychological Review, 113(3),439–460. doi:10.1037/0033-295x.113.3.439

Higgins, E. T., & Scholer, A. A. (2009). Engaging the consumer: The science and art of the valuecreation process. Journal of Consumer Psychology, 19(2), 100–114. doi:10.1016/j.jcps.2009.02.002

Hollebeek, L. D. (2011a). Demystifying customer brand engagement: Exploring the loyaltynexus. Journal of Marketing Management, 27(7–8), 785–807. doi:10.1080/0267257x.2010.500132

Hollebeek, L. D. (2011b). Exploring customer brand engagement: Definition and themes. Journal ofStrategic Marketing, 19(7), 555–573. doi:10.1080/0965254x.2011.599493

Hollebeek, L. D., Glynn, M. S., & Brodie, R. J. (2014). Consumer brand engagement in social media:Conceptualization, scale development and validation. Journal of Interactive Marketing, 28(2),149–165. doi:10.1016/j.intmar.2013.12.002

Houston, M. B. (2004). Assessing the validity of secondary data proxies for marketingconstructs. Journal of Business Research, 57(2), 154–161. doi:10.1016/s0148-2963(01)00299-5

Houston, M. B., & Johnson, S. A. (2000). Buyer–supplier contracts versus joint ventures:Determinants and consequences of transaction structure. Journal of Marketing Research, 37(1),1–15. doi:10.1509/jmkr.37.1.1.18719

Jerath, K., Ma, L., Park, Y.-H., & Srinivasan, K. (2011). A “position paradox” in sponsored searchauctions. Marketing Science, 30(4), 612–627. doi:10.1287/mksc.1110.0645

Kabadayi, S., & Price, K. (2014). Consumer–brand engagement on Facebook: Liking andcommenting behaviors. Journal of Research in Interactive Marketing, 8(3), 203–223.doi:10.1108/jrim-12-2013-0081

Karjaluoto, H., & Leinonen, H. (2009). Advertisers’ perceptions of search engine marketing.International Journal of Internet Marketing and Advertising, 5(1/2), 95–112. doi:10.1504/ijima.2009.021952

Katona, Z., & Sarvary, M. (2010). The race for sponsored links: Bidding patterns for searchadvertising. Marketing Science, 29(2), 199–215. doi:10.1287/mksc.1090.0517

Keller, K. L. (2010). The new branding imperatives. Cambridge, MA: MarketingScience Institute. Retrieved from http://exec.tuck.dartmouth.edu/downloads/350/msi_ff_10-700.pdf

Koschate-Fischer, N., Cramer, J., & Hoyer, W. D. (2014). Moderating effects of the relationshipbetween private label share and store loyalty. Journal of Marketing, 78(2), 69–82. doi:10.1509/jm.13.0075

Lin, S., Ross, W. J., & Liu, H. (2015). Does the social value of a brand matter? An empiricalinvestigation of the impact of brand social engagement on firm’s financial performance.Retrieved from http://ssrn.com/abstract=2670894

Malhotra, A., Malhotra, C. K., & See, A. (2013). How to create brand engagement on Facebook. MITSloan Management Review, 54(2), 18–20. http://sloanreview.mit.edu/article/how-to-create-brand-engagement-on-facebook/

Martin, D., Schouten, J., & McAlexander, J. (2006). Reporting ethnographic research: Bringingsegments to life through movie making and metaphor. Handbook of Qualitative ResearchMethods in Marketing, 361–370. doi:10.4337/9781847204127.00037

McAlexander, J. H., Schouten, J. W., & Koenig, H. F. (2002). Building brand community. Journal ofMarketing, 66(1), 38–54. doi:10.1509/jmkg.66.1.38.18451

Mervis, C. B., & Rosch, E. (1981). Categorization of natural objects. Annual Review of Psychology, 32,89–115. doi:10.1146/annurev.ps.32.020181.000513

Millen, D. R., & Patterson, J. F. (2002). Stimulating social engagement in a communitynetwork. In Proceedings of the 2002 ACM conference on Computer supported cooperativework (pp. 306–313). New York, NY: ACM Press. doi:10.1145/587078.587121

JOURNAL OF MARKETING MANAGEMENT 555

Miller, A. R., & Tucker, C. (2014). Health information exchange, system size and information silos.Journal of Health Economics, 33, 28–42. doi:10.1016/j.jhealeco.2013.10.004

Mollen, A., & Wilson, H. (2010). Engagement, telepresence, and interactivity in online consumerexperience: Reconciling scholastic and managerial perspectives. Journal of Business Research, 63(9–10), 919–925. doi:10.1016/j.jbusres.2009.05.014

Moorman, C., Ross, B., & Gorman, S. (2014). 12 tips for integrating social media into your marketingstrategy. Retrieved from http://cmosurvey.org/blog/12-tips-for-integrating-social-media-into-your-marketing-strategy/

Moorman, C., Ross, B., & Gorman, S. (2015). Measuring the impact of social media on your business.Retrieved from http://cmosurvey.org/blog/measuring-the-impact-of-social-media-on-your-business/

Moschis, G. P. (1976). Social comparison and informal group influence. Journal of MarketingResearch, 13(3), 237–244. doi:10.2307/3150733

Muñiz, A. M., Jr., & O’Guinn, T. C. (2001). Brand community. Journal of Consumer Research, 27(4),412–432. doi:10.1086/319618

Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “Like” button: The impact of merevirtual presence on brand evaluations and purchase intentions in social media settings. Journalof Marketing, 76(6), 105–120. doi:10.1509/jm.11.0105

Nelson-Field, K., Riebe, E., & Sharp, B. (2012). What’s not to ‘like?’: Can a Facebook fan base give abrand the advertising reach it needs? Journal of Advertising Research, 52(2), 262–269.doi:10.2501/jar-52-2-262-269

Pan, B., Hembrooke, H., Joachims, T., Lorigo, L., Gay, G., & Granka, L. (2007). In Google we trust:Users’ decisions on rank, position, and relevance. Journal of Computer-Mediated Communication,12(3), 801–823. doi:10.1111/j.1083-6101.2007.00351.x

Peracchio, L. A., & Luna, D. (2006). The role of thin-slice judgments in consumer psychology.Journal of Consumer Psychology, 16(1), 25–32. doi:10.1207/s15327663jcp1601_5

Phillips, B. J., & McQuarrie, E. F. (2010). Narrative and persuasion in fashion advertising. Journal ofConsumer Research, 37(3), 368–392. doi:10.1086/653087

Rangaswamy, A., Giles, C. L., & Seres, S. (2009). A strategic perspective on search engines: Thoughtcandies for practitioners and researchers. Journal of Interactive Marketing, 23(1), 49–60.doi:10.1016/j.intmar.2008.10.006

Rishika, R., Kumar, A., Janakiraman, R., & Bezawada, R. (2013). The effect of customers’ social mediaparticipation on customer visit frequency and profitability: An empirical investigation.Information Systems Research, 24(1), 108–127. doi:10.1287/isre.1120.0460

Rothbard, N. P. (2001). Enriching or depleting? The dynamics of engagement in work and familyroles. Administrative Science Quarterly, 46, 655–684. doi:10.2307/3094827

Rutz, O. J., & Bucklin, R. E. (2011). From generic to branded: A model of spillover in paid searchadvertising. Journal of Marketing Research, 48(1), 87–102. doi:10.1509/jmkr.48.1.87

Rutz, O. J., & Trusov, M. (2011). Zooming in on paid search ads – A consumer-level modelcalibrated on aggregated data. Marketing Science, 30(5), 789–800. doi:10.1287/mksc.1110.0647

Rutz, O. J., Trusov, M., & Bucklin, R. E. (2011). Modeling indirect effects of paid search advertising:Which keywords lead to more future visits? Marketing Science, 30(4), 646–665. doi:10.1287/mksc.1110.0635

Saks, A. M. (2006). Antecedents and consequences of employee engagement. Journal ofManagerial Psychology, 21(7), 600–619. doi:10.1108/02683940610690169

Sprott, D., Czellar, S., & Spangenberg, E. (2009). The importance of a general measure of brandengagement on market behavior: Development and validation of a scale. Journal of MarketingResearch, 46(1), 92–104. doi:10.1509/jmkr.46.1.92

Srivastava, R. K., Shervani, T. A., & Fahey, L. (1999). Marketing, business processes, and shareholdervalue: An organizationally embedded view of marketing activities and the discipline ofmarketing. Journal of Marketing, 63(Special Issue), 168–179. doi:10.2307/1252110

Stafford, J. E. (1966). Effects of group influences on consumer brand preferences. Journal ofMarketing Research, 3(1), 68–75. doi:10.2307/3149437

556 S. YANG ET AL.

Stanton, J. M., Sinar, E. F., Balzer, W. K., & Smith, P. C. (2002). Issues and strategies for reducing thelength of self-report scales. Personnel Psychology, 55(1), 167–194. doi:10.1111/j.1744-6570.2002.tb00108.x

Tadena, N. (2014, September 3). Social media spending is on the rise but impact is hard tomeasure. The Wall Street Journal. Retrieved from http://blogs.wsj.com/cmo/2014/09/03/social-media-spending-is-on-the-rise-but-impact-is-hard-to-measure/

Trusov, M., Bucklin, R. E., & Pauwels, K. (2009). Effects of word-of-mouth versus traditionalmarketing: Findings from an internet social networking site. Journal of Marketing, 73(5), 90–102. doi:10.1509/jmkg.73.5.90

Van Doorn, J., Lemon, K. N., Mittal, V., Nass, S., Pick, D., Pirner, P., & Verhoef, P. C. (2010). Customerengagement behavior: Theoretical foundations and research directions. Journal of ServiceResearch, 13(3), 253–266. doi:10.1177/1094670510375599

Vilpponen, A., Winter, S., & Sundqvist, S. (2006). Electronic word-of-mouth in online environments:Exploring referral network structure and adoption behavior. Journal of Interactive Advertising, 2,71–86. doi:10.1080/15252019.2006.10722120

Vivek, S. D., Beatty, S. E., & Morgan, R. M. (2010). Consumer engagement: Exploring customerrelationships beyond purchase. Journal of Marketing Theory and Practice, 20(2), 127–145.doi:10.2753/mtp1069-6679200201

Wang, A. (2006). Advertising engagement: A driver of message involvement on message effects.Journal of Advertising Research, 46(4), 355–368. doi:10.2501/s0021849906060429

Weinberg, B. D., & Pehlivan, E. (2011). Social spending: Managing the social media mix. BusinessHorizons, 54(3), 275–282. doi:10.1016/j.bushor.2011.01.008

Westerwick, A. (2013). Effects of sponsorship, web site design, and Google ranking on thecredibility of online information. Journal of Computer-Mediated Communication, 18(2), 80–97.doi:10.1111/jcc4.12006

Winer, R. S. (1999). Experimentation in the 21st century: The importance of external validity.Journal of the Academy of Marketing Science, 27(3), 349–358. doi:10.1177/0092070399273005

Witt, R. E. (1969). Informal social group influence on consumer brand choice. Journal of MarketingResearch, 6(4), 473–476. doi:10.2307/3150086

Witt, R. E., & Bruce, G. D. (1970). Purchase decisions and group influence. Journal of MarketingResearch, 7(4), 533–555. doi:10.2307/3149651

Xiang, Z., & Gretzel, U. (2010). Role of social media in online travel information search. TourismManagement, 31(2), 179–188. doi:10.1016/j.tourman.2009.02.016

Xiang, Z., Wöber, K., & Fesenmaier, D. R. (2008). Representation of the online tourism domain insearch engines. Journal of Travel Research, 47(2), 137–150. doi:10.1177/0047287508321193

Yang, S., & Ghose, A. (2010). Analyzing the relationship between organic and sponsored searchadvertising: Positive, negative, or zero interdependence? Marketing Science, 29(4), 602–623.doi:10.1287/mksc.1090.0552

Zhang, S., Jiang, H., & Carroll, J. M. (2011). Integrating online and offline community throughFacebook. In Proceedings of the 2011 international conference on Collaboration Technologies andSystems (pp. 569–578). Piscataway, NJ: IEEE Press. doi:10.1109/CTS.2011.5928738

Zinnbauer, M., & Honer, T. (2011). How brands can create social currency – A framework formanaging brands in a network era. Marketing Review St. Gallen, 28(5), 50–55. doi:10.1007/s11621-011-0063-8

JOURNAL OF MARKETING MANAGEMENT 557