+ All Categories
Home > Documents > Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of...

Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of...

Date post: 23-Oct-2015
Category:
Upload: johan-mark-berends
View: 220 times
Download: 10 times
Share this document with a friend
Description:
Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media settings. Journal of Marketing, 76(6), 105-120.
Popular Tags:
17
Rebecca Walker Naylor, Cait Poynor Lamberton, & Patricia M. West Beyond the "Like" Button: The Impact of Mere Virtual Presence on Brand Evaluations and Purchase Intentions in Social Media Settings By 2011, approximately 83% of Fortune 500 companies were using some form of social media to connect with consumers. Furthermore, surveys suggest that consumers are increasingly relying on social media to learn about unfamiliar brands. However, best practices regarding the use of social media to bolster brand evaluations in such situations remain undefined. This research focuses on one practice in this domain: the decision to hide or reveal the demographic characteristics of a brand's online supporters. The results from four studies indicate that even when the presence of these supporters is only passively experienced and virtual (a situation the authors term "mere virtual presence"), their demographic characteristics can influence a target consumer's brand evaluations and purchase intentions. The findings suggest a framework for brand managers to use when deciding whether to reveal the identities of their online supporters or to retain ambiguity according to (1 ) the composition of existing supporters relative to targeted new supporters and (2) whether the brand is likely to be evaluated singly or in combination with competing brands. Keywords: social influence, mere presence, social media, social networks, ambiguity W hile a decade and a half of work building on Hoff- man and Novak's (1996) analysis of computer- mediated environments has informed management of online media, much of this work suggests that consumers interact with brands online in ways similar to what they do offline. That is, consumers join online brand communities for many of the same reasons they join offline brand com- munities (e.g., Algesheimer, Dholakia, and Herrmann 2005; Muniz and O'Guinn 2001; Schau, Muñiz, and Amold 2009; Thompson and Sinha 2008). However, social media practitioners now seek best prac- tices for contexts in which brick-and-mortar research is largely inapplicable. Specifically, social media can make the identity of a brand's supporters transparent to prospec- tive consumers in ways that have no offline analog. Before the advent of social networking, consumers could only guess at the identities of other brand supporters on the basis of advertising or the identity of spokespeople. In contrast, in the social media world, consumers viewing a brand page are likely to see pictorial information about other people Rebecca Walker Naylor is Assistant Professor of Marketing (e-mail: [email protected]), and Patricia M. West is Associate Professor of Marketing (e-mail: west_284§fisher.osu.edu). Fisher College of Busi- ness, The Ohio State University. Cait Poynor Lamberton is Assistant Pro- fessor of Marketing, Katz Graduate School of Business, University of Pittsburgh (e-mail: cpoynor§katz.pitt.edu). Order of authorship is arbi- trary; all authors contributed equally to this article. The authors thank Dar- ren Dahl, Jeff Inman, Greg Allenby, and Andrew Hayes for their comments on various aspects of this research. who have voluntarily affiliated with the brand. We refer to the passive exposure to a brand's supporters experienced in such social media contexts as "mere virtual presence" (MVP). Little is known about if, or how, MVP affects con- sumers or how it can best be managed. Note that MVP takes many forms. For example, some Facebook brand pages display profile pictures of the brand's supporters. Companies may also use Facebook Connect, so that a user's Facebook profile picture is dis- played to other prospective users on their site (see, e.g., www.Groupon.com and www.Connect.Redbullusa.com). Other companies encourage consumers to post pictures of themselves using a brand either to their Facebook brand page (e.g., Talbots) or to a company-run social network (e.g.. Burberry's Art of the Trench website and the "How We Wear Them" section of Tom's Shoes' website). Although a 2011 study shows that more than 80% of Fortune 500 com- panies use some form of social media (Hameed 2011), prac- titioners recognize that a large number of "likes" does not necessarily translate into meaningful outcomes (Lake 2011). Given that consumers increasingly look to social media to form opinions about unfamiliar brands (Baird and Parasnis 2011; Newman 2011), how can managers use MVP to gen- erate substantive differences in brand evaluations and pur- chase intentions? We answer this question by exploring the effects of four distinct types of MVP on brand evaluations and purchase intentions. Note that in the pre-social-media world, the identity of a brand's supporters was largely unknown. The analog to this position in the social media world would be ® 2012, American Marketing Association iSSN: 0022-2429 (print), 1547-7185 (eiectronic) 105 Journal of Marketing Voiume 76 (November 2012), 105-120
Transcript
Page 1: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

Rebecca Walker Naylor, Cait Poynor Lamberton, & Patricia M. West

Beyond the "Like" Button: TheImpact of Mere Virtual Presence onBrand Evaluations and Purchase

Intentions in Social Media SettingsBy 2011, approximately 83% of Fortune 500 companies were using some form of social media to connect withconsumers. Furthermore, surveys suggest that consumers are increasingly relying on social media to learn aboutunfamiliar brands. However, best practices regarding the use of social media to bolster brand evaluations in suchsituations remain undefined. This research focuses on one practice in this domain: the decision to hide or revealthe demographic characteristics of a brand's online supporters. The results from four studies indicate that evenwhen the presence of these supporters is only passively experienced and virtual (a situation the authors term "merevirtual presence"), their demographic characteristics can influence a target consumer's brand evaluations andpurchase intentions. The findings suggest a framework for brand managers to use when deciding whether to revealthe identities of their online supporters or to retain ambiguity according to (1 ) the composition of existing supportersrelative to targeted new supporters and (2) whether the brand is likely to be evaluated singly or in combination withcompeting brands.

Keywords: social influence, mere presence, social media, social networks, ambiguity

While a decade and a half of work building on Hoff-man and Novak's (1996) analysis of computer-mediated environments has informed management

of online media, much of this work suggests that consumersinteract with brands online in ways similar to what they dooffline. That is, consumers join online brand communitiesfor many of the same reasons they join offline brand com-munities (e.g., Algesheimer, Dholakia, and Herrmann 2005;Muniz and O'Guinn 2001; Schau, Muñiz, and Amold 2009;Thompson and Sinha 2008).

However, social media practitioners now seek best prac-tices for contexts in which brick-and-mortar research islargely inapplicable. Specifically, social media can makethe identity of a brand's supporters transparent to prospec-tive consumers in ways that have no offline analog. Beforethe advent of social networking, consumers could onlyguess at the identities of other brand supporters on the basisof advertising or the identity of spokespeople. In contrast,in the social media world, consumers viewing a brand pageare likely to see pictorial information about other people

Rebecca Walker Naylor is Assistant Professor of Marketing (e-mail:[email protected]), and Patricia M. West is Associate Professorof Marketing (e-mail: west_284§fisher.osu.edu). Fisher College of Busi-ness, The Ohio State University. Cait Poynor Lamberton is Assistant Pro-fessor of Marketing, Katz Graduate School of Business, University ofPittsburgh (e-mail: cpoynor§katz.pitt.edu). Order of authorship is arbi-trary; all authors contributed equally to this article. The authors thank Dar-ren Dahl, Jeff Inman, Greg Allenby, and Andrew Hayes for their commentson various aspects of this research.

who have voluntarily affiliated with the brand. We refer tothe passive exposure to a brand's supporters experienced insuch social media contexts as "mere virtual presence"(MVP). Little is known about if, or how, MVP affects con-sumers or how it can best be managed.

Note that MVP takes many forms. For example, someFacebook brand pages display profile pictures of thebrand's supporters. Companies may also use FacebookConnect, so that a user's Facebook profile picture is dis-played to other prospective users on their site (see, e.g.,www.Groupon.com and www.Connect.Redbullusa.com).Other companies encourage consumers to post pictures ofthemselves using a brand either to their Facebook brandpage (e.g., Talbots) or to a company-run social network(e.g.. Burberry's Art of the Trench website and the "How WeWear Them" section of Tom's Shoes' website). Although a2011 study shows that more than 80% of Fortune 500 com-panies use some form of social media (Hameed 2011), prac-titioners recognize that a large number of "likes" does notnecessarily translate into meaningful outcomes (Lake 2011).Given that consumers increasingly look to social media toform opinions about unfamiliar brands (Baird and Parasnis2011; Newman 2011), how can managers use MVP to gen-erate substantive differences in brand evaluations and pur-chase intentions?

We answer this question by exploring the effects of fourdistinct types of MVP on brand evaluations and purchaseintentions. Note that in the pre-social-media world, theidentity of a brand's supporters was largely unknown. Theanalog to this position in the social media world would be

® 2012, American Marketing AssociationiSSN: 0022-2429 (print), 1547-7185 (eiectronic) 105

Journal of MarketingVoiume 76 (November 2012), 105-120

Page 2: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

choosing not to reveal the identity of a brand's online sup-porters, a case we call "ambiguous MVP." If a brand dis-plays the identity of its supporters, in social media settings,this information is typically conveyed through profile pho-tographs that display brand supporters' demographic char-acteristics.' Relative to a target consumer, displayed MVPmay be demographically similar or demographically dis-similar, or it may present a heterogeneous mix of similarand dissimilar consumers. We compare the effects of main-taining ambiguous MVP with that created by each of thesetypes of identified MVP. In doing so, we show when it ismore beneficial to reveal the identity of current brand sup-porters to prospective customers and when to retain ambi-guity about the brand's support base.

From a practical perspective, this research contributesto the limited academic research investigating how firmscan best configure their social networks to meet strategicobjectives. For example. Tucker and Zhang (2010) demon-strate that displaying the number of sellers and buyers inonline exchanges can change business-to-business listingand buying behavior. However, such findings provide lim-ited managerial guidance because they do not compare theeffects of displaying the number of members on a particularsite with the range of other actions a manager may considerwhen deciding how or whether to display online supporters.From a theoretical perspective, our work challenges socialinfluence theory (SIT; Latane 1981), which suggests thatvirtual exposure to unknown others should exert little socialinfluence. Furthermore, we provide novel insights intosocial influence effects created by heterogeneous groupsand ambiguous others, for which the traditional referencegroup literature (e.g., Bearden and Etzel 1982; Bearden,Netemeyer, and Teel 1989; Berger and Heath 2007;Bumkrant and Cousineau 1975; Childers and Rao 1992;Escalas and Bettman 2003) is largely silent. In addition, weshow the importance of joint versus separate evaluationmode (Hsee et al. 1999; Hsee and Leclerc 1998) as a mod-erator of the influence of ambiguous MVP. Finally, ourfindings yield a road map for brand managers to use whendeciding whether to reveal the identities of their online sup-porters or to retain ambiguity according to (1) the demo-graphic composition of existing supporters relative to tar-geted new supporters and (2) whether the brand is likely tobe evaluated singly or in combination with competingbrands.

Predicting Consumer Response toMVP

Building on past work in mere presence effects (e.g., Argo,Dahl, and Manchanda 2005), we use the MVP term todescribe the photographic presence of brand supporters inonline settings. This virtual exposure to other consumers

'Although consumers may use pictures of things other thanthemselves as their profile picture on social networking sites, in anonline survey we conducted of 307 Intemet users (M g = 28.7years), 97% of participants who reported having a Facebook pro-file (n = 274) indicated that they use a photograph of themselvesas their profile picture.

contrasts with the social influence a consumer mightencounter in an offline setting, where, for example, in aretail outlet, interpersonal comparison is more immediate,spatial crowding may occur, or future interaction is possible(e.g., Argo, Dahl, and Manchanda 2005).

Notably, classic theory suggests that MVP may have lit-tle effect on a consumer evaluating a new brand. Socialimpact theory (SIT; Latane 1981) contends that for socialinfluence to be manifest, individuals must be present inlarge numbers and be in close proximity to the target andthat the influence must be provided by an important or pow-erful source. In MVP, these conditions are not met. Rather,brand supporters are generally displayed in small groups(making them relatively few in number), are not in physicalproximity to the consumer, and, given that they arestrangers to the target consumer, are low in "sourcestrength."

However, we question whether the conditions of SIT arenecessary for MVP to exert influence. The reference groupliterature acknowledges that knowing who the other usersof a brand are may affect a consumer's reaction to thatbrand (e.g., Bearden and Etzel 1982; Bearden, Netemeyer,and Teel 1989; Berger and Heath 2007; Bumkrant andCousineau 1975; Childers and Rao 1992; Escalas andBettman 2003). Although this literature has not exploredvirtual presence of other users, it raises the possibility thatinformation about a brand's supporters may change brandevaluations even if it does not meet SIT's requirements. Wefirst use the reference group literature to discuss the likelyeffects of similar and dissimilar MVP. We then make pre-dictions about the impact of maintaining ambiguity asopposed to displaying different types of MVP. We also pre-dict consumers' responses to a heterogeneous group ofsimilar and dissimilar individuals, a topic that, while notaddressed in the reference group literature, becomes impor-tant when firms use social media platforms with potentiallyhighly diverse users.

Peas in a Pod: Simiiar Versus Dissimiiar MVPPeople tend to express affinity for those to whom they aresimilar (Lydon, Jamieson, and Zanna 1988; Morry 2007;Shachar and Emerson 2000). Furthermore, seeing similarothers supporting a brand will lead to greater affinity for thebrand (Berger and Heath 2007; Escalas and Bettman 2003;McCracken 1988). Target marketing relies on this idea, suchthat individuals are assumed to be more persuaded by adver-tising featuring those who are similar to the self (Aaker,Brumbaugh, and Grier 2000; Deshpandé and Stayman 1994).More recent work shows parallel effects in the context ofonline reviews, in which consumers infer shared tastes andpreferences from verbally provided descriptive information(vs. photos) about a reviewer, which in tum determine howpersuasive they find the reviewer's recommendation (Nay-lor, Lamberton, and Norton 2011). We refer to this infer-ence of shared preferences as "inferred commonality."

Note, however, that inferred commonality has primarilybeen considered in cases in which such inferences are ratio-nally based on provided information. For example, informa-tion provided in reviewer posts could rationally inform

106 / Journal of Marketing, November 2012

Page 3: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

inferred commonality; writing style, expression of priori-ties, or shared interest could prompt reasonable consumersto generalize to other facets of preference. In contrast, thepresent study considers the effect of the pictorial MVP ofconsumers. In this case, consumers have affiliated with thebrand but have done so without any persuasive intent.Moreover, they have not provided written product informa-tion or recommendations to try the brand that would groundinferences of commonality.

Despite these differences, we expect that similar MVPwill generate high levels of inferred commonality with abrand's user base. We base this expectation in managementresearch suggesting that demographic similarity (which werefer to simply as "similarity" and can be observed fromphotographs) leads to inferences of deeper-level commonal-ity. That is, even in the absence of any other information,demographically similar individuals are presumed to sharepersonality traits, values, and attitudes (Cunningham 2007).This inferred commonality prompts the individual to raisehis or her evaluation of the brand.

In contrast, previous literature has suggested that con-sumers exposed to dissimilar MVP will infer little common-ality with the brand's users and will express lower evalua-tions for the brand than for a brand with similar MVP.Importantly, the reference group literature suggests thateven if the dissimilar brand supporters are not explicitlydissociative (i.e., members of groups with whom consumersdo not want to be associated) (White and Dahl 2006,2007),consumers may avoid similar purchase patterns simply dueto demographic dissimilarity (Berger and Heath 2008).Work on non-target market effects also suggests that seeingdissimilar individuals can lead consumers to infer low lev-els of conunonality (Aaker et al. 2000). Thus, if MVP indi-cates that the brand is liked by people whom target con-sumers perceive as dissimilar, target consumers should inferless commonality between themselves and the brand's userbase and adjust their liking for the brand downward com-pared with when MVP is similar.

Is Ignorance Bliss? Ambiguous MVPGiven that displaying MVP dissimilar to a target consumermay lower brand evaluations in comparison with similarMVP, perhaps displaying ambiguous MVP is the firm'ssafest decision. Ambiguous MVP involves the display ofothers about whom no or very limited identifying demo-graphic information is provided. Thus, ambiguity may bemanifest by not showing any pictures of brand supporters,showing only supporters who have not provided a picture,or showing photos of brand supporters whose identity hasbeen obscured.

Prior research offers little guidance regarding the use ofambiguous MVP. Some literature suggests that when peopleencounter unidentified others, they infer little commonalitywith them. Sassenberg and Postmes (2002), for example,show that when people know nothing about other groupmembers, they report low levels of liking and low percep-tions of group cohesiveness. These authors ground theirfindings in social categorization theory (Turner et al. 1987),which argues that individuals who cannot be placed in a

person's in-group will be subject to any stereotypes com-mon to out-groups. However, recent research on ambiguityadopts an information-processing perspective and comes toa different conclusion. This work suggests that in theabsence of externally provided information about others,consumers anchor on the self to infer that ambiguous othersare like them. Because of these inferences, Naylor, Lamber-ton, and Norton (2011) demonstrate that an ambiguousonline reviewer is more persuasive than a dissimilarreviewer and equally as persuasive as a similar reviewer.

How does this research translate to the present context?Note that MVP does not involve extended formation of in-and out-groups or interaction among members. Rather, itinvolves only incidental, passive exposure to other con-sumers. Given the nature of MVP exposure, we proposethat the information-processing explanation, rather than thesocial categorization perspective, is likely to hold. In otherwords, when MVP is ambiguous, consumers will projecttheir own characteristics onto the brand's user base (thusinferring commonality), emerging with a level of affinitylike that generated by similar MVP but greater than that cre-ated by dissimilar MVP.

Heterogeneous MVPAlthough the difference between similar and dissimilarMVP can be predicted according to prior literature, existingtheory fails to explain responses to heterogeneous MVP.Understanding reactions to heterogeneous groups is impor-tant, because it is possible that a brand will not present sup-porters that are uniformly similar or dissimilar to the target,either because doing so is out of their control or becausetheir objectives include extension into previously uru-epre-sented market segments. Diverse groups do not form acohesive "reference group" in the traditional sense, and thusthe reference group literature has little to say on this point.Some prior research has suggested that diverse groups maybe interpreted in the same manner as a group perceived tobe uniformly dissimilar (i.e., that the group's preferences donot match the target's). For example, Jehn, Northcraft, andNeale (1999) argue that diversity in a workgroup cues indi-viduals to expect opinions and behaviors that diverge fromtheir own. Similarly, Chatman and Flynn (2001) show thatdemographic heterogeneity within a workgroup initiallyleads to low levels of cooperation. However, someresearchers advise broad inclusion of a wide range of con-sumers as members of social networking sites (Dholakiaand Vianello 2009), arguing that heterogeneity could indi-cate a brand's wide range of features or suggest broadappeal.

Consistent with these recommendations, we predict thatthe MVP of a heterogeneous mix of supporters can be astrength rather than a weakness for firms, albeit for differ-ent reasons than those Dholakia and Vianello (2009) pro-pose. We base our prediction in the idea that individualstend to be particularly sensitive to incidental similaritiesbetween themselves and others, showing more positive atti-tudes toward a product in the presence of even superficialsimilarities (e.g., Jiang et al. 2010). Furthermore, work onthe self-referencing effect shows the positive effect of self-

Beyond the "Like" Button /107

Page 4: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

relatedness for information processing (e.g., Perkins, Fore-hand, and Greenwald 2005), such that individuals showenhanced attention and cognitive fluency for informationperceived as self-congruent. If such effects hold in a socialmedia context, individuals will be more influenced by thepresence of even a small number of individuals in an MVParray who are similar to themselves (i.e., who are directlyself-relevant) than a small number who are dissimilar (andthus are less self-relevant). Therefore, we anticipate that aconsumer viewing a social media site with heterogeneousMVP will be particularly sensitive to the presence of thesimilar individual(s) in the array. In tum, brand evaluationswill be equivalent to those formed when consumers areexposed to similar MVP. Notably, given that we hypothe-size that ambiguous MVP will create brand evaluations likethose created by similar MVP, heterogeneous and ambigu-ous MVP should also produce equivalent brand evaluations.

A question that remains, however, is how much similar-ity must be present in a heterogeneous MVP array for it togenerate inferences and evaluations like homogenous simi-lar MVP. Note that in Asch's classic work on conformity(Asch 1955, 1956), social influence effects can be gener-ated by even a small number of individuals in a largergroup. That is, homogeneity among confederates was notnecessary to prompt study participants to alter their judg-ments of stimuli. Thus, while there is no theory to directlyguide predictions about heterogeneity in the social mediacontext, we propose that even a small proportion of similarindividuals in a heterogeneous MVP array may produceevaluations like those produced by homogeneous similarMVP. To test this, we empirically manipulate number ofsimilar individuals in an MVP array to range from zero (dis-similar MVP) to 100% (similar MVP).

Thus, we suggest that MVP will influence brand evalua-tions as follows:

H¡: Ambiguous MVP produces (a) equivalent brand evalua-tions to homogeneous similar MVP, (b) equivalent brandevaluations to heterogeneous MVP, and (c) significantlymore positive brand evaluations than homogeneous dis-similar MVP.

H2: The relationship proposed between MVP composition andbrand evaluations in Hi is mediated by inferences ofcommonality with the brand's user base.2

We flrst test these hypotheses across three studiesemploying different operationalizations of ambiguous MVPand similarity. Study la tests all parts of H| using age tomanipulate similarity. Study lb tests the parts of H] pertain-ing to ambiguity, similarity, and dissimilarity using genderto manipulate similarity. Then, given that Study 1 leavesunanswered questions about heterogeneity. Study 2 focusesprimarily on heterogeneity, providing a direct test of Hib.Studies lb and 2 both include tests of H2 (the mediation

2Note that because H|a and H|b predict equivalence, a media-tion test would not be able to explain variance in the dependentmeasure for these hypotheses. Thus, H2 predicts that ambiguousMVP leads to a higher level of inferred commonality than doesdissimilar MVP, which explains the difference in brand evalua-tions between these types of MVP predicted in Hj^.

hypothesis). Study 3 introduces our theorizing regarding themoderating effect of joint and single evaluation contextsand replicates results related to Hi^ and Hi ..

Study 1aStudy la compares participants' liking for an unfamiliarbrand when they observe different types of MVP. We useage to manipulate similarity given previous work by practi-tioners and academics highlighting the influence of agesimilarity on product preferences. For example, theYankelovich report on generational marketing argues thatdeterminants of product value are strongly influenced byage cohort, shared experiences, media icons, and life stage(Smith and Clurman 2009). Previous academic studies haveargued that other demographic characteristics drive attrac-tion between consumers, but note that age is likely to becorrelated with many of these characteristics. For example,Byme (1971) finds that shared job classification and maritalstatus make television media attractive to consumers. Simi-larly, Shachar and Emerson (2000) flnd that individualswho have families prefer watching shows about families.These characteristics are likely to be shared within at leastbroad age ranges, such that college students will differ frompeople 30-40 years of age, who will again differ frompeople older than 65 years. Thus, perceived age of the indi-viduals in an MVP array may act as a proxy for numerousother demographic characteristics that have been shown toinfluence similarity-based attraction.

Stimuii and ProcedureA total of 128 undergraduate students participating in thisstudy in exchange for extra credit were told that they wouldbe viewing an excerpt from the Facebook fan page createdby Roots, a Canadian clothing company. Participants readthe following information:

In this section of today's study, we'd like to you to look atexcerpts from an actual Facebook page for a real brand.This is the type of page where you can "become a fan" ofa company or brand .3 You will see excerpts from a pagemaintained by Roots, a real Canadian company interestedin expanding to the United States. Please look at the infor-mation featured on their Facebook page and respond tothe questions as honestly as possible.

Participants then viewed an excerpt from a simulatedRoots Facebook page (see the Web Appendix at www.marketingpower.com/jm_webappendix) and answered ques-tions about Roots clothing. As discussed previously, weoperationalized similarity using perceived age, holding gen-der constant. Participants indicated their gender before thestudy began so that all participants viewed fans matched totheir gender. All participants were told that there were thesame number of total fans regardless of MVP condition.Depending on condition, participants saw one of the follow-ing: (1) total number of fans and pictures of six fans that

3At the time we began this research, Facebook called brand sup-porters "fans" and brand pages "fan pages." The term "fan" hassince been replaced by the "like" button; consumers who werefans of a brand are now those that like the brand.

108 / Journal of Marketing, November 2012

Page 5: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

were the same age and gender as the participant^ (homoge-neous similar MVP condition), (2) total number of fans andpictures of six fans that were the same gender but a differentage than the participant (homogeneous dissimilar MVP con-dition), (3) total number of fans and three pictures of fansthat were the same gender and age and three pictures of fansthat were the same gender but a different age (heterogeneousMVP condition) as the participant, or (4) no fan pictures,only the total number of fans (ambiguous MVP condition).No specific direction was given to attend to the fans, andparticipants viewed the page as long as they liked. Otherinformation on the page was held constant across conditions.

We captured brand liking by asking, "Based on theinformation you just saw from their Facebook page, howmuch do you like Roots brand clothing?" on a nine-pointscale ranging from "do not like at all" to "like very much."As manipulation checks, we asked participants in the simi-lar, dissimilar, and heterogeneous MVP conditions to indi-cate how much they agreed that "Roots 'fans' are the sameage I am" and "Roots 'fans' are the same gender I am."Finally, we asked all participants whether they had heard ofRoots before the study and, if so, how familiar they werewith the brand on a nine-point scale (1 = "not at all famil-iar," and 9 = "very familiar").

ResultsSample and manipulation check. We first examined par-

ticipants' familiarity with Roots. Of the 128 participants, 27had heard of Roots before the study. Of these participants,15 indicated a familiarity score of five or above on the nine-point familiarity scale, and therefore we removed themfrom the data set. Of the remaining participants, two indi-cated that Roots fans were not the same gender they were(presumably because they did not follow instructions andindicated a different gender than their own before the begin-ning of the study); they were also removed from the dataset. This left a final usable sample of 111 participants (59men and 52 women) with an average age of 21 years. Themanipulation check revealed that participants in the similarcondition rated the fans as more similar to themselves inage (M = 7.81) than did participants in the heterogeneous(M - 5.11) and dissimilar conditions (M = 1.12; F(2, 81) =147.76,/? < .0001). We also note that there were no differ-ences in amount of time spent viewing the page across con-ditions (M = 26.7 seconds; F(3,107) = 1.13,/? = .34).

Liking for Roots clothing. We next examined partici-pants' liking for Roots clothing. Because MVP compositionwas a four-level variable, we used three orthogonal contrastcodes to compare the ambiguous condition with the otherthree conditions (i.e., these codes compared [1] ambiguousMVP vs. similar MVP [H|a], [2] ambiguous MVP vs. het-

^Because all participants in the subject pool at the universitywhere the studies were conducted were in their late teens to early20s, the fans used in the similar condition were also in this agerange. All fans in the dissimilar age condition were older, rangingfrom their 30s to their 60s. All pictures used were actual Facebookprofile pictures selected from the Facebook pages of individualswhose profile picture was public. We note that these picturesmanipulate perceived age, not objective age.

erogeneous MVP [HiJ, and [3] ambiguous MVP vs. dis-similar MVP [Hic]). This analysis thus indicates the effectof the managerial decision to reveal consumers' similar ordifferent demographic information to one another or toobscure it, maintaining ambiguity. Note that these contrastcodes partition the multivariate analysis of variance sums ofsquares into interpretable subsets, obviating the need forspecial alpha levels (Rosenthal, Rosnow, and Rubin 2000).Consistent with our hypotheses, only the contrast codecomparing the ambiguous MVP condition with the dissimi-lar MVP condition was significant. There was no significantdifference in liking between the similar (M = 4.74) andambiguous (M = 4.48) MVP conditions (F(l, 107) = .86,/? =.36) or between the heterogeneous (M = 4.89) and ambigu-ous MVP conditions (F(l, 107) =l.91,p= .17), in supportof Hja and H|b, respectively. We also note that there wereno differences in brand liking in the similar, heterogeneous,and ambiguous conditions when these conditions were con-sidered together in a separate analysis (F(2, 82) = .38,/? =.68). As Hic predicts, the only one of the three contrastcodes that was significant was the one comparing ambigu-ous MVP with dissimilar MVP: Participants liked Rootssignificantly less in the dissimilar (M - 3.81) than theambiguous (F(l, 107) = 5.04,p< .05) MVP condition.5

DiscussionThe result of Study la suggest that fans on a social net-working site do not need to directly interact with a targetconsumer or post comments about a brand to influence thebrand evaluations of a consumer new to the brand. Specifi-cally, MVP evokes equivalent levels of liking when it iscomposed of a homogeneous group of similar individuals,when it is composed of a heterogeneous group of dissimilarand similar individuals, and when brand supporters are leftdemographically ambiguous. In contrast, a homogenousgroup of dissimilar others produced significantly less brandliking. Thus, H] is supported in this context.

Study 1bIt is possible that part of the reason that ambiguity wastreated like similarity in Study la was because the numericrepresentation of fans in the ambiguous condition made itdifficult for consumers to consider the possibility that thesefans are different from themselves. It is also possible thatthe results we obtained could be unique to using age tomanipulate similarity. Therefore, in Study lb, we testedwhether our results hold using a different type of ambiguity(i.e., generic Facebook profile picture silhouettes) and whenmanipulating (dis)similarity using participants' gender

5We also analyzed these data using an alternate set of contrastcodes comparing (1) the similar, ambiguous, and heterogeneousMVP conditions with the dissimilar MVP condition; (2) the simi-lar and ambiguous MVP conditions with the heterogeneous MVPcondition; and (3) the ambiguous and similar MVP conditionswith each other. Consistent with our hypotheses, only the first ofthese alternate contrast codes had a signiflcant effect on brand lik-ing: Roots was liked significantly less in the dissimilar than in theother conditions (F(l, 107) = 5.04, p < .05); the other two contrastcodes had a nonsignificant effect on liking (both ps >.4O).

Beyond tiie "Like" Button/109

Page 6: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

(holding age constant). We also test our findings by alteringMVP through "fans of the day," an approach currently usedboth on Facebook and on the social portions of somebrands' own websites (e.g., the "Fan of the Month" featuredon Blackberry's blog).

Stimuli and ProcedureStudy lb once again asked participants about Roots brandclothing. All participants therefore first read the following:

As part of its social networking strategy, in addition to aFacebook page, a Twitter account, and a blog. Roots alsoinvites consumers to post photos of themselves to themain Roots website where they can be featured as "fans ofthe day."

Participants then saw a Screenshot of what was purportedlythe "Community" section of the Roots website, which fea-tured six fans of the day (see the Web Appendix at www.marketingpower.com/jm_webappendix). We held all ele-ments constant except the photos of the fans of the day.6 Thisstudy included three between-subjects conditions. Partici-pants in the ambiguous MVP condition saw six photos of theanonymous Facebook silhouettes shown when someone doesnot provide or make public a profile picture. Participants inthe similar MVP condition saw six photos of individualswho were the same age and gender that they were, whileparticipants in the dissimilar MVP condition saw six photosof individuals who were the same age but opposite gender.

After participants viewed the Roots website informa-tion, they indicated their level of agreement with the fol-lowing statement on a seven-point scale: "Roots is a brandfor me." Next, they answered the question "How likely areyou to join the Roots community so that you can have achance to be featured as one of the 'fans of the day'?" Tocapture inferred commonality to test our mediation hypoth-esis, we also asked all participants how much they agreedthat "I have a lot in common with the typical Roots shop-per." As a manipulation check, participants in the similarand dissimilar MVP conditions also indicated (yes/no)whether the fans of the day were the same gender theywere. Given that it could be argued that by manipulatingage in Study la, we also inadvertently manipulated attrac-tiveness, we also measured perceived attractiveness to ruleout this altemative explanation using the question, "Com-pared to the average person of their age and gender, howattractive were the Roots 'fans of the day'?" (1 = "signifi-cantly less attractive than average," and 7 - "significantlymore attractive than average"). Finally, all participants

stimuli for this study were adapted from the Roots web-site, which does feature a "Community" section (though this sec-tion does not actually include fans of the day). We note that thefans of the day are not actual Roots users, however, and that Rootsdoes currently sell clothing in the United States. We found thephotos of purported Roots fans by searching various websites onwhich consumers had posted public photos of themselves (e.g.,Facebook, HotorNot.com). Because all participants in the subjectpool at the university where the studies were conducted are in theirlate teens to early 20s and we held age constant in this study, allphotos used depicted people who would be perceived to be in thisage range.

answered the same questions about their familiarity withRoots used in Study la.

ResultsSample and manipulation check. Of the 116 undergrad-

uate students who participated in this study, 25 had heard ofRoots before the study. Of these participants, we removed 4from the data set because they reported being highly famil-iar with the brand. This left a final usable sample of 112participants (53 men, 59 women) with an average age of 21years. All participants in the similar MVP condition indi-cated that the fans of the day were the same gender theywere, and all participants in the dissimilar MVP conditionindicated that the fans of the day were the opposite gender.We also note that there were no differences in amount oftime spent viewing the page across conditions (M = 35.6seconds; F(2,107) =l.Ol,p= .37).

Brand liking and willingness to interact with the brandthrough social media. Given that the independent variable inthis study had three levels, we used two orthogonal contrastcodes (no special alpha levels required) to analyze the datacomparing (1) the ambiguous and similar MVP conditionswith each other (H|a) and (2) the ambiguous with the dis-similar MVP condition (H](,). We analyzed brand liking andwillingness to interact with the brand through social mediaseparately. Again, consistent with Hia, participants in theambiguous (M = 4.16) and similar (M = 3.91) MVP condi-tions expressed equivalent liking for Roots (F(l, 109) = .14,p = .71). Furthermore, consistent with Hi^, participants in thedissimilar MVP condition (M = 3.46) liked Roots marginallyless than participants in the ambiguous MVP condition (F(l,109) = 3.15,p = .06). The results for willingness to interactwith Roots through social media are similar. Participantsreported that they were equally likely to interact with thebrand in the ambiguous (M - 2.63) and similar (M - 2.57)MVP conditions (F(l, 108) = .90,/? = .35) and more likely tojoin the Roots community in the ambiguous than the dissim-ilar (M = 1.97) MVP condition (F(l, 108) = 4.50,p < .05).?

Mediation. Using the same contrast codes, we examinedwhether consumers' inferences that they had "a lot in com-mon with the typical Roots shopper" followed the same pat-tem, as H2 predicted. As we expected, participants reportedthat they had the same amount of commonality with thetypical Roots shopper in the ambiguous and similar MVPconditions (M^ biguous = 4.29, M,-,^,^^ = 3.85; F(l, 109) =

'We also conducted an alternate analysis of these data using twocontrast codes that compare (1) the similar and ambiguous MVPconditions with the dissimilar condition and (2) the similar andambiguous MVP conditions with each other. The results revealedthat the brand was liked marginally more in the similar andambiguous MVP conditions than in the dissimilar MVP condition(F(l, 109) = 3.75, p = .06) and was liked equally well in the simi-lar and ambiguous MVP conditions (F(l, 109) = 2.37, p = .13).Participants were also more likely to want to connect with thebrand through social media in the similar and ambiguous MVPconditions than in the dissimilar MVP condition (F(l, 109) = 4.50,p < .05) and were equally likely to want to connect with the brand inthe similar and ambiguous MVP conditions (F(l, 109) = 1.40,p =.24).

110 / Journal of Marketing, November 2012

Page 7: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

.05, p = .82) and more in common with the typical Rootsshopper in the ambiguous than in the dissimilar MVP con-dition (Mdi,5i i,ar = 3.29; F(l, 109) = 7.45,p < .01).

To explore whether inferred commonality mediated therelationship between the ambiguous versus dissimilar MVPcontrast code and liking for Roots clothing, we used a boot-strapping method (Preacher and Hayes 2008; Zhao, Lynch,and Chen 2010). A confidence interval (CI) that excludeszero for the indirect effect reveals that inferred commonal-ity with the typical Roots shopper mediates the relationshipbetween the ambiguous versus dissimilar MVP contrastcode and liking (95% CI [-.72, -.12]), consistent with Hj.

Given that we used photographs to manipulate MVP, itis possible that our results could be driven by the perceivedattractiveness of the specific fans shown, not perceptions ofhow similar the fans are to the participant (attractivenessand inferred commonality were moderately correlated; r -.38, p < .0001). To test whether this was the case, we exam-ined whether perceived attractiveness mediated the relation-ship between the focal contrast and liking. When attractive-ness is the only mediator in the model, it is a significantmediator (95% CI [-.38, -.004]). However, when we includedboth inferred commonality and perceived attractiveness inthe model as mediators, only inferred commonality mediatesthe relationship (95% CI [-.74, -.12]). The lower bound ofthe CI for perceived attractiveness is negative in this model,and the upper bound is positive (95% CI [-.05, .17]), indi-cating that attractiveness is not a significant mediator. Theresults for willingness to interact with the brand throughsocial media are substantively identical. Inferred common-ality mediates the relationship between the focal contrastand the dependent variable both when alone in the model(95% CI [-.37, -.05]) and when attractiveness is included inthe model (95% CI [-.36, -.04]). Attractiveness is a signifi-cant mediator when it is the only mediator in the model(95% CI [-.37, -.03]) but not when the model also includesinferred commonality as a mediator (95% CI [-.26, .05]).

DiscussionThe results of Study lb provide additional support for H],demonstrating that the results hold across a different opera-tionalization of ambiguity. In support of H2, consumersexpress greater liking for a brand and greater willingness tointeract with that brand through social media when thebrand displays ambiguous or similar MVP than when thebrand displays dissimilar MVP because of greater inferredcommonality with the brand's user base.^ Note that ambigu-

^Although not reported in the interests of brevity, we collectedadditional data in which we measured inferred commonality beforebrand evaluations. The results suggest that the effect of dissimilarMVP does not change regardless of whether it is made salientbefore brand evaluation questions are asked (F(l, 210) = 6.95,p <.01). As such, it appears that the effects of dissimilar and ambigu-ous MVP on brand evaluations are likely to be obtained eitherbelow or above the radar. However, when individuals are cued tonotice similar MVP, they appear to discount it when formingbrand evaluations (F(l, 210) = 4.50,p < .05). We would attributethis to the possibility that drawing attention to homogeneous simi-lar MVP activates persuasion knowledge, such that consumers tryto avoid being manipulated by the individuals presented.

ity in this format would be driven by consumers who havechosen not to upload pictures to their social media profile.Thus, this study suggests that if firms choose to select fansof the day, they can strategically choose to select or avoidindividuals who have opted to maintain their privacy.

Studies la and lb provide support for Hi using two dif-ferent manipulations of similarity (age and gender, respec-tively). These findings suggest that our results will beusable by a marketer who may only have access to eitherage or gender information based on consumers' past search-ing behavior or information provided by a social mediaplatform. We would anticipate that in some product cate-gories, the similarity-enhancing effect of gender and agecould be additive. That is, participants might infer greatercommonality if they saw the similar MVP of brand support-ers who were both the same age and same gender as them-selves than if they saw brand supporters who match only interms of age or gender. Whether age, gender, or bothtogether are most effective at raising inferred commonalityis likely dependent on the type of brands or products con-sumers consider. Clothing is a category in which both fac-tors are clearly important, as shown in these two studies.Further work could identify specific categories for whichone demographic factor or another is more central in deter-mining inferred commonality.

Study 2Although the numbers-only presentation used in Study laand the fans-of-the-day format used in Study lb are bothcommon ways to display MVP, firms also increasinglyallow consumers to upload pictures of themselves using orwearing a product to social media sites. If such pictures aredisplayed in ways that do not provide complete demo-graphic information, they would also present the consumerwith a type of ambiguous MVP. Therefore, Study 2explores whether photos that do not reveal all of a sup-porter's demographic characteristics create identical effectsto those created by revealing only the total number of fansor showing profile pictures that reveal no demographicinformation. Study 2 also explores the effect of heteroge-neous MVP in greater depth, testing the level of hetero-geneity required to create effects equivalent to those seenwith similar MVP. A secondary goal of this study was totest whether effects observed in prior studies extend to otherdownstream consequences of interest to managers beyondbrand liking and interacting with a brand through socialmedia.

Stimuli and ProcedureStudy 2 asked participants to react to an online clothingretailer called asos:

In this survey, we are interested in your opinions about areal brand's social networking presence. This brand, asos,is an online clothing retailer that sells both men's andwomen's clothing mostly in the U.K. As part of its socialnetworking strategy, in addition to a Facebook page and aTwitter account, asos also hosts "asos marketplace" on itscompany-owned website. Visitors to the website areinvited to join asos marketplace and to post photos ofthemselves wearing asos brand clothing.

Beyond the "Like" Button /111

Page 8: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

Participants then saw a screen shot of what was purportedlythe splash page Internet users would find if they went towww.asos.com. They then viewed the "Marketplace" sec-tion of the asos website, which featured user-posted photosof brand users wearing asos clothing (see the Web Appen-dix at www.marketingpower.com/jm_webappendix). Wemanipulated photos to create the different types of MVP,such that everything on the website was held constantexcept the user-posted photos . Participants indicated theirgender before the main study began so that all participantsviewed members of the asos marketplace who werematched to their own gender. We manipulated similarityusing perceived age, as in Study la.

Study 2 had eight between-subjects conditions in whichparticipants saw six photos of brand users. The number ofsimilar brand users ranged from zero of six (homogeneousdissimilar MVP) to six of six (homogeneous similar MVP)in seven of the conditions. The eighth condition displayedsix photos of brand users that showed only their clothing,not their faces (ambiguous MVP condition).

After participants viewed the asos website information,they responded to the following three questions on seven-point scales (1 = "very unlikely," and 7 = "very likely"):"How likely would you be to buy asos clothing if it wereavailable in the U.S.?" "Recently, asos has been consideringopening retail stores in addition to selling clothes online....How likely would you be to shop at an asos store if oneopened in your area?" and "If asos sent you a coupon to usein-store (for 20% off your total in-store purchase) to yourhome mailing address, how likely would you be to use thatcoupon?" We captured inferred commonality throughagreement (on a seven-point scale) with the statement, "Ihave a lot in common with the typical asos shopper."

As a manipulation check, participants in the similar anddissimilar MVP conditions rated how much they agreed that"The asos marketplace users are the same age I am" andindicated (yes/no) whether the users were the same genderthey were. Participants in all conditions responded to thequestion "Compared to other people in their age group, howattractive were the asos marketplace users?" on a seven-point scale anchored by "significantly less attractive thanaverage" and "significantly more attractive than average."

'We adapted the stimuli for this study from the asos website. Wenote that the users shown are not actual asos users, however, andthat asos does currently sell clothing in the United States. Weselected asos for this study because of its user-posted marketplacephoto section and because it was unfamiliar to the majority of thesubject pool at the university where the study was conducted. Wefound the photos of purported asos users through a Google imagesearch (thus, photos came from a variety of different websiteswhere people had posted public photos of themselves, includingBurberry's Art of the Trench website) and Facebook brand pageswhere users had posted photos of themselves wearing a particularclothing brand (e.g., Talbots), so the clothing used was not actu-ally asos clothing. As in Study la, because all participants in thesubject pool at the university where the studies were conductedwere in their late teens to early 20s, the photos used in the similarcondition depicted people who would also be perceived to be inthis age range. The purported asos users in the dissimilar age con-dition were older, with perceived ages ranging from their 40s totheir 70s.

Finally, all participants were asked whether they had heardof asos before the study and, if yes, how familiar they werewith the brand on a seven-point scale (1 = "not at all famil-iar," and 7 = "very familiar").

ResuitsSample and manipulation check. Of the 289 undergrad-

uate students who participated in this study for extra credit,17 had heard of asos before the study. All 17 indicated afamiliarity score of four or greater on the seven-point famil-iarity scale, and therefore we removed them from the dataset. Of the remaining participants, 12 indicated that the asosmarketplace users were not the same gender they were (pre-sumably because they did not follow instructions and indi-cated a different gender than their own before the beginningof the study) and were also removed from the data set. Thisleft a final usable sample of 260 participants (151 men, 109women) with an average age of 21 years. The age similaritymanipulation check revealed that participants in the similarMVP condition rated the asos marketplace members as moresimilar to themselves in age (Mg similar,0 dissimilar = 5.18) thandid participants in the heterogeneous (M5 in^w^, 1 dissimilar =4.07, M4 similar, 2 dissimilar = 3-74, M3 similar, 3 dissimilar = 3.75,^ 2 similar, 4 dissimilar = 2.74, Mj similar, 5 dissimilar = 2.38) anddissimilar MVP conditions (MQ similar, 6 dissimilar = 1 -88; F(6,233) = 23.71,/7<.0001).

Purchase intentions. Because the three purchase inten-tion measures were highly correlated (a = .90), we aver-aged them to form an overall purchase intention index. Ouranalysis uses seven contrast codes that compare purchaseintentions in the ambiguous MVP condition with everyother condition (with no need for special alpha levelsbecause the codes are orthogonal). Consistent with H ¡a andH]),, none of the contrasts comparing ambiguous MVP withthe homogeneous similar MVP condition or any of the het-erogeneous conditions were significant (all ps > .24). Theonly significant contrast (of the seven contrast codes com-paring the ambiguous MVP condition with every other con-dition) was the contrast code comparing the ambiguousMVP condition directly with the homogeneous dissimilarMVP condition. Consistent with H^^, participants in the dis-similar MVP condition were less likely to buy asos clothing(M - 3.60) than participants in the ambiguous MVP condi-tion (M = 4.60; F(l, 249) = 6.84,p < .Ol).io

To learn more about the effect of different levels ofheterogeneity, we conducted follow-up analyses comparingthe heterogeneous MVP arrays that contained the fewestnumber of dissimilar individuals with the homogeneous dis-similar MVP array. We found that one similar individual inthe MVP array was not enough to create purchase intentionssignificantly different from those generated by a homoge-neous dissimilar MVP array (F(l, 65) = 1.29, p - .26).However, when two of the six displayed individuals were

'"In a separate analysis (excluding the dissimilar MVP condi-tion), we also tested for differences across the homogeneous simi-lar and all heterogeneous MVP conditions. This omnibus analysisrevealed that there were no differences in purchase likelihoodacross these seven conditions (F(6, 228) = .ll,p = .60).

112 /Journal of Marketing, November 2012

Page 9: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

similar to the participant, purchase intentions were greaterthan for a dissimilar only MVP array (F(l, 64) = 3.75, /? =.06). Further analysis demonstrates the incremental impactof adding an additional similar individual to the mix usingregression. This analysis revealed a significant linear (b =.13, t = 2.15,p < .01), not curvilinear (b - -.03, t = -I.18,p =.24), relationship between the degree of similarity in theMVP array and purchase intentions (see Table 1).

Mediation. Using the same seven contrast codes com-paring each condition with the ambiguous MVP condition,we next examined whether consumers' inferences that theyhad "a lot in common with the typical asos shopper" fol-lowed the same pattem of results found for purchase inten-tions. As we expected, the only signiflcant difference wasagain that participants in the dissimilar MVP condition (i.e.,six dissimilar users and zero similar users) perceived lowercommonality with the typical asos shopper (M - 3.03) thanparticipants in the ambiguous MVP condition (M = 3.50;F(l, 252) = 3.85,i» = .05) (for means, see Table 1).

As in Study lb, we tested both inferred commonalityand perceived attractiveness (r = .34,p < .0001) of the asosmarketplace users as potential mediators of purchase inten-tions using a bootstrapping method (Preacher and Hayes2008; Zhao, Lynch, and Chen 2010). The results reveal thatinferred commonality with the typical asos shopper mediatesthe relationship between the contrast code comparing thedissimilar MVP condition with the ambiguous MVP condition(95% CI [.09, .72]). In contrast, when we tested attractive-ness as a mediator, the lower bound of the CI is negative, andthe upper bound is positive (95% CI [-.07, .34]), indicatingthat perceived attractiveness is not a signiflcant mediator.When we included both potential mediators in the model,inferred commonality remains a signiflcant mediator (95%CI [.10, .69]), and perceived attractiveness does not (95%CI [-.03, .15]).

DiscussionStudy 2 demonstrates that the brand liking effects in Studiesla and lb extend to purchase intentions and are consistentlymediated by inferred commonality with the brand's users.These results hold when manipulating ambiguity by showingphotos that concealed key demographic characteristics. Wenote that this is a more conservative test of our hypothesesthan that in Study la or lb, given that the photos used in Study2 conceal some, but not all, demographic characteristics.

Study 2's results also indicate that consumers mayrespond to heterogeneous MVP in the same way they do toambiguous and homogeneous similar MVP. Exposure totwo or more similar individuals in a set leads to purchaseintention levels not significantly different from ambiguousor homogenous similar MVP. We suspect, however, that theraw number of similar individuals in a heterogeneous MVParray is likely not as important as the proportion of similarindividuals. Would two similar individuals in a heteroge-neous MVP array of 100 individuals produce equivalentlevels of brand liking as an ambiguous or homogeneoussimilar MVP array? We leave identification of the absolutetipping point to further study but expect that two wouldlikely not be enough in this context.

Across the three studies reported thus far, we show thatleaving a brand's online supporters ambiguous has onlypositive consequences. However, Studies 1 and 2's effectswere viewed in the context of only one brand's presence—that is, in a separate evaluation context. In many cases, con-sumers will not evaluate a brand in isolation. Thus, the pre-scription to maintain ambiguity may need to be temperedfor firms that face a more rather than less competitivespace.

Therefore, Study 3 examines whether our effects hold ina joint evaluation context more similar to the experience aconsumer is likely to actually have on a social networkingsite. In previous research on interpretation of ambiguousothers. Nay lor, Lamberton, and Norton (2011) showed par-ticipants a set of different, verbally described reviewers(varying in identification and similarity) providing inputabout different products. In this setup, they found that anambiguous reviewer was slightly less persuasive than asimilar reviewer. This result diverged somewhat from flnd-ings in their other studies, in which participants viewedonly one reviewer, leaving an explanation for this "cost ofambiguity" for further research. We propose that whetherambiguous MVP creates liking equivalent to or less thanthat created by similar MVP will depend on whether abrand is evaluated alone (i.e., no competing brands or theirsupporters are viewed) or evaluated at the same time asother competing brands. From a practical standpoint, con-sumers may view only one brand's Facebook page (or thesocial component of only one brand's website) in a categorywith little direct competition or may view many brands'pages in a densely populated space.

TABLE 1Study 2: Means by MVP Composition

Condition MVP CompositionMean Purchase Intention

(Indexed Variable)

Mean Rating of How Muchin Common Participant Haswith Typical asos Shopper

12345678

6 ambiguous users6 simiiar, 0 dissimilar users5 simiiar, 1 dissimilar user4 simiiar, 2 dissimilar users3 simiiar, 3 similar users2 similar, 4 dissimilar users1 similar, 5 dissimilar users0 similar, 6 dissimilar users

4.614.384.484.254.214.233.963.54

3.633.943.753.493.353.453.213.00

Beyond the "Like" Button/113

Page 10: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

Viewing multiple competing brands at the same timecan be characterized as a case of joint evaluation, whereassingle brand viewing can be characterized as separateevaluation (Hsee et al. 1999; Hsee and Leclerc 1998). Pre-vious research has shown that people evaluate options dif-ferently in these two decision contexts, such that productsthat are evaluated highly under separate evaluation may beevaluated less positively under joint evaluation (Hsee andLeclerc 1998). This is because when an objectively attrac-tive product is evaluated singly, consumers rely on an inter-nal reference point to evaluate it. In contrast, when that tar-get product is evaluated alongside another objectivelyattractive product, consumers shift away from their intemalreference point. Instead, consumers focus more on the otherproduct, rather than an intemal source, as a reference point(Hsee and Leclerc 1998).

In the present context, consider separate evaluation of abrand with ambiguous MVP. We have argued that this brandwill be evaluated positively because consumers infer thatambiguous others are like themselves. Now imagine thatthe same brand is viewed alongside a brand with similarMVP. Rather than focusing exclusively on egocentricanchor-driven inferences of similarity, the consumer com-pares the ambiguous MVP directly with the similar MVP ofthe other brand. When ambiguity is directly compared withsimilarity, the consumer will still infer commonality with theambiguous MVP brand's user base, but this inferred infor-mation is likely to be weaker than the information observedwhen demographic similarity is displayed. Thus, in jointevaluation contexts, we predict that consumers will havehigher evaluations of a brand about which commonality ismore strongly indicated by the similar MVP displayed thana brand about which they have to make an inference (i.e., abrand whose supporters constitute ambiguous MVP), evenif the egocentric-anchoring-driven inference would have,under separate evaluation, led to equivalent levels of likingfor the brands with ambiguous and similar MVP. Formally,

H3: In a joint evaluation context, ambiguous MVP createsbrand evaluations that are less positive than those gener-ated by homogeneous, similar MVP but more positivethan those generated by homogeneous, dissimilar MVP.

Study 3Study 3 tests our hypothesis that the effect of ambiguitydepends on the evaluation context in which a brand isviewed. In addition. Study 3 shows that the results fromStudies 1 and 2 replicate in a different product category andin an additional social networking context.

Stimuli and ProcedureA total of 312 undergraduate students (178 men, 134women) with a mean age of 21 years participated in thisstudy for course credit. All participants read the followingintroduction:

Now we'd like to you to look at some social networkingwebsites developed by restaurants with locations nation-wide. To protect confldentiality, the names of the restau-rants have been changed to Restaurants X, Y, and Z [in thejoint evaluation condition; only Restaurant X was men-

tioned in the separate evaluation condition]. You will seesome of the material that appears on the first page of eachrestaurant's social networking site. After potential userssee this information, they can create an account to join arestaurant's site, which lets them post information thatother users can see, including comments and photographs.

Participants in the joint evaluation condition were told thatthe first pages of all three social networking sites that theywould observe featured a "welcome" to the site, pho-tographs of the restaurant, and photographs of members ofthe restaurant's site. Participants in this condition were thenshown the first page of the three restaurants' sites with thepictures of five website members varying such that onerestaurant featured members similar to the subject pool,"one featured dissimilar members, and one featured ambigu-ous members (using the same anonymous Facebook silhou-ettes used in Study lb). We rotated the order of type ofmember featured for each restaurant such that there wereactually three joint evaluation conditions that differed onlyby order. (In all conditions, participants saw three restau-rants, one with similar, one with ambiguous, and one withdissimilar members.) Participants in the joint evaluationconditions rated the restaurants sequentially.

In contrast, participants in the separate evaluation con-dition saw information about only one restaurant, whichwas manipulated between subjects to display similar, dis-similar, or ambiguous MVP. Participants in all conditionswere asked, "How much do you think you'd like the bararea at [restaurant name]?" (for each restaurant they saw)on a nine-point scale (1 = "would not like at all," and 9 ="would like very much").

ResultsSeparate evaluation. We first assessed whether our pre-

dictions held in the separate evaluation conditions usingtwo contrast codes that compare (1) the ambiguous andsimilar MVP conditions with each other and (2) theambiguous with the dissimilar MVP condition. When weregressed the liking variable on the two contrast codes, weagain observed support for Hja, as the bar area was likedequally well in the similar (M = 5.78) and ambiguous MVPconditions (M = 5.92; F(l, 144) = .55, p- .46). That is, asin Studies 1 and 2, there is no cost to ambiguity relative tosimilarity in a separate evaluation context. Consistent withHic and the results of the prior studies, participants liked thebar area more in the ambiguous MVP condition than in thedissimilar (M = 5.18) MVP condition (F(l, 144) = 4.11,p <.05). 12

"Member pictures used were the same used in Study la. Par-ticipants were again asked their gender before the study began sothat the fans shown matched their gender. We manipulated similar-ity using perceived age.

'2We also conducted an alternate analysis using two contrastcodes that compare (1) the similar and ambiguous MVP conditionswith the dissimilar condition and (2) the similar and ambiguousMVP conditions with each other. The results revealed that partici-pants liked the bar area more in the similar and ambiguous MVPconditions than in the dissimilar MVP condition (F(l, 109) = 5.24,p < .05) and equally well in the similar and ambiguous MVP con-ditions (F(l, 109) = .06,p = .81).

114 / Journai of Marketing, November 2012

Page 11: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

Joint evaluation. An omnibus repeated measures analy-sis of all three conditions revealed that there was not a sig-nificant between-subjects effect of order (in which therestaurants were viewed) on brand liking, so we droppedorder from further analysis (F(l, 163) = .65,p = .42).i3 Wethen used within-subject contrast codes analogous to theones used in the separate evaluation analysis to analyze par-ticipants' liking for the three restaurants. Consistent with H3and diverging from Hig, participants liked the bar area in therestaurant with similar MVP (M - 5.97) significantly morethan the bar area in the restaurant with ambiguous MVP (M =5.34; F(l, 164) = 13.22,/? < .001), showing a cost of ambi-guity relative to known similarity. However, even underjoint evaluation, ambiguity is stiU preferable to dissimilarity:Participants anticipated liking the bar area in the restaurantwith ambiguous MVP more than the bar area in the restau-rant with dissimilar MVP (M = 4.39; F(l, 164) = 30.01,p <.0001).i4

DiscussionStudy 3 examines participants' response to similar versusambiguous MVP in joint versus separate evaluation con-texts. In separate evaluation, ambiguous MVP leads to analmost identical response to that generated by similar MVP,as in Studies 1 and 2. However, in joint evaluation, ambigu-ous MVP leads to a significantly less positive response thandoes similar MVP. Still, ambiguous MVP generates likinggreater than that evoked by dissimilar MVP.

These findings shed light on the decision managersmust make about whether to reveal the identity of theirbrand's supporters: The decision must be determined notjust by whether the brand supporters shown are likely to beperceived as similar or dissimilar to a target consumer butalso by whether the consumer is likely to encounter the sup-porters in a joint or separate evaluation context, an issue weretum to in the "General Discussion" section.

General DiscussionAt the end of 2011, iMedia Connection published an articletitled "Why Facebook Fans Are Useless" (Lake 2011). In thearticle, the author notes that "on their own, Facebook 'likes'don't add any value." Yet research has shown that socialmedia can translate into increases in sales (Stephen and

i3We also analyzed the joint evaluation data using only eachparticipant's rating of the first restaurant they saw. Because par-ticipants viewed the restaurants sequentially, we expected theseresults to be consistent with the separate evaluation results. As weexpected, participants liked the bar area equally well in the similar(M = 5.57) and ambiguous MVP conditions (M = 5.88; F(l, 162) =.90,p = .34) but more in the ambiguous MVP condition than in thedissimilar (M = 4.66) MVP condition (F(l, 162)= 10.69, p< .01).

'''The results using the same alternate contrast codes used in theseparate evaluation analysis reveal that participants anticipatedliking the bar area in the restaurants with similar and ambiguousMVP more so than the bar area in the restaurant with dissimilarMVP (F(l, 164) = 63.47, p < .0001). However, in contrast to theseparate evaluation results, they liked the bar area in the restaurantwith similar MVP significantly more so than they did the bar areain the restaurant with ambiguous MVP (F(l, 164) = 13.22, p <.001).

Galak 2012). The present work shows how mini-connectionswith consumers created through social networking canindeed yield positive effects on brand evaluations and pur-chase intentions. A central proposition of our research isthat the decision to reveal a brand's fan base or to leavesupporters' identities ambiguous is important because thedemographic composition of the MVP presented affectsconsumers' reactions to the brand.

Importantly, the effects of revealing the identities of abrands' fan base vary depending on the demographic com-position of the individuals presented. In Study la, wedemonstrate that consumers respond as positively to abrand when the brand's supporters remain ambiguous(because no photos of supporters are displayed) as they dowhen the brand reveals the identity of supporters that theconsumer perceives to be similar to the self. Consumersalso respond as positively to the display of a heterogeneousgroup of similar and dissimilar brand supporters as they doto an ambiguously presented group. Importantly, ambigu-ous MVP produces significantly greater brand liking thanhomogeneous dissimilar MVP. Study lb replicates theseeffects using a different type of ambiguity (i.e., the genericFacebook silhouette profile picture) and a different manipu-lation of similarity (gender instead of age).

Perhaps surprisingly for many who assume that trans-parency is key in developing a brand's social networkingpresence, the results of both Studies la and lb suggest thatrevealing the identities of a brand's online supporters mayactually have negative consequences if the brand's support-ers are homogeneous and dissimilar to the target consumer.This may be the case when a brand initially extends into anew target market. In these cases, leaving a brand's fan baseambiguous may be a safer strategy because consumers willlike the brand as much when supporters are ambiguous aswhen at least some similar supporters are revealed.Notwithstanding expectations about the "social" nature ofsuch platforms. Studies la and lb show no negative conse-quences of choosing not to reveal the identity of a brand'ssupporters.

Study 2 uses a third type of ambiguity (photos of brandsupporters with key demographic characteristics obscured)and further explicates the impact of heterogeneous MVP,showing that a small proportion of similar supporters cancreate effects like that of ambiguous MVP. This findingsuggests that brands need not fear diversity on social net-working sites as long as they can anticipate that a targetaudience will make up a nontrivial proportion of the groupof supporters shown. Both Studies lb and 2 also documentthat the effects of MVP on brand liking are driven by con-sumers' inferences about how much they have in commonwith the brand's supporters.

While Studies 1 and 2 suggest that ambiguity is the pre-ferred strategy in separate evaluation contexts. Study 3shows that in joint evaluation contexts, ambiguity is not aspowerful as similarity in generating brand liking. Thesefindings suggest that if the consumer is likely to evaluate abrand in isolation, ambiguity may be the safest strategy. Incontrast, if the brand is likely to be encountered in a contextin which it is being compared with multiple other brands,managers may need to display information about online

Beyond the "Like" Button /115

Page 12: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

supporters, despite the potential risks, to compete withbrands with supporters whose demographic characteristicsare displayed.

Theoretical ContributionsThis work provides several novel theoretical insights. First,the concept of MVP offers a new framework for under-standing social influence. Although spatial proximity isabsent, exposure is only passive, typically just a handful ofindividuals are shown, and no future relationship is likely toexist among the consumers, we show that MVP still hassubstantial effects on consumers' brand evaluations andpurchase intentions. As such, the concept of MVP high-lights the ways that online social influence may have aneffect despite its difference from the offline presence of oth-ers and despite its failure to conform to the parameters ofSIT (Latane 1981).

Importantly, in contrast to traditional advertising orspokesperson contexts, we also note that MVP created bysocial media exposure is provided by individuals who vol-untarily affiliate with a brand, making it less likely thattheir action will be discounted by consumers due to reac-tance against marketer-driven recommendations (Fitzsi-mons and Lehmann 2004) or persuasion knowledge (Fries-tad and Wright 1994). Further research could explorewhether the effects of MVP hold if consumers do not trustthat the brand supporters presented are truly other con-sumers (as opposed to, e.g., employees of the brand "pos-ing" as supporters).

Furthermore, our investigation into consumers' responsesto ambiguous others may prompt deeper explorations ofinterpretations of interpersonal ambiguity. In the presentresearch, we show the equivalence of three types of ambi-guity: that created when (1) MVP is represented onlynumerically, (2) MVP is displayed using silhouette picturesthat suggest real individuals but obscure all their demo-graphic characteristics, and (3) MVP is displayed using pic-tures of real supporters that obscure some of their demo-graphic characteristics. Although these operationalizationsof ambiguity appear to have similar effects, further researchmay find additional nuances in the concept of ambiguityand may actually find that different types of ambiguity havevariant impacts on consumers.

Finally, by directly comparing consumer response toambiguous others in separate and joint evaluation, weexplain Nay lor, Lamberton, and Norton's (2011) findings.We suggest that the difference in ambiguity's effects acrossevaluation modes stems from the finding that the intemallyderived egocentric anchor is the determinant of similaritywhen ambiguity is encountered in a separate evaluationmode but that the importance of this intemal anchor isdiminished in joint evaluation. Therefore, in joint evalua-tion, a brand with ambiguous MVP will be less attractivewhen compared with a brand with similar MVP. Thus, inStudy 3, consumers evaluated a brand with identified similarsupporters more positively than a brand whose supporterswere not identified. This finding is consistent with work byForman, Ghose, and Wiesenfeld (2008), who find that con-sumers rate product reviews containing identity-descriptive

information more positively than those that lack identifyinginformation. In the same way, our work suggests that someidentification information may be better than none, particu-larly in joint evaluation contexts.

How Can This Research inform Practice?To determine how our findings can be used, we should firstnote the breadth of applicability for our studies. Note thatwe have removed consumers who are extremely familiarwith a brand from our analysis and that we focus on brandperceptions and purchase intentions as our outcomes ofinterest. This makes our findings most applicable to con-sumers who are relatively new to a brand and who, at leastpassively, have the goal of forming an opinion about thebrands to which they are exposed. This is a substantial seg-ment: Approximately 23.1 million consumers between 13and 80 years of age use social media to discover new brandsor products, and 22.5 million people use social media toleam about unfamiliar brands or products (Knowledge Net-works 2011). Thus, our results will be relevant to marketersseeking to reach this large segment of consumers but maynot be applicable for marketers of universally knownbrands.

Our research will also be easiest to use when marketerscan manipulate their displayed MVP in response to targetconsumer demographics. Emerging tracking and targetingtools can be used to do this. For example, Facebook ads areoften targeted only to certain demographic groups. In suchcases, marketers know that individuals who click on a linkto their social media sites will fit a certain demographicprofile and can adjust MVP accordingly. In other cases,consumers who remain logged in to social media sites whilebrowsing other Intemet sites may inadvertently provideaccess to age or gender information to the other sites theyvisit. Altemately, forms that have been filled out in oneonline location can provide information to other sitesthrough stored cookies. Using this information, companiescan tailor the MVP that a given consumer encounters whenhe or she visits a brand's social media page.

However, we recognize that this may not always be pos-sible. To help brand managers manage MVP both whenthey have granular demographic information about the spe-cific consumers visiting their site and when they do not, wedeveloped a decision framework based on two key factors:(1) the demographic composition of existing brand support-ers relative to targeted new supporters and (2) whether thebrand is likely to be evaluated singly or in combination withcompeting brands. This framework (presented in Figure 1)shows when brands should reveal the identity of theironline supporters and when ambiguity is preferable, and ithighlights cases in which managerial control over MVPcomposition is particularly crucial.

When brands should reveal the identity of social mediasupporters. If a brand has a social media support base thatis either homogeneous and similar to target consumers or aheterogeneous base that includes at least some supporterssimilar to a target consumer, our research suggests thatrevealing the identity of a brand's supporters is unlikely tohave adverse consequences for brand evaluations or pur-

116 / Journal of Marketing, November 2012

Page 13: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

FIGURE 1Framework for Decisions About Revealing the Identity of a Brand's Fan Base and Selection of a Social

Media Platform

Fan base ishomogeneous and

similar to targetaudience

Fan base isheterogeneous butincludes fans similarto target audience

Reveal Identity of fan base: May use anysocial media platform because no control

is needed over specific fans shov»/n.

Reveal Identity of fan base: Use onlysocial media platforms that a\\o\N control

over specific fans shown.

Fan base ishomogeneous and

different fromtarget audience

Fan base isheterogeneous and

includes no fanssimilar to target

audience

Maintain ambiguity of fan base: May useany social media platform because no

control is needed over specific fans shown.

chase intentions. This is true in both more and less competi-tive product categories. If the brand is being evaluated inisolation, revealing a fan base similar to a target consumer(or one that contains at least some similar supporters) islikely to prompt relatively positive evaluations. In jointevaluation contexts, revealing the brand's similar supporterbase may give the brand an edge over brands that provideno supporter information on their social media sites.

When a brand's supporters are better left ambiguous. Ifa brand's current supporters are likely to be perceived asdissimilar by new consumers, our results suggest thatrevealing the identity of a brand's existing supporters willundermine brand liking in both separate and joint evalua-tions. Thus, revealing supporters that are dissimilar to thetarget consumer is an inferior alternative to leaving support-ers ambiguous. One situation in which a brand's currentsupporters are likely to be dissimilar to targeted consumersis when a brand extends into new demographic segments. Inthese cases, it will be important to recruit new supportersfirst (who are perceived to be similar to the new target mar-ket). Managers should then ensure that these new support-ers are displayed either as a homogenous group when thenew target market visits the brand's social networking siteor mixed in with the old supporters to create heterogeneousMVP in which at least some supporters are shown that thenew target market perceives as similar.

When brand managers need control over the speciftcbrand supporters shown. Our research suggests that brandmanagers may want to control who is displayed on theirsocial media brand pages, particularly if their supporterbase is heterogeneous. If the supporters to be shown arerandomly selected each time a consumer visits the page, itis possible that supporters that the target consumer per-ceives as dissimilar will be shown. Giving the brandmanager the ability to tailor MVP makes revealing the iden-tity of a brand's supporters much less risky. Brand man-agers may choose to use existing social media sites such asFacebook more strategically (e.g., through a fans of the dayfeature such as that used in Study lb) or may create socialportions of their own website where they have the ability tocontrol exactly which consumers are featured in MVParrays. Fortunately for brand managers, the finding thatconsumers respond as positively to heterogeneous MVP asthey do to similar MVP suggests that the same heteroge-neous MVP array can enhance brand liking for multiple dif-ferent target segments.

Our research also suggests that brands may need to becautious in allowing consumers to post photos of them-selves using the brand. An examination of sites that allowusers to post these types of photos suggests that many brandmanagers do not realize the effect that failure to manage theMVP these photos create may have on brand evaluations or

Beyond the "Like" Button /117

Page 14: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

that this management can be done in ways that further theirstrategic objectives. Many brands appear to focus on inter-activity and inclusiveness, allowing any user-posted phototo stay up on their site (unless it is offensive), with nothought about how it may be influencing consumers new tothe brand. We hope that one of the outcomes of our researchis the importance of oversight of this type of content. Brandmanagers should continually monitor these types of sites,actively deleting photos or other content that does not cre-ate the desired form of MVP. If brands allow consumers topost such photos with no oversight, they may risk the crea-tion of MVP that is dissimilar from their target segment andsee brand evaluations suffer among new consumers as aresult.

The ethics of managing mere virtual presence. Compa-nies must also carefully consider the ethical implications ofcreating customized MVP arrays, as well as the potentialfor customer backlash, particularly if different arrays willbe shown to different consumers. A firm might be perceivedas willfully misleading its consumers if it presents a strate-gically selected MVP array as being exhaustive of a brand'ssupporters. However, as long as consumers are aware thatonly a segment of the brand's supporters are shown (whomay not be representative), they may interpret the displayof individuals like themselves as providing customized,diagnostic information, rather than as a dishonest tactic.Fans-of-the-day MVP presentations (a tactic successfullyused by Oreo on Facebook; Keath 2012) have the advan-tage of not purporting to be representative or random,which may decrease the potential for backlash.

It is also important to note that consumers do not seemto be aware of the influence that MVP has on their brandevaluations. In an online survey we conducted (for details,see footnote 1), we asked participants how important theidentity of other people who have already "liked" a brandon Facebook was in determining how much they them-selves like the brand, which they answered on a seven-pointscale (1 - "not at all important," and 7 = "extremely impor-tant"). Notably, respondents seemed basically unaware ofthe effect that the identity of other social media users has onthem, reporting a mean value of 3.13, significantly belowthe scale midpoint. Thus, studying MVP is important notonly because of the potentially powerful effect it can haveon brand liking but also because of consumers' apparentlack of awareness of its effects. Further research shouldexplore the possible interpretations of actively managedMVP to avoid the possibility that consumers will be misled.

Limitations and Avenues for Further ResearchThere are several factors that may moderate the effects ofMVP and that warrant additional research. First, socialmedia can be used to affiliate with brands with which a con-sumer already has a relationship. When consumers haveexperience with a brand, our effects may change. For exam-ple, existing brand users who encounter dissimilar MVPmight experience alienation or dissatisfaction because thisdemographic composition violates their expectations. Alter-nately, existing users may discount MVP information alto-gether, lacking the need to make any inferences about a

product's appropriateness for them. For these consumers,other major factors in a social media site (e.g., the preva-lence of coupons, prizes, discounts) may dilute any effectsof MVP. Further work could combine both MVP effects andother variations across social media platforms to determinewhether they lead to additive or interactive effects.

Next, our research suggests that consumers view dis-similar MVP as largely dissociative (White and Dahl 2006,2007). In the context of clothing and restaurant choices, aswell as other socially embedded consumption contexts, weanticipate that this is the case. In addition, for the youngconsumers who participated in our studies, homophily isstrong. However, it is also possible that for high-identity-confiict or low-self-esteem consumers or in certain productcategories (e.g., luxury goods), a demographically dissimi-lar group may be aspirational rather than dissociative. Forexample, Lambert-Pandraud and Laurent's (2010) study ofthe French perfume industry notes that although olderwomen are the target market for some fragrances, theadvertising for almost all brands features young actressesand models. It may be that, in such cases, dissimilar MVPwould have a positive effect on brand evaluations and pur-chase intentions.

Further research could also consider other operational-izations of similarity. We anticipate that gender and ageinformation are most likely to be available to marketersand, therefore, that understanding the effects of these typesof similarity or dissimilarity is of most utility. However, ifmarketers anticipate access to information about con-sumers' ethnicity, for example, further research could test todetermine whether this operationalization of similarity cre-ates parallel effects. Furthermore, the decision to alter thenumber of individuals displayed may also alter the effect ofMVP composition. For example, if there are only one ortwo brand supporters shown, consumers may recognize thatthe MVP composition is not diagnostic. As the number ofsupporters displayed increases, their composition may havegreater impact.

We also note that the inferences consumers make on thebasis of MVP may be false because demographic similarityoften diverges from psychographic or value-based similar-ity (Wells 1975). Thus, inferences associated with the MVPof similar others may inappropriately inflate expectations ofa "perfect fit" to a given consumer's preference, leading toproduct dissatisfaction (Diehl and Poynor 2010). Therefore,it would be worthwhile for further research to analyze theaccuracy of inferences stemming from MVP and how thisaccuracy affects brand evaluations, satisfaction, word ofmouth, and either present or future purchase intentions.

Finally, in our studies, we instructed participants to sim-ply look at a social networking site to form a perception ofa new brand. This might be considered a goal-directedbehavior, given that is extrinsically motivated, possesses aninstrumental orientation, and is part of a prepurchase searchprocess (Novak, Hoffman, and Duhachek 2003). However,consumers may approach social media sites with numerousgoals, some of which are more focused on the productionthan the consumption of content (see Hoffman and Novak2012), while some may be more experiential (Novak, Hoff-man, and Duhachek 2003). For example, consumers driven

118/Journalof Marketing, November 2012

Page 15: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

by intrinsic motivation or in pursuit of an experiential goalmay primarily base their affinity for a brand on its socialnetworking site's ability to help them create the desiredlevel of hedonic benefit. Here, SIT's prediction of null orweak effects may be correct. In contrast, if a consumer'sgoal is to affiliate with others, the effect of MVP may be

magnified. An exploration of consumers' goals when inter-acting with a brand through social media, the effect ofMVP, and the interaction between the two on the influenceof downstream variables (e.g., consumer desire to stay affil-iated with a brand over time) is an important avenue for fur-ther research.

REFERENCESAaker, Jennifer, Anne M. Brumbaugh, and Sonya Grier (2000),

"Non-Target Markets and Viewer Distinetiveness: The Impactof Target Marketing on Advertising Attitudes," Journal of Con-sumer Psychology, 9 (3), 127-40.

Algesheimer, René, Utpal M. Dholakia, and Andreas Hermann(2005), "The Social Influence of Brand Community: Evidencefrom European Car Clubs," Journal of Marketing, 69 (July),19-34.

Argo, Jennifer J., Darren W. Dahl, and Rajesh V. Manchanda(2005), "The Influence of a Mere Social Presence in a RetailContext," Journal of Consumer Research, 32 (September),207-212.

Asch, Solomon E. (1955), "Opinions and Social Pressure," Scien-tific American, 193 (5), 31-35.

(1956), "Studies of Independence and Conformity: AMinority of One Against a Unanimous Majority," Psychologi-cal Monographs, 70 (416).

Baird, Carolyn Heller and Gautam Parasnis (2011), "From SocialMedia to Social Customer Relationship Management," Strat-egy & Leadership, 39 (5), 30-37.

Bearden, William O. and Michael J. Etzel (1982), "ReferenceGroup Influence on Product and Brand Purchase Decisions,"Journal of Consumer Research, 9 (September), 183-94.

, Richard G. Netemeyer, and Jesse E. Teel (1989), "Mea-surement of Consumer Susceptibility to Interpersonal Influ-ence," Jourrml of Consumer Research, 15 (March), 473-81.

Berger, Jonah and Chip Heath (2007), "Where Consumers Divergefrom Others: Identity Signaling and Product Domains," Jour-nal of Consumer Research, 34 (August), 121-34.

and (2008), "Who Drives Divergence? Identity Sig-naling, Outgroup Dissimilarity, and the Abandonment of Cul-tural Tastes," Journal of Personality and Social Psychology, 95(September), 593-607.

Bumkrant, Robert E. and Alain Cousineau (1975), "Informationaland Normative Social Influence in Buyer Behavior," Journal ofConsumer Research, 2 (December), 206-215.

Byme, Donn Erwin (1971), The Attraction Paradigm. New York:Académie Press.

Chatman, Jennifer A. and Francis Flynn (2001), "The Influence ofDemographic Heterogeneity on the Emergence and Conse-quences of Cooperative Norms in Work Terms," Academy ofManagement Journal, 44 (5), 956-74.

Childers, Terry L. and Akshay R. Rao (1992), "The Influence ofFamilial and Peer-Based Reference Groups on Consumer Deci-sions," Journal of Consumer Research, 19 (September),198-211.

Cunningham, George B. (2007), "Perceptions as Reality: TheInfluence of Actual and Perceived Demographic Similarity,"Journal of Business and Psychology, 22 (1), 79-89.

Deshpandé, Rohit and Douglas M. Stayman (1994), "A Tale ofTwo Cities: Distinetiveness Theory and Advertising Effective-ness," Journal of Marketing Research, 31 (February), 57-64.

Dholakia, Utpaul M. and Silvia Vianello (2009), "The Fans KnowBest," Sloan Management Review/ Wall Street Journal Busi-ness Insights, (August 17), (accessed July 11, 2012), [availableat http://online.wsj.com/artiele/SB10001424052970204482304574222062946162306 .html].

Diehl, Kristin and Cait Poynor (2010), "Great Expectations?!Assortment Size, Expectations and Satisfaction," Journal ofMarketing Research, 47 (April), 312-22.

Escalas, Jennifer Edson and James R. Bettman (2003), "You AreWhat They Eat: The Influence of Reference Groups on Con-sumers' Connections to Brands," Journal of Consumer Psy-chology, \3 (3), 339-48.

Fitzsimons, Gavan J. and Donald R. Lehmann (2004), "Reactanceto Recommendations: When Unsolicited Advice Yields Con-trary Responses," Marketing Science, 23 (1), 82-94.

Forman, Chris, Anindya Ghose, and Batia Wiesenfeld (2008),"Examining the Relationship Between Reviews and Sales: TheRole of Reviewer Identity Disclosure in Electronic Markets,"Information Systems Research, 19 (September), 291-313.

Friestad, Marian and Peter Wright (1994), "The PersuasionKnowledge Model: How People Cope with PersuasionAttempts," Journal of Consumer Research, 21 (June), 1-31.

Hameed, Bilal (2011), "Social Media Usage Exploding AmongstFortune 500 Companies," Social Times (accessed January 17,2012), [available at http://socialtimes.eom/social-media-usage-exploding-amongst-fortune-500-companies_b35372].

Hoffman, Donna L. and Thomas P. Novak (1996), "Marketing inHypermedia Computer-Mediated Environments: ConceptualFoundations," Journal of Marketing, 60 (July), 50-68.

and (2012), "Why People Use Social Media: HowOnline Social Identity and Motivations Influence the Experi-ence of Being Connected," working paper, Sloan Center forIntemet Retailing, University of California, Riverside.

Hsee, Christopher K., Sally Blount, George F. Loewenstein, andMax H. Bazerman (1999), "Preference Reversals BetweenJoint and Separate Evaluations of Options: A Review andTheoretical Analysis," Psychological Bulletin, 125 (5), 576-90.

and Frederic Leclere (1998), "Will Products Look MoreAttractive When Presented Separately or Together?" Journal ofConsumer Research, 25 (September), 175-86.

Jehn, Karen A., Gregory B. Northcraft, and Margaret A. Neale(1999), "Why Differences Make a Difference: A Field Study ofDiversity, Conflict, and Performance in Workgroups," Admin-istrative Science Quarterly, 44 (December), 741-63.

Jiang, Lan, JoAndrea Hoegg, Darren W. Dahl, and Amitava Chat-topadhyay (2010), "The Persuasive Role of Incidental Similar-ity on Attitudes and Purchase Intentions in a Sales Context,"Journal of Consumer Research, 36 (February), 778-91.

Keath, Jason (2012), "Oreo Features Facebook Fans Daily," SocialFresh, (January 5), (accessed March 2, 2012), [available athttp://soeialfresh.com/oreo-facebook-fan-of-the-day/].

Knowledge Networks (2011), "Social Media Now InfluencesBrand Perceptions, Purchase Decisions of 38 Million inU.S." (accessed January 17, 2012), [available at http://www.knowledgenetworks .com/news/releases/2011/06141 l_soeial-media.html].

Lake, Amielle (2011), "Why Facebook Fans Are Useless," iMediaConnection, (accessed January 17, 2012), [available at http://w w w.imediaconnection .com/content/30235 .asp].

Lambert-Pandraud, Raphaelle and Gilles Laurent (2010), "WhyDo Older Consumers Buy Older Brands? The Role of Attach-ment and Declining Innovativeness," Journal of Marketing, 74(July), 104-121.

Beyond the "Like" Button /119

Page 16: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

Latane, Bibb (1981), "The Psychology of Social Impact," Ameri-can Psychologist, 36 (4), 343-56.

Lydon, John E., David W. Jamieson, and Mark P. Zanna (1988),"Interpersonal Similarity and the Social and IntellectualDimensions of First Impressions," Social Cognition, 6 (4),269-86.

McCracken, Grant (1988), Cuiture and Consumption: NewApproaches to the Symboiic Character of Consumer Goodsand Activities. Bloomington: Indiana University Press.

Morry, Marian M (2007), "Relationship Satisfaction as a Predictorof Perceived Similarity Among Cross-Sex Friends: A Test ofthe Attraction-Similarity Model," Journal of Social and Per-sonal Relationships, 24 (1), 117-38.

Muñiz, Albert M., Jr., and Thomas C. O'Guinn (2001), "BrandCommunity," Journal of Consumer Research, 27 (March)421-32.

Naylor, Rebecca Walker, Cait Poynor Lamberton, and David A.Norton (2011), "Seeing Ourselves in Others: Reviewer Ambi-guity, Egocentric Anchoring, and Persuasion," Journal of Mar-keting Research, 48 (June), 617-31.

Newman, Andrew Adam (2011), "Brands Now Direct Their Fol-lowers to Social Media," The New York Times, (August 4), B3.

Novak, Thomas P., Donna L. Hoffman, and Adam Duhachek(2003), "The Influence of Goal-Directed and ExperientialActivities on Online Flow Experiences," Journal of ConsumerPsychology, 13 {112), 3-16.

Perkins, Andrew W., Mark R. Forehand, and Anthony G. Green-wald (2006), "Decomposing the Implicit Self-Concept: TheRelative Influence of Semantic Meaning and Valence onAttribute Self-association." Social Cognition, 24 (4), 387-408.

Preacher, Kristopher J. and Andrew F. Hayes (2008), "Asymptoticand Resampling Strategies for Assessing and Comparing Indi-rect Effects in Multiple Mediator Models," Behavior ResearchMethods, 40 {3), 819-91.

Rosenthal, Robert, Ralph L. Rosnow, and Donald B. Rubin(2000), Contrasts and Effect Sizes in Behavioral Research: ACorreiationai Approach. New York: Cambridge UniversityPress.

Sassenberg, Kai and Tom Postmes (2002), "Cognitive and Strate-gic Processes in Small Groups: Effects of Anonymity of theSelf and Anonymity of the Group on Social Influence," BritishJournal of Social Psychology, 41 (3), 463-81.

Schau, Hope Jensen, Albert M. Muñiz Jr., and Eric J. Amould(2009), "How Brand Community Practices Create Value,"Journal of Marketing, 73 (September), 30-51.

Shachar, Ron and John W. Emerson (2000), "Cast Demographics,Unobserved Segments, and Heterogeneous Switching Costs ina Television Viewing Choice Model," Journal of MarketingResearch, 37 (May), 173-86.

Smith, J. Walker and Ann Clurman (2009), Generation Ageiess.New York: HarperCollins.

Stephen, Andrew T. and Jeff Galak (2012), "The Effects of Tradi-tional and Social Earned Media on Sales: A Study of aMicrolending Marketplace," Journai of Marketing Research,49 (October), 624-39.

Thompson, Scott A. and Rajiv K. Sinha (2008), "Brand Communi-ties and New Product Adoption: The Influence and Limits ofOppositional Loyalty," Journal of Marketing, 72 (November),65-80.

Tucker, Catherine and Juanjuan Zhang (2010), "Growing Two-Sided Networks by Advertising the User Base: A Field Experi-ment," Marketing Science, 29 (5), 805-881.

Turner, John C , Michael A. Hogg, Penny J. Oakes, Stephen D.Reicher, and Margaret S. Wetherell, eds. (1987), Rediscoveringthe Social Group: A Self-Categorization Theory. New York:Basil Blackwell.

Wells, William D. (1975), "Psychographics: A Critical Review,"Journal of Marketing Research, 12 (May), 196-213.

White, Katherine and Daren W. Dahl (2006), "To Be or Not Be?The Influence of Dissociative Reference Groups on ConsumerPreferences," Journal of Consumer Psychology, 16 (4),404--414.

and (2007), "Are All Out-Groups Created Equal?Consumer Identity and Dissociative Influence," Journal ofConsumer Research, 34 (December), 525-36.

Zhao, Xinshu, John G. Lynch Jr., and Qimei Chen (2010), "Recon-sidering Baron and Kenny: Myths and Truths about MediationAnalysis," Journal of Consumer Research, 37 (August),197-206.

120 / Journal of Marketing, November 2012

Page 17: Naylor, R. W., Lamberton, C. P., & West, P. M. (2012). Beyond the “like” button- the impact of mere virtual presence on brand evaluations and purchase intentions in social media

Copyright of Journal of Marketing is the property of American Marketing Association and its content may notbe copied or emailed to multiple sites or posted to a listserv without the copyright holder's express writtenpermission. However, users may print, download, or email articles for individual use.


Recommended