Está en la página 1de 9

Industrial Marketing Management 43 (2014) 873881

Contents lists available at ScienceDirect

Industrial Marketing Management

Should tweets differ for B2B and B2C? An analysis of Fortune 500
companies' Twitter communications
Kunal Swani a,, Brian P. Brown b,1, George R. Milne c,2
a
Raj Soin College of Business, Wright State University, 100 Rike Hall, Dayton, OH, 45435, United States
b
VCU School of Business, Virginia Commonwealth University, PO Box 844000, Richmond, VA 23284, United States
c
Isenberg School of Management, University of Massachusetts, 121 Presidents Drive, Amherst, MA 01003, United States

a r t i c l e i n f o a b s t r a c t

Article history: Business-to-business (B2B) marketers hesitate to embrace social media as a marketing tool, unlike their
Received 6 February 2013 business-to-consumer (B2C) counterparts, and they struggle to implement successful social media strategies
Received in revised form 29 September 2013 due to their limited understanding of the phenomenon. Drawing on communication and word-of-mouth theo-
Accepted 27 November 2013
ries, the authors investigate how marketers use Twitter differently across contexts and predict key factors likely
Available online 13 May 2014
to inuence the message strategies used in each. A longitudinal content analysis and logistic regression support
Keywords:
the assessment of a sample of more than 7000 tweets by Fortune 500 companies. Marketers in B2B and B2C set-
B2B advertising tings exhibit signicant differences in their branding and selling strategies; their use of message appeals; and the
Twitter use of cues, links, and hashtags to support information searches. Whereas B2B marketers tend to use more emo-
Longitudinal content analysis tional than functional appeals in their tweets, neither B2C nor B2B marketers have adopted hard sell message
Message appeals strategies.
Social media 2014 Elsevier Inc. All rights reserved.

1. Introduction descriptive and generally support ndings from practitioner studies.


For example, Michaelidou, Siamagka, and Christodoulides (2011) nd
The potential of social media as a marketing tool is undeniable. Yet that just more than 25% of B2B rms use social media to achieve their
social media research and stories of its effectiveness mainly focus on brand objectives and that non-users believe it is not important or do
business-to-consumer (B2C) examples in which social media leads not understand whether or how it might help their brands.
to increased brand awareness, loyalty, engagement, and sales (e.g., Social media tools have been lauded for their ability to enhance col-
Kumar & Mirchandani, 2012; Rapp, Beitelspacher, Grewal, & Hughes, laboration, encourage content sharing, and build community (Jansen,
2013; Taylor, Lewin, & Strutton, 2011; Wiersema, 2013). Before social Zhang, Sobel, & Chowdury, 2009; Michaelidou et al., 2011). Consumers
media marketing can be deemed an unqualied success though, mar- and businesses increasingly adopt social networking sites such as
keters need to better understand its role in business-to-business (B2B) Facebook, YouTube, LinkedIn, Pinterest, and Twitter as a result
settings; thus far, B2B marketers seem hesitant to adopt social media (Kietzmann, Hermkens, McCarthy, & Silvestre, 2011; Yang, 2012;
marketing strategies. According to one study, only 5% of B2B marketers Zaglia, 2013); Twitter is one of the most widely used social media
cite social media marketing as a well-optimized part of their marketing sites among Fortune 500 companies, such that approximately 73%
mix, 17% do not use social media at all, and 58% consider themselves in of them use it to communicate with followers (Barnes, Leschault,
the initial stages of developing their social media strategies (Giamanco & Andonian, 2012). This micro-blogging service allows marketers
& Gregoire, 2012). Similarly, 37% of marketers claim that they do to send tweets (i.e., text that is 140 characters or less with possible hy-
not know enough about social media sites to know where to begin perlinks and attached photos) to a network of associates (followers) at
(Emarketer, 2010). Marketing academics offer little guidance, because any point in time through a myriad of electronic devices (Twitter,
they have been slow to pursue research on social media phenomena 2013a). Its user base has grown exponentially since its inception,
in B2B settings (Wiersema, 2013). The few available studies tend to be attracting the attention of marketers who are interested in interacting
closely with their clientele. By tweeting their followers, marketers
have an opportunity to increase brand awareness, generate leads and
Corresponding author. Tel.: +1 216 212 1851; fax: +1 937 775 3545. revenues, foster relationships, and create brand loyalty (Kumar &
E-mail addresses: kunal.swani@wright.edu (K. Swani), bpbrown@vcu.edu (B.P. Brown),
milne@isenberg.umass.edu (G.R. Milne).
Mirchandani, 2012; Rapp et al., 2013).
1
Tel.: +1 804 827 3516. Many B2C companies already have used Twitter to their advantage
2
Tel.: +1 413 545 5669; fax: +1 413 545 3858. (Twitter, 2013c), but fewer success stories include B2B companies. In

http://dx.doi.org/10.1016/j.indmarman.2014.04.012
0019-8501/ 2014 Elsevier Inc. All rights reserved.
874 K. Swani et al. / Industrial Marketing Management 43 (2014) 873881

addition, both practitioner and academic research on micro-blogging Marketers can share information by sending tweets that might in-
websites are relatively scant (e.g., Michaelidou et al., 2011; Ramos, clude photos, videos, links, news, and stories. Although a tweet has
2008, 2009; Spekman & Dotson, 2009), such that B2B marketers appear 140-character limit, unique features available to marketers can organize
uncertain about how to take advantage of Twitter and similar sites. and share information effectively. Hashtags, a word beginning with the
Ramos (2009) notes that only 8% of B2B marketers using social media # symbol, can be added to tweets to organize conversations around
nd it highly effective in driving brand awareness, and only 5% consider specic topics and facilitate search. Mentions, or adding a Twitter
it effective for generating leads. Yet the rapid growth of and signicant username, also can be embedded in the tweet to grab a particular
investments in social media create the need for academics to explore follower's attention. Furthermore, tweets can include URLs, which are
and explicate methods for capitalizing on social media, across both con- automatically condensed to 19 characters to conserve space. With re-
sumer and business contexts. Without such guidance, B2B practitioners cently introduced features, such as expanded tweets, users can employ
may mistakenly assume that mimicking their B2C counterparts will links to preview content, view images, and play videos. This feature pro-
prove similarly rewarding. However, evidence suggests that the two vides more interactive experiences. Cashtags, which begin with the $
contexts differ enough to justify dedicated investigations and distinct symbol, also can be added to allow users to research stocks and compa-
approaches. ny performance.
The objective of this article is to explore whether B2B marketers A user can interact with a tweet by replying, retweeting (i.e.,
should adopt distinct Twitter message strategies. Considering the con- forwarding the message) to friends, or marking it a favorite tweet. A
ceptual distinctions of business and consumer markets, we investigate tweet can go viral quickly if followers start engaging by replying, favor-
how marketers use Twitter differently across these contexts, in an effort ing, and retweeting it. In addition, marketers can create buzz on Twitter
to explicate relevant social media phenomena. We use communication by promoting their tweets and accounts through various advertising op-
and word-of-mouth (WOM) theories to advance understanding of the tions. Promoted tweets are targeted to reach specic groups at the right
ow and process of communication in social media contexts. Further- time and place; promoted accounts focus on creating more followers
more, we contribute to B2B advertising and organizational buying liter- and inuencers (Twitter, 2013d).
ature by providing theoretical explanations and empirical support for
the differences in tweeting strategies between B2B and B2C contexts. 3. Social media communications theory
We thus begin with a general background of Twitter and its uses for
businesses. In discussing communication and WOM theories as a frame- Most B2B marketers use social media to increase brand loyalty, aware-
work for describing the ow and process of communication in social ness, and reputation; heighten buyer engagement; foster customer rela-
media contexts, we highlight relevant differences between B2B and tionships; and increase sales and protability (Kumar & Mirchandani,
B2C sectors, in accordance with organizational buying literature. Next, 2012; Michaelidou et al., 2011; Rapp et al., 2013). One way for marketers
we derive hypotheses from B2B advertising literature, followed by a dis- to fulll these objectives is to create brand communities and communi-
cussion of our methodology and results. We summarize our ndings cate with audiences through social media sites (Zaglia, 2013). Under-
before highlighting some managerial implications and directions for standing the ow of communication between businesses and their
research. customers thus is important.
Communication is the human activity that links people together
2. Overview of Twitter and creates relationships (Duncan & Moriarty, 1998, p. 2). In traditional
communication models (e.g., Hoffman & Novak, 1996; Lasswell, 1948;
The range of social media sites on the Internet (e.g., Facebook, Twit- Shannon & Weaver, 1949; Stern, 1994), a sender (source) encodes or
ter, YouTube, Google+, LinkedIn, and Pinterest) is vast, and each site creates a message, its content (message) gets transmitted through a
has its own unique architecture, culture, and norms that affect how medium, and the receiver decodes the sender's message. In more inter-
companies use it to execute their social media strategies (Kietzmann active communication processes, such as that of the social media
et al., 2011; Smith, Fischer, & Yongjian, 2012). Social media topics are environment, the sender and receiver change positions constantly
widespread but incompletely understood, with a variety of challenges while responding to each other's messages (Dennis, Fuller, & Valacich,
and opportunities in both B2B and B2C sectors (Kaplan & Haenlein, 2008; Labrecque, Zanjani, & Milne, 2011; Mueller, Garg, Nam, Berg, &
2010). In this research we focus on Twitter, the micro-blogging infor- McDonnell, 2011). That is, a feedback loop sends a receiver's response
mation network site, which hosts more than 200 million active users back to the sender.
who tweet approximately 400 million tweets daily (Twitter, 2013e). In Fig. 1 we depict the ow of communication between marketers
To exploit this tremendous user base, companies have begun to create and customers in a social media context. Both B2B and B2C marketers
brand-specic accounts and brand communities that can engage with encode and send social media messages through social media sites,
customers (Fournier & Avery, 2011; Kaplan & Haenlein, 2010). Approx- and customers receive and decode the messages. Marketers must de-
imately 73% of Fortune 500 companies use Twitter to connect with cus- cide which message elements to include in their communications to
tomers (Barnes et al., 2012). motivate their audiences to engage with their brand posts. Buyers or
Depending on their interests, Twitter users may follow certain busi- consumers receive the messages and must absorb and interpret them;
ness accounts or brand communities and eventually subscribe to com- if sufciently motivated, they also take action, such as continuing the di-
pany tweets. When a user becomes a follower, all subsequent tweets alogue, sharing the message with their network, storing the message, or
appear in his or her timeline. Marketers also can follow their users to ob- making a purchase. Marketers can track and monitor these actions, re-
serve their conversations and even induce two-way communication spond to comments, and continue the dialogue. Because of this response
ows to send direct messages to customers. Twitter thus adopts an capability, a continuous feedback loop exists.
asymmetrical follower model, such that businesses do not need to ap- To help build or create brand identity, meaning, responses, and rela-
prove their followers, nor do followers need to approve their businesses tionships (Keller, 2009), marketers also vary their communication strat-
(Twitter, 2013d). In this sense, Twitter is a highly viable means for mar- egies in their attempt to nd those that are likely to motivate audiences
keters to interact with customers interested in their offerings. It also to engage with the brand (Keller, 2009; Lovett, Peres, & Shachar, 2013).
enables them to gather real-time market intelligence, insights, and feed- In addition to selecting appropriate brand strategy approaches and mes-
back. Public tweets are available for anyone to read too, whether logged sage appeals, they decide whether they should encourage additional in-
in to Twitter or not. Thus, the option of open and public delivery of com- formation search and employ various selling strategies in social media
munications could benet marketers who need to spread their mes- announcements. The brand strategy approaches and message appeals
sages to a larger audience, beyond followers. employed seek to identify the brand with its customers and establish
K. Swani et al. / Industrial Marketing Management 43 (2014) 873881 875

SOCIAL MEDIA COMMUNICATION MODEL

Social
(Sender) Encode Decode (Receiver)
Media
Message
Marketers Buyer/Consumer
B2B/B2C
Decode Encode (Sender)
(Receiver)

Fig. 1. Social media communication model.

a brand association in customers' minds. Information search and selling 4. Hypotheses


strategies instead aim to create brand resonance and build customer re-
lationships. Overall, marketers must select appropriate communication Prior research generally has concluded that business and consumer
strategies to stimulate audiences to engage with their brand messages markets differ in their product offerings and decision-making processes
(Lovett et al., 2013; Swani, Milne, & Brown, 2013b). (Brown, Zablah, Bellenger, & Donthu, 2012; Mudambi, 2002; Webster &
Our proposed social media communication model (Fig. 1) suggests Wind, 1972). Because B2B offerings tend to be more technical and com-
that B2B and B2C marketers encode appropriate communication strate- plex, buyers use more formal, longer, group buying processes. Further-
gies, and then send social media messages as tweets to a corresponding more, the complexity of industrial products causes buyers to perceive
network of customers (buyers/consumers). These customers view, read, higher levels of performance and economic risk, such that they become
and decode the messages, then choose whether to respond. Marketers much more involved in the purchase decision. To mitigate these risks,
read and encode any responses and, if appropriate, respond back. We both buyers and sellers seek to establish long-term, collaborative
focus mainly on the one-way communication ow from marketers to relationships (Homburg, Klarmann, & Schmitt, 2010; Lynch & De
buyers/consumers on Twitter, so that we can evaluate the communica- Chernatony, 2004; Zablah, Brown, & Donthu, 2010). In addition, buyers
tion content (i.e., the encoding process used by marketers) and how it generally are involved and rational, using high levels of cognition during
relates to marketers' brand building strategies. their purchase decision process. In contrast, consumers tend to be
The utility of Twitter, or any social media site, hinges on substantial less involved and use low levels of cognition (Lothia, Donthu, &
numbers of people transmitting and sharing content; for marketers, Hershberger, 2003), such that they demonstrate impulsive buying be-
that content needs to enhance the brand's image (Berger & Milkman, havior toward some offerings (Minett, 2002; Swani et al., 2013b).
2012). Accordingly, marketers seek to understand which content is In response, B2B marketers generally use unique encoding processes
most likely to be tweeted (shared) among customers. Because tweets and pursue specic social media communication strategies to motivate
are the building blocks of communications on Twitter (2013d), the con- customer engagement (Brown et al., 2012) (Table 1). For example, B2B
tent of the tweet has particular importance as a primary determinant of marketers promote their corporate brand more than their individual
engagement (e.g., retweet, favor, reply), customer relationships, and product brands (Mudambi, 2002) and generally make more functional
other brand and nancial outcomes (Kumar & Mirchandani, 2012; appeals (Lohtia, Donthu, & Hershberger, 2003; Turley & Kelley, 1997).
Rapp et al., 2013). Marketers must be strategic in designing and execut- Furthermore, they likely include more informational cues and refrain
ing their tweets, to be able to inuence and motivate their followers, from commercialism cues in their communications to buyers. We use
based on their clear understanding of the factors likely to motivate cus- these criteria to develop hypotheses that denote differences between
tomers to engage with brand-related content. B2B and B2C tweets (Table 1).
Research on WOM and virality suggests that various motives insti-
gate content transmission (Berger & Milkman, 2012; Hennig-Thurau, 4.1. Brand strategy approach
Gwinner, Walsh, & Gremler, 2004; Lovett et al., 2013), including con-
sumers' psychological motivations for self-enhancement; desire to Our rst social media message criterion is based on an organization's
express their uniqueness, expertise, or emotions; goals of conversing decision to promote its corporate brand or its individual product brands.
with others; and efforts to reduce perceived risks. In turn, marketers Several studies note the importance of B2B branding (Brown et al.,
need appropriate communication strategies to induce these motiva- 2012; Kim, Reid, Plank, & Dahlstrom, 1998; Mudambi, 2002), and
tions and entice individuals to engage with brand messages through so- there is a growing stream of research that focuses on the importance
cial media. Customers might feel motivated to engage with messages of corporate branding (e.g., Brown, 1998; Brown & Dacin, 1997; Keller
that contain brand names to build their social identity (Lovett et al., & Aaker, 1998), for organizational buying in particular. Brown et al.
2013; Rapp et al., 2013). Furthermore, by engaging with emotional (2012) suggest that B2B brands can communicate both tangible attri-
and functional content or brand-related information, they might gain butes (e.g., product performance) and intangible attributes (e.g., rep-
a means to express their feelings, increase their self-enhancement, utation, distribution, support services), so B2B marketers focus on
and fulll their need to share information (Berger & Milkman, 2012; their corporate name branding strategies (e.g., Cisco, IBM, DuPont,
Lovett et al., 2013). GE, Intel), rather than the product name branding strategies adopted
These motivations to engage with social media messages should be by B2C marketers (e.g., Gillette, Clorox, Downy, Glade). They also
activated by appropriate message strategies by B2B and B2C marketers. tend to follow an umbrella branding approach, such that all products
Because marketers need appropriate social media communication take the name of the corporate brand (Michell, King, & Reast, 2001;
approaches, according to the characteristics of their followers, we Shipley & Howard, 1993). The use of such corporate brand names
predict that they use different encoding processes when tweeting in messages inuences the engagement of B2B buyers more than
to B2B buyers versus B2C consumers (Fig. 1). We thus review the it does that of B2C consumers (Swani et al., 2013b), so when B2B
differences between B2B and B2C contexts and explain their inuence marketers encode tweets in an attempt to motivate customers to en-
on customers' motivation to share content, specically in the context gage with brand messages, they likely emphasize corporate brand
of Twitter communications. names more than B2C tweets.
876 K. Swani et al. / Industrial Marketing Management 43 (2014) 873881

Table 1
Differences in B2B and B2C social media message strategies.

Message strategies B2B B2C Key citations

Brand strategy Corporate brand names Product brand names Brown et al. (2012), Mudambi (2002) and Swani et al. (2013b)
Message appeals Functional Emotional Turley & Kelley (1997) and Lohtia et al. (2003)
Selling strategy Few direct calls to purchase Many direct calls to purchase Rapp et al. (2013) and Spekman & Dotson (2009)
Information search High Low Gilliland & Johnston (1997) and Mudambi et al. (1997)

H1a. Corporate brand names appear more frequently in B2B tweets developing long-term relationships. Direct calls to purchase then
than in B2C tweets. may be more effective because consumers look for selling cues that
reward immediate purchases. When B2C companies use explicit cues,
H1b. Product brand names appear more frequently in B2C tweets than such as apply now, buy one, get one free, sale, or shop today,
in B2B tweets. in their tweets, they may entice consumers to make immediate
purchases. Thus, B2B marketers likely use different selling strategy ap-
proaches than B2C marketers when encoding tweets to motivate their
4.2. Message appeals customers to engage with their brand messages, and we hypothesize:

A basic element of advertising message development is the decision H3. Direct calls to purchase appear more frequently in B2C tweets than
to use a functional/rational appeal or an emotional appeal (Campbell, in B2B tweets.
Pitt, Parent, & Berthon, 2011; Lohtia et al., 2003; Turley & Kelley,
1997). Functional appeals refer to specic product specications, fea-
4.4. Information search
tures, performance, and other tangible cues. In contrast, according to
Kotler and Armstrong (1994, p. 468), Emotional appeals attempt to
Business marketers are charged with communicating critical, tech-
stir up either negative or positive emotions that can motivate purchase.
nological content to inform targeted buyers about an offering that
These include fear, guilt and shame appeals that [inuence] people to do
will fulll a specic, well-dened need (Minett, 2002). Gilliland and
things they should or stop doing things they shouldn't. We posit that
Johnston (1997) note that knowledge about the offering inuences
functional appeals in B2B tweets are more effective because the organi-
the extent of a buyer's information search. Because organizational
zational buying process involves assessing information related to com-
buyers tend to be experts in their elds, they engage more in analysis
plex offerings (Schiffman & Kanuk, 2004). Regardless of the purchase
during the purchase process to justify their decisions. Mudambi,
situation, organizational buying demands extensive problem solving
Doyle, and Wong (1997) show that buyers regard technical information
(Brown et al., 2012) and a cognitive decision-making process (Jensen
about products or services provided by the manufacturer as an impor-
& Jepsen, 2007). Thus, functional appeals should be particularly effec-
tant part of the buying process.
tive in B2B tweets.
The Internet provides an efcient channel for acquiring information
Emotional appeals instead tend to be more effective for consumer
that can support buying decisions by both buyers and consumers. Cor-
communications (Goldberg & Gorn, 1987; Page, Thorson, & Heide,
porate websites usually provide readily available, critical information
1990) mainly because the products are more value expressive
to their visitors; Twitter provides a means to post website links through
(Bruzzone, 1981). Message appeals should match the product type
tweets. Although marketers typically use links in tweets to increase in-
(Johar & Sirgy, 1991; Shavitt, 1990, 1992), implying that emotional ap-
bound trafc to their corporate website, or other related websites, em-
peals should be more effective for value-expressive products, whereas
bedding links also can provide buyers with more information about
functional/rational appeals work better for utilitarian (technical) offer-
the offering or company. In this sense, links might serve as sales funnels,
ings (Vaughan, 1980). Lohtia et al. (2003) nd that emotional appeals
driving prospects to corporate websites. Cisco, IBM, and Accenture in-
are more effective in B2C than in B2B banner ads. Because B2B mar-
clude cues such as learn more, read on, more information,
keters are likely to use different message appeals than B2C marketers
watch video, or new technology news in their tweets to link pros-
when encoding tweets to motivate customers to engage with their
pects to white papers, testimonials, videos, product specications, and
brand messages, we predict the following:
other potentially valuable information, in exchange for detailed person-
H2a. Functional appeals appear more frequently in B2B tweets than in al or company information. In the rational decision-making process for
B2C tweets. B2B settings, information search tends to be more extensive than it is
in B2C settings.
H2b. Emotional appeals appear more frequently in B2C tweets than in Furthermore, Twitter offers the unique hashtag feature, which com-
B2B tweets. panies can use to mark keywords or topics (e.g., #intel-powered,
#cisco-linksys, #ibm-db2, and #avaya_nes) and clearly categorize mes-
sages (Twitter, 2013b). Clicking on a hashtag allows users to see similar-
4.3. Selling strategy approach ly themed tweets, so they facilitate information search and should be
more prominent in business communications whose audiences search
Direct calls to purchase (i.e., hard sells) are explicit statements information more actively than consumer audiences. We predict then
encouraging prospective buyers to make an immediate purchase. that B2B marketers use different strategies for information sharing
Buyers tend to be reluctant to respond to such direct calls, due to when encoding tweets to motivate their customers to engage with
the rigor and complexity that they associate with their lengthy buy- brand messages.
ing processes. Thus they may lose interest in social media sites that
use explicit selling strategies (Rapp et al., 2013; Spekman & Dotson, H4a. Embedded links and cues for additional information search ap-
2009). Although the use of hard sell approaches in social media mes- pear more frequently in B2B tweets than in B2C tweets.
sages could deter engagement for both B2B buyers and B2C con-
sumers (Swani et al., 2013b), in B2C contexts, the emphasis often is H4b. Hashtags appear more frequently in B2B tweets than in B2C
on selling and encouraging impulsive, short-term sales rather than tweets.
K. Swani et al. / Industrial Marketing Management 43 (2014) 873881 877

4.5. Time effects 5.3. Regression models

Because social media is relatively less researched in B2B contexts To test our hypotheses, we ran a main effects logistic regression to
(Michaelidou et al., 2011; Rapp et al., 2013; Wiersema, 2013), B2B mar- compare message strategies across B2B and B2C tweets. The indepen-
keters likely are inuenced by B2C marketers in developing their social dent variables were the uses of corporate brand name, product brand
media communication strategies. Because the two contexts are so dif- name, functional appeals, emotional appeals, direct calls to purchase,
ferent, this mimicry may be inappropriate and ineffective for B2B mar- links and cues for information search, and hashtagsall dummy vari-
keters. As they increasingly experiment with and implement social ables, for which 1 indicated presence and 0 indicated absence. The de-
media methods, it becomes useful to verify whether B2B marketers pendent variable, communication type of each message, was coded as
adopt social media communication practices initiated by B2C mar- B2C (=1) or B2B (=0).
keters. Thus, we seek to determine whether B2B Twitter communica- To test for time effects, we also ran a hierarchical logistic regres-
tion practices converge with or diverge from B2C communication sion model in which we added the main effect of time (1 = time pe-
practices over time. riod 1; 0 = time period 2) and the interactions between time and
message strategies in the main effects model. Signicant interaction
5. Method effects suggest differences in the uses of message strategies across
B2B and B2C contexts over time.
To analyze the message strategies of Fortune 500 companies, we cap-
tured their tweets and submitted them to a longitudinal content analy- 6. Results
sis to assess their content. In addition, we applied logistic regression to
test our hypotheses. Content analysis approaches (Kassarjian, 1977) The coders identied 184 tweets as both B2B and B2C. We eliminat-
offer a frequently used methodology to study marketing content, and ed them from the data set, leaving a total of 3270 tweets from time pe-
particularly the practices and approaches that marketers use in their riod 1 and 3827 tweets from time period 2 for analysis.
marketing communications (Neuendorf, 2002; Swani et al., 2013b;
Turley & Kelley, 1997). In B2B contexts, content analyses have explicat- 6.1. Descriptive statistics
ed communications in several media, including print advertising (Clarke
& Honeycutt, 2000; Cutler & Javalgi, 1994; Jensen & Jepsen, 2007; Lohtia In Table 2 we report the descriptive statistics of message strategies
et al., 1995; Naccarato & Neuendorf, 1998; Turley & Kelley, 1997), banner for B2B and B2C tweets for time period 1, time period 2, and both pe-
ads (Lohtia et al., 2003), Facebook wall posts (Swani et al. 2013b), and riods. The data set featured 27.3% B2B tweets and 72.7% B2C tweets
tweets (Swani, Milne, Cromer & Brown, 2013a). (2(1) = 1461.83, p b .01). As predicted, the percentages of uses of
corporate brand names (B2B = 45.1%; B2C = 35.5%), functional
5.1. Data appeals (B2B = 24.2%; B2C = 13.6%), embedded links and cues for
additional information search (B2B = 80.7%; B2C = 75.7%), and
Our data reect 289 Fortune 500 companies' tweets, generated by hashtags (B2B = 56.4%; B2C = 44.4%) were consistently higher for
their ofcial Twitter accounts. The list of Fortune 500 companies with B2B tweets compared with B2C tweets. The percentages of uses of
Twitter accounts was based on Barnes (2010). We tracked the accounts emotional appeals (B2B = 30.3%; B2C = 59.5%) and direct calls to
at two points in time: during the week of March 29, 2011, which provid- purchase (B2B = 3.5%; B2C = 8.3%) instead were higher for B2C tweets
ed 3278 unique company tweets active during time 1, and then during compared with B2B tweets. Surprisingly though, the percentage of use
the week of September 29, 2011, which produced 4003 unique compa- of product brand names was similar across contexts (B2B = 17.7%;
ny tweets during time 2. B2C = 18.2%).

5.2. Content analysis


6.2. Main effects

Five research assistants, unaware of our research questions, helped


The overall main effects model was statistically signicant (2(7) =
code the data for times 1 and 2. All coders participated in four training
694.42, p b 0.01) and the model predicted percent correct was 0.73, suf-
sessions over a period of two weeks, to ensure that they understood
ciently higher than the 0.50 threshold. The pseudo R2 for the model
the key concepts and could apply the coding scheme efciently. Each
was 0.093. None of the correlations between exogenous variables
coder coded more than 60 messages during the training sessions. The
exceeded 0.32, reducing the potential for multicollinearity issues (type
intercoder reliability measure used a randomly selected subsample of
error) in our regression analysis. We report these results in Table 3,
100 messages from a separate data set (Lohtia et al., 2003; Neuendorf,
Panel a.
2002). We calculated intercoder reliability for the message strategy
We hypothesized that the corporate brand names would appear
variables and the communication type variable using Rust and Cooil's
more frequently in B2B tweets than in B2C tweets, and because the
(1994) proportional reduction in loss index (PRL); a value of 0.70
is acceptable, and 0.90 is desirable. All reliabilities were high (mean Table 2
PRL = 0.99) and above desired levels, suggesting that no meaningful Message strategy executions for B2B and B2C tweets.
differences contributed to the coders' evaluations. In coding the mes-
Time 1 Time 2 Time 1 and 2
sages from the original data set, three coders coded a separate, non-
overlapping portion of messages collected at time period 1, and the Message strategy B2B B2C B2B B2C B2B B2C
remaining two coders coded a separate, non-overlapping portion of Corporate brand name a
43.8 35.3 45.6 35.7 45.1 35.5
the messages collected at time period 2. Thus, each coder was fungible, Product brand namea 13.3 12.0 19.5 25.0 17.7 18.2
given the high level of intercoder reliability. Functional appealsa 7.8 1.7 31.2 26.5 24.2 13.6
Emotional appealsa 39.0 56.4 26.6 62.8 30.3 59.5
For the independent variables indicating strategy type, we used bi- Direct calls to purchasea 0.5 1.9 4.7 15.3 3.5 8.3
nary coding, where 1 indicated that the trait was present and 0 indicat- Information searcha 87.9 76.4 77.6 74.8 80.7 75.7
ed its absence. The coders also recorded the number of hashtags in each Hashtagsa 45.2 39.1 61.1 50.3 56.4 44.4
message so we had a binary variable indicating whether a hashtag was Total tweetsb 577 2693 1361 2466 1938 5159
present (1) or not (0). The communication context was classied into a
Percentage of tweets.
b
three categories: B2B, B2C, or both. Tweet counts.
878 K. Swani et al. / Industrial Marketing Management 43 (2014) 873881

Table 3
Logistic regression results.

a. Main effects model

Dependent variable: B2C tweets

Message strategy Estimate Odds ratio SE WALD Sig Hypotheses Probabilitya

Corporate brand name (1 = yes) 0.265 0.767 0.057 21.453 0.000 H1a supported 0.434b
Product brand name (1 = yes) 0.104 1.110 0.080 1.712 0.191 H1b not supported
Functional appeals (1 = yes) 0.576 0.562 0.075 58.764 0.000 H2a supported 0.360
Emotional appeals (1 = yes) 1.092 2.981 0.059 345.613 0.000 H2b supported 0.749
Direct calls to purchase (1 = yes) 0.786 2.195 0.142 30.913 0.000 H3 supported 0.687
Information search (1 = yes) 0.246 0.782 0.070 12.280 0.000 H4a supported 0.439
Hashtags (1 = yes) 0.482 0.618 0.057 72.218 0.000 H4b supported 0.382
Model t Pseudo pseudo R2 = 0.093 2(7) = 694.42, p b 0.01 Predicted percent correct = 73.6%

b. Interaction effects model

Dependent variable: B2C tweets

Message strategy Estimate Odds ratio SE WALD Sig Probability time 1c Probability time 2c

Corporate brand name (1 = yes) 0.324 0.724 0.075 18.556 0.000


Product brand name (1 = yes) 0.195 1.215 0.096 4.087 0.043
Functional appeals (1 = yes) 0.001 0.999 0.086 0.000 0.994
Emotional appeals (1 = yes) 1.455 4.286 0.078 351.757 0.000
Direct calls to purchase (1 = yes) 0.994 2.702 0.149 44.681 0.000
Information search (1 = yes) 0.008 0.992 0.088 0.008 0.930
Hashtags (1 = yes) 3.710 0.690 0.075 24.571 0.000
Time (1 = time period 1) 2.008 7.447 0.184 119.277 0.000
Corporate brand name time 0.061 1.063 0.122 0.248 0.619
Product brand name time 0.175 0.840 0.177 0.970 0.325
Functional appeals time 1.390 0.249 0.243 32.687 0.000 0.199d 0.499d
Emotional appeals time 0.856 0.425 0.124 47.672 0.000 0.645 0.810
Direct calls to purchase time 0.177 1.124 0.620 0.035 0.851
Information search time 0.758 0.469 0.164 21.406 0.000 0.317 0.498
Hashtags time 0.039 1.040 0.121 0.103 0.748
Model t Pseudo pseudo R2 = 0.136 2
(15) = 1034.39, Predicted percent correct = 75.5%
p b 0.01
a
Only probabilities for signicant main effects are shown.
b
Read as: 0.434 is the probability of the tweet being B2C when the corporate brand name is present.
c
Only probabilities for signicant interactions are shown.
d
Read as: 0.199 (0.499) is the probability of the tweet being B2C when functional appeals are present at time 1 (time 2).

variable representing corporate brand name was negative and signi- the 0.50 threshold. The pseudo R2 for the model was 0.136. We report
cant ( = 0.265, p b 0.01), we afrmed H1a. However, we cannot these results in Table 3, Panel b.
conrm H1b, because the presence of product brand names was not The interaction between time and the use of functional appeals was
more frequent in B2C tweets than in B2B tweets, according to the posi- signicant ( = 1.390, p b 0.01). This indicates that the percentage
tive but insignicant variable representing product brand name ( = use of functional appeals across B2B and B2C social media messages
0.104, p N 0.01). In line with H2a, functional appeals appear more converged over time (B2BT1 = 7.8%, B2CT1 = 1.7%; B2BT2 = 31.2%,
frequently in B2B tweets than in B2C tweets, according to the nega- B2CT2 = 26.5%). Furthermore, the interaction between time and the
tive, signicant variable representing functional appeals ( = 0.576, use of emotional appeals was signicant ( = 0.856, p b 0.01). The
p b 0.01). In H2b, we predicted that emotional appeals appear more fre- percentage use of emotional appeals across B2B and B2C social media
quently in B2C tweets than in B2B tweets, and in support of this predic- messages diverged even more over (B2BT1 = 39.0%, B2CT1 = 56.4%;
tion, the variable representing emotional appeals was positive and B2B T2 = 26.6%, B2C T2 = 62.8%). The interaction between time
signicant ( = 1.092, p b 0.01). In support of H3, direct calls to pur- and the use of cues and links for information search was signicant
chase appear more frequently in B2C tweets than in B2B tweets, with ( = 0.758, p b 0.01), so this tactic converged across B2B and
a positive, signicant variable ( = 0.786, p b 0.01). The use of links B2C social media messages (B2B T1 = 87.9%, B2C T1 = 76.4%;
and cues for information search also appeared more frequently in B2B B2BT2 = 77.6%, B2CT2 = 74.5%).
tweets than in B2C tweets, in that the variable representing the pres-
ence of information search cues was negative and signicant ( = 7. Discussion
0.246, p b 0.01), in support of H4a. Finally, in support of H4b, the
use of hashtags appeared more frequently in B2B tweets than in B2C Social media has the potential to enhance brand equity. It allows
tweets, according to the negative and signicant variable representing consumers and buyers to interact with their brands in ways that can
hashtags ( =0.482, p b 0.01). create brand identity and brand meaning, as well as encourage brand
responses and relationships (Keller, 2009). Furthermore, increasing
the brand's presence and reputation on social media sites can boost
6.3. Interaction effects sales performance and other nancial and brand outcomes (Kumar &
Mirchandani, 2012; Rapp et al., 2013). Thus it has become critical for
We added time and the interaction terms to our main effects model marketers to use appropriate communication strategies to motivate
and uncovered a signicant difference between the main effects model their audiences to engage with brand messages (Swani et al., 2013b).
and interaction effects model (2(8) = 399.97, p b 0.01). The overall in- With this study we have sought to explore whether differences be-
teraction effects model was signicant (2(15) = 1034.39, p b 0.01) and tween business and consumer markets justify the use of distinct social
the model predicted percent correct was 0.75, sufciently higher than media message strategies. Applying communication and WOM theories
K. Swani et al. / Industrial Marketing Management 43 (2014) 873881 879

to tweets, we contend that B2B marketers encode tweets differently 8. B2B implications
than B2C marketers so that they can motivate their unique customers
to engage with their brand messages. Building on B2B advertising and These ndings suggest that B2B marketers must continue to inte-
organizational buying literature, we also highlight several particularly grate social media strategies into their overall marketing communica-
relevant differences in brand strategy, message appeals, selling ap- tion strategies. Even if they feel uncertain of the role of Twitter or
proach, and information search between the B2B and B2C contexts. social media in general, these marketers keep increasing their usage
We tested seven hypotheses related to these differences and found sup- over time, as manifested in the 136% increase in the number of tweets
port for six of them. We also tested whether these differences persisted in time period 2 compared with time period 1 (time 1 = 577; time
over a six-month period. 2 = 1361). Prior research similarly indicates a growing presence of
Do B2B tweets differ from B2C tweets? Our results indicate some dif- B2B buyers on social media websites. In response, B2B marketers
ferences. In B2B tweets, corporate name branding is more frequent than have started to allocate more marketing resources to social media
it is in B2C tweets. We expected B2B marketers to highlight corporate and should try to take advantage of the social media sites that their
brand names in their tweets to their followers. Indeed, our results target buyers tend to frequent. Moreover, considering the impor-
show that the percentage use of corporate brand names is higher in tance of relationship building in organizational buying, social
B2B tweets compared with B2C tweets. A more intriguing nding is media offer a viable tool for B2B marketers to foster customer
that there appears to be no difference in the use of product brand brand relationships, especially if the companies have developed
names between the two contexts. We nd that both B2B (2(1) = brand reputations and preexisting customer relationships (Rapp
1217.00, p b 0.01) tweets as well as B2C (2(1) = 2770.00, p b 0.01) et al., 2013).
tweets have a higher presence of corporate brand names than product Yet B2B marketers also should keep the nuances of the organization-
brand names in their message strategies. This surprising result might al buying context in mind when developing their social media strate-
vary with industry or product type differences in our sample, consider- gies. Our results reinforce the notion that B2B marketers tend to focus
ing that Swani et al. (2013a) nd that B2B service companies tend to on promoting their corporate brands (45.1%) rather than their product
have a higher percentage of corporate brand names, whereas B2C ser- brands (17.7%) in their tweets. Companies with reputable brands can le-
vice companies have a higher percentage of product brand names in verage their brand image through social media, because customers who
their tweeting strategies. For B2B and B2C companies that offer prod- prefer to be associated with reputable brands are likely to engage with
ucts, they indicate no such difference (Swani et al., 2013a). We have them, such as by spreading positive WOM (Rapp et al., 2013). Our nd-
not analyzed tweets at the industry or product level, but these factors ing that emotional appeals are more common than functional appeals in
might inuence our results related to the use of product and corporate B2B tweets challenges established perspectives of traditional media
brand names in tweets. practices though, which point to the use of functional appeals as more
With regard to appeals, the ndings matched our predictions: B2B effective for B2B marketers (Turley & Kelley, 1997). Rather, it is
tweets featured a higher percentage of functional appeals, whereas B2C more consistent with growing evidence that the dynamics of social
tweets included a higher percentage of emotional appeals. Yet the data media advertising are distinct from those of traditional advertising.
also suggest that B2B tweets include more emotional appeals (30.3%) Emotional cues are important for B2B marketing too (Gilliland &
than functional appeals (24.2%; 2(1) = 13.19, p b .01; Table 1). This nd- Johnston, 1997; Jensen & Jepsen, 2007; Lynch & De Chernatony,
ing is consistent with recent organizational buying research that notes the 2004), and they may be especially effective in social media commu-
role of involvement and the signicance of brand emotions in decision- nications (Swani et al., 2013b). To differentiate themselves and
making processes (Gilliland & Johnston, 1997; Jensen & Jepsen, 2007; build strong brand relationships with their target customers, we rec-
Lynch & De Chernatony, 2004). Emotional content in B2B communica- ommend B2B marketers to integrate emotional appeals into their
tions can help foster customer relationships, add valuable dimen- tweets and other communications. Effective B2B brand communica-
sions to product or services, and lead to a sustainable differential tions emphasize both emotional and functional brand values to gain
advantage (Lynch & De Chernatony, 2004; Swani et al., 2013b), and sustainable, differential advantages over their competitors (Lynch &
the use of corporate brand names with emotional content has De Chernatony, 2004).
emerged as a very effective social media strategy for B2B companies Finally, hard sell tweets are infrequent (3.5%), but cues and links for
(Swani et al., 2013b). The use of emotional content in tweets thus can information search (80.7%) and the use of hashtags (56.4%) are relative-
increase customer engagement and strengthen brand relationships. ly common as B2B message strategies. The social media platform is not
Although B2C tweets use a higher percentage of direct calls to pur- well suited to serving as a selling tool; it is more applicable for building
chase than B2B tweets, it also is interesting to observe the relatively customer relationships (Rapp et al., 2013). Therefore, B2B marketers
low percentage of such calls in both contexts. That is, marketers do should avoid using explicit approaches; the frequency of information
not appear to use Twitter primarily as a selling tool. The dynamics of so- cues and hashtags suggests that social media already is effective mainly
cial media advertising suggest an approach that differs from more tradi- as an informational platform or resource. Thus companies can leverage
tional advertising, in which hard sells are less appropriate (Rapp et al., their customer relationships by providing audiences with useful brand
2013). Additionally, B2B tweets have a higher percentage of links and information.
cues for information search in their messages than B2C tweets, and
the percentage use of hashtags is higher in B2B tweets as well. Yet we 9. Limitations and further research
nd substantial usage of cues and links and of hashtags in both B2B
(80.7% and 56.4%, respectively) and B2C (75.7% and 44.4%, respectively) Several limitations of this study invite continued research. First, our
tweets. That is, regardless of the context, marketers use Twitter as an sample only generalizes to Fortune 500 rms' tweets; it should be ex-
information-sharing platform. tended to include messages from small businesses, specic industries,
Finally, our results show that the characteristics of B2B and B2C or different social media sites. Second, the differences we observed
tweets have shifted over time, sometimes diverging and sometimes were limited to a select group of message strategies. A better under-
converging. The recognition that uses of functional appeals and infor- standing of the uses of images, videos, or other message formats (e.g.,
mation search have converged over time indicates that Twitter has news, job postings, community information) in tweets and other social
grown into a platform for sharing product, service, and company infor- media communications would be useful.
mation across contexts. However, the use of emotional appeals has di- Additional research also could consider more intricate classications
verged over time, with a percentage increase in emotional appeals in of emotional (e.g., adventure, fear, humor, guilt) and functional (e.g.,
B2C tweets compared with B2B tweets. quality, performance, economic) appeals, to test for differences across
880 K. Swani et al. / Industrial Marketing Management 43 (2014) 873881

appeal types and contexts. Such deeper understanding of the use and ef- Jansen, B. J., Zhang, M., Sobel, K., & Chowdury, A. (2009). Twitter power: Tweets as elec-
tronic word of mouth. Journal of the American Society for Information Science and Tech-
fectiveness of brand emotions in B2B communications is needed. Addi- nology, 60(11), 21692188.
tional research also might consider other important inuences, such as Jensen, M. B., & Jepsen, A. L. (2007). Low attention advertising processing in B2B markets.
industry type and product type, on B2B and B2C social media communi- Journal of Business and Industrial Marketing, 22(5), 342348.
Johar, J. S., & Sirgy, J. M. (1991). Value-expressive versus utilitarian advertising appeals:
cation practices. Finally, we encourage researchers to explore social When and why to use which appeal. Journal of Advertising, 20(3), 2333.
media effectiveness and its impact on rm or brand performance in Kaplan, A. M., & Haenlein, M. (2010). Users of the world, unite! The challenges and oppor-
B2B contexts. In particular, we need a better understanding of how so- tunities of social media. Business Horizons, 53, 5968.
Kassarjian, H. H. (1977). Content analysis in consumer research. Journal of Consumer
cial media communications promote WOM and ultimately enhance Research, 4, 818.
marketing (e.g., brand awareness, brand relationships and brand equi- Keller, K. L. (2009). Building strong brands in a modern marketing communications envi-
ty) and nancial (e.g., sales, return on investment and stock price) out- ronment. Journal of Marketing Communications, 15(23), 139155.
Keller, K. L., & Aaker, D. A. (1998). Corporate level marketing: The impact of credibility on
comes. The prevalence of social media usage clearly demonstrates that
a company's brand extensions. Corporate Reputation Review, 1(1), 356381.
social media effectiveness is a research topic that requires continued ac- Kietzmann, J. H., Hermkens, K., McCarthy, I. P., & Silvestre, B. S. (2011). Social media? Get
ademic attention. serious! Understanding the functional building blocks of social media. Business
In conclusion, our research has explored differences in the tweeting Horizons, 54(3), 241251.
Kim, J., Reid, D. A., Plank, R. E., & Dahlstrom, R. (1998). Examining the role of brand equity
strategies used in B2B and B2C contexts. We identify signicant differ- in business markets: A model, research propositions, and managerial implications.
ences in these strategies, specically as they relate to the use of branding Journal of Business-to-Business Marketing, 5(3), 6589.
and selling strategies, message appeals, and cues, links, and hashtags that Kotler, P., & Armstrong, G. (1994). Principles of Marketing (6th ed.). Englewood Cliffs, NJ:
Prentice-Hall.
signal additional information in their Twitter communications. We en- Kumar, V., & Mirchandani, R. (2012). Winning with data: Increasing the ROI of social
courage our fellow academics to build on this understanding of social media marketing. MIT Sloan Management Review, 54(1), 5561.
media in both settings, particularly in the generally under researched Labrecque, L., Zanjani, S. H. A., & Milne, G. R. (2011). Authenticity in online communica-
tions: Examining antecedents and consequences. In Angeline Close (Ed.), Online Con-
B2B setting. sumer Behavior: Theory and Research in Social Media, Advertising, and E-Tail. : Taylor
and Francis Group.
Lasswell, H. D. (1948). The structure and function of communication in society. In L.
Acknowledgments Bryson (Ed.), The Communication of Ideas: A Series of Addresses (pp. 3235). New
York: Cooper Square Publishers.
We thank Professor Bruce Skaggs and Spencer Ross for their com- Lohtia, R., Donthu, N., & Hershberger, E. K. (2003). The impact of content and design ele-
ments on banner advertising click-through rates. Journal of Advertising Research,
ments and insights on prior versions of this article; we thank our coders 43(4), 410418.
for their tireless assistance. Lohtia, R., Johnston, W. J., & Aab, L. (1995). Business-to-business advertising: What are the
dimensions of an effective print ad? Industrial Marketing Management, 24(5),
369378.
References Lovett, M., Peres, R., & Shachar, R. (2013). On brands and word-of-mouth. Journal of
Marketing Research, 50(4), 41274444.
Barnes, N. G. (2010). The Fortune 500 and Social Media: A Longitudinal Study of Blogging, Lynch, J., & De Chernatony, L. (2004). The power of emotion: Brand communication in
Twitter and Facebook Usage by America's Largest Companies. Available at. http:// business-to-business markets. Journal of Brand Management, 11(5), 403419.
www.umassd.edu/cmr/ Michaelidou, N., Siamagka, N. T., & Christodoulides, G. (2011). Usage, barriers and mea-
Barnes, N. G., Leschault, A. M., & Andonian, J. (2012). Social Media Surge by the 2012 For- surement of social media marketing: An exploratory investigation of small and medi-
tune 500: Increase Use of Blogs, Facebook, Twitter and More. Available at. http:// um B2B brands. Industrial Marketing Management, 40, 11531159.
www.umassd.edu/cmr/socialmedia/2012fortune500/ (Accessed October 8, 2012) Michell, P., King, J., & Reast, J. (2001). Brand values related to industrial products.
Berger, J., & Milkman, K. L. (2012). What makes online content viral? Journal of Marketing Industrial Marketing Management, 30(5), 415425.
Research, 49(2), 192205. Minett, S. (2002). B2B Marketing: A Radically Different Approach for Business-to-Business
Brown, T. J. (1998). Corporate associations in marketing: Antecedents and consequences. Marketers. London: Pearson.
Corporate Reputation Review, 1(3), 215234. Mudambi, S. M. (2002). Branding importance in business-to-business markets: Three
Brown, T. J., & Dacin, P. A. (1997). The company and the product: Corporate associations buyer clusters. Industrial Marketing Management, 31(6), 525533.
and consumer product responses. Journal of Marketing, 61(1), 6884. Mudambi, S. M., Doyle, P., & Wong, V. (1997). An exploration of branding in industrial
Brown, B. P., Zablah, A. R., Bellenger, D. N., & Donthu, N. (2012). What factors inuence markets. Industrial Marketing Management, 26(5), 433446.
buying center brand sensitivity? Industrial Marketing Management, 41(3), 508520. Mueller, K., Garg, S., Nam, J. E., Berg, T., & McDonnell, K. T. (2011). Can computers master
Bruzzone, D. (1981). New evidence on when to use mood and message. BRC Newsletter, the art of communication? A focus on visual analytics. Computer Graphics and
45. Applications, IEEE, 31, 1421.
Campbell, C., Pitt, L. F., Parent, M., & Berthon, P. R. (2011). Understanding consumer con- Naccarato, J. L., & Neuendorf, K. A. (1998). Content analysis as a predictive methodology:
versations around ads in a web 2.0 world. Journal of Advertising, 40(1), 87102. Recall, readership, and evaluations of business-to-business print advertising. Journal
Clarke, I., & Honeycutt, E. D. (2000). Color usage in international business-to-business of Advertising Research, 1933 May-June.
print advertising. Industrial Marketing Management, 29(3), 255261. Neuendorf, Kimberly A. (2002). The Content Analysis Guidebook. Thousand Oaks, CA: Sage.
Cutler, B. D., & Javalgi, R. G. (1994). Comparison of business-to-business advertising: The Page, T. J., Thorson, E., & Heide, M. P. (1990). The memory impact of commercials varying
United States and the United Kingdom. Industrial Marketing Management, 23(2), in emotional appeal and product involvement. In Stuart J. Agres, Julie A. Edell, & Tony
117124. M. Dubitsky (Eds.), Emotion in Advertising (pp. 255268). New York: Quorum Books.
Dennis, A. R., Fuller, R. M., & Valacich, J. S. (2008). Media, tasks, and communication pro- Ramos, L. (2008). Making social media work in B2B marketing. Forrester Report.
cesses: A theory of media synchronicity. MIS Quarterly, 32(3), 575600. Ramos, L. (2009). B2B marketers: Tap into social networking sites to energize community
Duncan, T., & Moriarty, S. (1998). A communication based marketing model for managing marketing. Forrester Report.
relationships. Journal of Marketing, 62(2), 113. Rapp, A., Beitelspacher, L. S., Grewal, D., & Hughes, D. E. (2013). Understanding social
Emarketer (2010). Complimentary eMarketer Report: Seven Guidelines for Achieving media effects across seller, retailer, and consumer interactions. Journal of the
ROI from Social Media. Available at. http://www.emarketer.com/blog/index.php/ Academy of Marketing Science, 41(5), 547566.
complimentary-emarketer-report-guidelines-achieving-roi-social-media/ Rust, R. T., & Cooil, B. (1994). Reliability measures for qualitative data: Theory and impli-
#qLcSBtK2HzEdrxHA.99 (accessed April 4, 2012) cations. Journal of Marketing Research, 31, 114.
Fournier, S., & Avery, J. (2011). The uninvited brand. Business Horizons, 54, 193207. Schiffman, L., & Kanuk, L. (2004). Consumer Behavior (8th ed.). Upper Saddle River, NJ:
Giamanco, B., & Gregoire, K. (2012, JulyAugustt). Tweet me, friend me, make me buy. Pearson Education.
Harvard Business Review, 8893. Shannon, C. E., & Weaver, W. (1949). The Mathematical Theory of Communication. Urbana:
Gilliland, D., & Johnston, W. (1997). Toward a model of business-to-business marketing University Illinois Press.
communications effects. Industrial Marketing Management, 26(1), 1529. Shavitt, S. (1990). The role of attitude objects in attitude functions. Journal of Experimental
Goldberg, M. E., & Gorn, G. J. (1987). Happy and sad TV programs: How they affect reac- Social Psychology, 26(2), 124148.
tions to commercials. Journal of Consumer Research, 14(3), 387403. Shavitt, S. (1992). Evidence for predicting the effectiveness of value-expressive versus
Hennig-Thurau, T., Gwinner, K. P., Walsh, G., & Gremler, D. D. (2004). Electronic word-of- utilitarian appeals: A reply to Johar and Sirgy. Journal of Advertising, 21(2), 4751.
mouth via consumer-opinion platforms: What motivates consumers to articulate Shipley, D., & Howard, P. (1993). Brand-naming industrial products. Industrial Marketing
themselves on the internet? Journal of Interactive Marketing, 18(1), 3852. Management, 21(1), 5966.
Hoffman, D. L., & Novak, T. P. (1996). Marketing in hypermedia computer-mediated envi- Smith, A. N., Fischer, E., & Yongjian, C. (2012). How does brand-related user-generated
ronments: Conceptual foundations. Journal of Marketing, 60(3), 5068. content differ across YouTube, Facebook, and Twitter? Journal of Interactive
Homburg, C., Klarmann, M., & Schmitt, J. (2010). Brand awareness in business markets: Marketing, 26(2), 102113.
When is it related to rm performance? International Journal of Research in Spekman, R. E., & Dotson, E. (2009). Using social media in the B2B context. Harvard
Marketing, 2(3), 201212. Business Review (Prod. #: UV2973-PDF-ENG).
K. Swani et al. / Industrial Marketing Management 43 (2014) 873881 881

Stern, B. B. (1994). A revised communication model for advertising: Multiple dimensions Webster, F. E., & Wind, Y. (1972). A general model for understanding organizational buy-
of the source, the message, and the recipient. Journal of Advertising, 23(2), 515. ing behavior. Journal of Marketing, 36(2), 1219.
Swani, K., Milne, G. R., Cromer, C., & Brown, B. P. (2013a). Fortune 500 companies use of Wiersema, F. (2013). The B2B agenda: The current state of B2B marketing and a look
Twitter communications: A comparison between product and service tweets. ahead. Industrial Marketing Management, 42(4), 470488.
International Journal of Integrated Marketing Communications, 5(2), 4756. Yang, T. (2012). The decision behavior of Facebook users. Journal of Consumer Information
Swani, K., Milne, G. R., & Brown, B. P. (2013b). Spreading the word through likes on Systems, 52(3), 5059.
Facebook: Evaluating the message strategy effectiveness of Fortune 500 companies. Zablah, A. R., Brown, B. P., & Donthu, N. (2010). The relative importance of brands in mod-
Journal of Research in Interactive Marketing, 7(4), 269294. ied rebuy purchase situations. International Journal of Research in Marketing, 27(3),
Taylor, D. G., Lewin, J. E., & Strutton, D. (2011). Friends, fans, and followers: Do ads work 248260.
on social networks? How gender and age shape receptivity. Journal of Advertising Zaglia, M. E. (2013). Brand communities embedded in social networks. Journal of Business
Research, 51(1), 258275. Research, 66(2), 216223.
Turley, L. W., & Kelley, S. (1997). A comparison of advertising content: Business-to-
business versus consumer services. Journal of Advertising, 26(4), 3948.
Kunal Swani (Ph.D., University of Massachusetts Amherst), Assistant Professor of Market-
Twitter (2013a). What is Twitter? Available at. https://business.twitter.com/en/basics/
ing, Raj Soin College of Business, Wright State University, Kunal.swani@wright.edu.
what-is-twitter/ (Accessed September 10, 2013)
Twitter (2013b). Glossary. Available at. https://business.twitter.com/en/basics/glossary/
Brian P. Brown (Ph.D., Georgia State University), Associate Professor of Marketing, Virginia
(Accessed September 10, 2013)
Commonwealth University, bpbrown@vcu.edu.
Twitter (2013c). Case Studies. Available at. https://business.twitter.com/en/optimize/
case-studies/ (Accessed September 10, 2013)
George R. Milne (Ph.D., University of North CarolinaChapel Hill), Professor of Marketing,
Twitter (2013d). Twitter for Small Businesses. Available at. https://business.twitter.com/
Ph.D. Program Director, Isenberg School of Management, University of Massachusetts
pdfs/Twitter_Smallbiz_Guide.pdf (Accessed September 10, 2013)
Amherst, milne@isenberg.umass.edu.
Twitter (2013e). How Twitter Can Help your Businesses. Available at. https://business.
twitter.com/how-twitter-can-help-your-business (Accessed September 10, 2013)
Vaughan, R. (1980). How advertising works. Journal of Advertising Research, 20, 2733.