Follow Me! Global Marketers' Twitter Use
Eun Sook Kwon
University of Georgia
Yongjun Sung
University of Texas at Austin
Abstract
Social media have grown into a powerful marketing communications tool in the global economy. Marketers dedicate their time and resources to build trust and rapport with consumers through social media, yet a dearth of academic research addresses their use of Twitter. The present research undertakes an exploratory content analysis, focusing on 44 global brands' Twitter use. The results indicate that marketers try to imbue human personality into their brands by using human representatives, personal pronouns, and verbs in the imperative form. In addition, information such as brand names and redirecting cues is frequent in brands' tweets. Overall, marketers tend to exhibit their brand presence and personalities in their Twitter accounts, thereby building relationships with current and potential consumers. This article concludes with some social media marketing implications and future research directions.
Keywords: Social media, Twitter, brand anthropomorphism, content analysis
Social media have grown into a powerful tool, attracting millions of users, many of whom have integrated these new means of mass communication into their daily lives (boyd and Ellison 2007). Although the first adopters were mainly teenagers, a growing population of 25-34-year-olds and white-collar professionals are social media users who build and maintain existing social relationships (Haythornthwaite 2005) and find new connections based on common interests, political perspectives, or activities (boyd and Ellison 2007). Increasing "connectivity" has led to the dramatic growth of the online social media market. ACNielsen (2010a) reports users of active social network sites (hereafter, SNS) increased nearly 30%, from 244.4 million in February 2009 to 314.5 million in February 2010, and they spent more than 5.5 hours on social networks in February 2010 (or 1.4 hours per week). A recent study shows that Twitter, with around 20 million visitors, enjoyed a 900% increase in visitors, up from 2 million in 2008 (ComScore 2009).
Sensing a variety of potential benefits, marketers also have ventured into the world of social media to use them for sales, customer service, promotions, and human resource tactics (Sung et al. 2010; Verna 2010). Williamson (2010) suggests that brand presence in social media attracts consumer attention, generates brand awareness and familiarity, and enables marketers to reach consumers directly. Starbucks was the tenth most popular marketer-run page on Facebook in October 2010, with nearly 16 million fans, and Coca-Cola ranked fifteenth with around 15.5 million fans (PageData 2010). Various industries and companies also have had success with Twitter. Through Twitter accounts, Dell generated $6.5 million in revenue during 2009 in orders for personal computers, accessories, and software. In turn, the number of Dell account users following Dell's tweets (the term for single Twitter posts or messages) increased 23% over the span of three months (Guglielmo 2009).
Previous research has examined virtual brand communities and other fan pages, focused mainly on the context of Facebook or other SNSs (e.g., Sung et al. 2010; Trusov, Bucklin, and Pauwels 2009). By surveying a sample of 333 virtual brand community members in Korean social network sites, Sung and colleagues (2010) show that members of brand communities have six social and psychological motives (e.g., interpersonal utility, entertainment seeking, information seeking) and that members of consumer- versus marketer-generated brand communities reveal similarities and differences in terms of the extent of their community participation. Noting that SNSs send out electronic invitations to existing members' networks, Trusov, Bucklin, and Pauwels (2009) discover that word-of-mouth (WOM) referrals in SNSs have a stronger impact on new customer acquisition than that of average advertising or media appearances.
However, there is a marked difference between the virtual brand communities on SNSs such as Facebook and those found on Twitter. Facebook users connect with their friends (boyd and Ellison 2007), and by becoming friends with other users, they gain access to their friends' content and can post information about personal issues, send birthday wishes, or comment on status updates (Parr 2010). Twitter users instead share observations on their surroundings, information about an event, or opinions regarding a certain topic (Smith 2010), and they choose what information they want to receive (e.g., news, links) and which brands to follow (Parr 2010). When they follow another account, Twitter automatically pushes single tweets by a consumer or marketer onto the users' main home page with text-like messages such as those communicated by mobile telephones (Kaplan and Haenlein 2010). A recently added feature also allows Twitter users to locate full conversations by clicking on a tweet with a chat bubble icon (Siegler 2010), similar to a feature on Facebook. Twitter further enables the public to see what people are saying and follow conversations centering on a particular topic, using hashtags (e.g., #trendingtopics, #tangled) before the word or phrase to indicate trends (boyd, Golder, and Lotan 2010), which can be useful in the context of announcing or promoting events or products and monitoring conversations among consumers (Chang 2010). By beginning a sentence with conventions such as "@username" in reply to another user or "RT" on a tweet, Twitter users both directly address specific users and refer to them by dispersing messages throughout a network of interconnected users (boyd, Golder, and Lotan 2010).
Consequently, the basic structure of each medium differs: Facebook consumers can interact with brands and other consumers as a collective but in a constrained context within bounded spaces or groups. Twitter consumers instead can interact individually with brands but often with little contact between pools of followers. Twitter followers may be more interested in what a brand wants to communicate rather than what fellow admirers of the brand have to say. Therefore, this research focuses on the social medium of Twitter rather than Facebook to examine how brands use Twitter and what they tweet.
Twitter is a relatively unexplored area for marketing researchers. From a marketer's standpoint, Twitter can be relatively faster, less expensive, and geographically unbounded (Dickey and Lewis 2009). As a real-time information network, Twitter connects consumers to the latest information about topics they find relevant and interesting (see Twitter.com). These benefits enable marketers to connect online with consumers who are motivated to engage with their brands. In a sense, Twitter enables marketers to personify brands and help build and maintain consumer relationships by engaging in conversations. Despite its fast-growing popularity and potential as a marketing communications tool, few empirical studies examine brand use and consumer interaction on Twitter though. Therefore, the primary goals of this study are as follows: (1) to examine corporate Twitter account profile pages, (2) to analyze the use of brand anthropomorphism on Twitter, and (3) to investigate marketing messages on Twitter. The findings reveal that Twitter enables brand anthropomorphism, which may influence the formation of company-customer interactions and thus encourage long-term, personal relationships while also providing an optimal vehicle for disseminating corporate information. To that end, this study uses content analysis methods to examine tweets by the top global brands with Twitter accounts in the United States.
Social Media and Twitter
Social media have evolved through Web 2.0, a term coined to describe a new wave of Internet innovation that enables users to publish and exchange content (Kaplan and Haenlein 2010). Social media encompass a wide range of electronic WOM (eWOM) forums, including blogs (e.g., the unofficial Apple weblog, Cnet blog), microblogs (e.g., Twitter), social networking sites (e.g., Facebook), creative work-sharing sites (e.g., YouTube), business networking sites (e.g., LinkedIn), collaborative websites (e.g., Wikipedia), and virtual worlds (e.g., Second Life) (Mangold and Faulds 2009). Among them, SNSs and microblogs are the most popular, accounting for 22.7% of all time spent online in the United States (ACNielsen 2010b). Superseding e-mail, SNSs have become the top spots on which Americans spend most of their time online (ACNielsen 2010b).
Twitter launched in 2006 as a free microblogging social network that enabled users to post short messages known as tweets that could be viewed by other subscribers, more commonly referred to as followers. Tweets of no more than 140 characters can be sent from and received by almost any kind of electronic equipment, including desktop computers, laptops, BlackBerrys, iPhones, and other mobile devices (Farhi 2009). Users broadcast messages to the masses, with no need to visit a particular person's profile to do so. By following other users, people automatically receive messages on their own Twitter home pages tweeted by those whom they are following. Even though some Twitter accounts are kept private, and some users require that they offer approval to people who wish to follow them, messages exchanged on this microblog are public by default, unlike status updates often restricted on SNSs such as Facebook. Thus everyone can read and comment on a Twitter message (Kaplan and Haenlein 2010).
Boosted by the appearance of celebrities such as Britney Spears, Ashton Kutcher, and Oprah Winfrey, Twitter's growth has been phenomenal. Visitors increased 1,382%, to 7 million in February 2009, up from 475,000 in February 2008, making it the fastest growing social media site for that month (McGiboney 2009). Currently, Twitter has nearly 200 million registered users and processes more than 110 million tweets per day-averaging nearly 1,300 tweets per second (Chiang 2011). Message brevity, in contrast with lengthy e-mail marketing messages, has been hailed as an asset of microblogs; it enables consumers to browse a large amount of updates efficiently (Zhao and Rosson 2009). Aided by increases in smartphone sales and the continued rollout of Internet and mobile network infrastructure, Twitter is poised for even more growth (Wauters 2010).
Twitter and New Marketing Communications
In recent years, more consumers have been using different types of social media to search for information that influences their purchasing decisions (Razorfish 2009). They thus not only extend their social networks to people they have never met in person to gather opinions about products and services but also become fans on SNSs to communicate with marketers and other peer consumers. The interconnectedness of Twitter users provides a distinctive channel for marketing communication. The structural and interactive features of Twitter, as opposed to those of a traditional corporate website, foster ongoing conversations between marketers and consumers for all three stages of the marketing process: prepurchase (i.e., marketing research), purchase (i.e., sales promotions), and postpurchase (i.e., customer services) (Kaplan and Haenlein 2010). For example, during the introduction of VIA instant coffee in October 2009, Starbucks actively listened to and participated in conversations about VIA on Twitter. This communication helped get samples of the new product into people's hands for trials, and it made it possible for consumers to move beyond their common initial reaction that instant coffee must be distasteful (Kaplan and Haenlein 2010).
Because Twitter can be used in conjunction with other media outlets, marketers potentially achieve a synergistic effect that increases brand awareness and drives more traffic to the brand or product. Stelzner (2010) suggests that marketers use social media mainly for generating exposure for their business and increasing traffic to their corporate websites, rather than for selling products and services. Heaps (2009) also indicates that distributing information through Twitter can increase traffic to company websites, which eventually allows the companies to "tell their story" better and connect at a deeper level through more detailed content and information about events and alerts. Likewise, marketers have used Twitter as a communication tool to connect directly with consumers (Williamson 2010; Zhao and Rosson 2009). Therefore, while establishing their presence on social media and engaging in more conversations, marketers ultimately try to enhance brand-consumer relationships and increase sales.
Brand Anthropomorphism on Twitter
Anthropomorphism is a psychological process of seeing the human in nonhuman forms and events (Guthrie 1993). Attributing human characteristics to nonhuman entities is pervasive among all peoples (Caporael and Heyes 1997). In marketing research, the psychological process of imbuing brands with personalities also is referred to as brand personality (Aaker 1997). As a logical extension of the anthropomorphism of nonhuman entities, research on the uses and perceptions of technology shows that people often apply social norms of reciprocity to their human-computer interactions. For example, Moon (2000) reveals that when people encounter a technology that possesses characteristics associated with human behavior, such as language, turn taking, and interactivity, they often respond by exhibiting social behaviors and making social attributions that guide their interpersonal behavior.
By creating brand presence in social media, marketers can give an extra boost to people's tendency to anthropomorphize products or brands as part of their overall marketing strategies (Aggarwal and McGill 2007). Social media enable brands to open a dialogue, provide prompt feedback, communicate with a sense of humor, and admit mistakes. Kelleher and Miller (2006) find that blog participants are more conditioned to perceive an organization's "conversational human voice" than those who read an organization's website. In addition, Cho (2006) observes that marketers post their brand or industry-related stories on blogs, which imbues the brands with personality traits based on their use of informal language. Through such brand-consumer interactions in social media, brands can foster ongoing interactions and conversations between brands and consumers, as well as among consumers (McAlexander, Schouten, and Koenig 2002).
To summarize, by having a presence on Twitter and inviting consumers to engage in the discourse, marketers can display both their personalities and the brand's human characteristics. Therefore, engaging in communication or conversations may be regarded as brand behaviors in the virtual community. As Fournier (1995) has suggested, these outcomes help generate trait inferences that collectively summarize consumer perceptions about the brand, elevating the brand to the status of a relationship partner. We therefore seek to measure the extent to which global brands encourage anthropomorphism on their Twitter accounts. When consumers interact with representatives of that brand, they likely rely on social relationships to guide their interactions (Aggarwal 2004). Therefore, we ask:
RQ1: To what extent do global brands use human representatives on Twitter?
In addition, to examine whether brands engage in conversations with consumers, we analyze tweet types and the use of both personal pronouns and verbs in the imperative form. Personal pronouns implicitly claim relationships between brands and their audiences (Pollach 2005). The use of first-person pronouns such as "I" or "we" usually helps establish relationships with readers by stating information as beliefs rather than facts. The use of second-person pronouns such as "you" and "your" typically invites readers into the conversations (Pollach 2005). Both methods therefore reduce the impersonality of mass communication (Fairclough 1989). Pollach (2005) and Insche (2008) also suggest that verbs in the imperative form involve readers in the discourse. Thus, identification of these variables may indicate whether brands are engaging and invite consumers into conversations, leading to our second research question:
RQ2: To what extent do global brands engage in conversations with consumers?
The use of nonverbal cues in messages also helps brands reveal verbal nuances or emotions. Computer-mediated communication (CMC) researchers note the lack of social context cues, such as verbal nuances (e.g., gaze, body language), physical context (e.g., meeting sites, seating arrangements), and observable information about social characteristics (e.g., age, gender, race) during interpersonal communication in an online environment (Brown, Broderick, and Lee 2007). To overcome this impersonality, CMC conventions have emerged to express user actions, emotions, moods, humor, sarcasm, and irony (Pena and Hancock 2006). Well-known conventions include sideways emoticons formed with keyboard characters resembling facial expressions (e.g., :) or -_0), abbreviations (e.g., "OMG" for "oh my god"), repeated punctuation (e.g., "Woo hoo, Friday!!!!!", "Is it really raining again?!?!?!"), and intentional misspellings (e.g., "sleeeeeping," "riiiight") for emphasis (Nastri, Pena, and Hancock 2006). With such nonverbal cues, brands also can convey emotions and verbal subtleties to consumers.
RQ3: To what extent do global brands use nonverbal cues to reveal verbal nuances or emotions?
Finally, we explore these research questions across product/service categories and tweet types (i.e., original tweets, replies, and retweets).
RQ4: How does brand anthropomorphism differ by (a) business category and (b) tweet type?
Information Types
The emergence of the Web has enabled marketers to provide unlimited amounts of information and increase consumer brand knowledge using various methods, such as banners, pop-ups, and corporate websites. Twitter and other social media in general also are emerging as an alternative marketing communications tool for distributing information in seconds. Resnik and Stern (1977) identify 14 evaluative information criteria of product characteristics, including price, quality, performance, components, availability, and special offers. Choi and colleagues (2006) update and refine Resnik and Stern's approach for the new advertising information environment, adding five informational cues: website address, toll-free number, disclaimer, mail-in address, and brand name/advertisers. By propagating diverse types of industry- and brand-related information, marketers can affect consumer perceptions of their brands and increase sales. The amount and types of information also may differ across product categories (Abernethy and Franke 1996). For example, ads for durable goods tend to contain more information than those for nondurables, and ads for rational products convey more information than those for emotional products (e.g., Choi et al. 2006). Therefore, we developed the following questions:
RQ5: What types of information do global brands disseminate frequently on Twitter?
RQ6: How do the types of information differ by (a) business category and (b) tweet type?
Method
Sample
We selected our sample from the 2010 "World's Most Valuable Brands" list, produced by Interbrand with BusinessWeek. First, we examined the main homepages of all 100 global brands listed to locate their official Twitter account names. If we found none, we searched for the company name on Twitter's People Search. If there were multiple accounts, we selected the one clearly identified as the company's official Twitter feed and the one with the largest number of followers. As a result, we identified 77 Twitter accounts for consideration. Second, we singled out active Twitter accounts that had tweeted (1) more than 500 times and (2) more than 100 times during the two-month period from October 1 to November 30, 2010. This method yielded 44 Twitter accounts. Third, we randomly selected 50 tweets from each of the 44 identified accounts, which yielded 2,200 unique tweets.
Coding Scheme
We employ content analysis as our primary research method. We coded brand anthropomorphism and information types along several dimensions. In particular, we determined whether human representatives appeared on the main homepages (e.g., profile pictures) and bio sections of the brands. We also analyzed the use of personal pronouns, verbs in the imperative form (Insche 2008; Pollach 2005), and types of actions. To generate potential items for action types, we randomly selected and analyzed tweets that were not included in the main study. After identifying a list of potential items, we conducted a pilot study to ensure that they fit the scope of this study. These tactics resulted in five categories for the main study: asking for relationship, asking to provide feedback and contact customer service, redirecting to other media, sales, and eWOM. The categories for the nonverbal cues came from prior literature (i.e., Nastri, Pena, and Hancock 2006).
Regarding information types, we adopted the 14-item typology (Resnik and Stern 1977), together with the 5 additional informational cues developed (Choi et al. 2006) to reflect recent changes in the media environment. We also added a few information types to the coding instrument, according to the results of the pilot study. This process resulted in five categories: product-related, company-related, source, redirecting, and simple brand name mentions. The coding for each measurement item used two nominal categories: 1 indicated it was available or depicted, and 2 indicated not available or not depicted, as we summarize in Table 1.
Table 1. Operationalization of Brand Anthropomorphism and Information Types
|
Dimensions |
Operationalization |
|
|||
|
BRAND ANTHROPOMORPHISM |
|
||||
|
Human representatives |
Pictures of marketers, celebrity endorsers, or their information (name, contact information), as well as signature of marketers on tweets |
|
|||
|
Personal pronouns |
First-person (e.g., "I," "my," "me," "myself," "we," "us," "our") |
|
|||
|
Second-person (e.g., "you," "your," "yours," "yourself") |
|
||||
|
Verbs in imperative forms |
Asking for Relationship |
|
|||
|
|
Follow, stay tuned, become a fan on other social media, sign up/register, join us at an event, support the company (e.g., vote for us) |
||||
|
Asking for Feedback & Contact Customer Service |
|
||||
|
|
Let us know what you think, submit your suggestions, e-mail us, send us a direct message, text us, contact customer service |
|
|||
|
Redirecting to Other Media |
|
||||
|
|
Check out the links, watch new ads, learn about something |
|
|||
|
Sales |
|
||||
|
|
Use/purchase products, preorder, download, enter sweepstakes or contest |
|
|||
|
eWOM |
|
||||
|
|
Tweet, share tweets, retweet |
|
|||
|
Nonverbal cues |
Abbreviations (e.g., LOL, OMG) |
|
|||
|
Emoticons (e.g., :), :( ) |
|
||||
|
Repeated punctuation (e.g., !!!, ...) |
|
||||
|
Capitalization (e.g., FOLLOW US, ) |
|
||||
|
Sound mimicking (e.g., Ahhhhhh, ooopsy) |
|
||||
|
INFORMATION TYPES |
|
||||
|
Product |
Price, quality, performance, components, availability, special offers, taste, packaging, nutrition, new ideas, download |
|
|||
|
Company |
Job-related, sales-related, sponsorship/affiliation, investor relations, education |
|
|||
|
Source |
Independent research, company-sponsored research, executives, celebrities, media |
|
|||
|
Redirecting |
800 number, URL, mail-in/e-mail address |
|
|||
|
Simple brand mention |
Brand name |
|
|||
Coding Procedure and Intercoder Reliability
Four coders analyzed each tweet's content. They first reviewed coding categories, previewed samples of tweets, and practiced using the coding scheme. Unclear and disputed items were clarified, with relevant changes if necessary. The coders then conducted a pilot test on 50 tweets of two companies for the reliability test. Intercoder reliability, computed as the percentage of agreement, reached 95.51% on average overall, ranging from 87.69% to 100%. Information types had the lowest reliability at 87.69%, ranging from 84% to 92%, followed by action types of verbs in the imperative (92.33%) and repeated punctuation of nonverbal cues (92.33%). We resumed the coding process following the reliability test. Each coder received 550 tweets; we randomly selected 11 brands for each coder.
Results
Sample Characteristics
Of the 44 brands, 31 (70%) addressed their presence on Twitter through their brands' website. The U.S.-based brands were dominant, with more than 59% (26 brands), followed by Japan (9%, 4 brands) and Germany (6%, 3 brands). In Table 2 we list all brands analyzed, which averaged 142,227 followers each. Google was the top brand with 2,604,020 followers, followed by Starbucks (1,150,803), MTV (963,911), and BlackBerry (173,888). On average, the brands also followed 9,285 other Twitter users. Starbucks ranked first, with 80,426 users, whereas Oracle followed only 39 users.
Table 2. Study Brands (total followers)
|
Brand |
Country of Origin |
Category |
Total Followers |
Total Following |
Total Tweets |
Oct. 1-Nov. 30, 2010 |
|
|
U.S. |
Internet Services |
2,604,020 |
306 |
1,977 |
141 |
|
Starbucks |
U.S. |
Restaurants |
1,150,803 |
80,426 |
6,274 |
366 |
|
MTV |
U.S. |
Media |
963,911 |
2,796 |
11,472 |
2,047 |
|
BlackBerry |
Canada |
Electronics |
173,888 |
1,933 |
2,491 |
902 |
|
Coca-Cola |
U.S. |
Beverages |
166,071 |
65,595 |
18,794 |
3,157 |
|
Microsoft |
U.S. |
Computer Software |
149,040 |
289 |
2,298 |
326 |
|
Disney |
U.S. |
Media |
98,341 |
75 |
1,182 |
112 |
|
American Express |
U.S. |
Financial Services |
78,251 |
8,414 |
4,962 |
911 |
|
H&M |
Sweden |
Apparel |
77,424 |
52 |
1,235 |
378 |
|
McDonald's |
U.S. |
Restaurants |
68,078 |
9,013 |
4,178 |
711 |
|
Yahoo! |
U.S. |
Internet Services |
51,667 |
1,378 |
5,761 |
852 |
|
Pepsi |
U.S. |
Beverages |
48,617 |
39,697 |
2,068 |
186 |
|
Nokia |
Finland |
Electronics |
46,145 |
71 |
1,104 |
230 |
|
Nintendo |
Japan |
Electronics |
42,628 |
1,271 |
1,080 |
256 |
|
Cisco |
U.S. |
Business Services |
42,464 |
2,792 |
3,154 |
256 |
|
Ford |
U.S. |
Automotive |
41,312 |
34,468 |
6,165 |
254 |
|
Pizza Hut |
U.S. |
Restaurants |
39,033 |
26,856 |
1,707 |
106 |
|
Nike |
U.S. |
Sporting Goods |
37,446 |
95 |
6,654 |
2,275 |
|
Lancôme |
France |
FMCG |
36,032 |
10,519 |
900 |
285 |
|
Adobe |
U.S. |
Computer Software |
32,220 |
241 |
3,240 |
365 |
|
Samsung |
South Korea |
Electronics |
31,631 |
22,433 |
7,698 |
876 |
|
Gucci |
Italy |
Luxury |
31,146 |
186 |
579 |
131 |
|
Toyota |
Japan |
Automotive |
29,732 |
18,741 |
1,850 |
197 |
|
Sony |
Japan |
Electronics |
27,482 |
2,221 |
7,538 |
848 |
|
Oracle |
U.S. |
Business Services |
25,445 |
39 |
1,483 |
103 |
|
KFC |
U.S. |
Restaurants |
20,125 |
20,137 |
2,220 |
228 |
|
eBay |
U.S. |
Internet Services |
19,509 |
250 |
1,288 |
297 |
|
Accenture |
U.S. |
Business Services |
17,166 |
663 |
1,028 |
149 |
|
Audi |
Germany |
Automotive |
15,415 |
11,848 |
1,521 |
339 |
|
Smirnoff |
U.K. |
Alcohol |
13,854 |
15,222 |
1,695 |
380 |
|
L'Oréal |
France |
FMCG |
9,845 |
676 |
1,508 |
139 |
|
Volkswagen |
Germany |
Automotive |
9,369 |
3,465 |
752 |
118 |
|
Honda |
Japan |
Automotive |
8,182 |
6,272 |
2,358 |
219 |
|
Avon |
U.S. |
FMCG |
8,039 |
473 |
1,679 |
339 |
|
Hyundai |
South Korea |
Automotive |
6,958 |
2,642 |
1,893 |
280 |
|
Gillette |
U.S. |
FMCG |
6,596 |
7,150 |
4,256 |
514 |
|
Thomson Reuters |
Canada |
Media |
6,194 |
83 |
1,313 |
168 |
|
Siemens |
Germany |
Diversified |
5,769 |
1,489 |
2,835 |
212 |
|
GE |
U.S. |
Diversified |
5,737 |
2,506 |
2,462 |
114 |
|
Sprite |
U.S. |
Beverages |
4,369 |
87 |
1,400 |
265 |
|
Citi |
U.S. |
Financial Services |
2,801 |
3,031 |
3,222 |
655 |
|
Xerox |
U.S. |
Electronics |
2,633 |
531 |
2,271 |
112 |
|
UPS |
U.S. |
Transportation |
1,508 |
110 |
1,062 |
256 |
|
Zurich |
Switzerland |
Financial Services |
1,081 |
1,997 |
1,168 |
180 |
Notes: FMCG = fast moving consumer goods category.
Regarding the total number of tweets, we found an average of 3,222, with a maximum of 18,794 (Coca-Cola) and a minimum of 579 (Gucci). Coca-Cola was also the top account in terms of tweeting activity during the sampling period, with 3,157 tweets (approximately 53/day), followed by Nike and MTV with 2,275 (38/day) and 2,047 (35/day) tweets, respectively. Brand logos were frequently used in both profile pictures (86%, 38 brands) and background images (43%, 19 brands), whereas product images were less frequent in profile pictures (6%, 3 brands) or background images (30%, 13 brands). In the user information section on the right-hand side of the page, brands provided information such as website links and company biographies. In the website section, 37 brands (84%) posted Web addresses for their official corporate website, and 6 brands (16%) listed their other social media sites.
Brand Anthropomorphism
We identified 24 brands with a human representative on their Twitter accounts. Two included the images of celebrity endorsers, and 14 brands provided marketer information such as name, picture, contact information, and company position. Fifteen brands showed marketer signatures on their tweets (300 counts). Coca-Cola left a signature on 96% of its tweets, followed by BlackBerry (90%), Pepsi (84%), and Sprite (72%).
Of the 2,200 tweets, 47.5% (1,046) were original marketer-initiated messages, whereas 37.5% (825) were replies in which marketers responded to consumers. Retweets were the least common (15%, 329), when marketers would disseminate what other tweeters said. Approximately 54% (1,178) of the tweets contained personal pronouns: 736 with second-person and 650 with first-person pronouns. Verbs in the imperative form accounted for 26.8% (589) of the total tweets. Among them, 32.3% redirected readers to other media, 26.5% asked to initiate relationships (i.e., follow, sign up for events), 19.4% asked followers to contact the brand for service and feedback, 8.7% asked for sales, and 8.1% asked to broadcast a message through eWOM.
Our results further indicate that 564 tweets (25.6%) contained at least one type of nonverbal cue. Abbreviations were the most frequent (315), followed by emoticons (123), repeated punctuation (93), capitalization (63), and words mimicking sounds (52).
The use of human representatives was most common in automotive industries (83.3%), followed by beverages/restaurants (62.5%), electronics/computer-related (50%), services (46.2%), and fashion/beauty (42.9%) sectors. As we show in Table 3, the use of pronouns varied by business category. Pronouns were prevalent in the fashion/beauty category (63%), followed by beverages/restaurants (61.1%), automotive (56.7%), electronics/computer-related (48.8%), and service (45.8%) industries (c2 = 43.69, p < .01). The use of imperative verbs was most frequent in services (32.8%), followed by fashion/beauty (26.6%), electronics/computer-related (26.4%), automotive (23.7%), and beverages/restaurants (20%) sectors (c2 = 22.8, p < .01). Nonverbal cues appeared in messages from beverages/restaurants companies (34%), followed by services (29.2%), automotive (26%), electronics/computer-related (23%), and fashion/beauty (12.9%) firms (c2 = 50.91, p < .01). Beverage/restaurant brands used more emoticons (40.4%) than did fashion/beauty (26.7%), electronics/computer-related (21.6%), and services (11.6%). This difference is significant: c2 = 44.86 (p < .01).
In addition, the use of personal pronouns was most prevalent in replies (72.4%), followed by original tweets (42.3%) and retweets (42.2%) (c2 = 187.92, p < .01). First-person pronouns appeared most in replies (59.8%), followed by retweets (56.8%) and original tweets (48.4%) (c2 = 13.48, p < .01). Second-person pronouns were identified in 72% of the reply tweets, 56.6% of original tweets, and 40.3% of retweets (c2 = 59.02, p < .01). No significant difference emerged for the use of verbs in the imperative form across tweet types (c2 = 3.27, p = .20), but reply messages exhibited a higher tendency to ask consumers to contact the brand for feedback and customer service, whereas original tweets redirected customers to other media venues. Replies contained more nonverbal cues (emoticons, 35.4%) than other types of tweets (retweets = 21.9%; original tweets = 19.1%; c2 = 66.93, p < .01). Abbreviations were most frequent in retweets (70.8%) and original tweets (67.5%) but accounted for only 44% of replies (c2 = 34.16, p < .01).
Table 3. Brand Anthropomorphism by Business Categories and Tweet Types
|
|
Automotive |
Electronics/Computer |
Services |
Fashion/ Beauty |
Beverages/ Restaurants |
Original Tweets |
Replies |
Retweets |
Total |
|||||||||
|
Personal Pronouna,c |
170 |
56.7% |
244 |
48.8% |
298 |
45.8% |
214 |
61.1% |
252 |
63.0% |
442 |
42.3% |
597 |
72.4% |
139 |
42.2% |
1178 |
|
|
|
First-persona,c |
101 |
59.4% |
111 |
45.5% |
170 |
57.0% |
131 |
61.2% |
137 |
54.4% |
214 |
48.4% |
357 |
59.8% |
79 |
56.8% |
650 |
|
Second-persona,c |
91 |
53.5% |
159 |
65.2% |
194 |
65.1% |
120 |
56.1% |
172 |
68.3% |
250 |
56.6% |
430 |
72.0% |
56 |
40.3% |
736 |
|
|
Imperative Verbsa |
71 |
23.7% |
132 |
26.4% |
213 |
32.8% |
93 |
26.6% |
80 |
20.0% |
296 |
28.3% |
203 |
24.6% |
90 |
27.4% |
589 |
|
|
|
Redirectingc |
26 |
36.6% |
50 |
37.9% |
66 |
31.0% |
29 |
31.2% |
19 |
23.8% |
124 |
41.9% |
32 |
15.8% |
34 |
37.8% |
190 |
|
Relationshipb |
18 |
25.4% |
35 |
26.5% |
43 |
20.2% |
32 |
34.4% |
28 |
35.0% |
81 |
27.4% |
51 |
25.1% |
24 |
26.7% |
156 |
|
|
Contact branda,c |
12 |
16.9% |
14 |
10.6% |
64 |
30.0% |
7 |
7.5% |
17 |
21.3% |
15 |
5.1% |
96 |
47.3% |
3 |
3.3% |
114 |
|
|
Salesa,c |
2 |
2.8% |
6 |
4.5% |
23 |
10.8% |
15 |
16.1% |
5 |
6.3% |
36 |
12.2% |
5 |
2.5% |
10 |
8.6% |
51 |
|
|
eWOMc |
4 |
5.6% |
15 |
11.4% |
16 |
7.5% |
9 |
9.7% |
4 |
5.0% |
29 |
9.8% |
7 |
3.4% |
12 |
13.3% |
48 |
|
|
Nonverbal Cuesa,c |
78 |
26.0% |
115 |
23.0% |
190 |
29.2% |
45 |
12.9% |
136 |
34.0% |
200 |
19.1% |
292 |
35.4% |
72 |
21.9% |
564 |
|
|
|
Abbreviationa,c |
58 |
74.4% |
60 |
51.7% |
131 |
68.9% |
24 |
53.3% |
42 |
30.9% |
135 |
67.5% |
129 |
44.0% |
51 |
70.8% |
315 |
|
Emoticona,c |
9 |
11.5% |
25 |
21.6% |
22 |
11.6% |
12 |
26.7% |
55 |
40.4% |
10 |
5.0% |
110 |
37.5% |
3 |
4.2% |
123 |
|
|
Repeated punctuationb,c |
13 |
16.7% |
24 |
20.7% |
20 |
10.5% |
12 |
26.7% |
24 |
17.6% |
33 |
16.5% |
45 |
15.4% |
15 |
20.8% |
93 |
|
|
Capitalizationb,c |
3 |
3.8% |
14 |
12.1% |
28 |
14.7% |
0 |
0.0% |
18 |
13.2% |
36 |
18.0% |
23 |
7.8% |
4 |
5.6% |
63 |
|
|
Sound mimicking |
5 |
6.4% |
10 |
8.6% |
11 |
5.8% |
5 |
11.1% |
21 |
15.4% |
15 |
7.5% |
34 |
11.6% |
3 |
4.2% |
52 |
|
|
Total |
300 |
500 |
650 |
350 |
400 |
1046 |
825 |
329 |
|
|||||||||
ac2 test, p < .01 for business category.
bc2 test, p < .05 for business category.
cc2 test, p < .01 for tweet type.
Information Types
Informational cues were identified in 73.6% of the total tweets, including brand names (68.8%), redirecting informational cues (65.9%), product-related cues (26.3%), and company-related cues (6.1%).
The use of these informational cues varied by business category (c2 = 138.88, p < .01), accounting for 81.8% of electronics/computer-related, 81.2% of services, 76.3% of fashion/beauty, 69.7% of automotive, and 51.8% of beverages/restaurants messages. Brand name was most frequently found in electronics/computer brand messages (72.4%), followed by automotive (72.2%), beverages/restaurants (71%), fashion/beauty (68.2%), and services (62.9%) (c2 = 12.52, p < .05). Redirecting cues appeared mainly in services (83.3%), followed by electronics/computer (68.2%), automotive (66%), fashion/beauty (49.4%), and beverages/restaurants (37.6%) industries. The difference is significant, at c2 = 177.85 (p < .01). Among all informational cues, brand name was mentioned most frequently, followed by redirecting cues, product-related cues, source, and company-related cues. In contrast, services contained more redirecting cues (83.3%), followed by brand name (62.9%), product-related (25.4%), source (11.6%), and company-related (9.1%) information.
Informational cues varied by tweet type (c2 = 508.16, p < .01), accounting for 90.9% of retweets, 89.8% of original tweets, and 46.3% of replies. Brand name was the most frequent information across tweet types, but it appeared most in retweets (75.9%), followed by original tweets (66.9%) and replies (66.2%) (c2 = 9.66, p < .01). Redirecting cues were often used in original tweets (77.5%) and retweets (79.6%) but accounted for only 26.4% of replies (c2 = 346.0, p < .01). Product-related information emerged most in replies (34%), followed by original tweets (24%) and retweets (23.7%) (c2 = 15.44, p < .01). Company-related information appeared most in original tweets (7.5%), followed by retweets (5.7%) and replies (3.1%) (c2 = 8.92, p < .05). Source information accounted for 15.4% of retweets, 14.1% of original tweets, and 1.6% of replies (c2 = 47.96, p < .01).
Table 4. Information Types by Business Categories and Tweet Types
|
|
Automotive |
Electronics/ Computer |
Services |
Fashion/ Beauty |
Beverages/ Restaurants |
Original Tweets |
Replies |
Retweets |
Total |
|||||||||
|
Informationa,c |
209 |
69.7% |
409 |
81.8% |
528 |
81.2% |
267 |
76.3% |
207 |
51.8% |
939 |
89.8% |
382 |
46.3% |
299 |
90.9% |
1620 |
|
|
|
Brand nameb,c |
151 |
72.2% |
296 |
72.4% |
332 |
62.9% |
182 |
68.2% |
147 |
71.0% |
628 |
66.9% |
253 |
66.2% |
227 |
75.9% |
1108 |
|
Redirectinga,c |
138 |
66.0% |
279 |
68.2% |
440 |
83.3% |
132 |
49.4% |
78 |
37.7% |
728 |
77.5% |
101 |
26.4% |
238 |
79.6% |
1067 |
|
|
Producta,c |
45 |
21.5% |
91 |
22.2% |
134 |
25.4% |
93 |
34.8% |
63 |
30.4% |
225 |
24.0% |
130 |
34.0% |
71 |
23.7% |
426 |
|
|
Sourcea,c |
16 |
7.7% |
37 |
9.0% |
61 |
11.6% |
60 |
22.5% |
10 |
4.8% |
132 |
14.1% |
6 |
1.6% |
46 |
15.4% |
184 |
|
|
Companya,d |
11 |
5.3% |
24 |
5.9% |
48 |
9.1% |
12 |
4.5% |
4 |
1.9% |
70 |
7.5% |
12 |
3.1% |
17 |
5.7% |
99 |
|
ac2 test, p < .01 for business category.
bc2 test, p < .05 for business category.
cc2 test, p < .01 for tweet type.
dc2 test, p < .05 for tweet type.
Discussion and Implications
Social media in general have transformed the exchange of information and induced behavioral changes in consumer-brand relationships. To attract consumer attention and generate brand awareness and familiarity, an increasing number of marketers have created brand presences through social media. In particular, Twitter is a fast-growing social medium that has attracted millions of users. Despite the suggestions of marketing researchers and practitioners that Twitter enables marketers to provide customer services, increase sales, and connect with consumers (e.g., Kaplan and Haenlein 2010; Sung et al. 2010), little empirical research has examined interactions between consumers and brands on Twitter. The interactive features of Twitter enable brands to listen to consumers, initiate a dialogue, reply to consumers, and communicate with a sense of humor-all of which represent "behaviors" enacted by the brand (Fournier 1995). Along with this postulate, we focus on how top global brands use Twitter to engage with consumers. This study thereby advances understanding of brand anthropomorphism and information distribution in a virtual environment.
In particular, the findings establish that about half of the analyzed brands had human representatives. As suggested by Aggarwal (2004), seeing a more human image, rather than interacting with faceless accounts, may encourage consumers to fall back on social relationships to retain human-to-human interactions. Furthermore, replying to messages and using personal pronouns or imperative verb forms imply that brands are listening to their consumers and want to invite them to engage in discourse. Although relating emotions and verbal nuances were less significant in our study, the identification of such nonverbal cues indicates that brands still attempt to reduce impersonality. Our investigation of imperative verbs further suggests that global brands use Twitter as a tool to initiate and maintain relationships with consumers. By telling consumers to "follow the brand," "come by the booth," "join us at the event," or "sign up" for a planned occasion, brands attempt to initiate relationships. In doing so, they not only save marketing costs but also may develop and cultivate long-term consumer-brand relationships.
Consumers expect a dialogue in social media, in which brands listen to what they have to say rather than simply pushing promotional marketing messages without taking into account what customers think, feel, and want (Millward Brown 2010). By communicating and interacting with consumers, marketers can convey both their personalities and the brand's human characteristics. Their engaging communication or conversations thus indicate brand behaviors in the virtual community. This process may help generate trait inferences that collectively summarize consumer perceptions about brands while also elevating the brand to the status of a contributing, relational partner (Fournier 1995).
Regarding information type, our results suggest that product- or company-related information is less frequent, whereas brand names appear often in tweets, followed in frequency by redirecting cues. These findings align with previous research that shows Twitter is useful in generating exposure to brands and driving more traffic to brand websites, rather than for selling products or services (e.g., Heaps 2009; Stelzner 2010). That is, Twitter can generate buzz around brands and help them engage with current and future consumers. In addition, redirecting consumers to different channels was a common tactic; all the brands with Twitter accounts provided their corporate or social media web addresses on Twitter, and tweets often contained embedded links. This practice may be similar to traditional banner ads, which motivate consumers to click the URLs to obtain more detailed information (Kaplan and Haenlein 2010). Our investigation of imperative verbs also suggests that brands ask consumers to visit or refer them to other media, such as corporate websites or social media sites, as well as to watch televised commercials.
The variables we examined tend to vary by business category but do not exhibit any significant patterns. Regarding tweet types though, more anthropomorphism variables occurred in replies than in original tweets and retweets. Informational cues were more common in original tweets and retweets. This finding may be related to the function of each tweet type. With replies, marketers respond directly to consumers, answering their questions. Therefore, brands are likely to use more personal pronouns, imperative verbs, and nonverbal cues to emphasize continuing relationships. However, with original tweets (i.e., when marketers initiate a message) and retweets (when marketers forward other tweeters' messages), the marketers are trying to target an unspecific number of consumers. Thus, these tweet types focus more on brand information, sources, or references.
In summary, this research fills a gap in extant literature related to brands' use of social media in the Web 2.0 environment, underscoring the importance of Twitter as a vehicle to build and maintain relationships between consumers and brands. We argue that by attributing human characteristics to brands through conversations and interactions on Twitter, marketers animate and humanize their brands. Such an approach may influence consumers to elevate the status of the brand from a passive object to a full-fledged relationship partner (Fournier 1995). This transformation may result in unique brands with distinctive images and meaningful personalities, which should increase consumer preference for their brands and a sense of differentiation from competitors (Aaker 1997). By using Twitter in conjunction with other media outlets, marketers can attract consumer attention and drive more traffic to corporate websites or SNSs. Thus, marketers reach consumers with personally relevant information while increasing and shaping consumer brand knowledge using cohesive messages across various media venues.
Limitations and Further Research
Our exploratory content analysis reveals some interesting trends, but its generalizability is limited. First, the sample is restricted to 44 global brands and 50 tweets related to each brand, conducted over only a two-month period. Therefore, additional content analytic research should expand on these findings by examining a more extensive set of brands and tweets. Future studies would do well to incorporate tandem analyses of follower/tweet data to ascertain whether any brand relationships exist.
Second, though content analytic studies provide a wealth of descriptive data about social media marketing, research also needs to help marketers and brand managers understand how consumers respond to such practices. Further research into Twitter's potential for serving as a brand communication tool should adopt methodologies that are more conducive to determining consumer perceptions and reactions, such as surveys and in-depth interviews. Academic research focusing specifically on Twitter, to explain the motivations for consumers' decisions to engage with a brand, has been scarce. Understanding these motivations and their possible variations is essential for marketers attempting to build up their presence on Twitter. Further research should investigate how information technology not only assists and enhances relationship formation but also repairs damage-and why consumers might terminate relationships by opting out of such memberships (Sheth and Parvatiyar 1995).
Third, the increasing popularity of cross-media promotions among advertisers demands studies that focus on this topic. Additional research might explore how consumers integrate information on social media with information on other online and offline media, such as magazines and televised ads.
References
Aaker, Jennifer (1997), "Dimensions of Brand Personality," Journal of Marketing Research, 34 (August), 347-56.
Abernethy, Averby M. and George R. Franke (1996), "The Information Content of Advertising: A Meta-Analysis," Journal of Advertising, 25 (2), 1-17.
ACNielsen (2010a), "Global Audience Spends Two Hours More a Month on Social Networks Than Last Year," March 19, blog.nielsen.com, http://blog.nielsen.com/nielsenwire/global/global-audience-spends-two-hours-more-a-month-on-social-networks-than-last-year/.
--- (2010b), "What Americans Do Online: Social Media and Games Dominate Activity," August 2, blog.nielsen.com, http://blog.nielsen.com/nielsenwire/online_mobile/what-americans-do-online-social-media-and-games-dominate-activity/.
Aggarwal, Pankaj (2004), "The Effects of Brand Relationship Norms on Consumer Attitudes and Behavior," Journal of Consumer Research, 31 (1), 87-101.
--- and Ann L. McGill (2007), "Is That Car Smiling at Me? Schema Congruity as a Basis for Evaluating Anthropomorphized Products," Journal of Consumer Research, 34, 468-79.
boyd, danah m. and Nicole B. Ellison (2007), "Social Network Sites: Definition, History, and Scholarship," Journal of Computer-Mediated Communication, 13, 210-30.
---, Scott Golder, and Gilad Lotan (2010), "Tweet, Tweet, Retweet: Conversational Aspects of Retweeting on Twitter," Proceedings of the 43rd Hawaii International Conference on Social Systems. IEEE: Kauai, HI.
Brown, Jo, Amanda J. Broderick, and Nick Lee (2007), "Word of Mouth Communication within Online Communities: Conceptualizing the Online Social Network," Journal of Interactive Marketing, 21 (3), 2-20.
Caporael, Linnda R. and Cecilia M. Heyes (1997), "Why Anthropomorphize Folk Psychology and Other Stories," in Anthropomorphism, Anecdotes, and Animals, R.W. Mitchell, N.S. Thomson, and H.L. Miles, eds. Albany, NY: University of New York Press, 59-73.
Chang, Hsia C. (2010), "A New Perspective on Twitter Hashtag Use: Diffusion of Innovation Theory," Proceedings of the American Society for Information Science and Technology, 47 (1), 1-4.
Chiang, Oliver (2011), "Twitter Hits Nearly 200M Accounts, 110 M Tweets Per Day, Focuses On Global Expansion," January 9, forbes.com, http://www.forbes.com/sites/oliverchiang/2011/01/19/twitter-hits-nearly-200m-users-110m-tweets-per-day-focuses-on-global-expansion/.
Cho, Soyoen (2006), "Interpersonal Communication between Brands and Consumers: A Self-Presentation Study of Corporate Blogs," New Media Conference, University of Minnesota.
Choi, Sejung M., Nora J. Rifon, Carrie S. Trimble, and Bonnie B. Reece (2006), "Information Content in Magazine, Television and Web Advertising: A Comparison and Update," Marketing Management Journal, 16 (1), 188-203.
ComScore (2009), The 2009 U.S. Digital Year in Review, February.
Dickey, Irene J. and William F. Lewis (2009), "Consumer Generated Media: Evolving Marketing Opportunity for Consumer Engagement," Society for Marketing Advances Proceedings, 181-85.
Dwyer, Robert F., Paul H. Schurr, and Sejo Oh (1987), "Developing Buyer-Seller Relationships," Journal of Marketing, 51, 11-27.
Fairclough, Norman (1989), Language and Power. London: Pearson.
Farhi, Paul (2009), "The Twitter Explosion," American Journalism Review, 31 (3), 26-31.
Fournier, Susan (1995), "The Brand-As-Relationship Partner: An Alternative View of Brand Personality," in A Brand as a Character, a Partner and a Person: Three Perspectives on the Question of Brand Personality, Jennifer Aaker and Susan Fournier, eds. Advances in Consumer Research, Vol. 22, Provo, UT: Association for Consumer Research, 391-95.
Guglielmo, Connie (2009), "Dell Rings up $6.5 Million in Sales Using Twitter," December 8, Bloomberg.
Guthrie, Stewart (1993), Faces in the Clouds: A New Theory of Religion. New York: Oxford.
Haythornthwaite, Caroline (2005), "Social Networks and Internet Connectivity Effects," Information, Communication and Society, 8 (2), 125-47.
Heaps, Darrell (2009), "Twitter: Analysis of Corporate Reporting Using Social Media," Corporate Governance Advisor, 17 (6), 18-22.
Insch, Andrea (2008), "Online Communication of Corporate Environmental Citizenship: A Study of New Zealand's Electricity and Gas Retailers," Journal of Marketing Communications, 14 (2), 139-53.
Kaplan, Andreas M. and Michael Haenlein (2010), "The Early Bird Catches the News: Nine Things You Should Know about Micro-Blogging," Business Horizons, 54 (2), 105-13.
Kelleher, Tom and Barbara M. Miller (2006), "Organizational Blogs and the Human Voice: Relational Strategies and Relational Outcomes," Journal of Computer-Mediated Communication, 11 (2), 395-414.
Mangold, W. Glynn and David J. Faulds (2009), "Social Media: The New Hybrid Element of the Promotion Mix," Business Horizons, 52, 357-65.
McAlexander, James H., John W. Schouten, and Harold F. Koenig (2002), "Building Brand Community," Journal of Marketing, 66 (January), 38-54.
McGiboney, Michelle (2009), "Twitter's Tweet Smell of Success," March 18, blog.nielsen.com, http://blog.nielsen.com/nielsenwire/online_mobile/twitters-tweet-smell-of-success/.
Millward Brown (2010), "Brands Need a More Human Face to Avoid Being ‘Foreigners' in Social Media: Highlights Research," November 18, http://www.millwardbrown.com.
Moon, Youngme (2000), "Intimate Exchanges: Using Computers to Elicit Self-Disclosure from Consumers," Journal of Consumer Research, 26 (4), 323-39.
Nastri, Jacqueline, Jorge Peña, and Jeffrey T. Hancock (2006), "The Construction of Away Messages: A Speech Act Analysis," Journal of Computer-Mediated Communication, 11 (4).
PageData (2010), "PageData: Independent, Accurate Page Metrics and Trends from Inside Network," http://pagedata.insidefacebook.com.
Parr, Ben (2010), "Facebook, Twitter and The Two Branches of Social Media [OP-ED]," October 11, mashable.com, http://mashable.com/2010/10/11/facebook-twitter-social/.
Pena, Jorge and Jeffrey T. Hancock (2006), "An Analysis of Socioemotional and Task Communication in Online Multiplayer Video Games," Communication Research, 33 (1), 92-109.
Pollach, Irene (2005), "Corporate Self-presentation on the WWW: Strategies for Enhancing Usability, Credibility and Utility," Corporate Communications: An International Journal, 10 (4), 285-301.
Razorfish (2009), August, http://feed.razorfish.com/feed09/the-data.
Resnik, Alan and Bruce L. Stern (1977), "An Analysis of Information Content in Television Advertising," Journal of Marketing, 41, 50-53.
Sheth, Jagdish N. and Atul Parvatiyar (1995), "Relationship Marketing in Consumer Markets: Antecedents and Consequences" Journal of the Academy of Marketing Science, 23 (4), 255-71.
Siegler, MG (2010), "New Twitter Now Showing Full Conversations in the Side Pane," November 20, techcrunch.com, http://techcrunch.com/2010/11/20/twitter-conversations-pane/.
Smith, Aaron (2010), "Who Tweets?" December 9, pewresearch.org, http://pewresearch.org/pubs/1821/twitter-users-profile-exclusive-examination.
Stelzner, Michael A. (2010), "Social Media Marketing Industry Report: How Marketers are Using Social Media to Grow Their Businesses," socialmediaexaminer.com, http://www.socialmediaexaminer.com/social-media-marketing-industry-report-2010/.
Sung, Yongjun, Yoojung Kim, Ohyoon Kwon, and Jangho Moon (2010), "An Explorative Study of Korean Consumer Participation in Virtual Brand Communities in Social Network Sites," Journal of Global Marketing, 23, 1-16.
Trusov, Michael, Randolph E. Bucklin, and Koen Pauwels (2009), "Effects of Word-of-Mouth versus Traditional Marketing: Findings from an Internet Social Networking Site," Journal of Marketing, 73 (5), 90-102.
Verna, Paul (2010), "Where Does Social Media Fit within an Organization?" February, eMarketer.
Wauters, Robin (2010), "Twitter Growth: Happening All over the Globe (Graphs)," June 23, TechCrunch.com, http://techcrunch.com/2010/06/23/twitter-international-growth/.
Williamson, Debra A. (2010), "Brand Interactions on Social Networks," June, eMarketer.
Zhao, Dejin and Mary B. Rosson (2009), "How and Why People Twitter: The Role that Micro-Blogging Plays in Informal Communication at Work," Proceedings of the ACM 2009 International Conference on Supporting Group Work, 243-52.
About the Authors
Eun Sook Kwon is a doctoral student in the Department of Advertising and Public Relations, Grady College of Journalism & Mass Communication, University of Georgia (e-mail: eskwon@uga.edu). Yongjun Sung is an assistant professor in the Department of Advertising, University of Texas at Austin (e-mail: yjsung@mail.utexas.edu).
