The winners take it all: A comparative study of Twitter campaigns under pressure
First Monday

The winners take it all: A comparative study of Twitter campaigns under pressure by Lene Pettersen and Anders Olof Larsson

This paper presents a comparative study on two Twitter campaigns that came under pressure when recipients started to engage in the campaigns in unexpected ways. Despite two organizations employing different engagement strategies with users, both campaigns played out in somewhat similar ways. Users with an abundance of retweets tend to be male, holding privileged positions in society, with large networks of followers.


Literature review and theoretical approach
Discussion and conclusion




Online campaigns are launched for a variety of marketing or strategic communication purposes — for instance, to engage with target groups or audiences. Ideally, as recipients spread the content provided by a campaign, the original message becomes ‘viral’ and reaches beyond intended target groups. Granted, while individual Twitter users might not be able to gain traction outside of the Twittersphere, a topic gone viral through a series of retweets is likely to spill over into traditional media outlets, thereby reaching even larger audiences (e.g., Bruns and Burgess, 2012; Hermida, 2010).

This brief overview could arguably be seen as the ideal outcome for a social media campaign. As this paper will demonstrate, however, online campaigns sometimes have more obstacles in their way. Indeed, while studies of comparably successful campaigns have been published (e.g., Gallaugher and Ransbotham, 2010; Jarvenpaa and Tuunainen, 2013), rather few insights have been provided into similar scenarios that could perhaps be diplomatically described as less than successful (e.g., Veil, et al., 2015).

This paper presents a comparative study on two Twitter campaigns that came under pressure when recipients began to engage in unexpected ways. First, we focus on the #gladfor (Norwegian for “happy about”) campaign, launched by the Norwegian political party Høyre, to raise awareness of their accomplishments one year into their governing period. Second, we detail the #detnære (Norwegian for “the small things” or “the familiar things”) campaign that was launched by Norwegian chocolatier Freia, urging consumers to share memories associated with their products.

By comparing two different campaigns, this study provides useful insight into professional online conduct during two different campaigns largely gone wrong. Besides focusing on the professionals behind these campaigns, we further attempt to tease out the inner workings of the events by assessing how users used hashtags #gladfor and #detnære. We examine how perspectives altered as the campaigns took unintended directions.

First, we look at the engagement strategies that two organizations (Høyre and Freia) employed when their online campaigns came under pressure. Second, we look at the structural characteristics of those Twitter users who dominated the campaigns, both in terms of being most active and of being most retweeted.

Basing itself on a complete data set of hashtagged tweets for both campaigns, the study is theoretically grounded in the social theories of Bourdieu (1986; 1984). Bourdieu sees society as a social space characterized by class structures. Social fields or spheres are present within this space, having unique types of capital (Aaksvaag, 2008). Bourdieu (1986) distinguishes between three kinds or types of capital: social (relations, relationships), cultural (cultural codes, education, intellect), and economic (monetary, stocks etc.). Capital is power and the exclusive property of elites. From a Bourdieusian approach, then, the actor Freia holds a dominant market position with the largest share in the chocolate market in the social space of Norway, and thus possesses large economic capital, while the Norwegian political party Høyre — conservative party, currently holding power — have amassed social, cultural, and economic capital.

Moreover, a Bourdieuan approach will help reveal characteristics of the active and most retweeted Twitter users in the two campaigns. Thus, we gain insights into whether social media platforms actually offer new engagement opportunities for citizens, or if dominant views in the Twittersphere simply reflect traditional hierarchies of power, gender representation, and other factors present in society’s social space.



Literature review and theoretical approach

Social media use by commercial organizations functions as “facilitating relational and dialogic models of communication” [1]. As such, services like Twitter would appear to be suitable for organizational employment — a suggestion that has been proven true in a series of cases involving comparably successful user involvement (e.g., Gallaugher and Ransbotham, 2010; Jarvenpaa and Tuunainen, 2013; Bayus, 2013). From an international perspective, failed social media campaigns initiated by the likes of McDonald’s (Hill, 2012) and the New York Police (Hathaway, 2014) have gained plenty of media attention — showing how hashtags can become so-called ‘bashtags’.

It would appear from the research available that the key issue for organizations experiencing difficulty with social media is whether to engage with protesters. For example, Veil, et al. (2015) studied an activist ‘hijacking’ of a Kraft Foods social media site over the use of genetically modified organisms (GMOs). As pointed out by Veil, et al., “Kraft suffered through a six-month social media onslaught” [2] before making demanded changes to their product — changes that could have been made earlier and would have spared Kraft from apparent difficulties. These findings stress the need to engage more clearly and earlier with any complaints as well as activists.

In a similar study, Ott and Theunissen (2015) uncovered how a series of different organizations chose to communicate when their respective online presences were put under pressure. For example, an incident involving the seemingly unjust dismissal of an Applebee’s employee was inadvertently publicized on Facebook, where it received a great deal of attention. The strategy of the organization in response was apparently to delete negative comments and to otherwise engage as little as possible in any dialogue, focusing instead on copying and pasting elements from a corporate statement. These tactics appeared to have enraged users further, and the campaign against the company only receded after they themselves stopped engaging with the situation. Applebee’s “flawed understanding of dialogue fanned the crisis” [3].

As for our non-commercial case — the political party Høyre — party leader Erna Solberg proved to be the third most active on Twitter of all party leaders during the last month before the latest national Norwegian elections. Her activity on Twitter was mostly geared towards communicating with other users through @mentions. Specifically, 281 out of 293 (96 percent) tweets sent from the @erna_solberg account during the campaign were directed in this way (Larsson and Ihlen, 2015). Maintaining a proactive rather than reactive approach (e.g., Veil, et al., 2015) might be easier said than done, but it would seem that the former provides better results in the long run.

With a Bourdieusian focus, we can study new phenomena (Twitter campaigns or citizen engagement on the platform) with respect to structural characteristics at play in social contexts or spaces (e.g., values, power distribution) and structural characteristics that actors in these contexts hold (e.g., different kinds of capital, societal position, gender). Moreover, we can study how these characteristics interplay with one specific social field, namely the Twittersphere where social capital (number of followers and connections) is field specific. The field is a sphere of relations of dominance, sub-ordinance or equivalence, rooted in the types and amounts of capital actors possess (Aakvaag, 2008; Ihlen, 2007).

To Bourdieu, social capital is “the sum of resources, actual or virtual, that accrue to an individual or a group by virtue of possessing a durable network of more or less institutionalized relationships of mutual acquaintance and recognition” [4]. Bourdieu observes three types of capital in society: economic (monetary, property, stocks), cultural (education, knowledge, language, values, mastery of cultural codes and symbols), and social (relationships — family, colleagues, friends, contacts from schools and organizations). At the same time, the three capital types can be perceived as symbolic capital (prestige, honor), and the types have sub-types. For example, a subgroup of economic capital is knowledge capital, that is, insight into how political processes or the media works (Bourdieu, 1986; Ihlen, 2007). News events should resonate with widely held cultural values (Ihlen, 2007). Key values in the Norwegian context are trust, self-expression, political democracy, equality among genders and all citizens regardless social rank, intellectual freedom and religious membership, freedom of speech, economic and social equality, solidarity (World Values Survey, 2010–2014; New in Norway, 2016). Over time, Norwegians have become more concerned with environmental and climate changes (Listhaug and Jakobsen, 2008).

Capital is convertible, meaning it can be exchanged into different kinds of capital. For example, economic capital can be exchanged with education. The value of social capital depends on the number of connections that a given person can mobilize (Bourdieu, 1986). To Bourdieu, actors are defined by the combination of capital that they can articulate through social relations. That capital includes the value of social networks, which Bourdieu demonstrated can produce or reproduce inequality among citizens in society. In his discussion on social space, Bourdieu leaves out social capital, which suggest he reckons economic and cultural capital to be more important than social capital (Aaksvaag, 2008). Nonetheless, the forms of capital are unequally distributed among members in society, the exclusive property of elites.

Social media is often argued to bring opportunities for ordinary citizens to influence societal matters. The degree to which the Internet offers such opportunities, or if it simply reflect traditional hierarchies of power, share of voice and influence, is nevertheless subject to scholarly debate (Hargittai, 2008; Dutton, 2013; Norris, 2003). Dijk (2005) argues that socioeconomic status and digital resources are reciprocal. Such international findings also reflect the Scandinavian Twittersphere, where research has found high-end users to be elite, urbanized, highly educated and media savvy (Larsson and Moe, 2013).

From the perspective of Bourdieu, social spaces and fields are reproduced by actors and their actions (Aaksvaag, 2008). Thus, actors have the ability to form and change their social surroundings, although change is difficult due to powerful structural forces (Aaksvaag, 2008). This accord well with studies on the interplay of off-line and online spaces, which find that the power of off-line political elites shapes both the content and sentiment of contributions within the online political communication ecosystem (Dutton, 2013). Gender wise, men still dominate leadership roles (Joy, 2008) as well as civic and political participation (Putnam, 2000). Those voices heard off-line are largely heard online. Off-line sources are found to be replicated online — white, male, highly educated and politically active (Hindman, 2009). Indeed, a recent study (Brandtzaeg, 2015) of Facebook and civic engagement found that the gender differences in civic engagement that exist off-line to a large degree are replicated and reinforced on Facebook. Such results mirrors the characteristics of Scandinavian Twitter users, and questions the potential that social media platforms offer for enabling those without high amounts of Bourdieu’s capital types opportunities for influencing societal matters.

To sum up the theoretical discussion that forms this study’s starting point — social media supposedly provide new opportunities for organizations to engage with the audience and thus gain more voters or sales. Moreover, with social media, citizens now have an opportunity to affect societal agendas. However, previous research has demonstrated that structural patterns off-line are replicated online, thereby reproducing capital inequalities at play in the off=line social space.




Data collection and analysis

The comparative approach employed here allows for insights derived from two contextually different, yet thematically similar, Twitter campaigns. The #gladfor campaign arguably belongs in a political context, as it was launched by the incumbent Norwegian conservative party Høyre approximately one year into their tenure as the major party of the Norwegian parliament, the Stortinget. Using the #gladfor heading, the party posted a message to several social media platforms reading: “One year since the election! Thank you to all who voted for us. Which of our accomplishments are you #gladfor?” The message was posted to Facebook on 9 September 2014 [5], and our analyses will focus on this initial period. As the results will show, while the original message was posted to Facebook, the hashtagged discussion soon spilled over into Twitter.

The #detnære campaign emanated from a commercial context — specifically, an attempt by Freia, one of the largest chocolate manufacturers in Norway, to allow for their customers and fans to share their memories associated with their products. Besides launching the #detnære hashtag, Freia also provided tools on their Web site where interested customers could design their own badge, describing their thoughts and memories of the small things in life related to the Freia brand itself. At the click of a button, this badge could then be shared on a variety of social media platforms. The campaign was launched March 2015; March 2015 was the time span for data collection for this specific campaign.

Data collection was performed by means of Texifter (Shulman, 2014). Providing access to historical Twitter data, tweets containing the hashtags of interest were gathered according to the timespans described earlier.

Data analysis was performed by means of quantitative as well as qualitative methods. First, a series of quantitative procedures provided an overview of the identities of the top users for each hashtag. While such highly active users can be defined and sought out in different ways, we focused on assessing the most active users overall (i.e., those users who tweeted the most using either the #gladfor or #detnære hashtags) and those users who became the most popular by means of having their posted messages redistributed through retweets. By focusing on both the most active and the most popular users, quantitative analysis helps to assess the overall structure of each campaign on Twitter.

Second, for qualitative analysis, the top 20 most active users and the 20 most retweeted users were focused on further for more in-depth scrutiny. Specifically, structural data regarding each of these top users was first gathered from their publicly available Twitter profile pages. When the user’s bio did not provide sufficient information of their societal role, further Internet searching was performed. The next step in the textual analysis was to classify the users’ structural characteristics into categories (for example, based on professional information provided by users).

Gender and the number of followers of the top 20 held were thoroughly mapped. Finally, we performed a textual analysis on the content provided by these top users, attempting to detail references to key values for Norwegians (as described earlier), and to see different ways in which users engaged with both hashtags — such as irony, humor, references to political decisions or environmental facts etc. Text analysis was inductive and characterized by moving between data and theory in a holistic circle (Wadel, 2014). As the process proceeded, interesting patterns evolved. Some patterns led to rearrangements of the first categories, and to new data gathered (e.g., information about gender).





While the #gladfor campaign was initiated on Facebook on 9 September, it immediately made its way onto Twitter. This is clearly shown in Figure 1, as the hashtag launches directly into considerable traffic on that same day. Indeed, 75 percent (N = 1,823) of the total of 2,435 tweets sent during the period were sent during this first day, only to decrease rapidly on 10 April (362 tweets), 11 April (128 tweets) and 12 April (38 tweets posted). As such, while interest in #gladfor was initially comparably large, those engaged quickly disbanded.


Time-line graph depicting Twitter activity for the #gladfor hashtag
Figure 1: Time-line graph depicting Twitter activity for the #gladfor hashtag, 1–30 September 2014.


Turning to assessing the most active and the most retweeted users within #gladfor, Table 1 presents the aforementioned group.


Most active users of the #gladfor hashtag
Table 1: Most active users of the #gladfor hashtag.


As this table illustrates, the most active actor using the #gladfor is a geek/activist, a male describing himself as a ‘code-nerd’ and a ‘cultural Marxist’, thereby clearly holding political views opposite to Høyre. The second most active user identifies himself as a communications consultant who airs support for Høyre, providing 30 tweets. The third most active user, then, appears to be another critical user — a teacher who does not support Høyre.

A closer analysis sorts the 20 most active users roughly into eight categories based on similarity. The categories should not be seen as mutually exclusive, yet the distinctions made here are meaningful in terms of their societal roles. Also, it should be noted that the blogger category overlaps with engaged citizens. For example, bloggers are also engaged citizens, yet not all engaged citizens write blogs. Because of this distinction, we have chosen to separate these into two categories.

When the most active users originally presented in Table 1 are presented with regards to their apparent roles, engaged citizens are on the top (N = 6), with a balanced distribution of gender and with twice as many participants as the next most active category, journalists (N = 3). From here, the number of participants is divided among politicians (N = 2), bloggers (N = 2), geeks (N = 2), teachers (N = 2), environmentalists (N = 2), and lastly, one communication consultant — the most active user as noted earlier. Table 2 provides these user groups based on their combined amount of Twitter followers.


Most active users within the #gladfor hashtag sorted by network size
Table 2: Most active users within the #gladfor hashtag sorted by network size.


As mentioned, the volume of social capital brings opportunities (Field, 2003). Such a measurement of popularity provides us with some insights into which groups of users that reached large numbers of others with their message — and which groups did not. While users with the role of ‘engaged citizens’ (N = 6) dominate Table 2, their combined number of followers emerges as the fourth lowest (N = 4,759), outperformed by Høyre politicians (boasting 29,800 followers), journalists (11,024), bloggers (11,304), and the communication consultant (14,200 followers).

Of note is also the result that the account associated with Høyre is actually the official party account — @hoyre. Høyre is at the top in terms of social capital, with more than twice as many followers than the second rank. The politician from the largest opposition party, however, is at the bottom.

In contrast to the results derived from the number of followers enjoyed by the most active users of #gladfor as discussed earlier, contrasting findings are found when we turn to assess the most retweeted users (Table 3).


Most retweeted users within the #gladfor hashtag
Table 3: Most retweeted users within the #gladfor hashtag.


The most retweeted user is a male journalist (109 retweets), followed by a politician from the Labour party (AP) (75 retweets), and Ski Norway, part of the non-governmental organization (NGO) Norges Skiforbund (Norwegian Ski Federation) (65 retweets) — the latter account tweeting about prioritizing climate. Six persons are present in both samples. These are a consultant, the chairman of an environmental NGO, a teacher, two journalists, and a politician from the opposition party AP. It is also worth mentioning that among those most retweeted, only one user could be clearly identified as a woman. Also, as the official @hoyre account is not visible in Table 3, the party itself did not succeed in getting their views across in this way.

Focusing on the roles of the identified users, Table 4 presents these groups of users according to their combined number of followers.


Most retweeted users within the #gladfor hashtag sorted on network size
Table 4: Most retweeted users within the #gladfor hashtag sorted on network size.


When assessing the number of followers, Table 4 clearly shows how NGOs take the lead, followed by politicians from the opposite parties and then by engaged citizens. However, although both NGOs and citizens are represented by three accounts each, the NGOs have more than double the amount of followers, thus enjoying a large network size. Clearly, the most retweeted users share that they hold a network size that is well beyond the reach of ordinary or even engaged citizens.

As for the content provided by the most retweeted users, these were mainly formulated by men holding societal positions of power: political stakeholders, journalists, public figures and non-governmental organizations. In terms of genre and content, the most retweeted tweets all illustrate characteristics of irony, sarcasm and humor. These express emotions in relation to issues of social justice value, often touching upon the ways that the reign of Høyre has actually made Norway a less financially equal society. For example, the most retweeted tweet, originally posted by a journalist, comments on social and economic equality: “I am #gladfor the opportunity to inherit as much as I want without paying a single dime to the community. You don’t have rich parents? Poor you!” Economic and social inequality were featured in one of the most retweeted tweets from a politician of the Socialist Left Party: “I am #gladfor the 3200 kroners [US$410] I get in tax cut (we earn well at the Parliament!), while a nurse will have to do with 39 øre [US¢40] a day. Hurray!”

Others are quite more sarcastic and ironic, such as this example from one of the most retweeted bloggers: “I am #gladfor this government doesn’t do anything else besides hanging around at Twitter, making hashtags. The less action, the less harm, after all.” Also, ironic tweets with a touch of humour is popular, especially when they are sent from known public figures such as the following, from a respected author, journalist and editor: “I am so #gladfor having a #government which is so #cool and #youthful that it uses #hashtags and #SoMe, #asif.”

To sum up, the most active users of the #gladfor have been described here as geeks, journalists and politicians. Yet, when we collapse the top users into categories, engaged citizens come out on top as the most active group of users. This picture changes slightly when we look at the most retweeted, where NGOs, politicians from opposition parties and engaged citizens dominates. Moreover, while 30 percent of the most active sample are females, only one of the 20 most retweeted were sent by a female. As such, common to most of those who are most retweeted is that they are men, enjoy large numbers of followers on Twitter and that they play some key societal role. Having large networks trumps being active, yet again, having large networks follows with capital and rank outside of the Twittersphere.



Time-line graph depicting Twitter activity for the #detnaere hashtag
Figure 2: Time-line graph depicting Twitter activity for the #detnære hashtag, 1–31 March 2015.


In comparison to the #gladfor time-line graph, the activity surrounding the #detnære does not appear to start in medias res. Rather, a few early posts are provided that are more in line with the organization’s intention. The build-up as visible in Figure 2 appear to start on 20 March (392 tweets), less than two weeks before Easter, reaching its peak the day after (752 posted tweets) and declining in the following days. As will be discussed below, the peak on 20 March appears to be related to Norwegian media reports that the badge functionality on the Freia Web site was subject to censorship. Specifically, Freia attempted to filter out certain terms from the application. These efforts were apparently geared not only towards profanity and other slurs, but also to bar any mention of palm oil, a particularly environmentally unfriendly ingredient featured in one of Freia’s most popular products, chocolate Easter eggs. This led to a series of tweets in which badges from the Web site that featured various attempts to trick the filter were shared. For example, engaged users employed misspelled words to use Freia’s own Web page to raise a critical voice regarding their environmental policies. Table 5 provides an overview of the most active #detnære users.


Most active users of the #detnaere hashtag
Table 5: Most active users of the #detnære hashtag.


Interestingly, the most active actor using the #detnære is the same person identified as most active user in the #gladfor campaign, more specific a male geek/activist, describing himself as a ‘code-nerd’ and a ‘cultural Marxist’. The second most active is a critical blogger, who provides 33 tweets containing the hashtag. The third rank is a fake Freia account that was set up, presumably by some activist, during the campaign. With regards to gender balance, there are five females among the most active users: two engaged citizens, a media professional, and two communication advisors. As such, while the balance is perhaps not as tilted as for the #gladfor campaign, we see a clear overrepresentation of males taking part in the hashtagged discussion.

Much like for our analysis of #gladfor, we now move on to see what groups of users appear to have had the highest impact factor — in terms of having large networks of followers. Table 6 provides one illustration of the sample, based on network size.


Most active user groups of the #detnaere hashtag sorted by number of followers
Table 6: Most active user groups of the #detnære hashtag sorted by number of followers.


Similar to the most active users and most retweeted in the #gladfor campaign, engaged citizens can be identified among the top user groups in terms of combined network size. However, media professionals dominate; combined, these users enjoy almost five times the network size of the second rank, the communication advisors.

When we look at who is dominating the #detnære hashtag by enjoying plenty of retweets, the top users appear as somewhat shifted yet still similar to the results presented in Table 5. Table 7 provides an overview of the most retweeted users within the hashtag at hand.


Most retweeted users within the #detnaere hashtag
Table 7: Most retweeted users within the #detnære hashtag.


While the most active user in #detnære (the male geek/activist) is not present in the most retweeted sample, the second most active — a critical blogger — emerges as the most retweeted. Among the most retweeted, new categories based on role are revealed, namely enthusiasts (Twitter users that typically tweet in a special genre related to a specific topic such as music, football, humoristic/weird, or political satire), humanists and environmental NGOs, and one political engaged lawyer working in the government. Much like as before, there is a gender bias among the most retweeted; only two females appear to have reached popularity in this way — an enthusiast and an engaged citizen, none of which could be identified among the most active users of #detnære.

Turning to assessing groups of users of their combined number followers, Table 8 provides further insights.


Most retweeted user groups within the #detnaere hashtag sorted on network size
Table 8: Most retweeted user groups within the #detnære hashtag sorted on network size.


When sorting on social capital, media professionals switch rank with their 25,300 followers, with citizens down to seventh rank. The humanist and environmental NGOs keep their second place, while the only journalist in the sample climbing to third. Characteristic of the top two are males with large amount of social capital.

When we turn to the content of tweets, the most retweeted messages are similar to those discussed for the #gladfor case in that they are ironic, sarcastic and humorous, closely related to key values that Norwegians hold, especially those related to issues of environmental protection, which has become important to Norwegians (Listhaug and Jakobsen, 2008). As previously earlier, badge functionality on the Freia Web site was subject to censorship, and indeed, all the most retweeted messages are related to values concerning environmental issues and censorship.

As an example, the most retweeted tweet is from a critical blogger:

“Look what #Freia slipped through the censorship #detnære“ [with an image of his produced badge enclosed — saying] “To whisper sensually in your girlfriends ear; baby, I want to do to you what Freia is doing to the cacao farmers.”

Another highly retweeted message is from a liberal humanist with a history of positions from NGOs working with humanistic youth, human ethical perspectives and councils for religious and life stance issues: “This is how we get rid of the rain forest. #detnære.” Another highly retweeted tweet in which illustrates how censorship was not well received by the Norwegians and the importance of transparency as a key value, is from a media professional: “Here is the list with words that Freia doesn’t like #detnære [enclosed by a list containing many of the terms that Freia censored]”



Discussion and conclusion

To return to the engagement strategies that the two organizations (Høyre and Freia) employed when their online campaign took an unintended direction, the analysis revealed that while Høyre was present in the campaign with their large number of followers, they engaged little in conversations with users, nor were they present in the most retweeted sample. A few Høyre supporters with large networks were present in the hashtagged activities, but overall, the opposition parties, journalists and engaged citizens boasted higher degrees of activity.

While interest in #gladfor was initially comparably large, those engaged quickly disbanded engagement through the hashtag over the four days that the hashtag was trending. A total of 2,435 tweets were sent during this first day, only to decrease rapidly over the following three days. However, in the case of #detnære and Freia, this organization was not present in the Twittersphere at all. Thus, the option to engage with users was not possible. The build-up of #detnære appears to have started on 20 March (392 tweets), reaching its peak the day after (752 posted tweets) and declining over the following days. Interestingly, both backfired events had approximately the same duration of time.

Surprisingly, the backfire on #detnære perhaps was started with a blog post from a Norwegian pink blogger Sophie Elise posted 13 March, encouraging a boycott of Easter chocolate eggs from Freia. However, a documentary about food suppliers in Norway on 17 March revealed that other, smaller chocolate producers were excluded from food retailers by Freia, in order to maintain a market position in the Norwegian market. When we studied the Twitter data at hand, the peak on 20 March appeared to be closely related to Norwegian media reports of censorship on the Freia web site. From our analysis, the documentary and the media report of censorship seemed to be triggering factors, rather than the blog post alone. Transparency is one of the key values that Norwegians hold, so censorship clearly provoked Norwegians.

Moreover, when traditional media set the matter on the societal agenda, #detnære escalated. We cannot say if the two backfired events harmed Høyre and Freia’s reputation, but the analysis suggest that despite two different engagement strategies from two quite different organizations, the Twitter campaigns tended to be over quickly. This stands in contrast with findings by Veil, et al. (2015), which found that Kraft suffered through a six-month social media onslaught. Based on their research on Kraft, Veil, et al. (2015) stressed the need to engage more clearly and rapidly with complaints or activists.

Freia holds primarily economic capital, while Høyre holds a combination of social, cultural and economic capital within the Norwegian space. Freia and Høyre were not harmed by critical comments in Twittersphere or in traditional media. Despite press on Freia’s censorship, Freia did not experience any decrease in the sale of Easter eggs (Tobiassen, 2015). This supports Bourdieu’s notion that among the three types of capital, the economic type is the most powerful and important in social space (Bourdieu, 1984; Aakvaag, 2008).

When we turn our discussion to the structural characteristics of the Twitter users who dominated the campaigns, our analysis found, to a great extent, that those who already hold more privileged societal positions dominated the samples of most retweeted in both campaigns. Thus, elements in the social space spill over into the two campaigns in the Twittersphere. For example, although Høyre holds governmental power and the largest number of followers (29,800 followers) in the most active sample in #gladfor, they are not present in the most retweeted sample. Here the most retweeted are, when sorted on roles, politicians from the opposition parties AP, SV and the Oslo party, followed by journalists, NGOs, and engaged citizens.

Similarly, when we categorize the most retweeted tweets based on network size, the NGOs are at the top, followed by politicians from the opposition parties, engaged citizens, and a media professional.

The picture was somewhat different in #detnære, where a media professional held the largest network (14,300 followers). When we sorted the most retweeted in #detnære based on roles, enthusiast, humanist, and environmental NGOs were on the top, followed by engaged citizens and consultants. When sorting the most retweeted sample in #detnære based on their amount of social capital, a media professional took the lead.

Bourdieu (1986; 1984) argues that the value of social capital depends on the number of connections that an individual can mobilize. A consistent pattern is that the most retweeted users in both campaigns have huge networks, well beyond ordinary citizens. The most retweeted actors thus posit large social, cultural, but also economic capital. With a large number of followers actors’ messages can reach a larger audience than those with smaller network. Citizens with large social networks also have more opportunities to reach a larger audience than those with smaller networks, yet few ordinary citizens hold large networks. This accord with Bourdieu’s (1986) argument that social capital is the exclusive property of elites.

Similarly, gender-wise, despite at least a more balanced gender distribution among the top active tweets in both campaigns, female perspectives were outperformed by males and their larger social networks. Woman were underrepresented in retweets. Also, the most retweeted tweeters retweeted each other’s tweets. While six of the 20 most active users of the #gladfor were woman, only one female was present among the most retweeted. Similarly, among the most retweeted in #detnære, only two women were present, none of which were the most active users of #detnære. With few exceptions, in both campaigns the most retweeted sample were men holding key societal roles off-line. This results agrees with earlier research that off-line sources are replicated online (Dutton, 2013).

Our findings agree in part with research into Facebook and civic engagement, where gender differences that exist off-line to a large degree are replicated and reinforced on Facebook (Brandtzaeg, 2015).

Our study suggests that although the most active users were dominated by males (70 percent), there was a small segment (30 percent) of women present among the most active. This distribution is recognizable in previous, similar research projects (Christensen, 2013; Larsson and Moe, 2013). Interestingly, the only woman who was retweeted held a large social network (6,191 followers). Among the most active users, a female blogger/writer counted 11,100 followers, while a female journalist had 4,244. The female engaged citizens and a teacher held in total only 1,283 followers. This suggests that even when women were engaged, their views and presence were large ignored. It also supports previous assumptions that those who already hold power off-line replicate it online. The societal gender gap in leading and influential positions could explain why so few woman were retweeted.

Moreover, the content analysis found that most retweeted tweets in both campaigns were key values that Norwegians hold; equality among genders and all citizens regardless social rank, societal transparency, economic and social equality, freedom of speech and political view, and environmental concerns. Freia’s censorship in the #detnære clearly sparked negative feedback.

Our study is not without limitations. We have only analyzed data from two Norwegian campaigns and the findings cannot be generalized. Nevertheless, we did study two different campaigns where qualitative analysis revealed interesting patterns present in both samples about structural elements that accord well with Bourdieu’s (1986; 1984) social theory. Future research should reveal why so few women tend to be retweeted.

We can conclude that practitioners could adapt Høyre’s social media strategy by growing a large network of followers. When planning campaigns in social media, it is important to keep in mind key values of the targeted audience, and to realize that campaigns are part of a larger social space. However, if there are unexpected issues on Twitter, organizations need to keep calm and not be overly reactive. End of article


About the authors

Lene Pettersen is an Associate Professor at the Faculty of Management of the Westerdals Oslo School of Arts, Communication and Technology in Norway.
E-mail: lene [dot] pettersen [at] westerdals [dot] no

Anders Olof Larsson is an Associate Professor at the Faculty of Management of the Westerdals Oslo School of Arts, Communication and Technology in Norway.
E-mail: larand [at] westerdals [dot] no



1. Macnamara and Zerfass, 2012, p. 288.

2. Veil, et al., 2015, p. 106.

3. Ott and Theunissen, 2015, p. 100.

4. Bourdieu and Wacquant, 1992, p. 119.

5. Please refer to



G.C. Aakvaag, 2008. Moderne sosiologisk teori. Oslo: Abstrakt Forlag.

B.L. Bayus, 2013. “Crowdsourcing new product ideas over time: An analysis of the Dell IdeaStorm community,” Management Science, volume 59, number 1, pp. 226–244.
doi:, accessed 20 October 2016.

P. Bourdieu, P. 1986. “The forms of capital,” In: J.G. Richardson(editor). Handbook of theory and research for the sociology of education. New York: Greenwood Press, pp. 241–258.

P. Bourdieu, 1984. Distinction: A social critique of the judgement of taste. Translated by R. Nice. London: Routledge & Kegan Paul.

P. Bourdieu and L.J.D. Wacquant. 1992. An invitation to reflexive sociology. Chicago: University of Chicago Press.

P.B. Brandtzaeg, 2015. “Facebook is no ‘Great equalizer’ A big data approach to gender differences in civic engagement across countries,” Social Science Computer Review.
doi:, accessed 20 October 2016.

A. Bruns and J. Burgess, 2012. “Researching news discussion on Twitter,” Journalism Studies, volume 13, numbers 5–6, pp. 801–814.
doi:, accessed 20 October 2016.

C. Christensen, 2013. “@Sweden: Curating a nation on Twitter,” Popular Communication, volume 11, number 1, pp. 30–46.
doi:, accessed 20 October 2016.

J. van Dijk, 2005. The deepening divide: Inequality in the information society. London: Sage.

W.H. Dutton (editor), 2013. The Oxford handbook of Internet studies. Oxford: Oxford University Press.

J. Field, 2003. Social capital. London: Routledge.

J. Gallaugher and S. Ransbotham. 2010. “Social media and customer dialog management at Starbucks,” MIS Quarterly Executive, volume 9, number 4, pp. 197–212.

E. Hargittai, 2008. “The digital reproduction of inequality,” In: D.B. Grusky (editor). Social stratification: Class, race, and gender in sociological perspective. Boulder, Colo.: Westview Press, pp. 936–944.

J. Hathaway, 2014. “NYPD’s Twitter outreach backfires in most predictable way possible,” Gawker (22 April), at, accessed 22 May 2016.

A. Hermida, 2010. “Twittering the news: The emergence of ambient journalism,” Journalism Practice, volume 4, number 3, pp. 297–308.
doi:, accessed 20 October 2016.

M. Hindman, 2009. The myth of digital democracy. Princeton, N.J.: Princeton University Press.

K. Hill, 2012. “J#McDStories: When a hashtag becomes a hashtag,” Forbes (24 January), at, accessed 22 May 2016.

S.L. Jarvenpaa and V. K. Tuunainen. 2013. “How Finnair socialized customers for service co-creation with social media,” MIS Quarterly Executive, volume 12, number 3, at, accessed 22 May 2016.

Ø. Ihlen, 2007. “Building on Bourdieu: A sociological grasp of public relations,” Public Relations Review, volume 33, number 3, pp. 269–274.
doi:, accessed 20 October 2016.

L. Joy, 2008. Advancing women leaders: The connection between women board directors and women corporate officers. New York: Catalyst.

A.O. Larsson and Ø. Ihlen. 2015. “Birds of a feather flock together? Party leaders on Twitter during the 2013 Norwegian elections,” European Journal of Communication, volume 30, number 6, pp. 666–681.
doi:, accessed 20 October 2016.

A.O. Larsson and H. Moe. 2013. “Twitter in politics and elections: Insights from Scandinavia, In: K. Weller, A. Bruns, J. Burgess, M. Mahrt, and C. Puschmann (editors). Twitter and society. New York: Peter Lang, pp. 319–330.

O. Listhaug and T.G. Jakobsen, 2008. “Verdiundersøkelsen 2007. Norske meninger om miljø — lokalt og globalt,” Samfunnsspeilet, 1/2008, pp. 31–40, at, accessed 22 May 2016.

J. Macnamara and A. Zerfass. 2012. “Social media communication in organizations: The challenges of balancing openness, strategy, and management,” International Journal of Strategic Communication, volume 6, number 4, pp. 287–308.
doi:, accessed 20 October 2016.

New in Norway, 2016. “Values that are part of the Norwegian mindset,” at, accessed 22 May 2016.

P. Norris, 2003. “Preaching to the converted? Pluralism, participation and party Websites,” Party Politics, volume 9, number 1, pp. 21–45.
doi:, accessed 20 October 2016.

L. Ott and P. Theunissen, 2015. “Reputations at risk: Engagement during social media crises,” Public Relations Review, volume 41, number 1, pp. 97–102.
doi:, accessed 20 October 2016.

R.D. Putnam, 2000. Bowling alone: The collapse and revival of American community. New York: Simon & Schuster.

S. Shulman, 2014. “Texifter is plugged into gnip” (12 June),, accessed 26 January 2016.

M. Tobiassen, 2015. “Rasende bloggere stoppet ikke Freia-salget,” Dagens næringsliv (21 May), at, accessed 22 April 2016.

S.R. Veil, J. Reno, R. Freihaut, and J. Oldham, 2015. “Online activists vs. Kraft Foods: A case of social media hijacking,” Public Relations Review, volume 41, number 1, pp. 103–108.
doi:, accessed 20 October 2016.

C. Wadel,2014. Feltarbeid i egen kultur. Oslo: Cappelen Damm akademisk.

World Values Survey, 2010–2014. “Inglehart–Welzel cultural map,” at, accessed 22 May 2016.


Editorial history

Received 2 July 2016; accepted 20 October 2016.

Creative Commons LicenseThis paper is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

The winners take it all: A comparative study of Twitter campaigns under pressure
by Lene Pettersen and Anders Olof Larsson.
First Monday, Volume 21, Number 11 - 7 November 2016

A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2017. ISSN 1396-0466.