Introduction

Election campaigning is argued to be highly professional, influenced by trends in commercial marketing and dedicated to selling a candidate, party and platform to citizens using sophisticated communication media and features. The adoption of digital technologies is a feature of professionalisation but challenge the paradigm of electioneering that focuses purely on persuasion. The features of Web 2.0 and the social Web, embedded within platforms widely used by citizens, permit a range of interactions which can be features of any web presence should the designer see value in embedding a more interactive dimension within campaign communication. This paper explores the extent to which campaigns are becoming more interactive using two schematics. Firstly, the analysis first utilises McMillan’s concept of directional flow of communication. The second schematic operationalises Howard’s concept of the hypermedia campaign which focuses more on strategic political communication. These models allow us, through a case study of presidential campaigns in France and the USA, to assess how campaigns direct visitors and permit their input into the campaign and the extent this fits to strategic communication objectives. The data shows campaigns are, albeit mainly through the colonisation of social media platforms by candidates, taking on some elements of interactive communication. However, communication tends to be horizontal between visitors and not involving the candidate or official voice of the campaign and campaigns largely seek to gather data and harness supporters to working for the candidate rather than empowering them to shape the campaign.

Online Campaigning: From Shovelware to the Social Campaign

The online dimension of election campaigning is the latest phase in the professionalisation of political communication, and one that some argued to have revolutionary potential. Although early uses of digital technologies involved the creation of fairly simple Websites that hosted material created for offline distribution such as leaflets, flyers and manifestoes, the more recent societal uptake of social media have led to significant innovations in election campaigning. Websites have evolved from a static, read-only page format to incorporate rolling news feeds, Weblogs functions and feeds both in and out to profiles on third party sites. In particular, the adoption of Facebook, Twitter and YouTube by campaigners has seen increased opportunities for the visitor to a political party or candidate Website to be able to leave behind comments that can be read and responded to by the host or by other users.

However, largely we find the potential for re-imagining the terms of engagement between campaigners and the broader citizenry unrealised. Campaigners are keen to harness the elements of digital technologies that permit direct communication from the campaign to potential voters, with the sole intention of securing and maintaining support. Largely studies have found that tools which permit communication from potential voters, or communication that is interdirectional between voters as well as between voters and the campaign, are seldom adopted or when their interaction is strictly controlled (Foot and Schneider, 2006; Kluver et al, 2007; Stromer-Galley, 2014), and even in the era of social media few campaigns truly harnessed the interactive (Lilleker and Jackson, 2011) or co-creational (Lilleker, 2013) aspects of digital technologies. Like Websites, Facebook, Twitter and YouTube are mainly used for one-way persuasive communication, from the campaign outwards; although it is not always possible to prevent interaction occurring campaigns eschew becoming involved in conversations because of concerns over being drawn off message and having to elaborate on core promises (Stromer-Galley, 2000).

Where there is variation within the utilisation of the more interactive features offered by communication technology this is found to support a representative function. Studies have found more innovative uses of social platforms by, for example, UK Members of Parliament (Jackson and Lilleker, 2010), Members of the European Parliament (Koc-Michalska and Lilleker, 2013) and local candidates when seeking election (Graham et al, 2013) suggesting that some representatives extend the personal and accessible dimension of their role into social media. Yet, these practices tend not to be extended into campaigning. However, in recent years the campaigns of Segolene Royal (runner-up in the 2007 French presidential race) and Barack Obama (who ran a successful outsider campaign for the US presidency in 2008) have been argued to be highly innovative in harnessing some interactive aspects afforded by digital technology. In both cases, and in subsequent contests, media commentators and academics alike have posed questions about the extent to which each is a Web 2.0 election.

Technology, it is suggested, can facilitate a win-win zone for campaigners and citizens. Campaigns can communicate to and gain feedback on dimensions of their campaign, including policy, which engages citizens and increases the likelihood of them converting to being loyal supporters (Lilleker, 2015). Loyal supporters are in turn likely to extend the reach of a campaign via social media, through any actions (likes and shares) and interactions being visible within the network of the citizen and thus in turn citizens play the role of advocate (Norris and Curtice, 2008). Citizens feel more connected and empowered and in theory participation rates increase as more citizens are exposed to political campaign communication, engaged to some extent as well as being more informed (de Zuniga et al, 2009). Hence, the extent campaigning moves to a more interactive and empowering form of campaigning the more the campaign might reap rewards that equally have a positive impact on democratic life more broadly.

Interactivity and Campaign Strategy

Campaigns can utilize digital environments to enhance their effectiveness in a number of ways. For citizens, there is readily accessible information from the convenience of any location with some form of network access. For organisations, it provides a direct channel from the organisation to customers and so from political campaigns to potential voters. However, what has been termed the social media revolution, defined as the dramatic uptake of the use of social platforms such as Facebook, Twitter and YouTube, has dramatically changed the way the Internet is used. For the masses it is not simply a resource for locating content, rather content creation, adaption, dissemination and commentary are all activities that most are able to carry out. In particular, these activities are facilitated by easy to use platforms such as Blogger or Tumblr, Google+ as well as the more generalist Facebook or Twitter platforms that permit sharing thoughts, comments, images and videos. It is true that while it may be easy to have voice online it is not easy to ensure being heard (Morozov, 2012), so questioning the extent that the ability to create is in itself empowering or democratising; however, most studies agree social media has opened up a new dimension in interactivity between users, and between the ordinary user and a range of corporate and not-for-profit organisations so altering the norms of communication (Burgess, 2014).

Studies have consistently found that in a political context interactivity tends to be highly controlled and it is more important to appear to be interactive and accessible than to actually communicate with citizens; the so-called domain of Web 1.5 (Jackson and Lilleker, 2009). It is therefore useful to question to what extent if any political communication has entered into a more interactive paradigm. McMillan (2002) recognised differing variants of interactivity, and made a link between communication strategy and the experience of users. Her typology suggests three levels of interactivity: interaction with the page, the system or with other users. The most basic, page-level, form of interaction is the choice of accessing a file, reading, listening or watching or it can involve choosing which links to follow. Page-level interaction potentially offers each user a unique experience but is related only to accessing information and is a long way from the ideals of interactivity that relate to the facilitation of dialogue (Rafaeli, 1988). The second level of interactivity for McMillan is an ability to interact with the system. Here the user is able to choose what to access and how to access it, for example, choosing to have news feeds delivered to the desktop via RSS, or taking part in file sharing activities. System-level interaction involves users working together to some extent. While system-level interaction lays the foundation for co-creation, most studies of interactivity offer the view that user-to-user interactivity is the most engaging and democratising. Where user-to-user interactivity is offered file sharing can be accompanied by conversation and co-production of music, video and other users’ experiences. User-to-user interactivity fits well with the ‘big ideas’ of Web 2.0, is a by-product of the social media revolution, and is compatible with notions of reimagining democracy for the social media age.

Operationalising McMillan’s typology draws on the Information Traffic Model (Bordewijk and Kaam van, 1986; McMillan and Downes, 2000) and the Excellence Model for Public Relations practice (Grunig and Grunig, 1992) to develop a four part interactivity model. More recently further adapted to create a six part model (Ferber et al, 2007). In order to develop a series of classifications for the type of communication facilitated by features embedded within Web presences we adapt these to create a schematic that details five directions in which communication may flow. First, downward communication, this is normally in the form of a monologue from the creator, the organisation, to site visitors. Second, upward communication that permits data collection about visitors to sites and associated presences online but is not in itself interactive. Third, horizontal communication facilitated by hyperlinks allowing travel from one site to another. Fourth, asynchronous communication, which is interactive but not public or in real time, for example, e-mail or any private feedback system where there may be a delay in responding or even no response at all. Fifth, synchronous communication, or public interaction in real time, which can be facilitated by the features of social media platforms, chat rooms, forums and similar features. This categorisation strategy permits us to know the extent of experiences are allowed, with downward and horizontal allowing interaction with the page only, upward and asynchronous allowing interaction with the system but only asynchronous permitting user-to-user interaction. Furthermore, the strategy allows an average performance score (dividing the number of features utilised by the total policy) that allows simple comparability of prioritisation of feature use by a campaign and comparison between campaigns.

The uptake of features by political candidates and party Websites that permit differing forms of interaction is not determined by the availability of technology. Rather innovations in electioneering are dependent upon assessments driven by campaign logic. Any campaign related activity is selected purely on an assessment of its utility in meeting the objectives of the campaign and therefore any shift to a more interactive paradigm will depend on its perceived value to the campaign (Tenscher et al, 2012). This is particularly true in what Howard (2006) defined the hypermedia age. Technology, Howard argues, serves four functions for a campaign. First, technology facilitates the transmission of information from the campaign out to prospective voters, so replicating a mass media model of communication but using all available media, from television advertising to video games to Facebook. Second, technology facilitates the targeting of communication, aiming specific messages at specific target groups of the electorate to whom they are most relevant and attractive. Third, and in order to facilitate targeting, technology facilitates the harvesting of data. Whether it is captured from the Facebook profiles of supporters, the collection of data on the owners of e-mail addresses captured by the campaign, or data on which type of voters are most likely to watch certain popular television programmes, visit certain Websites, or are most likely to be concerned about certain political issues or causes. Data harvesting permits both designing and delivering messages in order to maximise impact. Gaining a following on social media, which is argued to be a concomitant of frequently posting to and interacting with a vibrant community (Koc-Michalska et al, 2015), hence meeting the strategic objectives linked to data harvesting may involve re-orienting the nature of campaign communication. Harvesting can also involve the collection of donations that also permits the capturing of personal data from supporters. Fourth, technology facilitates interactive communication with members and supporters, to enhance a mobilisation strategy through recruiting volunteers as well as encouraging turnout. Campaigns also ruthlessly target wavering supporters to firm up their support and with undecided voters in order to persuade through dialogue with the campaign. Howard’s work was seen as a reflection on the embedding of technology within the US Democrat party campaign machine and especially the Obama campaign of 2008 though as we shall see certain aspects were accentuated to a greater extent with technology’s potential not fully exploited for all functions.

Howards’ communication functions represent campaign logic interpretations of McMillans’ typology of forms of interactivity. Transmission is downward communication, harvesting an expanded form of upward communication, targeting both downward but also asynchronous as it represents communication designed on a reactive basis resulting out of data collection, interaction represents both asynchronous (email-based and other forms of two-way communication) and synchronous communication. Assessing the use of technology to meet these functions as well as McMillan’s typology allows a more holistic and nuanced perspective of campaigning within the social media age.

Methodology

The research adopts the traditional method of content analysis, counting features within the Web presences of the final two candidates, who went forward to the second round of voting or won the respective party primaries in France 2007 and 2012; the United States 2008 and 2012. The Web presences were analysed during the closing phase of the contest, within 7 days of the final vote. Choosing the final candidates may overlook innovation among candidates with little chance of winning; however, these are the candidates that set the tone of a campaign and who utilised the online environment with a view to winning; minority candidates may have other purposes for running a campaign such as promoting issues or building support for a particular ideological stance (Rohrschneider, 2002). Assessing the use of the online environment in the final days allows us to see the ultimate iteration of online strategy, as opposed to seeing innovations tried and possibly abandoned. Every measure is imperfect, however, it is proposed that this analysis allows the measurement of the most sophisticated and strategic campaigns within the respective nations, each designed with the purpose of contributing to victory in the campaign.

Initially, the presence or absence of 106 features was counted in order to determine the differences between the architectures of each Website and linked presences (in most cases Facebook and Twitter profiles and YouTube channels or other similar social media platforms). The features are firstly categorised as per McMillan’s schematic, outlined above, following the categorisation strategies employed in a number of recent studies (Gibson and Ward, 2000; Lilleker and Jackson, 2011; see appendix 1 for full list of features counted). A second categorisation strategy aligns the features specifically to indicators for a hypermedia campaign. For each set of categorisations, we develop an average performance score (calculated as a percentage score by dividing the number of features appearing by the maximum possible for each feature category), this is a standard measure that allows direct comparability between parties and across time (Farmer and Fender, 2005; Schweitzer, 2008; Vaccari, 2008; Larsson, 2011). Cumulatively we are able to determine the extent to which parties adhere to the principles of communication within Web 2.0 environments, utilising the standard McMillan schematic. Second, the extent to which they adhere to the core tenets of the hypermedia campaign, utilising an adapted schematic based on Howard’s core descriptors. The comparison of each campaign Web presence involved coding features present on the official Website and live links to other online presences based on their type, in particular social media sites. Given the standardisation of social media profiles of organisations, each of which permit users to share, like and comment on content, these platforms were not coded as discrete and containing additional features. Our analysis permits an assessment of national differences at each election, given their close temporal proximity, as well as the evolution of Web campaigning between and within nations and between ideological blocs.

Online Political Communication: A Move Towards Interactivity?

Table 1 demonstrates the similarities and differences between the campaigns displaying the average online performance in each of McMillan’s categories using average performance percentages. Further analytical measures for proving differences are not appropriate given the small sample for each contest. These simple but clear numbers permit easy comparison of each campaign Web presence across the four contests and show the ebb and flow of campaign Website design from contest to contest and nation to nation.

Table 1 Communication flows (AOP by candidate and year)
Full size table

Across the campaigns there was reasonable equanimity over the importance of informing visitors, using downward communication flows, to the Web presences of candidates. The differences between the numbers of informational features used are dependent on the type and style of communication and how the Websites integrate aspects of the campaign that take place offline. For example, some sites included large archives of press releases while others had a more bespoke news archive. Strategically though information provision was a universal priority and all candidates utilise more than 50 per cent of informational features, with the exception of Francois Hollande at 48 per cent, with the overall average being around the 60 per cent mark.

While it would be fairly incomprehensible to find candidates not supplying information given its persuasive power, allowing feedback is a strategic decision that involves greater risk but one often balanced against the importance of building a database of contact details and data on visitors. In France in 2007, the extent to which feedback was allowed was via online feedback forms and simplistic private communication tools; little evidence of data collection could be found. The transition from the Web 1.0 campaigning era to that of Web 2.0 and the uptake of social media is demonstrated by the increased deployment of features that allow feedback and for visitors to leave data behind them through private and, to an extent, public channels. Obama in particular led the way in his use of blogtools within his Website, which in itself had the architecture of a social network that involved each user having to create a profile. Obama also had profiles on every major social networking site targeting Hispanic, Gay and Bisexual, younger and older demographics. Joining his network, liking or following his profiles, all allowed data to be left behind. Even his YouTube channel permitted any visitor to comment on videos with little evidence of censorship. It was the weight in numbers of supporters who utilised the comments facility that caused negative comments to be hidden with many being responded to by other visitors. His opponent, the self-admittedly less technology-savvy John McCain, despite his deficiencies had a robust set of mechanisms for data collection, though less for feedback, but by 2012 for both US contests utilising upward communication flows appeared to be the norm. Hollande replicated the US model, Sarkozy was more circumspect, concentrating only on getting visitors to sign-up and join but not leave comments perhaps as a result of the perceived high likelihood that public feedback channels could be hijacked by his opponents. The Obama versus Romney contest saw both candidates utilising all available features to permit upward communication flow perhaps suggesting Obama showed the way for harvesting data while also being unable to roll back letting citizens comment as it would have been a news story in itself and Romney’s campaign felt obliged to follow suit. Allowing feedback, in particular, may indicate that campaigns have progressed to a more interactive paradigm.

Hyperlinking suggests a desire to connect with a wider online ecosystem, be that a support network, a media network or linking to sites to reinforce arguments. Separating out partisan, ideological, movement, media and factual sites we find that in France candidates mainly link to partisan or ideological sites, only Sarkozy used hyperlinks in 2012 to reinforce his arguments by linking to government produced official statistics. This is a sharp contrast to the United States where there are links out to media, though these are usually partisan media for reinforcement or exposing the excesses of opponents; for example, Obama referencing scare stories on Fox News on his own news blog. But in each contest it seems greater strategic value was perceived in having sites with lots of links embedded from stories and from the site itself. Arguably this may be the greater openness and inclusivity that we find in American campaigns where there is greater pageantry and public display of partisan affiliation. The public nature of campaigning may therefore also be represented in the use of hyperlinks to a range of sites.

Greater disparity is found when identifying features that permit asynchronous communication. While there are myriad ways in which visitors may leave comments, only the Obama campaign embedded more than 50 per cent of features that permitted two-way conversation across his Web presences. Yet, his opponents are not far behind when embedding tools that permitted asking questions, providing ideas or feedback. Therefore, we find a shift to a more interactive campaigning paradigm, and approaches differed in sophistication rather than potentiality. In other words, some campaigns offered only a few mechanisms, 4–5, others slightly more, but actually these features all performed very similar tasks.

For McMillan and many cyber optimists, synchronous interactivity is the gold standard of communication. For greater accessibility, engagement and empowerment, politics needs to be more interactive and in particular between elected representatives or those seeking election and the broader citizenry (Coleman, 2005). That is something of an ideal that has largely been unrealised, and remains the case across these campaigns to an extent. In fact, leaving Obama to one side as an outlier, the innovations witnessed in the 2007 French election stand out as representing a unique moment in electoral campaigning. Ségolène Royal’s invitation to supporters to co-author her ‘Notebooks of Hope’, effectively a manifesto for France, has not been replicated. Similarly, her encouragement of supporters to contribute to the Segosphere, a network of supportive but independent Weblogs, remains unique. Sarkozy, meanwhile, was dabbling with second life technologies, creating his own world on Ile de France where his leadership could be tested out. These practices were not repeated in subsequent contests, and in themselves did prove risky to some extent (see Lilleker and Malagon, 2010).

The Obama campaign offered every possible means for the campaign to be based on a conversation, yet in many ways it epitomised the description of controlled interactivity (Stromer-Galley, 2014). The Obama news Weblog received significant numbers of comments, mostly adulation from within the network of supporters who signed up to MyBarackObama his social network. Those joining had to provide a real name and US zip code in order to have freedom across his site; they were also targeted with requests to support the campaign through free labour or donations. The Obama site may have been a big conversation but there was very little political discussion and the network was not used to test out political ideas or to suggest policy. The Obama utilisation of the online environment was about securing victory using what may be seen as relational marketing tools (Lilleker, 2015) not about building an interactive political movement despite appearances.

Furthermore, Obama did not universalise a new paradigm for election campaigning. Sign-ups, donation tools and features that supported upward communication flows were utilised by campaigns that followed that of Obama. But, there was a Web 2.0 feel with Web 1.0 communication style. All candidates but Obama allowed interactivity only within spaces where it was unavoidable. As with other post-2008 European contests (Lilleker and Jackson, 2011), social media and Facebook in particular was the place where interactivity was permitted. This pattern of hiving off these activities was matched by Obama’s opponents in the United States. Therefore, arguably we find campaigns utilising the Internet purely for campaigning ends, the most innovative in terms of the political or democratic implications were the campaigns in France in 2007, the most sophisticated uses of technology were those of Obama in 2008 that were replicated in 2012. What we do not witness is a paradigmatic shift towards a more co-created campaign among these candidates. Campaigns appear to concentrate on informing and collecting information, communication between visitors and to the campaign are eschewed. Using Howard’s schematic we can view how campaign professionalism incorporates the online environment and how strategy evolved 2007–2012.

Campaign Professionalism in a Hypermedia Environment

As the Howard hypermedia campaign model focuses on campaign objectives the features are recategorized and there is a degree of conflation. Therefore, Table 2 offers an alternative perspective on digital campaigning and further insights into strategic design. In terms of transmission we find few differences, so again we find informing a priority for all candidates. Interestingly, some features that were prevalent in the 2007 and 2008 contests were not present in 2012. In particular, large archives of press releases, something that dominated the 2008 McCain site in terms of size, and other materials not originally designed for online publication. What this suggests is an end to the shovelware design, where offline materials are simply placed into the Website, and a more bespoke communication strategy tailored to the demands of the online environment and its users. In particular, when considering how the online environment is used to further the aims of campaign strategy, the use of Facebook and Twitter as newsfeeds demonstrates that even tools and platforms designed for interaction can be used for transmission alone.

Table 2 Adherence to the hypermedia campaign logic
Full size table

Yet, there were also indications of fairly low sophistication, in particular when creating content targeted at particular social groups. In the United States, the basic targeting involved a microsite for Hispanic speakers, and that was de rigueur for all candidates. Royal had a microsite for the young, this largely linked to the youth part of the Segosphere which in itself was segmented in various ways such as the large Women for Royal blogging group. Obama offered information tailored for numerous groups, even including those who were not his natural allies such as veterans. However, his statements for many minority groups in particular reflected the Democrat focus on more liberal policies towards same sex couples, ethnic minorities as well as specific appeals to women voters. Sarkozy in 2007 and 2012 and Hollande showed no evidence of targeting whatsoever. While this may appear of low importance, explicit inclusion of groups can suggest a candidate is reducing social barriers or inequities; having a uniform approach can suggest speaking to every citizen but may not appeal to groups who seek specific cues to form a bond with a candidate.

Conversely, an untargeted communication strategy may actually reflect a belief that having overtly targeted messages is inappropriate, and so targeted communication flows are under the radar via e-mail. Hence, data harvesting should be a central feature of any organisation’s use of the online environment. Every Facebook profile, sign-up offer, follow, or share, sends data back to the organisation and subsequently permits analysis and targeting on the doorstep (Lilleker, 2013). Obama utilising the knowledge of the Facebook team and Blue State Digital was most sophisticated in collecting data using most sign-up opportunities and harvesting e-mails and cell phone numbers. Obama’s strategy built to a significant extent on the activities pursued by Royal and Sarkozy in 2007, for Sarkozy, Hollande, McCain and Romney data harvesting was largely a single sign-up offer on the Website or join via social networking sites. The contest in 2007 in France saw no use of Facebook, therefore, candidates had to find other means of signing up and building communities, such as the Segosphere and Second Life. As social media became widely used campaigns saw less reason to innovate, equally data became available to purchase from Facebook or Google with the resources available to the US candidates. Commentary on campaign tactics show Obama was the most ruthless in connecting together analysis from big data and below-the-line targeting, using e-mail, texts and doorstep campaigning and so won the data harvesting and utilisation battle (Scherer, 2012).

Interaction via hypermedia conflates elements of upward communication with asynchronous and synchronous communication. Using the hypermedia categorisation, we see Obama remain an outlier but most campaigns offered myriad ways for visitors to respond to a campaign through private and public channels, all of which may be empowering and give a sense of accessibility to the visitor. Owing to the use of Weblog tools, multiple social networking sites and other sites where videos and pictures can be posted, Obama offered the greatest number of opportunities for visitors to comment back to the campaign or enter into dialogue with other visitors. In 2008, there was significant dialogue between official campaign organisers and members within the MyBarackObama network and on the news Weblog within the Website; this type of activity was witnessed to a much lower extent in 2012. The social networking sites were largely utilised as a newsfeed and attracted significant numbers of comments from fans and followers, but there is no evidence of the host responding. There may have been responses to e-mails or via private channels, but public dialogue was missing between the campaign and citizens who visited and left feedback, showing that even where synchronous dialogue is permitted it largely occurs only between visitors and not between them and the official campaign staff. Therefore, despite the range of interactions permitted across multiple platforms we find few of these opportunities result in meaningful dialogue between host and user. In terms of interactivity it is likely that those who chose to interact with the candidates in France in 2007 actually gained the most rewarding experience.

Discussion: Online Election Campaigning, Old Wine in New Bottles?

The comparisons of the campaigns of the six candidates who stood to lead their respective nations show interesting similarities and differences in innovation, sophistication and strategy. First, and perhaps most obviously, the campaigns by the US presidential candidates had more sophisticated online presences with wider functionality. This finding may be a symptom of resources or design but there was a clear disparity in overall terms in the various functions available between the nations and the extent of feature use that permitted certain functionalities. Perhaps demonstrating the greater marketization of the US politics over recent years, the Web presences of US presidential candidates also demonstrated greater segmentation with bespoke areas, policy documents and use of social media to reach specific voter groups. Although these were fairly broad groups, identified by gender, sexuality and ethnic roots it demonstrates some message targeting, in particular by the Obama campaign, to indicate these groups were of specific interest to the campaign and a desire to talk to them directly. French presidential candidates spoke to, and perhaps viewed their constituency, as the whole nation and, therefore, presented a single message to all visitors alike. The US presidential candidates also have Web presences that are linked into a network to a far greater extent than their French counterparts. Both Obama campaigns, and those of McCain and Romney, showed a much greater use of hyperlinks, which largely directed visitors to a wide variety of social media as well as mainstream media pages. The US candidates also linked to a number of Websites of political action committees which are semi-independent organisations that operate along partisan lines, as well as pursuing their own campaigns. These links specifically demonstrate the nature of the network of a presidential candidate and demonstrates how the partisan and ideological aspect of the campaign is maintained within the online environment.

While the data suggests French candidates were generally online laggards in terms of their innovation and sophistication, this does mask the important developments that were spearheaded in France in 2007 leading one analyst to claim that had it not been for the campaign of Segolene Royal Obama’s campaign would not have been as interactive (Serfaty, 2010). While one cannot discount the significant influence of Howard Dean, both as a presidential candidate in 2004 and subsequently as Chairman of the Democratic Party (Kreiss, 2012), nor the fact that Obama sought to surround himself with a young, innovative team of campaign advisors with specialisms in social media marketing (Harfoush, 2009), there is some indication that the Royal style of campaigning did inform Obama’s strategy. Independent of the influence flowing across the Atlantic, it was clear that Francois Hollande built on Royal’s experiences, positive and negative. Hollande eschewed allowing public voices to clutter or interfere with his campaign message but learned from Obama alternative, lower risk, ways to build a supportive network. In particular he developed the harvesting of data collection and was keen to permit feedback via private channels to the campaign including various sign-ups and encouraging supporters to work for the campaign offline and online.

While we might explain the differences in communication styles ideologically, with more progressive candidates being more interactive and inclusive, we can also argue that there were differing styles for incumbents and challengers. The interactive approaches in France in 2007 reflected the positioning of two untested presidential hopefuls. Similarly, in 2008, Obama was an outsider and challenger. The 2012 Sarkozy was more circumspect, defending his record using statistics and avoiding giving space for attacks. Obama, while retaining the interactive approach, ensured that he did not give space to opponents and so the campaign heavily controlled the ways and forms of communication allowed to visitors. Therefore, we find that even though interactive campaign communication is permitted candidates limit interaction to asynchronous or private forms of communication to the candidate, allowing synchronous communication to take place on their social media platforms where their visitors and followers can have a voice but do not have to be responded to officially. Hence campaigns remain in the realm of controlled interactivity independent of the extent they permit visitors to interact.

We therefore find that digital technologies have become a virtual shop window, a way of displaying the colours, images, key messages as well as the style of communication. The latter is argued to be an indicator of the character of the brand with i-branding, or interactive branding, literature arguing that offering ways for visitors to interact gives the impression of accessibility, inclusivity and openness (Simmons, 2007). In this respect many candidates attempted to harness supporters to create a multi-voiced, polyphonic brand character. Royal’s Website presented the words of myriad voices, a community that contributed to the Notebooks of Hope and the one that constituted the Segosphere of Weblogs. Sarkozy was supported by a polyphonic choir through NSTV, Nicolas Sarkozy television, a series of videos from experts proclaiming why he had the right policies for France. Obama, to use more American parlance, established a personal social network of cheerleaders. Obama therefore introduced the idea of having mobilised and active online supporters who would work for the campaign.

The latter became a feature of the social media campaigns in 2012 with supporters being encouraged to like and share content in order to spread the reach of the candidate, as well as have constant visible supportive endorsements transmitted that any user of the social networks might accidentally be exposed to. Hence, much interaction is related to onwards transmission with each action undertaken by supporters leading to enhanced visibility among the followers of a candidate’s supporters. Harnessing supporters to work for the i-brand became a crucial objective in 2012. Unlike the loose alliances built around the Segosphere in 2007, Obama’s 2008 campaign, as well as all 2012 campaigns gave visitors an invitation to join the campaign. Front and centre of the Websites were requests to donate to the campaign, join the party, and become a team member actively taking a role in the campaign online and offline. The only aspect that separated the candidates was a subtle difference in semantic style that appeared ideologically driven. The candidate of the left tended to use ‘WE’, so expressing being at the head of a movement but representing the people in that movement, candidates of the right generally used ‘I’ denoting their personal power over policy and their possession of a vision for the future of their country. Only Obama risked i-branding through synchronous interactivity, which provides the most concrete impression of accessibility, yet he remained an aloof non-participant. Yet Obama was by no means unique in 2012, social media meant that all candidates can develop an accessible i-brand co-created to some extent (Lilleker, 2013) and by not personally interacting, and there being no official responses to visitor comments by the campaign, all candidates remained distant and aloof from the ‘big conversation’.

Conclusion: Interactive Campaigning and the Reconfiguration of Democratic Engagement

In some senses Ségolène Royal paved the way for a new paradigm in election campaigning, one that exploited new technologies to meet key campaigning objectives. The lesson from her campaign was that if the opportunities were provided, supporters could be attracted and harnessed to the campaign. Building on the campaigns of Dean and then Royal, Obama offered a more inclusive style through his use of synchronous communication tools. He may not have attracted mainstream America, but in 2008 he did build a younger, more politically radical support base and mobilised them to work for his campaign. The year 2012 saw to a large extent the normalisation of these practices. Community building became the key function of Websites and getting community members to donate and be active supporters, harnessed to promoting the candidate’s i-brand, were key to the communication strategies of the candidates in France and the United States. Therefore, one can expect that election campaigns will see candidates and parties for election using digital technology to mobilise their supporters, increasingly finding ways to convert supporters into activists.

The challenge election strategists face is channelling the enthusiasm of activists away from trying to influence the design of policy and the campaign and towards simply accelerating messages through online platforms. It would be unwise for a candidate to eschew the like and share functions but synchronous communication is likely to be avoided as far as it is possible. Rather, supporters will be targeted with messages seeking their support only and online resource generation functions will be prioritised. The Obama campaigns in 2008 and 2012 relied heavily on small donations from supporters, much of that was generated online. In the United States, and possibly elsewhere also, earning money is likely to be a key function of the online campaign. The question is whether this can be achieved through simple ‘donate now’ buttons, even when they are backed by targeted and persuasive e-mail campaigns and social media advertising aimed at aligning the candidate with the concerns and desires of voter groups. The question is, and this question is virtually impossible to answer, whether the perception of accessibility accelerated loyalty levels among supporters and so drove the momentum of the Obama campaign. If so, his achievement was one other candidate in the United States or France managed.

Campaigning online may lead to the creation of communities that resemble grassroots, mass membership organisations. However, they will fit to a greater extent to Margetts’ (2001) cyberparty concept with a fluid membership, rather than Johansen’s (2012) traditional party where members are cognitively involved and connected into a real-world social network. Members will be encouraged to join an online community, give up their data and be contacted. Requests will encourage signing up to other social media outlets, liking and sharing and so promoting it through the online environment. The community will also be encouraged to work for the campaign in other ways, both online and offline with tools provided to support this within the community pages of the campaign Website. Smaller parties may opt for a looser, social-media-based, community where recruitment is easier and cheaper. The greater the resources the more emphasis will be put on data harvesting, database management, message tailoring and delivery in order to elicit further resources. What these four campaigns indicate is that the online environment has become an important battleground and, despite differences in sophistication, there is a clear convergence around key functions offered by digital communications technology. However, this is not a new interactive paradigm of campaigning but largely faux interaction, in Stromer-Galley’s (2014) terminology ‘controlled interaction’, designed to bind supporters to a campaign and create a short-term active community to aid with electoral victory.

One can argue that to some extent this may revitalise democratic engagement, encourage the conversion of those who join the online communities into full-time and active members and who become advocates. However, the counter argument would be that if the community is held together by a desire for key pledges to be implemented, or perhaps for the promise of a vaguer ‘new politics’ to be instituted, and this does not come into fruition then the community will soon disperse. Disengagement may equally result if candidates fail to communicate with those communities after Election Day, to explain policy directions and to continue to harness the support of their active supporters. If communities are held together only by short-termist objectives, and only willing to be involved in small donations and, more likely, occasional clicks to show support, then the revitalisation of democratic engagement is likely to be short lived without continued nurturing. These campaigns demonstrate that it is possible to have an interactive communication strategy, and that in the United States in 2008 and both 2012 campaigns the candidate who is most interactive may also gain some form of i-brand dividend. This dividend may result from the acceleration of their messages online, or through perceived accessibility, but it appears to be the case that the candidates who used a broad suite of interactive features, synchronous or asynchronous, emerged as the victor in 2008 and beyond. Although it is impossible to test the hypothesis that interactivity equates to votes, and so control for all other campaign and public opinion variables, it may be the case that i-branding can win over some latent supporters. Therefore, campaigns might be well advised to become more interactive, allow supporters to communicate with them, perhaps respond strategically, while also harnessing the enthusiasm of their supporters. The candidates and parties may find that this earns them further support and, in doing so, they might contribute to the reinvigoration of democratic engagement, participation and political debate more broadly within their nations.