Social representations of marketing work: advertising workers and social media

Social media has become a core tool in marketing. It is used in marketing research, retail and advertising. Recruiters for marketing posts even state that they want to appoint “digital natives”. Fortunately, marketing workers “love social media” such as Facebook and Twitter and spend more time communicating, socialising and sharing than other professional groups (Mortensen, 2013, p. 27). The purpose of this study is to explore how marketing workers represent marketing work with one another on social media.

It has been argued that contemporary workers discuss their work on social media for two purposes. One school of thought tells us that they do it to build their personal brands. Here, workers are likely to engage in impression management (Marder et al., 2016b; Marder et al., 2016c) to build cultural capital and extend their networks (Gandini, 2016). In the case of marketing workers, such findings may reflect the increasingly precarious working conditions of contemporary marketing work. Another school of thought tells us that workers’ engagements with social media “represent an employee-driven urge to make sense, to analyse and to exchange impressions of what labour is for, why we do it and what it means to us” (Schoneboom, 2011, p. 135). Rather than illustrate contemporary trends, this perspective sees social media use as a manifestation of a long-standing desire among workers to imbue their work with personal, social and cultural meaning.

This study uses social representation theory (SRT) in an attempt to integrate these two perspectives. This theoretical tradition, which forms a cornerstone of social psychology and yet has only had a limited impact in marketing (Penz, 2006), tells us that all groups must develop shared images, attitudes and “ways of understanding the world which influence action” (Potter, 1996, p. 168). These emerge as group members reflect on what they are doing and why they are doing it – allowing them to “feel a common identity since they have a common ‘world-view’” (Breakwell, 1993, p. 202). Viewed through this perspective, social media posts about marketing work may provide a parsimonious way to understand what marketing work is and how workers relate to it.

Starting from this perspective, the study investigates how advertising workers represent their work on social media through a structured thematic analysis of over 1,000 Twitter messages, including the associated images and external links. Advertising workers’ representations of marketing work are particularly interesting, as they experience many of the tensions inherent in all marketing work to heightened degrees (Kover and Goldberg, 1995; McLeod et al., 2011; Moeran, 2005). Indeed, existing studies show that advertising workers engage in a range of discursive practices to deal with these tensions, including joking (Kenny and Euchler, 2012), bullying (Hackley, 2000) and, importantly, discussing what it means to be an advertising worker in the first place (Grabher, 2001).

The analysis of advertising workers’ Twitter messages is presented through two themes. The first theme illustrates how marketing work is defined through representations of working practices, clients and colleagues. The second theme investigates different evaluations of marketing work by exploring how working practices, clients and colleagues are critiqued by marketing workers. One of the most surprising findings from the data is the extent to which marketing workers reflect on complex issues uncovered in previous studies such as power, sexuality and impression management and resistance. The key point for this study is not that these issues exist but that workers themselves represent them on social media in full view of clients, colleagues and managers. This suggests that, as a group, marketing workers use social media to build shared understandings that strengthen their abilities to work together and that social media posts may provide a parsimonious way for marketing managers to understand the nature of marketing work.

The paper proceeds as follows. First, the theoretical motivation for the study is set out, and the literature on marketing work and social media is reviewed. Second, the methods used in the study are described and justified. Third, findings are presented through two themes. Finally, the paper discusses the findings in relation to existing theory and offers conclusions, limitations and areas for further study.

Across industrial, service and knowledge settings, workers cultivate social connections with one another and share their experiences of working life (Korczynski, 2003; May, 1999; Stroebaek, 2013). Through these connections and experiences, they make work personally, socially and culturally meaningful (Braverman, 1974; Roethlisberger and Dickson, 1939; Roy, 1959). As a result, there is a long history of different work groups producing cultural representations of their experiences – including songs, poems, literature and art (Korczynski et al., 2013). Such representations allow workers to bond around shared experiences even though they might not share the same work environments (Payne et al., 2017).

Surprisingly little research has investigated marketing workers in this regard. Little is known about the “‘private’ concerns and lives of those who do marketing work” (MacLaran and Catterall, 2000, p. 643). Marketing management is theorised instead as “something beyond human life and social practice” (Svensson, 2007, p. 272). This is not to say marketing work is ignored completely. The marketing practice literature provides descriptive accounts of firm-level activities (Brodie et al., 2008). Ethnographers, organisational researchers and cultural sociologists have explored the nature of advertising work (Alvesson, 1994, 1998; Grabher, 2001; Koppman, 2014; Moeran, 2005). Here, though, the social aspects of marketing work tend to be viewed as something that benefits either an individual worker’s career, their organisation or team (McLeod et al., 2011). For example, “periods of idleness” in-between projects have been viewed as an opportunity for knowledge sharing and network building between advertising workers (Grabher, 2001, p. 368). Likewise, interpersonal relationships have been viewed as a way to improve employee retention and, through this, company success (Crutchfield et al., 2003). We need to go back to Elliott and Margerison (1977) to find a sustained analysis of marketing work in and of itself.

Social media provides workers with the opportunity to augment the traditional forms through which workers have represented their work. Nielsen (2013) observes that, as workers become “accustomed to relying on social networking sites and smartphones in everyday life, it would require a conscious choice to organize in ways that were not, in part, reliant on these tools” at work (2013, p. 175). Can such digital representations of marketing work be used to understand marketing work? Emerging literature has begun to explore how workers represent their work on social media. Two perspectives have developed here: an individual perspective and a social perspective.

Social media research developed in the early 2000s. It explores how workers use social media within their everyday work activities. It suggests, for example, that workers might use social media to share knowledge among virtual communities of practice or communicate more efficiently with colleagues in different locations (Riemer and Richter, 2010). As a result, it has been argued that digital representations of working life provide researchers with “an unmediated glimpse into the world of work” (Schoneboom, 2011, p. 133). They confer upon “the private realm of daily experience a public audience” (Zappavigna, 2012, p. 38). Rogers (2013), for instance, proposes that social media posts to Twitter can now be interrogated as a naturalistic data set (Ferro et al., 2012). Based on this proposal, Shami et al. (2014) suggest that workers should be encouraged to use social media, as it can allow managers to unobtrusively monitor their performance.

However, reports on actual social media use in specific organisations conclude that, as a productive tool, social media may be limited (Archambault and Grudin, 2012). Indeed, researchers criticise the idea that social media data provide an etic perspective on working life. Research with consumers suggests that social media posts are performances (Kerrigan and Hart, 2016). Rather than show realistic accounts, conscious of the public nature of their posts, workers edit, style and censor their posts for particular audiences and manage their impression to others (Marder et al., 2016c; Stibe et al., 2011; Zhang et al., 2014). Marder et al. (2016a), for example, suggest the most “powerful group” determines self-presentation styles on social media – labelled the “strongest audience effect”. Likewise, Zhang et al. (2010) report on the use of a proprietary microblogging service within an unnamed Fortune 500 company and conclude that workers engaged in virtual “impression management”.

To explain why individual workers use social media, researchers have drawn on Goffman’s (1956, 1997) theory of self-presentation. Workers’ use of social media platforms such as Twitter has been accepted as an instance of self-branding (Hearn, 2010; Labrecque et al., 2011; Page, 2012; Zhao and Rosson, 2009). Gandini (2016) argues that social media posts by knowledge workers should be seen as “digital work” centred on the “acquisition of a reputation”. Based on interviews, he interprets social media posts as “an investment in social relationships with an expected return” (2016, p. 124).

In sum, the individual perspective suggests that marketing managers can look at social media posts to understand how workers perform their working identities, build networks and develop their careers. This suggests that marketing managers could use social media posts to evaluate new staff or monitor individual members of their teams. However, it leads us to treat representations of work with suspicion and assumes that any images workers offer will be edited in some way to make the worker look good.

A second stream of research within the field has explored the ways workers use social media to form relationships with others. That is, it focusses on the “affiliative function” of social media posts (Zappavigna, 2011, p. 799). Zappavigna (2012), for instance, notes that people use social media generally to bond “around collective complaints about life’s little daily irritations” (2012, p. 5). In particular, by adding a hashtag, certain aspects of a message become “‘hyper-charged’ with an additional semiotic pull that may be likened to a gravitational field” (Zappavigna, 2011, p. 801). This attracts like-minded users to offer their own contributions and builds a “copresent, impermanent, community […] around evolving topics of interest” (Zappavigna, 2011, p. 800).

Early work here found that, irrespective of whether posts were realistic or not, microblogging was valued by workers as a way to keep in touch, share common experiences and communicate one’s feelings. It thus increased social connectedness among groups of workers (Zhao and Rosson, 2009). Expanding this, and informed by work sociology, it has been claimed that social media posts by workers represent the long-standing “employee-driven urge to make sense, to analyse and to exchange impressions of what labour is for, why we do it and what it means to us” (Schoneboom, 2011, p. 135). Indeed, research explores how social media is used by workers to access professional groups beyond their specific workplaces with studies investigating how unions engage with social media to mobilise worker solidarity (Upchurch and Grassman, 2015).

So, in contrast to individualistic explanations of workers’ use of social media, which focus on the function of social media posts for an individual’s career or identity, social explanations focus on the ways groups form around social media posts. Rather than relying on individuals’ rationalisations for their social media posts, such research tends to explore the posts themselves and often focusses on negative depictions of working life. This implies that marketing managers could use social media posts to understand common issues shared by marketing workers. But, because the posts tend to focus on negative depictions, marketing managers might well be left with an overly pessimistic image of marketing work.

One problem with these two perspectives is that they explain different types of social media posts in different ways. Individualistic accounts assume that social media posts tell us something about workers. They look at workers’ own explanations for positive depictions of their work – which are viewed as edited and stage-managed. Social accounts assume that social media posts tell us something about work. They focus on negative depictions – which are seen as the most effective way to form social connections.

In an attempt to integrate these two perspectives and provide a theoretically derived justification for utilising social media posts to learn about marketing work, this study introduces a different theoretical tradition: SRT. It is a cornerstone of social psychology and underpins many popular analytic techniques in discourse analysis (Potter and Wetherell, 1987). It is based on the idea that, for any group to function, it must construct shared understandings of the objects it works with. Irrespective of whether these understandings are edited or realistic, once they are shared they can form a basis of action. Analytically, such work is necessary before groups can work together (Potter, 1996). However, practically, it forms an ongoing part of action, as group members reflect on and represent what they are doing and thinking (Howarth, 2006). Indeed, it is a point of debate whether social representations – strictly speaking – are produced apart from action or are an ongoing part of action.

SRT was originally set out by Moscovici (1963). He defines social representation “as the elaborating of a social object by the community for the purpose of behaving and communicating” (1963, p. 251). In other words, social representation is both a process and the outcome of that process. As a process, social representation occurs as groups objectify elements of the material and social world – taking them as real things – and as groups create anchors that allow individuals to relate those objects. In this sense, social representations tell us about an object, a subject that represents it and the social group to which the subject is positioned.

A key aspect of SRT is an acceptance of cognitive polyphasia. It “implies that different and incompatible cognitive styles and forms of knowledge can coexist within one social group and can be employed by one and the same individual” (Voelklein and Howarth, 2005, p. 434). This challenges cognitive theories of attitudes which depend on individuals possessing internally consistent attitudes towards a given object (Gaskell, 2001, pp. 228-41). It also means that social representations research is less interested in issues of measurement and, more often, is concerned with establishing precisely what people think about, how they think and what they are able to do because of their representations (Voelklein and Howarth, 2005, p. 437). Moscovici (1963), for example, argues that social representations do not always become manifest in language. Rather, they may work through art, photography, newspaper articles and other cultural media. Other researchers have explored how issues of power and ideology affect social representations (Howarth, 2006) and how conflict and cooperation occur in social representations (Wagner et al., 2000; Potter and Billig, 1992).

The tenets of SRT have consequences for marketing managers seeking to understand the experiences of their workers. First, SRT suggests that effective action occurs when groups are able to engage in the social-representation process. Through this, they can develop social representations that help them to do things. Second, understanding how and why people act in concert requires an understanding of their social representations. Third, social representations are best accessed through naturalistic data. As such, SRT provides a theoretical foundation that integrates the individualistic and social perspectives of social media and work. It suggests that when workers choose to share images of their working lives, they are engaging in a social representation of their work. They are objectifying it and anchoring – irrespective of whether they are individually motivated to do so for the benefit of their career or identity or because they want to access a wider professional group for support. This provides theoretical justification for the idea that marketing workers’ representations of marketing work on social media may provide a parsimonious way for marketing managers to understand the nature of marketing work and the experiences of marketing workers. To explore this, the study is driven by the following research question: How do marketing workers represent marketing work on social media?

This study aims to explore how marketing workers socially represent their working lives on social media. In this regard, it is notable that advertising workers developed a hashtag to mark social media posts that concerned their working lives: #agencylife. This hashtag emerged during the spring of 2013 with the advertising industry press reporting that advertising workers were increasingly using Twitter to share their experiences of marketing work. Ad Age stated: “If you’re an ad person, and you’re on Twitter, you’ve probably caught at least a few #AgencyLife tweets in your stream”, the hashtag “caught on like wildfire […] in the span of just a few hours, with hundreds of tweets by ad people eager to poke some fun at themselves and blow off some steam” (Parekh, 2013). The Huffington Post (2013) ranked the hashtag as one of the top business-related hashtags “for its insight into the thinking of the media and agency world”.

Posts to #agencylife provide an opportunity to explore how marketing workers socially represent their working lives on social media. Advertising workers have long been used to research marketing work. It is thought that their creative working practices and institutional arrangements mean they not only experience but also actively discuss many of the inherent tensions of marketing work on a day-to-day basis. For instance, they must be artistically creative yet produce business results (Kover and Goldberg, 1995; Michell, 1984; Moeran, 2005). They provide an expert business service but lack professional qualifications to prove their worth (Alvesson, 1994; McLeod et al., 2011). They work with clients to produce effective communications but frequently feel abused, limited and insulted by them (Alvesson, 1998). Moreover, advertising workers are particularly interesting because the creative nature of much advertising work leads them to use a range of discursive tactics to reflect on, describe and deal with these tensions. Studies show us that advertising workers joke (Kenny and Euchler, 2012), sexualise (Hackley, 2000), critique (Pratt, 2006) and, importantly, discuss what it means to be an advertising worker (Grabher, 2001). Given this research background, #agencylife is an appropriate case to enrich the understanding of marketing work.

Social media data can be analysed using computational approaches such as network analysis and natural language processing (Noguti, 2016; Tinati et al., 2014). These techniques either rely on well-structured metadata or focus exclusively on the linguistic content of messages. For instance, natural language-processing techniques ignore other rich data included in many social media posts such as photos, videos and external weblinks. In keeping with social representation theory, the study was designed to take account of the domain-specific nature of the discussions and the rich, intertextual nature of the messages. Computational methods were used to sensitise research to the data. An interpretivist approach was used to allow representations to emerge from the data.

From October 2013, tweets marked with #agencylife were harvested in real time using the Twitter developer console (https://dev.twitter.com/console). At the time of the study, this was freely available to use. A bespoke algorithm was created to capture tweets from Twitter’s “Gardenhose”. This is a randomly sampled stream taken from the larger “Firehose stream” of all public tweets. Across 12 months, three batches totalling 17,553 messages were collected (Table I). Non-English-language posts and bot-generated tweets were manually removed from the data set. Each tweet was indexed with a sequential identification number. The sample is illustrated in Table I.

A structured thematic analysis as set out by Braun and Clarke (2006) was used to parsimoniously summarise the main representations in the data. To begin the analysis, the first 100 messages from the first sample were open-coded to create an initial coding structure. To do this, two coders were asked to summarise each message. Both coders were familiar with the advertising industry and were native English speakers. When tweets included mentions, hashtags and weblinks, these were included as the content of the tweet. Coders followed all links to contextualise the meaning of the tweet. Importantly, they ignored sentiment, irony, sarcasm and tone in all of the analyses. For example, positive and negative references to co-workers were treated equally when coded. The reason for ignoring sentiment and tone is a pragmatic one. The coders could not know categorically if a reference was ironic or sarcastic. Multiple codes were allowed for each tweet if deemed necessary by the coders. After 100 tweets, saturation was achieved with no new codes added after Tweet 70. This initial code scheme included 24 codes. Following this, the coders reviewed a second extract of 350 messages. This sample comprised all of the messages that had been “retweeted” at least once during the first two sample periods. This confirmed the validity of the initial coding structure and tested whether retweeted messages were substantially different from others. During this round of coding, the scheme was revised to produce a more stable list of 17 codes (Table II Frequency codes). No new codes were added. Again, coders were instructed to ignore sentiment, use multiple codes where necessary and draw on all information conveyed through a tweet. A third sample of 674 tweets was selected randomly. In total, 1,014 unique tweets were investigated (including retweeted messages, the total sample analysed was 2,101 tweets). At this point, theoretical saturation was reached. These tweets were produced by 583 individual users across five continents. Within this sample, only seven users tweeted more than ten times. The lead user tweeted 45 times.

A final round of axial and relational coding was undertaken by the researchers to parsimoniously summarise the dominant themes in the data. This was structured through SRT. First, axial coding focused on the ways marketing work was objectified. Second, axial coding focused on the ways marketing work was anchored. Each theme is illustrated in detail below with pertinent examples from the data interpreted to demonstrate the nature of the theme.

All usernames, aside from celebrities and institutions, have been anonymised using the conventional @user profile name and all links replaced with http://link. All images were taken from publicly available websites linked through Twitter whose terms of service inform users that images not marked as private may be used for research purposes. The relevant research ethics committee at the authors’ universities approved this research. The images and text of messages were made available to reviewers during the peer-review process. However, because of the publisher’s copyright policies, images have been removed from the published version of the manuscript.

The first theme developed from the analysis stands in harmony with many of the existing studies of advertising workers. It involves advertising workers using their posts to represent what it means to be an advertising worker. Traditionally, such representations are viewed through a lens of individual identity work (Alvesson, 1994). It is argued that it is necessary for advertising workers to invest time and effort proving themselves to others – specifically, their clients and colleagues (Pratt, 2006). In so doing, they must carefully balance their individuality, creativity and cultural distinction (Kover and Goldberg, 1995) against the need to conform to industry standards, professional norms and client needs (Alvesson, 1998; Hackley, 2000). In this section, such images are interpreted instead as social representations through which advertising workers collectively objectify what it means to be an advertising worker.

Posts to #agencylife highlight everyday experiences in advertising organisations. Workers represent projects they are working on and offer a real-time commentary on things happening in their offices. In general, they present their work as fun, creative and innovative but demanding, rigorous and stressful. For example, in Tweet 1668, the text reads: “Going home. #agencylife”. This is combined with an image of an in-car entertainment system. It displays a clock reading 11:46 p.m. and shows that the driver is listening to Seu Jorge’s Portuguese-language version of David Bowie’s song “Rebel Rebel”. The text and image seem to demonstrate the sender’s commitment to advertising work. They are going home at nearly midnight. But they are not rushing home. They have time to capture the moment and share it on social media. Indeed, it is notable that the image shows them listening to a song about rebellion, sexuality and gender performance. However, this post can also be seen as a claim about the nature of advertising work itself – not just the activity or commitment of a single worker. The combination of the text and image with #agencylife means that, as much as this is a potential identity performance, it is also a statement about the nature of advertising work in general. That is, living the #agencylife means enjoying driving home late at night and seeing this as something unique and rebellious not a chore. Other posts repeat this kind of sentiment such as: “13 hour days? Childsplay when your passion is what you wake up for every morning. #advertising #agencylife” and “Sneaking in some work at #Interact14! #agencylife http://link”. This links to an image of two workers sitting on a floor surrounded by crates of drinks working on laptops. It represents advertising workers as always working and always enjoying it.

Messages also represent the working environments of advertising organisations as fun yet productive spaces. Posts show people playing ping-pong, pool, table football and video games in designated spaces in their workplaces. In Tweet 1920, there is an image of two colleagues at a desk playing a toy game where the players control plastic boxing figures. It is accompanied by text reading: “Always find time to fit in a little fun. #office #agencylife #planit #toy http://link”. Likewise, an image of a pool table with balls dispersed across it in Tweet 2096, is combined with text reading: “We’re giving our brains a lunch break courtesy of our #office pool table. #AgencyLife http://link”. An image of advertising workers sat on modern sofas playing a computer game in front of a large screen in Tweet 1220, is followed by text reading: “Work hard, game hard. This is how we get down at the #Companyname office. #AgencyLife”. In each of these examples, the workers can be said to not only perform identity work but, as they do so, they also represent advertising work as a particular form of productive play. It is hard work but always involves some element of fun. Indeed, in the last example, gaming is, itself, presented as an intrinsic part of hard work. Advertising workers have fun but work hard – and work hard by having fun.

One reason advertising workers may need to represent what they do as fun but productive work is that clients value agencies that offer them something different, but they also demand that they produce results. Indeed, as represented on social media, relationships with clients form a key component of advertising work. Advertising workers post messages to #agencylife to demonstrate their commitment to their clients. Clients are objectified as partners who help produce successful advertising. Messages project creativity, taste and distinction onto clients. They show lists of clients displayed on agency walls, advertising workers preparing gifts for the clients and staged photos showing “client visits” to agency offices.

Likewise, client approval is objectified as a key reward of advertising work. An illustrative message here reads: “My clients are my everything”. Another states: “When a client calls and says we ‘blew it out of the water’ – that’s like the best thing #agencylife #pr #design”. Such messages may be intended for specific clients. However, the public nature of the messages means that other clients and workers may see them as well. Therefore, rather than limiting our interpretation to digital branding, we can also interpret them as social representations for other advertising workers. Specifically, they objectify client relationships as a key part of advertising work.

Just as messages objectify their relations with clients as part of advertising work, they also objectify relations with colleagues as an essential part of advertising work. In particular, advertising workers contribute messages that demonstrate their successful socialisation into an advertising community. But, in sharing these images, they also construct that community. For example, many messages comment on shared fashions and clothing styles among agency workers. In so doing, they not only define which styles identify an advertising worker but also emphasise the importance of fashion itself for the community. In Tweet 1768, for instance, two agency workers state that they dress identically because they work together so often. The message shows two women in similar-coloured clothes. The text reads: “We work together too much. #Matchymatchy #agencylife #dailygrind #notplanned #workit #ootd”.

Seen through the lens of individual identity performances, such images can be said to demonstrate individual workers’ cultural distinction (Alvesson, 1998). In many of these examples, the messages include #agencylife and #ilovemyjob or #lovemyjob, which suggests that workers intend their posts to demonstrate their socialisation into and passion for advertising work. However, the messages also demonstrate that advertising work is not just an individual identity. It is a community brought together by its shared emphasis on fashion.

Further emphasising this is a series of messages which represent advertising workers as a “family”. Tweet 1810 depicts an agency worker blowing out birthday candles; the wall behind this activity is filled with images of people, presumably agency workers, socializing and participating in activities. This is also coupled with #lovemyjob: “At work we’re like a family and birthdays are kind of a big deal. #agencylife #lovemyjob”. Posts show how these work families support one another in the execution of advertising work. They celebrate birthdays, holidays and festivals. They develop shared rituals such as “#fridaybeercart theme is ‘Shots for a shot at the Mystery Box of Goodies’. @companyname #agencylife http://link” displayed in Tweet 1559. It is combined with an image which depicts the bird’s-eye view of a trolley with a bucket containing bottles of beer, wine and other beverages. On the trolley there are also slices of lime and a bottle of tequila laying on a very large bowl of ice-cubes. Some messages even show workers sharing childcare duties with one another at work, along with pets and other family members visiting the workplace. The notion of a family is telling: a family is a single social unit but it contains different roles and identities and is made up of individuals.

Social media posts about marketing work allow advertising workers to objectify what advertising work is. In particular, they objectify it through a series of productive differences. Advertising workers present potential sources of tension and conflict as things that actually help them to do their work. Through this they show that advertising work is different from other forms of knowledge work. Yet, these representations do not focus exclusively on differences. They might represent their differences from their clients, for example, but also make a point of praising their clients. Similarly, advertising workers use the hashtag to objectify the community of advertising workers as a single social group – a family. Often, they do so by representing their place within that family. Yet, at the same time, as they construct their own work identities, they objectify a wider community and distinguish it from other groups.

A second theme revealed through the analysis emerged by considering the range of attitudes represented in the posts. This interpretation focussed less on what was being represented and more on how it was being evaluated within the messages. In this sense, they define common elements of advertising work that are good and bad and define acceptable justifications for what makes them good and bad. This section demonstrates how particular elements of advertising work and relations with clients and colleagues are presented negatively through mundane, reflective or critical images posted to the stream. Such activities have been theorised through a lens of resistance (Kenny and Euchler, 2012). Here, they are interpreted as instances of anchoring. In reality, of course, many posts both objectify and anchor at the same time. These two processes are separated here for illustrative purposes. Equally, advertising work is not only anchored negatively. Many positive images are offered above.

Some posts linked to #agencylife represent the frustrations, mistakes and catastrophes involved in advertising work. They not only describe what advertising workers do but also represent particular aspects of advertising work negatively. For instance, a number of messages include links to a visualisation of the creative process which splits the work into five stages. First, work begins. Then there is a long period labelled “Fuck off”. This is followed by a short period labelled “Panic”. This is followed by a shorter period labelled “All the work while crying”. Finally, there is the “Deadline”.

Likewise, many messages present advertising work as a dull office job and advertising workers as stressed and damaged people. They present a direct contrast to the fun work activities represented above. Messages such as Tweet 1722, which shows an agency worker sitting at a laptop at a mundane desk, suggest that agency life is not unique. Many other messages show people working at computers, looking like any other office worker. Several include images of empty offices, indicating that workers were working alone out of hours, and include comments such as “Office alone #agencylife”. Tweet 1343, for instance, shows a photo of a conference table without people around it. It is littered with laptop computers, cables, pens and food wrappers. It reads: “This is just a sad sight. But thanks @ItsuOfficial for the sustenance […] #agencylife”.

Some messages represent responses to other messages about the nature of advertising work. For example, commenting on the time commitments involved in advertising work, one post includes an image of a wall calendar and centres on text reading “MLK Day (closed)” – under which someone has added “Liar”. The text simply reads “#agencylife”. This post can be interpreted as critiquing the rewards of late nights objectified above. It emphasises that when the office is closed, workers feel expected to work. As such, it can be taken to represent advertising work as a source of stress, anxiety and precarity.

Messages also present contradictory images of the spaces and organisations around advertising work. In place of the stylish office spaces objectified above, messages share photos of wires running across offices and projectors not working. There were messages showing people crammed into confined spaces. Tweet 1341 shows an advertising worker hunched over a desk, sitting on a stool; the tweet reads: “I don’t think @companyname is promoting good posture […] #agencylife”. Another tweet includes an image of brown liquid pouring from the ceiling of an advertising agency office – which it claimed was faeces.

Posts routinely criticise clients for impeding and misunderstanding the creative process. In contrast to the idea that clients are partners in advertising work and that client approval is the ultimate reward in advertising work, here it is individual freedom that is emphasised. A large number of #agencylife tweets include #thingsyouhearinagencies. Posts combining these hashtags include short quotes that represent typical occurrences in advertising work. Some are celebratory but many represent impediment and the challenges that advertising workers face. Many are critical of clients for their ignorance, cowardice or short-sightedness. A number of messages link to a website explaining the different ways that clients “killed” good ideas. These range from the influence of a new but uninformed marketing manager to the misuse of a focus group. Others link to a blog called “This Advertising Life”, which uses “internet memes” to express daily the frustrations of advertising work. They include .gifs (graphic internet formats which are usually small, low-quality digital animations) of angry characters from popular culture texts with added comments such as “When I’m told to shorten a 25 character headline”. Similarly, many tweets point to an art project by Sharp Suits – two advertising workers who made posters from “typical” client comments, including “The sandwich needs to be more playful”, “I really like the colour but can you change it” and “Do you have a Mac or PC? I have a Dell”.

Although advertising work is objectified as a community activity – a family – this does not mean that it is always harmonious. Just as existing studies have identified power relations and resistance at work within agencies (Hackley, 2000), so too, advertising workers used the hashtag to share images with one another that emphasise how their colleagues exert power over them often to the detriment of their work. In this regard, a number of messages are linked to a Tumblr site called “Hovering Art Directors. http://link #designer #problems #agencylife” (Hovering Art Directors, 2013). This site encourages agency workers to post photos of their managers standing behind their colleagues. They show managers leaning over to correct people’s work, watching over them with notepads or simply holding meetings behind people trying to work. The coupling of these images with “#designer” and “#problems” implies that such oversight is not represented as a good thing.

Social media posts allow everyone to share complaints and frustrations while being aware that these performances are available for public consumption and that such posts may be staged or edited. In this case, advertising workers adopt this discursive practice to criticise clients, colleagues and management. One interpretation of these posts is that advertising workers are sharing their negative experiences as a form of catharsis. However, they also perform an important collective task. They allow advertising workers to collectively agree on what is good and bad about their work. Interpreted through an SRT perspective, such posts can be said to allow advertising workers to anchor a range of attitudes towards advertising work. While advertising workers objectify their work as fun, creative and rewarding, they also acknowledge that this is not always true. It is not the case that one of these representations is more real than any other. Rather, they allow workers to develop a shared understanding of the positive and negative aspects of advertising work.

This study set out to explore how marketing workers represent marketing work on social media. Specifically, it focusses on Twitter messages related to work, including the associated images and external links. Existing theory suggests that workers will share images of work for two reasons. Either they present stylised images for personal branding or share critical images to bond with others. In contrast, the study started from an assumption that an element of any individual’s representation of their work has a social function. It allows them to collectively define the objects they are working with and it allows them to share common attitudes toward those objects.

This study has explored how advertising workers objectify advertising work as a set of activities and tasks that take place in particular types of organisational settings and as a set of relationships with clients and colleagues. These tasks, settings and relationships are represented positively and negatively within the posts. This, in turn, allows workers to share common understandings and to justify what is good and bad about advertising work. For example, the study illustrates how relationships with clients can be taken as a rewarding feature of advertising work and also an impediment to successful work. In short, the relationships are both good and bad.

This cognitive polyphasia may be particularly useful for advertising workers, given the precarious and fluid working conditions in the industry. Research demonstrates that advertising workers need to develop an ability to move into different teams (Grabher, 2001), shift identities (Hackley, 2000; Pratt, 2006) and adapt new styles of working (Nyilasy and Reid, 2009) on a daily basis. Possessing shared ideas about what they are doing is an essential glue that holds these relations together. Moreover, it allows individual advertising workers to justify success and failure without giving up. If a worker can blame their client, working conditions or colleagues for undermining their ability to produce good advertising one minute and celebrate them the next, it provides them with powerful discursive flexibility.

This recalls Potter and Wetherell’s (1987) analysis of academics. They would decry flaws in peer reviews to justify negative assessments of their studies and celebrate the rigour of the process when their other research was published. For Potter and Wetherell (1987), the ability to move between these two discourses – peer review is flawed, peer review is not flawed – was essential to academic communities. It allows individual academics to take credit for their successes and blame others for their failures. This works because both discourses are accepted within their academic communities. In other words, academics know that others will call them out neither when they highlight flaws in the peer-review process nor when they emphasise its strengths. Both are taken to be true. Examining the representation of advertising work on social media allows us to see advertising workers supporting similar contradictory ideas.

Along with helping to bring together positive and negative depictions of advertising work within a single explanation, another benefit of this theoretical approach is that it does not require us to reject the individual or social perspectives. Rather, it allows us to accept that, in many cases, individuals might be consciously attempting to build their brands or access a wider community, but, as they do so, they objectify and anchor marketing work. They social represent marketing work.

Indeed, one question which has occupied social theorists, including marketing theorists, concerning social media is to what extent the things people share are realistic. Research within the individualistic perspective suggests that images of work offered by workers on social media should be treated with caution. People tend to edit how they appear through impression management. In contrast, research within the social perspective says that workers are motivated to share their experiences with others to give work meaning. While they might not offer literal representations, their representations will resonate with their experiences on some level.

From an SRT perspective, this question misses the point. SRT assumes that social objects are real in the sense that they form the basis of cognition and communication. Put simply, marketing work is what marketing workers agree it is. So, instead of asking workers why they share images of their work, emphasis should be placed on exploring how they represent their work. A consequence of this is that the ways workers represent their work on social media may have important insights for marketing managers about the nature of marketing work and marketing workers.

This study demonstrates that social media offers a way for marketing managers to understand the experiences of marketing workers. Building on social representation theory, it assumes that when marketing workers share images of their work on social media they are engaging in objectification and anchoring. That is, by sharing images of marketing work on social media, they define what marketing work is and how they can relate to it. Through these representations, marketing workers build shared understandings that strengthen their abilities to work together. The point is not that one of these shared images is more correct than any other but that marketing workers need them to do marketing work.

A key implication of this study is that marketing managers can use social media data produced by their workers in marketing management. The study shows that marketing workers are willing to share common aspects of work in a public forum in full view of clients, colleagues and managers. Indeed, they portray many critical elements of marketing work that have been identified in existing academic research. Exploring these social representations of marketing work allows marketing managers to integrate a better appreciation of marketing work and marketing workers into their practice. This marks a key difference from recommendations offered in existing studies which encourage managers to focus their attention on using social media to monitor and evaluate specific marketing workers, consumers or marketplaces.

The study opens up a number of avenues for further research. First, the analysis considered the social media posts in isolation. The analysis gave equal weight to all posts when, in reality, social media are viral in nature. Some posts effect the direction of a discussion, and others are ignored. Some senders have the power to influence discussion or post with more freedom. Further research could investigate the patterns of consumption of social media posts and the networks between marketing workers and even present posts back to marketing workers to explore how they account for the things they share with each other. Second, the study focussed on a hashtag generally associated with advertising workers. Further research could look at other forms of marketing work, for example, market researchers or data analysts. Third, the study considered the personal information presented on a user’s profile, but it is possible that some posts included were not produced by marketing workers. For marketing managers to learn through social media, they would need to develop ways to filter such information. This raises issues relating to privacy and the potential of feedback loops – where workers change their offline and online behaviours because they know or expect their managers are monitoring them.