Brand Awareness through Co-creational Storytelling on Instagram
A Case Study on the Zalando #shareyourstyle campaign
Master Thesis
Graduate School of Communication, University of Amsterdam
Master Corporate Communication (MSc)
Sibylle Mittrach Student-ID 11106336
Supervisor: Dr. Pytrik Schafraad Submitted: 24.06.2016
Sibylle Mittrach Brand awareness through co-creational storytelling
Abstract
This study examines to what extent co-creational storytelling with hashtags on the Social Media platform Instagram can increase brand awareness. The analysis was performed as a case study on the online retailer Zalando. Using as theoretical foundation, the concept of user-generated content and electronic word-of-mouth, as well as theories on innovation diffusion and storytelling, co-creational storytelling in Social Media is comprehended and the
applicability of it through hashtags becomes the focus of attention in this study. With a mixed-method approach of automated semantic networking analysis, visibility analysis, and manual content analysis, all Instagram posts tagged with #shareyourstyle and posted between August 31st
, 2015 till March 15th
, 2016 were examined. Results show, first, hashtags are used to create stories on Instagram. Second, awareness for a brand can be generated through co-creational storytelling, as long as the brand takes efforts to synchronize the co-created stories into a brand narrative. Third, hashtags help users on Instagram to take part in conversations and network with each other. The findings of this study contribute to the gap in academic knowledge about co-creational storytelling, hashtag diffusion, and awareness building on Instagram, as well as to practical knowledge on the utilization of co-creational storytelling by corporations.
Introduction
The Social Media platform Instagram is no insider’s tip for sharing nice pictures anymore. With more than 400 Million active users and more than 80 Million new photos every day, the photo-sharing mobile phone application is one of the most used social networks worldwide (Instagram/press, 2016). Although the platform did change their algorithm recently, the app is largely free from advertisement in the sense of visible ads on other Social Media websites like Facebook. However, the number of corporate accounts on Instagram is increasing because corporations recognize the possibility of boosting awareness for the brand and start utilizing the platform as a commercial tool.
Brand awareness, which is defined as “the extent to which costumers are familiar with distinctive qualities or image of a particular brand” (Barreda, Bilgihan, Nusair, & Okumus, 2015, p. 600), can make an important contribution to brand image, brand equity1
, as well as market share. When consumer are aware of a brand, they recall the brand much faster in a situation they want to purchase a certain product, or they recognize the brand easier, when faced with different brands (Barreda et al, 2015). On Social Media platforms, like Instagram, awareness for a brand can also be built up in collaboration with consumer. User-generated content (UGC) on Social Media sites is a way for users to express themselves and to
communicate with others. It can be defined as content that is published online on any form of Social Media platform and outside of professional context (Smith, Fischer, & Yongjian, 2012).The UGC that emerges on Social Media websites encourages interaction with other users and recognition in the peer (Nuttavuthisit, 2010). The utilization of this content by brands and the inclusion of user-content into the branding process, can lead to more favorable consumer responses as well as to a brand that is matching better to the target group (Barreda
1 Brand equity can be defined as the „differential effect that brand knowledge has on consumer response to the
marketing of that brand“ (Washburn, Till, & Priluck, 2004, 488).
Sibylle Mittrach Brand awareness through co-creational storytelling 2
et al, 2015).Similar as to Twitter- or Facebook-hashtags, hashtags on Instagram “can be viewed as topical markers, an indication to the context […] or as the core idea expressed” (Tsur & Rappaport, 2012, p. 643) in the post. But hashtags can also be used as hyperlinks to tell a story to fellow users or to contribute to a story from fellow users. For brands this is attractive, as brands can become meaningful to consumers when a link between the brand and the consumer’s self-concept develops, which can happen through shared stories (Escalas, 2004). Consumers map a corporate story onto stories in their memories and get closer to the brand and its narrative. The idea of establishing branded or brand-related hashtags, that users use to connect the user-generated stories with corporate image or reputation to show their affiliation for the brand, raises by corporations the hope to build awareness for their brand by co-creational storytelling with consumers.
So far the scientific knowledge, if brands can really benefit from taking part in platforms like Instagram, is fairly vague. This is precisely because Instagram is built around the idea of consumers sharing content with their peers, rather than businesses advertising their products to consumer. Current academic literature broadly addresses the relevance of stories for organizational purpose, but so far does not recognize the co-creational power of Social Media or the use of hashtags to consumers in the storytelling process (Singha & Sonnenburg, 2012). The presence and strength of user-generated content (UGC) on the one hand (Leung, 2009; Smith, et al., 2012; Van Dijck, 2009), and the values of co-creation for brands
(Halliday, 2016; Nuttavuthisit, 2010; See-To, & Ho, 2014) on the other hand, are extensively studied, but applying both research streams on storytelling is less available. Furthermore, research so far disregards the Social Media platform Instagram. Studies about storytelling on Twitter and other Social Media platforms (Papacharissi, & Oliverira, 2012; Hermida, 2010; Dayter, 2015), as well as studies on co-creation in Social Media (Nuttavuthisit, 2010; Spear, & Roper, 2013), can easily be found, but not so when searching for the same academic topics studied on Instagram.The intention of this study is therefore, to contribute to the knowledge
about co-creational storytelling on Instagram, the role of hashtags in the process, and its utilisation by corporations. Findings of this study may give implications for corporate practice to help brands to better assess the possibilities of co-creational storytelling on Instagram.
Since fall 2015, the online retailer Zalando is trying to use Instagram to build together with consumers stories around style, individuality and community by using the hashtag shareyourstyle. This campaign focusing on co-creational storytelling using hashtags, and the success of Zalando as one of the leading online retailers in 15 European markets with more than 17 Million active customers (Zalando/press, 2016), makes the Zalando #shareyourstyle campaign an excellent case for this explorative study. Thus, the research question for this study is the following:
RQ: In how far, is the awareness for the Zalando brand built up through the co-creational storytelling with #shareyourstyle?
Theoretical Framework
User-generated content and electronic word-of-mouth
As from the definition in the introduction of this study inferable, the concept of UGC allows user to take an active part in the public discourse on the Internet. When UGC is “interpersonal communication (positive and negative) about a company, brand or product between a receiver and a communicator, who is perceived as non-commercial” (Hutter, Hautz, Dennhardt, & Füller, 2013, p. 345), it is referred to as electronic-word-of-mouth (eWoM). In Web 2.0 consumers and corporations have multiple roles and touch-points with each other and the boundaries between information, fun, leisure and commerce can get blurred (Van Dijck, 2009). However, not all users generate content as well as electronic word-of-mouth. According to van Dijck (2009), active contribution to the generation of content only comes from a small part of the Internet community. Most of the users are rather passive spectators, who just view, watch and read the content online. Thus, the possibility of
Sibylle Mittrach Brand awareness through co-creational storytelling 4
generating content online challenges the communication of corporations. Corporate messages can get easily lost in the noise produced by users that not only talk about brands, but also talk back to content produced by corporate professionals. According to studies on the topic, eWoM can be more influential than advertisement and marketing, because when consumers generate positive eWoM they potentially identify more with a brand (Hütter et al., 2013). Moreover, supported by research on consumers’ trust and value co-creation, eWoM is supposed to increase purchase intention when it is positive, and decrease purchase intention when it is negative (See-To & Ho, 2014). From studies on the topic inferable, motivation for users to generate content online can be on the one hand, to share with others out of genuine concern (Van Dijck, 2009) or for social interaction (Leung, 2009), and on the other hand, to be recognized by peers, as well as express or enhance themselves (Halliday, 2016). According to the Consumer-Culture Theory, consumers in Web 2.0 are “digitally empowered to use brands as resource for life projects” (Halliday, 2016, p. 143). Therefore, the creation of content and its sharing with others on the Internet can become part of self-identity creation (Halliday, 2016). When consumers start producing eWoM and see brands in the process as relationships, dialogues and interaction partners, which whom they share experience, values and ideas, they are involved in a creation process (Nuttavuithisit, 2010). In the
co-creational process users attach with a company, emotionally through the identification with the brand, and physically thought the provision of their produced content. Hence, the co-creational process has the potential to improve and even extend the consumer-company relationship (Nuttavuithisit, 2010). However, the co-creation process potentially offers advantages and disadvantages for brands. On the one hand, next to the enhancement of the consumer-company relationship, content that is produced by consumers can be categorized as earned content. Earned content is related to a brand, but not owned by it and therefore more likely perceived as less persuasive and commercial by the audience (Xie & Lie, 2015). On the other hand, brands risk being exposed to negative eWoM, if they promote co-creation and
consumers misuse this possibility and start to mistrust the company, identify less with it or even dislike it in the process (See-To & Ho, 2014).
Co-creation on Instagram with hashtags
Following the definition of Kaplan and Haenlein (2010), Social Media is “a group of Internet-based applications that build on the ideological and technological foundations of Web 2.0, and that allow the creation and exchange of user-generated content” (p. 61). The photo-and video-sharing platform Instagram is such a Social Media application. Focusing on visual information, Instagram provides brands with a space to easily attract consumer with their branded- and owned content, as well as with content that is co-created by consumer independently. The application offers similar connectivity as Twitter by connecting users as so-called follower, and the network is asymmetric as well, meaning that a person can follow another user, but does not automatically get followed back by this same user (Hu, Manikonda, & Kambhampathi, 2014). However, Instagram is much more than other Social Media
applications, a platform where users create and share their own content. Users mainly post their own pictures and reposting is less common. Apparent from the definition in the introduction, with hashtags users annotate their content with keywords behind the hashtag symbol to connect it to other content, ideas, conversations, other users, or even brands (Kotsakos, Sakkos, Katakis, & Gunopulos, 2015). Thus, hashtags “add valuable meta-knowledge to a particular piece of text” (Kotsakos et al., 2015, p. 27), or in the case of Instagram, picture. Furthermore, users add hashtags to their content to appear on a public timeline when searched for the hashtag, because “any string of characters, which is preceded by a hash symbol becomes a hyperlink” (Scott, 2015, p. 12). Therefore, using hashtags increases on the one hand, the possibility of identifying and connecting to other users or brands with similar interest or taste (Manikonda, Hu, & Kambhampati, 2015), and on the other hand, increases the possibility of joining a conversation or story, as well as to contribute
Sibylle Mittrach Brand awareness through co-creational storytelling 6
to a co-creation (Scott, 2015). Thus, tagging an Instagram post with a hashtag, that displays a brand, is connected to a conversation or co-creation, or is branded by the brand, can be categorized as eWoM, earned by the brand (Scott, 2015).
Diffusion of user-generated content
To be successful, innovations, which can be ideas, practices, or objects, have to diffuse into society through communication via media channels or peer-to-peer interaction over time (Rogers, 2003). This process is called diffusion and is characterized by certain stages of adoption that can be illustrated graphically in the shape of an s-curve (see Figure 1).
Figure 1. Rate of adoption of an innovation over time (Rogers, 2003).
According to Rogers (2003), in the beginning, the innovation is only adopted by a few people, which are often opinion leaders in their peer or got the idea from marketing or
advertisement initiatives. As can be seen in Figure 1, the adoption rate during the first stage is low and slow, indicated by almost a straight line around zero. Then, eventually when a critical mass is reached, which is defined as a minimal number of adopters that are needed for the adoption “to be self-sustaining” (Mahler & Rogers, 1999, p. 721), a large number of people adopt the innovation and the diffusion process increases rapidly. A critical mass is especially crucial for the adoption process of an interactive innovation, like communication technology or networking innovations, to proceed. These kinds of innovations have an increasing utility when adopters can interact and communicate through and with the innovation. Therefore, the
Alternative methods of data gathering have been little utilized, even as a means to supplement the predominant approach of survey data gathering and quantitative
methodologies of data analysis. One wonders why ethnographic methods like in-depth interviews and observation have not been utilized more widely, especially in the organizational innovation studies—many of which are conducted by organizational communication scholars and by students of organizational behavior, both of whom increasingly utilize ethnographic methods. The dominant style of diffusion investigations is thus the quantitative analysis of data gathered by survey interview methods from large samples. The overall effect of these dominant research methods has been to emphasize an understanding of the diffusion process as the product of individual decisions and actions. Interpersonal influences on individuals in the diffusion process have been
underemphasized because of the research methods used. Perhaps the approach to studying diffusion formulated by Ryan and Gross has become overly stereotyped.
However, in recent years, several communication scholars have investigated the critical mass and individual thresholds in the diffusion process, especially for the spread and adoption of interactive innovations such as electronic mail or fax in an organization or in some other system (Markus, 1987; Kramer, 1993). At a certain point in the diffusion process for any innovation, the rate of adoption begins to suddenly increase at an
inordinate rate. This take-off in the rate of adoption creates the S-curve of diffusion (see Figure 26-1).
Figure 26-1. The Diffusion S-Curve
Reprinted from Diffusion of Innovations (4th ed., p. 11) by E. M. Rogers, 1995, New York: Free Press. Copyright 1995 by E. M. Rogers. Reprinted with permission of the author. For innovations that are essentially a means of interactive communication, however, such as the new communication technologies of fax and e-mail, a critical mass
higher the actual and perceived number of other adopters is, the higher the utility of the innovation for all adopters gets (Mahler & Rogers, 1999). Lastly, the adoption rate either slows down again and keeps a certain level, because the system is saturated, or decreases, because the innovation gets replaces by a new one (de Nooy, Mrvar, & Batagelj, 2005). Whether people adopt an innovation depends on several factors, like awareness, perceived values, and perceived risks (Füller, Schroll, & Hippel, 2013). In particular, awareness, which is in context of brands the consumer’s familiarity with the qualities or the image of a brand (Barreda et al., 2015, p. 600), is the first step in the users’ decision-making step. If a person aware of a brand or an innovation he/she is able to identify it within a mass of brands or innovations. Following the AIDA2
hierarchy of effects model, awareness followed by interest and desire leads to action, or adoption in the case of the diffusion process (Hutter et al., 2013). Furthermore, according to research on brand awareness in online social networks, awareness is positively linked to the generation of word-of-mouth as well (Barreda et al., 2015). This means, eWOM is generated when users are aware of an innovation or brand, but also influences the awareness of other users. Inferring from that, next to marketing and
advertisement, eWoM can increase awareness for a innovation and its adoption in society, heighten the perceived values and lower the perceived risks by motivating others, providing information and help, and enable peer gratification (Kawakami, Kishiya, & Parry, 2013). According to Hutter et al. (2013), it is possible to deduce the strength of awareness for a brand from how active consumers engage with the Social Media activities of a brand. This means, the more consumers engage in the Social Media pages of brands or the more they engage in co-creating brand content, the higher their awareness for this brand is.
According to the findings by Hutter et al. (2013), this study assumes that those users who engage actively with a brand on Instagram by tagging content with a brand related
2 AIDA is an acronym for the four stages of the hierarchy of effects model, namely awareness,
Sibylle Mittrach Brand awareness through co-creational storytelling 8
hashtag are aware of the brand and use the hashtags to show their relation to the brand and other users in the community.Moreover, it is expected that the use of the brand related hashtag increases over time, as with a growing number of users applying the hashtag to their posts the awareness of others about the campaign as well as the gratification in the peer for the users presumable increases. As the nature of this study is exploratory, sub-research questions are answered, instead of hypothesis verified. Therefore, the following first sub-research question of this study asks, in how far it is visible that awareness for Zalando increases over time when Instagram user generate content that is linked through a hashtag to the brand.
RQ 1.1: In how far is an increase in user awareness for the Zalando brand in the posts
tagged with #shareyourstyle visible?
Further, in this study it is expected that the progress of the diffusion theory is
recognizable when applied to explain the adoption of a hashtag in a Social Media network as well. A hashtag is an idea or practice in the Social Media network and can emerge through innovators or marketing and advertisement efforts of brands (Chang, 2010).Taking the knowledge about the adoption course of the innovation diffusion process into consideration, it is reasonable to expect that in the hashtag adoption process at least three phases are
recognizable. Hence, these phases can be distinguished through the extent of the adoption rate.Consequently, this study assumes that the hashtag adoption starts with a low and slow rate of adoption in the first phase, in which advertisement and marketing of a brand is gradually launched and only view users are already aware of the campaign. In the second phase, this study assumes that a rapid increase of adoption is recognizable, as a certain critical mass of people started applying the brand related hashtag to their content and the awareness for the campaign in a wider public did grew through external marketing and advertisement efforts. Lastly, it is expected that a stagnated adoption rate is noticeable in a third phase, as the network is saturated and only a few new users apply the hashtag to their content or
advertisement and marketing efforts for other brands or campaigns divert the user’s attention. In how far this theoretical reflection of the innovation diffusion process accounts for the hashtag diffusion in a Social Media network as well, will be examined in this study with the following second sub-research question.
RQ 1.2: In how far is an adoption process of the shareyourstyle hashtag comparable to
the innovation diffusion process visible over time?
Co-creational storytelling on Instagram
Storytelling in organizational context is often used to describe the company’s culture, vision, or unique selling point to stakeholders internally and externally. As stories can travel easily within an audience, they create conversation and interaction between the company and its stakeholders, as well as between the different stakeholders (Singha & Sonnenburg, 2012). Moreover, stories can be an instrument to convey emotions and values that touch upon peoples’ individual experiences. Corporate stories activate the narrative processing of individuals, who incorporate the story into stories of their own life (Escalas, 2004). Thus, stories can be a means for corporations to increase consumer’s awareness, empathy and recognition of the brand (Singha & Sonnenburg, 2012).
However, Social Media challenges the corporations’ power over their stories and enable consumers to not just map the brand story onto their own, but to create brand stories of their own. Social Media provide users with networks, relations and interactions, which are “the three ingredients central to co-creation“ (Singha & Sonnenburg, 2012, p. 190). In this way consumer-generated stories get told for and about brands, through which brands have to navigate to keep their narrative under control. Captured by Singha and Sonnenburg (2012) with the metaphor of an improv theatre, co-creational storytelling is the on-going interlinked process of content generation. In this process, corporations and consumer continuously react to each other and interchange the narrator- and listener role.
Sibylle Mittrach Brand awareness through co-creational storytelling 10
Figure 2. Co-creation of brand narrative (Singha & Sonnenburg, 2012, p. 192).
As visible in Figure 2, co-creational storytelling means for a brand’s narrative to be in an on-going co-creative process. In the best case, the narration gets passed along between narrators based on the corporate’s script and evolves over time in the direction intended by the corporation (Singha & Sonnenburg, 2012). In the improvisational co-creational
storytelling process the brand’s narration may appear blurry through the noise the user-generated stories produce. However, brand stories combined or extended with consumers’ stories can firmly fix the brand’s narrative in the mind of consumers (Escalas, 2004).
Derived from theory on implicit framing, this study tries to apply the knowledge from the field of media framing to corporate storytelling and identify some sort of implicit stories that emerge within the co-creational process. According to Hellsten, Dawson, and
Leydesdorff (2010), implicit narrations are “embedded in latent dimensions of the
communication, and they are generated because of spurious correlations between word (co-) occurrences in communications” (p. 5). This means, although the story is not explicitly linked to a brand and may seem independent from it, it is possible that the content of the story is implicitly correlated to a brand and it’s narration. Furthermore in the case of this study, the stories are not written out narratives with explicit order and narrative rhetoric. The stories are told by a concatenation of hashtags and without stopwords like “and”, “or”, etc. that connect the hashtags in a systematic order. However, as the meaning of words is always dependent on
involved one (participating in the idea of real beauty) and even when the performance was more involved, different audience members chose to play different roles that varied from spectator to actor, like protagonist or hero. For example, the self-esteem movement consists of toolkits and videos that encourage young girls to embrace their inner beauty (spectator) to initiating activities online and offline (protagonist or hero).
While these three elements are essential for an improv theater performance, they account for neither co-creation nor
the continuity of the co-creation. Brand owners have to keep the performance going as co-creation can abate because of uninteresting stories. Therefore, owners may need to reanimate the brand performance by fostering diversity of opinions and stories that challenge one another (Sonnenburg 2004). Diversity of opinion and challenging stories refuel consumers' waning interests and maintain the consumers' engagement in the brand performance, which is imperative for the performance to even exist, let alone evolve. The next scene discusses these co-creation issues and the factors that set a brand performance into action and keep it going.
However, before we proceed, it is important to clarify some-thing about storytelling. According to the field of narratology, story is the content and the process of telling the story is the narrative (Genette 1980; Richardson 2000). Since storytelling in social media is a continuous on-going and collaborative process, made up of interlinked content, we conceptualize brand perfor-mances in social media as brand narratives co-created from interrelated stories.
Scene 2: The Plot Twist
The adapted model of co-creation in improv theater (Sawyer 2003) shown inFig. 3helps explain the idea of a co-created brand narrative that is ongoing (N(t)) and changes with each story. Each story provided by a participant (P(1)) depends on two conversational forces to be a part of the co-creation: the premise emerging from the current brand narrative (N(1)) and the reactions of the other participants to the processed story of the participant (P(1)). These forces play a crucial part in how each story is integrated into the brand narrative (N(2)) (Sawyer 2003). The co-creation of the brand narrative depends on the type of perspective participants choose to offer in the narrative community (e.g. fan who praises the brand, evangelist who preaches the brand, critic who challenges the brand, and hacker who slanders the brand), the forum (e.g. blog, social networking, Fig. 2. Dove print ads for Real Beauty.
Source: Unilever 2008.
Fig. 3. The co-creation of the ongoing brand narrative (based onSawyer 2003).
its context (Hellsten et al., 2010), this study assumes that through mapping the hashtag
discourse similar to the way it is done for research of implicit frames, it is possible to form an impression about the stories by seeing the hashtags in combination to each other. Further, it is necessary to assume that the stories told by users in Social Media will never fully align with the brand narrative, as the co-creational storytelling process between users and the company occurs in a fast and partly uncontrolled fashion (Singha & Sonnenburg, 2012). However, with growing awareness about the purpose of a hashtag and its potential connection to a topic or brand, it is possible that the users’ stories increasingly align with the narrative intended by the brand within a certain corporate concept or script, over time (Singha & Sonnenburg, 2012).
Taking the implicit nature of the stories told with hashtags into consideration and knowing about the interlinked structure of the co-creational storytelling process, this study asks the following two final sub-research questions.
RQ 1.3: How is the hashtag #shareyourstyle used to tell implicit stories?
RQ 1.4: In how far is story development and diversification over time visible in the
posts tagged with #shareyourstyle, and in how far do these stories align with the corporate concept for the campaign?
Method
Research design
In order to analyse the content tagged with the shareyourstyle hashtag, the Social Media platform Instagram was in focus. Due to the large amount of #shareyourstyle posts available on the platform and their relevance for the campaign case, this study used a mixed-method approach, with an automated content analysis as main mixed-method. Furthermore, as Instagram is a picture-driven network and the hashtags are only one part of the Instagram posts, a manual content analysis with a smaller sample was conducted to analyse the pictures. Moreover, a visibility analysis was executed to evaluate the hashtag diffusion process.
Sibylle Mittrach Brand awareness through co-creational storytelling 12
With the automated content analysis, this study chooses to follow an inductive
technique, with which rather than through the directed search with beforehand defined coding schemas, patterns can be found in the data itself (Boumans & Trilling, 2015). Due to the implicit nature of the stories told with hashtags, the automated content analysis allows this study to seek patterns in the large data set and the findings are not disturbed by limitations of a human researcher (Boumans & Trilling, 2015). With the sub-research question, the findings from the automated and the manual analyses are explored to describe, map and explain the course of the hashtag diffusion, the content of the posts, as well as the implicit stories of the discourse around the hashtag shareyourstyle.
Data collection
The data for the automated analysis, the manual analysis, as well as the visibility analysis was extracted from the Social Media platform Instagram. Following the instructions by Hellsten et al. (2010) and the Pajek manual by Vlieger and Leydesdorf (2010) for the automated content analysis, several steps were needed to implement the automated semantic network analysis. First the data had to be downloaded. As Instagram provides no tool to download historical data from their platform and the restriction of the Instagram Application Programming Interface (API) limit the download possibilities additionally, a key for the API, available for private persons only under certain circumstances, had to be used in connection with a tool called InstaR from the statistical software R. Hence, it was possible to download the content tagged with the hashtag shareyourstyle into an excel file. This excel file was used for the visibility analysis. Additionally, from this excel file, separate text files for every post were created through the software Kutools for Excel, which is freely available, for the automated analysis. Lastly, screenshots were used for the manual analysis.
Research units
Out of the total amount of posts tagged with #shareyourstyle (N = 48.418), only the content tagged in the time-phase between August 31st
, 2015 till March 15th
, 2016 was used (n = 5.154) for the purpose of this study. For the purpose of the automated content analysis and the visibility analysis a census sample, including all posts within the time-phase, was
gathered. However, the sample was cleared of posts that did not use the shareyourstyle hashtag in the caption or the first comment under the caption, which reduced the sample considerably (n = 2.076). For the examination of the hashtag diffusion process through the visibility analysis, the excel file with all posts tagged with #shareyourstyle was cleared from all information other than the date of the post and the content of the caption and first
comment. Moreover, as the study intends to map the adoption and discourse of the
shareyourstyle hashtag closely to the evaluation of the Zalando campaign, the text-file sample was split into three time phases for the automated semantic network analysis. Therefore, the posts were organized into three different folders for each period. The periods were divided along the timetable of the Zalando #shareyourstyle campaign.
The first phase, included all Instagram posts tagged with the hashtag shareyourstyle from August 30th till November 14th, 2015. During this phase Zalando started their campaign with a television commercial (TVC), different PR measures, like blogger/influencer Co-operations, interviews, print campaigns, etc., and advertisement on the Zalando shipping box. To capture this phase and analyse it with the automated content analysis, the census sample included, after cleaning, 452 posts.The second phase included all Instagram posts tagged with #shareyourstyle from November 15th, 2015 till January 31st, 2016. In this time slot the
#shareyourstyle campaign was extended with a collaboration with the brand Calvin Klein, using now also the hashtag shareyoursexy to tag pictures in connection with the campaign. The extension was accompanied with different PR measures and influencer marketing. Again, the sample was cleaned and contained afterwards 309 posts. Lastly, the third phase included
Sibylle Mittrach Brand awareness through co-creational storytelling 14
all Instagram posts tagged with #shareyourstyle from February 1st
till March 15th
, 2016. During this phase no new PR measures were implemented and the extent to which the hashtag is used, without the pushing effect of PR and marketing, is in focus. The sample for this phase included 1316 posts. Moreover, from each phase’s sample 100 posts were randomly selected, through the randomizing function of Excel. From the randomly selected posts, screenshots of the picture with the caption, the comments, and the tags were made to analyse the posts manually. Within the sample for the automated analysis the distribution of the amount of posts in each phase is quite divergent. However, the hashtag stories are still comparable, because for the automated semantic networking analysis the same amount of most-used-words within every time-phase was included into the analysis. Further, the content in every text file was cleared from everything other than hashtags, like tagged persons, location, etc., because for the automated analysis only the hashtags were of interest for the automated analysis. This was done to avoid mistakenly use of these words by the computer program, which is not able to distinguish between text and hashtags. Moreover, the hashtags were cleared from the hyperlinks and symbols unreadable by the computer program were deleted.
Automated semantic-network analysis
As mentioned before, the automated content analysis, in this study a semantic-network analysis, finds similarity in patterns of word co-occurrence. By analysing the occurrence of words in the text files, the automated analysis calculates the strength of associations between the words in the posts (Vlieger & Leydesdorff, 2011). However, for the computer program to do so several steps had to be followed. These steps are explained in detail in the manual for the automated semantic network analysis by Vlieger and Leydesdorff (2010).
Starting with the preparation for the software program FrequencyList, the text files had to be combined into a single text file for each period separately. From these text files, three separate lists of 75 of the most-used-words were obtained through the FrequencyList
program. The number of 75 words is recommended by literature to remain able to visualize the discourse later in a single map (Vlieger & Leydesdorff). As the text in the files consisted only of hashtags, it was not necessary to filter with a standard stopwordlist before using FrequencyList, as recommended by Vlieger and Leydesdorff (2010). In a next step, for every period separately, the lists produced from the FrequencyList were integrated with the text files and uses as the input for a software program called FullText. FullText generates data files for each research period that can be imported into SPSS for a factor analysis and used for the network visualization with the program Pajek.With Pajek the discourse of the stories built around #shareyourstyle were visualized in semantic maps. In these maps the nodes are the hashtags and the lines between the nodes are the correlations between the hashtags (Vlieger and Leydesdorff, 2010). With this program and the output of a SPSS factor analysis, the semantic maps show networks, in this case implicit hashtag stories, graphically. With the help of this visualization the time-phases were compared to each other.
Manual codebook measures
For the manual coding of the smaller sample of pictures, a codebook ensured the valid measurement of the variables. The codebook was divided into two parts. First, general
information about the posts, and second, information about the pictures were collected. To measure awareness for the Zalando brand, one variable in the general information part of the codebook asked for the use of Zalando related hashtags in the caption next to the picture (e.g. #zalando, #zalandostyle, #shareyoursexy). Another variable asked for the tagging of the Zalando brand’s Instagram account, either directly on the picture or in the caption next to it. Derived from theory, an active commitment of users with a brand that manifests itself for example in publishing eWoM for a brand is an indication for the user’s brand awareness (Hutter, et al., 2013). Therefore, in this study the active assignment of a Zalando related hashtag and/or the tagging of the Zalando brand’s Instagram account were equated in this
Sibylle Mittrach Brand awareness through co-creational storytelling 16
study with awareness for the brand. Moreover, the assignment of the post to one of the three time-phases was included into the general information part of the codebook.
Furthermore, in the information about the pictures part of the codebook two variables were included into the codebook. First, one variable categorized the type of picture based on the content, and second, one variable classified the content of the caption. The eight
categories proposed by Hu et al. (2014) for Instagram pictures, which are self-portraits, friends, activities, captioned photos, food, gadgets, fashion, and pets, were extended in this study for the differentiation of a self-portrait into three categories (self-portrait with whole body visible, self-portrait with only face visible, self-portrait with only body visible). Further, the applicability of Zalando related content in the picture, namely the Zalando shipment box or Zalando advertisement, was included into the type of picture variable. And lastly, the possibility to categorize the type of picture as a video or something other than the before mentioned types was included into the codebook variable. For the content of the caption variable, the coding schema from Naaman, Boase, and Lai (2010) was adopted and adjusted to the usual content of Instagram captions in the way that the variable contained in the end nine characteristics (Information sharing, self-promotion, opinion or statement, feelings, question to follower, anecdote about self, no text, text in other language than English or German, and others).
After constructing the codebook, it was uploaded into the research platform Qualtrics, which helped to reliable code the content for the manual content analysis.
Inter-coder-reliability
For the manual content analysis, the inter-coder-reliability was tested after
constructing the codebook. Therefore, two coders coded the same 10% of the sample (n = 30). With this data, the common measure for inter-coder-reliability Krippendorff’s alpha was calculated. Krippendorff’s alpha was chosen as the suitable measure, as it “generalizes across
scales of measurement; can be used with any number of observers, with or without missing data; and it satisfies all of the important criteria for a good measure of reliability” (Hayes & Krippendorff, 2007, p. 78).
For all variables except the variables type of picture, and content of the caption the Krippendorff’s alpha showed a value of 1.000 for perfect reliability. For the variable type of picture Krippendorff’s alpha reached .879, and for the variable content of the caption it reached .703, which indicates according to Hayes and Krippendorff (2007) good to modest reliability and could be accepted for the further analysis in this study.
Results
Brand awareness process
To answer the first sub-research question (RQ 1.1), which asks in how far an increase in the user’s awareness for the Zalando brand is visible in the posts tagged with
#shareyourstyle, the data from the manual content analysis was used. Therefore, the presence of hashtags that indicate a connection to Zalando was operationalized as awareness for the Zalando brand.
Sibylle Mittrach Brand awareness through co-creational storytelling 18
As visible in the graph (see Figure 3), from 300 analyzed posts, 71% of the posts were not tagged with a Zalando related hashtag. Unfortunately, it is not possible to infer from these posts, if the users are aware of the Zalando brand or the connection of the
shareyourstyle hashtag to the Zalando campaign. Although, it is possible that the users tagged their posts knowingly with #shareyourstyle in the course of the Zalando campaign, only from the posts tagged with a hashtag that is related to Zalando it is possible to concluded that people purposely used #shareyourstyle to spread eWoM about Zalando. However, 12% of the posts were tagged with #zalando, 13% with #zalandostyle, and 3% with #shareyoursexy. Only a view more posts were tagged with Other Zalando related hashtags (1%), like #ZalandoXElle, or #zalandobloggerawards.
To examine, in how far an increase in posts tagged with at least one Zalando related hashtag and the Zalando brand’s Instagram account is visible over time, the use of Zalando related hashtags and Zalando Instagram account was graphically pictured over the three time-phases (see Figure 4).
Figure 4. Course of Zalando related hashtags and Zalando brand’s Instagram account used for posts tagged with #shareyourstyle in percent, over time (n=300).
As visible from Figure 4, the number of posts tagged with Zalando related hashtags decreased over time. Beginning with 35% of posts being tagged with at least one hashtag that indicates a connection to Zalando, the number decreased to 29% in the second phase and dropped to 19% in the third time-phase(Chi2
= 6.529, p<.038). Further, when looking at the distribution of Zalando related tags on the posts over time, it is observable that considerable more times the Zalando brand’s Instagram account was tagged on the posts in the second time-phase (30%). Between the times the Zalando account was tagged in the first (18%) and third time-phase (19%) no crucial difference is visible (Chi2
= 5.112, p<.048). Concluding from the findings visible in Figure 4, this study assumes that although the use of Zalando related hashtags decreased over time, the awareness for the Zalando brand at least seemed to grow further in the second phase, as more people started to tag the Zalando’ Instagram account on their pictures or in their captions. Nevertheless, the awareness, if through hashtag or account tagging, did not increase further in the third phase like expected, as it rather decreased.
Hashtag diffusion process
With the second sub-research question (RQ 1.2), this study intended to learn in how far the adoption process of the shareyourstyle hashtag over time is comparable to the standard course of the innovation diffusion process. Therefore, the excel file with all posts tagged with #shareyourstyle and their posting dates was used for a media visibility analysis. With the help of a manual, available online
(http://www.polcomm.org/amsterdam-content-analysis-lab/manuals/), the excel file was imported into SPSS. Following the instructions of Vliegenthart (2012) for the media visibility analysis, the data was scoured for the hashtag shareyourstyle and all zalando related hashtags. Binary variables were created, indicating the absence of the hashtags with 0, and the presence of it with 1. After that, the variables were used to illustrate the adoption of the hashtags over time with a viability graph.
Sibylle Mittrach Brand awareness through co-creational storytelling 20
Figure 5. Course of shareyourstyle hashtag adoption and Zalando related hashtag adoption over time (n=2.076).
From Figure 5 visible, the use of the hashtag shareyourstyle did increase in the first month of the Zalando campaign (September and October), kept a certain level during November, and decreased in December. The same course is visible for the Zalando related hashtags, which underlines the finding from the section before (see Figure 4). This process of adoption, recognizable from both lines in the graph of Figure 5, resembles the innovation diffusion process. Inferring from the innovation diffusion process, a reason for the stagnation of the hashtag adoption during November and the decrease in December could be network saturation and/or a replacement of the hashtag. However, only the use of the Zalando related hashtags with #shareyourstyle decreased permanently after December (see Figure 5). In contrast, the use of the shareyourstyle hashtag alone leaped after December. This
development was already observable from the difference of sample size for the automated semantic networking analysis between the first two time-periods and the third period, mentioned in the method section.
Implicit storytelling with hashtags
To investigate how #shareyourstyle is used to tell implicit stories (RQ 1.3), and in how far the stories develop and diversify over time (RQ 1.4), from the text-files of the three
time-periods, the variance of the hashtags was calculated and three separate factor analysis were performed afterwards. Some hashtags showed a variance close to zero for the first time-phase (see Appendix A). Although it is suggested to exclude words with a variance of zero (Vlieger & Leydesdorff, 2010), this study decided to keep all hashtags for the factor analysis of time-phase one in the analysis. Explanation for the low variance could either be, that these hashtags were used in almost all posts in the sample or, that these hashtags were used in almost none of the posts. The variances for the variables in the second and third time-phase showed no critical value. In the end, all variables in all three time-phases were used for three separate factor analyses. The number for the components in the factor analysis was set on six to facilitate an unproblematic visualization at a later stage (Vlieger & Leydesdorff, 2010). Each component represented a network of words, which form an implicit story. As suggested from literature, factors with loadings less than .40 were excluded from the stories, because the statistical significance of these factors is low (Field, 2013). Cronbach’s Alpha (α) was calculated for each story to test the reliability. Due to the absence of a more objective way, labels were assigned to the stories by the researcher (Leydesdorff & Weber, 2011). Further, the positioning of the stories within each time-phase was estimated by the Eigenvalue (EV) measure. Finally, word occurrence matrixes were visualized with a cosine-normalized matrix produced by Pajek, which were adjusted afterwards based on the results of the factor analysis (Vlieger & Leydesdorff, 2010). In the following three semantic maps the words, representing hashtag, are assigned to colors referring to the story they correlated in the factor analysis.
Sibylle Mittrach Brand awareness through co-creational storytelling 22
Figure 6. Semantic map of story discourse around #shareyourstyle in phase 1.
As visible from Figure 6, one strong main core, consisting of one main narration dominated the discourse around the hashtag in the first time-phase (see Appendix B for factor loadings). Marked with black color in the center of the semantic map and labeled as fashion story, the strongest discourse around #shareyourstyle was told with hashtags such as
fashionpost, outfits, fashionblog, follow, and inspo (black color, EV = 12.725, α = .949). At the lower side of the center, another strong narration was labeled as beauty story, which was told with the hashtags tattoo, beauty, like4like (dark grey color, EV = 4.865, α = .814). Furthermore, an outfit story was told with hashtags like girl, lookoftheday, outfitoftheday (light grey color, EV = 3.766, α = .682). Stories labeled as men fashion story were formed with the hashtags instagay, fashionmodel, menswear (white color, EV = 2.925, α = .659) The outfit story and the men fashion story formed stories more in the periphery, but thematically closely connected to the central stories as well. Stories less connected to the central core, were posts tagged with hashtags drmartensph, standforsomething, and drmartens and labeled as Dr. Martens story (blue color, EV = 2.718, α = .981), and posts tagged with zalando,
antwerp, ellepartyatw15, party, and zalandoxelle and labeled as Zalando party story (orange color, EV = 2.567, α = .708).
Figure 7. Semantic map of story discourse around #shareyourstyle in phase 2.
Looking at the discourse in the next time-phase (see Figure 7), the distinction between the core stories was less clear. Almost all networks were closely related in their narration (see Appendix C for factor loadings). The strongest story, labeled as shopping story, contained hashtags such as onlineshopping, personalshopper, whatiwore, and musthave (black color, EV = 11.932, α = .948). Another story, labeled fashion story, contained the hashtags münchen, love, instastyle (dark grey color, EV = 10.916, α = .922). Stories more in the periphery of the map were around the hashtags meanswear, mensfashion, streetstyle, labeled as men fashion story (light grey color, EV = 5.067, α = .854), and around the hashtags fashionblog,
outfitinspiration and zalando labeled as zalando story (orange color, EV = 3.414, α = .700). Lastly, stories labeled as german blogger story (white color, EV = 2.539, α = .736) and shareyoursexy story (light grey, EV = 7.147, α = .802) are visible above the core stories.
Sibylle Mittrach Brand awareness through co-creational storytelling 24
Figure 8. Semantic map of story discourse around #shareyourstyle in phase 3.
In the third time-phase, the networks spread further (see Appendix D for factor loadings) into a strong core and a smaller periphery (see Figure 8). Again stories labeled as outfit story (grey color, EV = 4.271, α = .754), fashion story (light grey color, EV = 3.995, α = .760), and blogger story (dark grey color, EV = 4.947, α = .785), dominated the narration. However, stories labeled as men fashion story (black color, EV = 7.145, α = .834) grew visible in importance. From a rather peripheral position in time-phase one and two, the stories around manlike topics appeared to move into the core of the narration. Moreover, stories around the hashtags drmarten, and standforsomething reappeared in the narration (blue color, EV = 3.528, α = .948).
Summing up, implicit stories around fashion, shopping and blogging with secondary story-lines around the brand Dr. Martens (time-phase 1 and 2) and a party by Zalando in cooperation with the woman magazine Elle (time-phase 1), were readable from the hashtags in the three semantic maps.
Table 1
Stories of posts tagged with #shareyourstyle in all three time-phases.
Stories
Phase 1 Fashion Beauty Outfit Men Fashion
Dr. Martens Zalando Party
Phase 2 Shopping Fashion Shareyoursexy Men Fashion
Zalando German
Blogger
Phase 3 Men Fashion
Blogger Outfit Fashion Dr. Martens German
Blogger
Dominance High ---> ---> ---> ---> Low
Note. Labels were given by the researcher. Dominance was inferred from the EV of the factor loadings.
When looking at the story labels in all the time-phases and their dominance within the phases (see Table 1), it becomes clear that in phase one and two Zalando played a role in the implicit stories around #shareyourstyle, although the dominance of the stories was rather medium (shareyoursexy story) to low (zalando party story and zalando story). Looking at stories found in phase 3, it appears that Zalando disappeared from the six most dominant narrations around #shareyourstyle, and stories around the brand Dr. Martens seem to be the only directly brand related content. Over the course of the three time-phases, stories around men fashion developed from a rather low dominance in the discourse to a core narration around #shareyourstyle. In addition, stories around German bloggers emerged in phase two and three, which indicates that #shareyourstyle was getting used in combination with Instagram networking, in this case from German blogger communities.
Story alignment over time
Finally, the question, in how far the Instagram user’s stories, found with the semantic-networking analysis, align with the Zalando corporate narrative concept over time (RQ 1.4), was in focus of the analysis. The message of the Zalando #shareyourstyle advertisement claim, “Show us who you are. Pick your signature style. Wear your favorites” (Zalando/press, 2016),was operationalized as brand narrative around individuality, style, and fashion. Visible
Sibylle Mittrach Brand awareness through co-creational storytelling 26
in all semantic maps (Figure 6 till Figure 8, and Table 1), stories around hashtags concerning outfits, blogging, and fashion formed the core narration around #shareyourstyle.
Figure 9. Distribution type of picture over time in percent (n = 300).
Furthermore, the results for the type of picture variable from the manual content analysis were regrouped into four main categories of pictures, namely pictures with persons, fashion, zalando, and others (see Figure 9). When analyzing these types of pictures over time, an increase of pictures with fashion and a decrease of pictures with persons (Chi2 = 19.707, p<.003) was recognizable. Pictures showing directly Zalando related content, either
advertisement or shipment boxes, did increase slightly from the first (1%) to the second time-phase (5%), but did drop down to zero in the last time-phase. Overall, the stories users wanted to tell with their #shareyourstyle Instagram posts appeared to be indirectly aligned with the Zalando brand narrative concept.
Discussion
The objective of this study was to gather knowledge about co-creational storytelling on Instagram, examine in how far hashtags are useful for that, and estimate if brands, in this case Zalando, can benefit from it to increase the awareness for their brand. Hence, this study
focused on the implicit stories observable from hashtags and their direct or indirect co-creational relation to the Zalando brand. Concluding from the findings, the question of this research, asking in how far awareness for the brand Zalando is built up through the co-creational storytelling with #shareyourstyle, has to be answered in four steps.
First,the awareness for the Zalando brand seemed to exist and grow in the first two time-phases and stagnate and decrease in the third-time phase. At least, during the time Zalando was still investing in advertisement and marketing, awareness for the brand was gained. The massive presence of advertisement in the first periods of the campaign ensured that users were motivated to generate eWoM and take part in the co-creational process of the campaign. Second, a diffusion process was to some extent recognizable in the course of the hashtag adoption. However, it turned out the framework of the innovation diffusion could rather help to understand the driving factors behind the adoption of hashtags, like the need for awareness and time. Particular time is essential in the diffusion process (Rogers, 2003). Hashtags, however, suffer from their short-lived nature (Chang, 2010). Although enough adopters on strategic hubs in the network, like Instagram influencers or bloggers, potentially adopted the hashtag and the awareness for the hashtag’s purpose reached enough users, it is possible that a new hashtag already diffused into the network and diverted the attention. Zalando did not invent, but adopt, the hashtag shareyourstyle. The hashtag was used before the Zalando brand launched its’ campaign and it will also be used independently from Zalando in the future. Evidence for this is already noticeable in two ways. Firstly, the
adoption process of the shareyourstyle hashtag did grow considerable further in the third-time phase, even though the adoption with Zalando relation decreased. This underlines that
#shareyourstyle evolved independently from Zalando. Secondly, users applied
#shareyourstyle for the content of other brands, like Dr. Martens, as well. Therefore, it is possible that a sub-diffusion process of #shareyourstyle in relation to Zalando was visible during the first and second time-phase of this study, but the actual diffusion process of the
Sibylle Mittrach Brand awareness through co-creational storytelling 28
entire #shareyourstyle adoption is still in an on-going process, independently of Zalando. Third, this study could find implicit stories told with hashtags. The hashtags around #shareyourstyle are used to systematically generate content that is connected to topics of fashion, inspiration, outfits etc. Moreover, even without narrative order and linking words, it is possible to form an impression about the story by seeing the hashtags in combination to each other. The results of this study strengthen the argument that hashtags form a “tightly connected cluster of meanings” (Papacharissi & Olivereria, 2012, p. 268), with which users connect their content to stories, discussions, or overall topics in the network. Fourth, the stories analysed in this study are to some extent aligned with the narrative of Zalando. But, #shareyourstyle implies literally to share stories about style, which suggests stories around fashion and other Zalando related topics, rather than stories about topics, like politics and sports. Furthermore, only in the first and second time-phase of the analysis a direct connection to the Zalando brand, through branded-hashtags, could be found in the semantic maps as well as in the analyses of the manual coded content. Moreover, in the first, as well as in the third time-phase stories around the brand Dr. Martens and their campaign around the hashtag standforsomething were found. This seems to be evidence for an overlapping of the Zalando narrative with other brand’s stories, which contradicts a full alignment of the narration around #shareyourstyle.
Taking all the points from before in consideration, it is important to notice that the course of the hashtag adoption and the development of the implicit stories told with the hashtag are in two ways a contribution to the gap in current literature on co-creation and corporate storytelling on Instagram. On the one hand, the results underline the values of stories for brands and the possibility of co-creational storytelling with consumer on Instagram. User-generated stories can contribute to a brand narrative, when some kind of input from the brand helps to “synchronize the (…) stories into a (more or less) coherent whole” (Singha & Sonnenburg, 2012, p. 193). As long as brands direct their attention towards
the process and take part in the conversation, users will be motivated to generate content that is related to the brand. The moment Zalando left the users alone with the co-creation around #shareyourstyle, the application of the hashtag got diverted away from Zalando. Thus, it requires constant “nurturing of the parameters” (Singha & Sonnenburg, 2012, 196) to manage the co-creation process in the brand’s interest on Instagram. On the other hand, this study could prove that hashtags have an important purpose for the Instagram network. Through the use of the hashtag a kind of “networking system of social awareness” (Papacharissi &
Olivereria, 2012, p. 268) could be established. However, in the case of this study it was rather a social awareness for the hashtag shareyourstyle itself, than for the Zalando brand’s
campaign.Indication for that is on the one hand, the explosive use of the hashtag
independently of Zalando, and on the other hand, that a lot of the hashtags used together with #shareyourstyle are used by users to be found, followed, liked, or connected to blogger networks, or their peers (e.g. #follow4follow, #like4like).In the course of the time-period analyzed in this study, #shareyourstyle seemed to develop in the same direction.The hashtag evolved into an instrument for the Instagram community to take part in conversations and co-create content independently from a brand for themselves and the community.
Summing up, the awareness for the Zalando brand did to some extent increase through the co-creational storytelling with #shareyourstyle. And, as this campaign was just one step in a large-scale repositioning of the Zalando brand, the brand proved to be able to generate content together with their consumers in order to tell stories and spread the word about their brand in the Instagram community. However, the #shareyourstyle campaign is just one case and it is possible that the findings are not generalizable for other brands. Future research could try to replicate the idea of this study and compare between different kinds of brands and diverse corporate sectors. Despite some limitations, this study can contribute to the theoretical fields of co-creation, storytelling, and Social Media, as well as the research area of the
Sibylle Mittrach Brand awareness through co-creational storytelling 30
semantic-networking analysis, and the findings can further be used to give some implications for corporate practice.
The first implication drawn from this study is that before establishing a hashtag for a brand, it is important to decide for an appropriate hashtag. It could be more beneficial to use a hashtag that is branded with the name of the company, for example #mycalvins the hashtag of a successful Calvin Klein campaign. In this way, consumer generate eWoM that is directly related to purchased products of the brand, rather than to a more abstract brand story. Second, marketing and advertisement efforts are needed to establish a hashtag for a brand. In classical television and print advertisement, or social media influencer marketing, the generation of paid or owned content before receiving earned content is crucial to promote a campaign to a wide public. Third, continuous effort is needed to earn eWoM over a longer period of time. A brand has to continuously make an effort to navigate through the eWoM tagged with their hashtag and keep the narrative under control or direct along the lines of their narration. Otherwise, the story can get blurred through the noise of the consumers’ stories or the narrative could get carried away from the brand. Fourth, the co-creational storytelling approach seems practicable for Instagram. Pictures appear to be a good mediator of stories, when they are put in context with the caption and the hashtags of the posts. However, not for all brands Instagram is suitable. Companies with products, that are easily shareable in the picture-driven network, have a better chance of success in this kind of Social Media.
All in all, the possibilities Instagram offers to companies, to build up relationships to their consumers and increase awareness for the brand without the side effects of classical advertisement, like resistance, make the platform an interesting tool for corporations. However, when entering the platform with a corporate account or a commercial campaign, brands have to accept the rules and conditions of the network and adopt their efforts to them. Nevertheless, Instagram is much more a suitable platform to earn content through good relations with customers than to just own it or pay for it.
References
Barreda, A., Bilgihan, A., Nusair, K., & Okumus, F. (2015). Generating brand awareness in Online Social Networks, Computers in Human Behavior, 50, 600-609.
Boumans, J. W., & Trilling, D. (2015). Taking Stock of the Toolkit, Digital Journalism, 4 (1), 8-23.
Chang, H.C. (2010). A New Perspective on Twitter Hashatg Use: Diffusion of Innovation Theory. Unpublished manuscript, Department of Informatics, College of Computing and Information University at Albany, New York, United States.
Dayter, D. (2015). Small stories and extended narratives on Twitter. Discourse, Context and Media, 10, 19-26.
Dijck, van, J. (2009). User like you? Theorizing agency in user-generated content. Media Culture & Society, 31 (1), 41-58.
Escalas, J. E. (2004). Narrative Processing. Building Consumer Connections to Brands, Journal of Consumer Psychology, 14 (1/2), 168-180.
Field, A. (2013). Discovering Statistics Using SPSS (3rd Edition). London: Sage Publications Ltd.
Füller, J., Schroll, R., & Hippel, van, E. (2013). User generated brands and their contribution to the diffusion of user innovations. Research Policy, 42, 1197-1209.
Halliday, S. V. (2016). User-generated content about brands: Understanding its creators and consumers. Journal of Business Research, 69, 137-144.
Hayes, A. F., & Krippendorff, K. (2007). Answering the Call for a Standard Reliability Measure for Coding Data. Communication Methods and Measures, 1 (1), 77-89. Hellsten, I., Dawson, J., & Leydesdorff, L. (2010). Implicit media frames: Automated
analysis of public debate on artificial sweeteners. Public Understanding of Science, 19, 590-608.
Sibylle Mittrach Brand awareness through co-creational storytelling 32
Hu, Y., Manikonda, L., & Kambhampati, S. (2014). What we Instagram: A First Analysis of Instagram Photo Content and User Types. Paper presented at the International AAAI Conference on Web and Social Media, Retrieved from
http://www.aaai.org/ocs/index.php/ICWSM/ICWSM14/paper/view/8118 Hutter, K, Hautz, J., Denhardt, S., & Füller, J. (2013). The impact of user interactions in
social media on brand awareness and purchase intention: The case of MINI on facebook. Journal of Product & Brand Management, 22 (5/6), 342-351. Instagram (2016). About Instagram. Retrieved from https://www.instagram.com/press/ Kaplan, A.M., & Haenlein, M. (2010). Unsers of the world, unite! The challenges and
opportunities of Social Media. Business Horizons, 53 (1), 59-68.
Kawakami, T., Kishiya, K., & Parry, M.E. (2013) Personal Word of Mouth, Virtual Word of Mouth, and Innovation Use. Journal of Product Innovation Management, 30 (1), 17-30.
Kotsakos, D., Sakkos, P., Katakis, I., & Gunopulos, D. (2015). Language agnostic meme-filtering for hashtag-based social networking analysis. Social Networking Analysis and Mining, 5 (1), 1-14.
Leung, L. (2009). User-generated Content On The Internet: An Examination of Gratifications, Civic Engagement and Psychological Empowerment. New Media & Society, 11 (8), 1327-1347.
Leydesdorff, L., & Welbers, K. (2011). The semantic mapping of words and co-words in contexts. Journal of Informetrics 5 (3), 469–475.
Naaman, M., Boase, J., & Lai, C. I. (2010). Is it really me? Message content in social awareness streams. Paper presented at ACM conference 2010.
Nooy, de, W., Mrvar, A., & Batagelj, V. (2005). Exploratory social network analysis with Pajek. Cambridge: Cambridge University Press.
Nuttavuthisit, K. (2010). If you can’t beat them, let them join: The development of strategies to foster consumer’s co-creation practices. Business Horizons, 53, 315-324.
Mahler, A., & Rogers, E.M. (1999). The diffusion of interactive communication innovations and the critical mass: the adoption of telecommunications services by German banks. Telecommunications Policy, 23, 719-740.
Manikonda, L., Hu, Y., & Kambhampati, S. (2015). Analyzing User Activities,
Demographics, Social Network Structures and User-Generated Content on Instagram. Unpublished manuscript, Department of Computer Science, Arizona State University, Arizona, United States of Amerika.
Papacharissi, Z., & Oliveira, M.F. (2012). Affective News and Networked Publics: The Rhythms of News Storytelling on #Egypt. Journal of Communication, 62, 266-282. Rogers, E. M. (2003). Diffusion of innovations (5th Edition). NewYork: The Free Press. Scott, K. (2015). The pragmatics of hashtags: Inference and conversational style on Twitter.
Journal of Pragmatics, 81, 8 – 20.
See-To, E.W.K., & Ho, K.K.W. (2014). Value co-creation and purchase intention in social network sites: The role of electronic Word-of-Mouth and trust – A theoretical analysis. Computer in Human Behavior, 31, 182-189.
Singha, S., &, Sonneburg, S. (2012). Brand Performance in Social Media. Journal of Interactive Marketing, 26, 189-197.
Smith, A. N., Fischer, E., & Yongjian, C. (2012). How Does Brand-related User-generated Content Differ across YouTube, Facebook and Twitter. Journal of Interactive Marketing, 26 (2), 102-113.
Spear, S., & Roper, S. (2013). Using corporate stories to build the corporate brand: an impression management perspective. Journal of Product & Brand Management, 22 (7), 491-501.
