AN ATTITUDINAL IMPACTS ANALYSIS OF SOCIAL MEDIA PLATFORMS AND BRAND RELATIONSHIP QUALITY AT MUSIC FESTIVALS

The purpose of this study is determination of ways in which music festival organizers can target their social media communication with greater certainty toward younger generations (i


Introduction
As the number and type of festivals grow and consumer choice increases, consumer loyalty is an increasingly significant issue for music festival Delivered by Ingenta IP: 146.176.247.82On: Sun, 29 Nov 2020 19:02:14 Article(s) and/or figure(s) cannot be used for resale.Please use proper citation format when citing this article including the DOI, publisher reference, volume number and page location.

STRAND AND ROBERTSON
network analysis (SNA) framework that is introduced here prior to an extended discussion in the research methodology section.

Context to Study: Social Influence and Value of Music Festivals
The social and cultural impact of music festivals is an area of increasing discussion in the literature (Bourdieu, 1997;Pavluković, Armenski, & Alcántara-Pilar, 2017;Putnam, 1995).Indeed, concomitantly, some research suggests that music festival attendees have more interest in the social and cultural value dimensions of the festival than they do in the economic ones (Chaney & Martin, 2016).Much of this focus relates to the potential for music festivals to have either a negative or positive impact on the community.
Negative impacts include changes in community values and patterns, environmental damage, higher prices of basic services, resident exodus, noise and crowds, unsafe sexual behavior, use of alcohol and drugs, and conflicts with festival goers (Pavluković et al., 2017).The perceptions of these impacts can play a crucial role in the community's acceptance or rejection of a music event (Pavluković et al., 2017).
Positive impacts of music festivals include the potential to affirm or contest the social order and to reinforce the social cohesion and trust within the community, which produces more social networks and shared experiences (Hyslop & Ewig, 2015;Mair & Whitford, 2013;Pavlukovic et al., 2017).Furthermore, for communities of young people, there is a particular social significance in these shared experiences and in the sense of trust brought about by music festivals (Robertson, Hutton, & Brown, 2018).The positive outcome of this form of engagement includes the creation of mutual benefits, the reinforcement and increase of related social assets, and the facilitation of future collaboration (i.e., as a counter to potential negative impacts) (Stevenson, 2016).
However, despite the positive social values that can be gained from attending festivals, Chaney and Martin (2016) suggested that festival managers are facing an increasing number of challenges in assuring consumer loyalty.Further, a large body of research emphasizes that managers and researchers alike will benefit from the understanding that research is required to understand how brand communications work so as to increase brand equity (Astuti & Setiyadi, 2017;Schivinski & Dabrowski, 2015)-that is, make certain that the assets that add value to the brand are in the minds of attendees (Aaker, 1991;Ross, 2006;Ross, James, & Vargas, 2006).Moreover, it is important for organizers to use social media brand communications positively to influence the individuals' perception of the music festival brand (Schivinski & Dabrowski, 2015).
Despite these needs, research that looks at the ways in which social media can build engagement with younger attendees is limited, both in number and in their depth.There is a lack of discussion as to how festival attendees perceive social networking sites as a platform for value cocreation in the context of music festivals (Hoksbergen & Insch, 2016;Hudson, Roth, Madden, & Hudson, 2015).Therefore, the purpose of research here is to contribute toward filling this gap in dialogue and aiding digital managers in the use of brand communities to increase not only customer loyalty but also the social aspects of music festivals (e.g., interaction and collaboration).Accordingly, this knowledge can aid and enhance the engagement of the digital native generation (Duffet, 2017) as well as their experience of music festivals.Thus, the article proposes that social media communication is a brand relationship quality (BRQ) (Fournier, 1998;Xie, Poon, & Zhang, 2017) mediation factor for music festivals.
The purpose of this study is a determination of ways in which music festival organizers can target their social media communication with greater certainty toward younger generations (i.e., Generation Z or Centennials, also called "digital natives").The research has two core objectives: the first is to investigate how music festivals' use of social media can affect their BRQ with their audience, and the second is to determine how connecting to online brand communities prior to, during, and after the music festival affects the satisfaction and loyalty of attendees.

Literature Review
The following factors are central to this study: social influence and value, brand experience, brand equity and brand equity through sponsorship, and BRQ.Each of these has significance for the social stimuli that are associated with a brand's design, identity, packaging, communications and environments" (Brakus et al., 2009, p. 53).Brand experience can have direct effects on customer citizenship behavior toward other customers and toward organizations (Hur, Kim, & Kim, 2018;Xie et al., 2017).

Brand Equity
To increase consumer loyalty and satisfaction of the festival brand, it is important for music festivals to consider how brand communication works to increase brand equity.Schivinski and Dabrowski (2015) suggested that the value of brand equity, through brand communication, is that it can increase the probability of the brand being incorporated into the customer's consideration and shortening their decision-making process, resulting in the brand becoming their choice of habit.When considering brand loyalty and BRQ, many authors refer to Keller's (1993) consumer brand equity model (CBBE) to measure them (Astuti & Setiyadi, 2017;Sadek, Elwy, & Eldallal, 2018;Schivinski & Dabrowski, 2015).
In their analysis of marketing communication at a music festival in Spain and its relation to the generation of brand equity, LLopis-Amorós et al. (2018) concluded that brand equity has a direct effect on behavioral intention "as well as an effect mediated by satisfaction" (p.839).Their research utilized off-site interviews with 622 festival attendees from the Arenal Sound Festival.In a follow-up examination of this data, Llopis-Amorós et al. ( 2019) highlighted the distinctiveness of Generation Z, as compared to older market groups, signaling that they are largely impervious to advertising, that their interactions are spread irregularly through different media channels and that they can, thus, be a challenge to communicate with.

Brand Equity Through Festival Sponsorship.
Sponsorship creates a mutually beneficial way to communicate with target markets, thereby creating a greater event experience for the attendees (Cummings, 2008;Eales, 2016;Luonila, 2016).It is an important element of brand building for festivals (Luonila, 2016), and communication with target markets via sponsorship at music festivals can include the provision of festival necessities, such brand experience is key to create consumer loyalty (Brakus, Schmitt, & Zarantonello, 2009).This is no less the case for managers of festivals (Chen, King, & Lee, 2018;Schivinski & Dabrowski, 2015).Correspondingly comprehension of the communication tools available that may facilitate reach between music festival organizers and attendees is a vital component of that knowledge (Llopis-Amorós, Gil-Saura, Ruiz-Molina, & Fuentes-Blasco, 2019).

Festival Integrated Marketing Communication
There is some consensus that marketing communication adds value to the brand and has a significant function in developing a relationship with stakeholders (Llopis-Amorós, Gil-Saura, & Molina, 2018).It is widely posited that social media has changed traditional marketing communications and has, along with other online (internet) platforms and tools, gradually shaped brand communications rather than it being controlled and administered by marketers (MacKay, Barbe, Van Winkle, & Halpenny, 2017;Schivinski & Dabrowski, 2015).
Combining a positive music festival experience with proactive social networking could help festival brands and sponsors build long-term relationships with music fans (Hudson et al., 2015;MacKay et al., 2017).Both festivals and social networks wish and need to build communities (Hede & Kellet, 2011;MacKay et al., 2017).Integrated marketing communications (IMC) can be used to increase brand loyalty and the social value of festivals.IMC promotes consistency among the elements of the marketing mix, all of which are integral to the overall success of marketing communications (Hede & Kellett, 2011).
However, prior to addressing the role of social media in expediting or influencing social values, further discussion is given here as to how social values can affect the brand equity of festivals-that is, brand equity influences how customers develop more favorable associations with the brand, typically based around the brand's image (Leenders, 2010).Accordingly, if the festival brand is associated with positive experiences in the past, the attendees are more likely to stay loyal and visit the festival again (Hudson & Hudson, 2013;Leenders, 2010)

STRAND AND ROBERTSON
whereby the sum of items is related to each factor; where a high weight combination of the five factors is evidenced it is then possible to measure the level of brand relationship quality in a broad sense (Zhou, 2007).Although Fournier (1998), Duffet (2017), andXie et al. (2017) offered variants of these components for BRQ, the context of each study differs markedly.Accordingly, because of their stability, Zhou's (2007) five components are applied here.These are discussed further in the research findings.
In their hypothesis on the positive effect of brand experience on BRQ, Xie et al. ( 2017) identified the major dimensions of brand experience-sensory, affective, behavioral, and intellectual-as positively affecting relationship quality.These are also applied in the research analysis and discussed further in the research findings.

Social Value Through Social Media and Cocreation
Millennials and Generation Z actively engage in online discussions and have various options for cocreation, which gives consumers an opportunity to be a group of active creators and influencers (Williams, Inversini, Buhalis, & Ferdinand, 2015).This study proposes that the most popular social media tools used by festival attendees are the festival's website, Facebook, Twitter, and YouTube (Hudson et al., 2015;MacKay et al., 2017).The importance for festivals organizers to engage with networked consumers is widely agreed (Flinn & Frew, 2013;Hudson & Hudson, 2013;MacKay et al., 2017) and includes online word-of-mouth (eWOM) (Williams et al., 2015), cocreation (Kennedy, 2017;MacKay et al., 2017), and content marketing (Hudson & Hudson, 2013;Pulizzi, 2012).A brief summary of each is shown below.
eWOM can ensure interest by creating not only online narratives of customers and event attendees but also a form of online promotion (Williams et al., 2015).Attendees can share positive or negative sentiments from the large quantity of personal data available online (e.g., Twitter or online review sites).These data can be extracted through different computer linguistic techniques (Vujic & Zhang, 2018).
Cocreation can be described as "the process where more than one party systematically joins forces to interact, learn and share information to create value" as water, a place to relax, and the opportunity to go backstage (Cummings, 2008).Despite this, there is a lack of literature regarding how sponsorship works in effecting image change (Smith, 2004) and the extent to which sponsorship provides an additional platform for fans to engage with the sponsor brand (Tsordia, Papadimitriou, & Parganas, 2017).
Although the research by Llopis-Amorós et al. ( 2018), Hudson andHudson (2013), andHudson et al. (2015) referred to the importance of the sponsorship in affecting brand equity for music festivals, neither includes it in their analysis.Thus, this study's inclusion of sponsorship further contributes to the literature.

BRQ
Although brand relationships can be complicated, this study proposes that the competitive future of festivals will depend on the quality and extent of the reciprocation of positive and enduring relations between festival consumers and festival owners (Todd, 2015).Accordingly, as Xie et al. ( 2017) proposed, brand relationships can be explained in the form of a metaphor in which "if we are in a close relationship with other people, we would like to help them out of trouble and share happiness with them" (p.268).Thus, BRQ is defined as "a customerbased indicator of the strength and depth of customerbrand relationship" (Hudson et al., 2015, p. 68) or simply as the degree in which we help and share happiness.This relationship has two parts: one is the participation in a relationship, and the other is the nature of the relationship (Zhou, 2007).
To measure how social media can affect the brand relationship and brand equity, Williams et al. (2015Williams et al. ( , 2017) ) suggest Social Network Analysis (SNA), which measures the relationships between social networks that transmit information, distribute resources, coordinate activities, and manage social norms with different entities, such as families, communities, companies, and countries (Scott, 2007;Williams et al., 2015).
Three hypotheses arose from the questions that concluded the literature review.The hypotheses to be tested are as follows: H1: Social media positively influences brand experience (sensory, affective, behavioral, and intellectual dimensions) and positively influences the customer's relationship with music festivals.H2: Social media integration with sponsor brands positively influences attendance and BRQ.H3: Social media has a positive impact on BRQ between the attendees and the music festival.
The hypotheses are tested through primary research, and the analysis utilizes a SNA framework based on Williams et al. (2015).SNA is concerned with the configuration of relationships and determining the outcomes of the aforementioned entities; in contrast, quantitative research seeks to explain outcomes in terms of the entity's characteristics (Scott, 2007;Williams et al., 2015).Scott (2017) described SNA as a broad approach comprising sociological analysis as well as a set of methodological techniques that aim to describe and explore the patterns apparent in the social relationships that individuals and groups form with one another.The reference to "patterns" suggests that (Kennedy, 2017, p. 68).It can also be understood as when the roles of the brand and the consumer converge (Hoksbergen & Insch, 2016;Kennedy, 2017) and when the active involvement between the customer and the situation creates value (Van Winkle & Bueddefeld, 2016).The process of content marketing can be instrumental in that value generation.
Content marketing, also known as "storytelling" (Hudson & Hudson, 2013), is a branding method in which the brand itself creates and distributes relevant and valuable content to entice and involve the target audience (Du Plessis, 2017).This method generates positive behavior in terms of the consumer's perspective of the brand (Pulizzi, 2012).For example, the Tomorrowland Festival focuses on storytelling in the form of high-quality YouTube videos produced after the festival and sharing them on different social platforms (Tomorrowland, 2018).As Anderson (2011) opined, for festivals to be successful, they should further encourage eWOM through engaging the consumer in the brand experience, and cocreative festival activity in the festival.
Research Questions Arising From the Literature Review Does social media influence Brand Experience dimensions (Xie et al., 2017)?If it does, can this lead to an increase in brand relationship quality between the festival brand, a music festival, and the festival attendees & sponsors?
Are the five factors of BRQ (Commitment/ Connection; Attachment/Attention; Familiarity/ Awareness; Trust/Respect; Association/Recognition) (Zhou, 2007) represented online through social media, between target participants, a music festival, and individual attendees and festival sponsors?
Can corporate sponsorships transfer the brand image of a music festival towards the sponsor brand through social media WOM at a music festival (Smith, 2004)?

WOMM, Netnographic Study, and Social Network Analysis
Netnography is a method of market research that investigates computer-mediated communications STRAND AND ROBERTSON relationships and connections between the case festival and the attendees using the social media platform Twitter.Three samples were collected over time: prior to, during, and after the festival.This collection procedure saves both time and money compared to more traditional methods; moreover, it has the capacity to reach many users, and large samples can be seen to represent communities more accurately (Saunders & Lewis, 2012).

The TRNSMT Festival
The TRNSMT Festival was chosen as the event case for the hypothesis testing.The festival is a city center-based music festival, located in Glasgow, Scotland (Hanley, 2017a).It first operated in 2017 over the weekend of July 7 to 9, with a capacity for 50,000 attendees (Hanley, 2017a).The date of the event is significant, as the first weekend of July was when T in the Park, the largest music festival in Scotland, had been held most years since its inauguration in 1994.The former T in the Park organizers, DF Concerts, created TRNSMT when T in the Park ended in 2016 (Hanley, 2017a(Hanley, , 2017b)).In 2017, the festival was estimated to have added at least £10m to the local economy in Glasgow (Gates, 2017).This research occurred in 2018 when TRNSMT took place in a public park, Glasgow Green, and was spread over the 2 weekends from June 28th to July 8th (Hanley, 2017b;TRNSMT Festival, 2018).

The Social Media Platform
Twitter (www.twitter.com) is an online social media networking platform for microblogging (Huberman, Romero, & Wu, 2008;MacKay et al., 2017).The term "microblogging" is used because of the original 140-character word limit for all Twitter messages (Huberman et al., 2008), which was then increased to 280 characters in 2017 (BBC, 2017).Twitter creates a source of organic eWOM and can offer a great variety in opinions expressed (McGregor, Mourão, & Molyneux, 2017).In a large volume, it is possible to extract from Twitter users their sentiments for a specific purpose (O'Connor, Balasubramanyan, Routledge, & Smith, 2010;Vujić & Zhang, 2018).The terms used to identify specific activities on twitter, and referred to in this article are: SNA is particularly interested in the construction of pictures and diagrams that disclose the patterns that are not generally apparent to human observers (Scott, 2017).In SNA, entities are modeled as nodes, and relationships are modeled as connections (Jarman, Theodoraki, Hall, & Ali-Knight, 2014;Williams et al., 2015).Nodes can be described as individual parts of a larger structure, such as families, cities, companies, or countries (Bastian, Heymann, & Jacomy, 2009;Williams et al., 2015).To identify nodes and connections, this study uses software to analyze the data, including NodeXL (Williams et al., 2015), UCINET (Borgatti, Mehra, Brass, & Labianca, 2009), and Gephi (Bastian et al., 2009).
These types of data software analyze the data through creating network graph sociograms (Hansen, Shneiderman, & Smith, 2011), which is the most common way to visualize networks and usually consists of two primary building blocks: verticals (entities, nodes, or agents) and edges (connections or relationships) (Hansen et al., 2011).The connections are the ties between the nodes that classify by their similarity, by their relationship, by their interactions, and by their flow.For example, these connections could be communities of interest on online platforms, such as Twitter, in which the nodes would be Twitter accounts and the connectors would be the eWOM interactions of retweets, replies, and mentions (Williams et al., 2015).To identify nodes and connections, SNA should be combined with content analysis in order to analyze the key concepts or literature that can identify themes to be examined (Du Plessis, 2017).By analyzing which brand content consumers (nodes) engage with on their social media platforms through likes, shares, and comments (connectors), one can analyze customer engagement as well as the spread of eWOM (Du Plessis, 2017;Williams et al., 2015).Du Plessis (2017) used a content analysis method to identify key concepts, processes, conversations, platforms, and channels in social media content and then analyzed them with program-assisted analysis software (QDA Miner Software).
Accordingly, it is possible to conclude that the research employs both qualitative and quantitative text analysis of the content of eWOM at the music festival studies (below).The data were collected through e-research, which investigates the Tweets are archived 1 month prior to the festival (May 29, 2018 to June 6, 2018), during the festival (June 29 to July 8, 2018), and 2 weeks after the festival (July 17 to 22, 2018).The data collection focuses around the search term "TRNSMT."

Stage 2: Analysis of the Relationship of the Community
In the second stage of analysis, the collected tweets are filtered to identify underlying information about the directed relationships between the users in form of replies, retweets, and mentions (Scott, 2007;Williams et al., 2015).These forms of interactions are modeled as three unweighted directed networks (prior to, during, and after the festival) using the open source application Gephi.
Gephi is freeware software that can be used for graph and network analysis (Bastian et al., 2009).It uses a clustering algorithm to identify and model underlying clusters and to efficiently identify subgroups in large network data sets (Bastian et al., 2009;Clauset, Newman, & Moore, 2004;Williams et al., 2015).

Stage 3: Content Analysis of the COI
In the final stage of the SNA, the collected data text is analyzed (Saunders & Lewis, 2012;Williams et al., 2015).The text analysis performed on the content of the tweets within the clusters identified in Stage 2 is measured through both descriptive text Tweet = short message/update/post Replies = when another user replies to something tweeted by a user Retweet/RT = when someone shares a tweet from another user Mention = when someone communicates directly with a specific user using @ Data Analysis Method To gather information about how the music festival brands were viewed by consumers online and offline, a multimethod data collection method was selected, using SNA and content analysis.Utilizing the procedure of William et al. (2015William et al. ( , 2017) ) (Fig. 1), this study uses SNA prior to, during, and after the event, which facilitates the understanding of relational data (Scott, 2017)-that is, how the connections relate to one another (Jarman, 2018;Zhou, 2007;Zhou, Su, Zhou, & Zhang, 2016;Zhou, Zhang, Su, & Zhou, 2012).By combining SNA (Borgatti et al., 2009;Scott, 2007Scott, , 2017) and text analysis it is possible to examine the community of interest (COI) for the event and destination, which for this research will concern the younger attendees of the TRNSMT festival.

Stage 1: Identifying the COI
This first stage of the SNA is central to the whole data analysis process (Williams et al., 2015).The primary data are collected through selecting search terms to be archived using the online tool Tweet Archivist (www.tweetarchivist.com/).

STRAND AND ROBERTSON
weeks of the festivals, June 29 to July 8, 2018.The total number of tweets about TSNSMT during this was 46,916.The number increased from less than 1,000 tweets per day 1 week before the festival, and peaked at around 6,000 tweets per day during the first weekend of the festival.The volume of tweets decreased to 1,000 tweets per day during the following week but increased to 4,000 tweets for the last weekend of the festival.
The last set of data were collected 2 weeks after the festival, between July 17 and 22, 2018.The number of tweets mentioning TRNSMT decreased to 629 tweets in total, with 139 tweets per day as the highest number (Fig. 2).
The festival tweets were then consolidated, and duplicates in any of the categories were removed.

Relationship Structure of the COI
The next stage of the SNA required inputting of the collected Twitter spreadsheets of archived tweets in the application Gephi to identify the underlying clusters and subgroups in the network data sets (Bastian et al., 2009).

Identifying COI/Clusters
The clusters (communities) were identified by measuring the distinctiveness of the clusters (Williams et al., 2015) and by using modularity statistics based on the community detection algorithms of the network (Bastian et al., 2009).Modularity explains the degree in which networks can be segmented or divided to indicate clusters (Borgatti et categories and the content of the text (Saunders & Lewis, 2012;Williams et al., 2015).A keyword frequency analysis is performed on the Twitter content within these clusters.
When identifying the frequent keywords, it is possible to get a better understanding of the nature and intent of discussions around them (Leech & Onwuegbuzie, 2007;Williams et al., 2015).Consequently, in this case, this has given capacity to analyze the perceived brand experience of consumers related to sensory, affective, behavioral, and intellectual dimensions (Xie et al., 2017).

Data Description
Tweet Archivist enables data collecting by tracking and archiving tweets containing terms and hashtags during a chosen timeframe (Williams et al., 2015).The advantage of tracking both terms and hashtags is that it provides a wider range of collected tweets than just using one these alone and, thus, better represents the communities' conversation (Saunders & Lewis, 2012;Vujić & Zhang, 2018)."TRNSMT" was chosen as both term and hashtag in the Tweet Archivist tool.

Data Collection
The first set of data were collected between May 29, 2018 and June 6, 2018.The total number of tweets during this time was 2,174, increasing from 73 tweets per day to 401 tweets per day.The second set of data were collected during the most active 2 clustered around fan accounts @1975_Tour/@ sgranigerphotos/@DIYmagazine; group 4 (G4: yellow) clustered around individual user @city_sa; and group 5 (G5: orange) clustered around media account @Sundaypost.
During the Festival (see Clusters in Fig. 4).The festival network consisted of 13,190 nodes (Twitter users), and 6,324 communities talking about TRN-SMT were identified, with an average modularity of 0.793.This number indicates that there was a high degree of clustering in the network.The five largest subgroups in the TRNSMT Festival network were identified.Group 1 (G1: purple), @YouTube, was the biggest cluster, group 2 (G2: green), @TRN-SMTFest, was the second largest, group 3 (G3: blue) was clustered around performing artist @ AdamLambert, and group 4 (G4: yellow) was clustered around performing artist @LiamGallagher.The fifth largest group (G5: orange) was clustered around the media platforms @ScottishSun and @ BBCScotland.BBC Scotland is one of the sponsors of the festival.Blondel, Guillaume, Lambiotte, and Lefebvre's (2008) algorithm to detect the modularity of the number of connected components in a network (Bastian et al., 2009).Zhou et al.'s (2012) research indicated that a modularity degree of 0.4 is an appropriate metric for identifying clusters.The research also suggests that clusters beyond 0.6 do not show further meaningful distinctiveness (Williams et al., 2015;Zhou et al., 2012).Accordingly, this SNA used 0.4 as a basis for accepting that meaningful clusters exist and 0.6 to indicate a high degree of clustering.

al., 2009). Gephi used
Prior to the Event (Fig. 3).The festival network consisted of 2,131 nodes (Twitter users), and 615 clusters of communities speaking about TRNSMT were identified, with an average modularity of 0.913.This number indicates that there was a high degree of clustering.Five main clusters were identified: group 1 (G1: purple) clustered around @ TRNSMTFest, the official festival account; group 2 (G2: green) clustered around individual users @ mirrenmckinlay & @erinkane; group 3 (G3: blue) BBCScotland and @CarlsbergUK.The location of the different accounts also varied.However, the individual accounts, the media, and the organizers were predominantly based in Scotland, and the performers were mainly from the UK; the fan accounts, in contrast, were widespread and international, with some accounts based both in China and Peru.
Table 1 displays and compares the characteristics and location of the different clusters prior, during, and after the event.

Content Analysis of the COI
Using the tweets archived, the duplicates were removed so that a quantitative and qualitative text analysis of the content could be executed.Within the clusters identified, a keyword frequency analysis was performed to identify the frequently used keywords related to the TRNSMT Festival and, thereby, understand the nature and intent of After the Event (see Clusters in Fig. 5).The festival network decreased significantly 2 weeks after the festival, with 713 nodes (Twitter users) and 228 communities detected.The average modularity was 0.911, which indicates that there was still a high degree of clustering after the event.The largest subgroups identified were as follows: group 1 (G1: purple) clustered around @YouTube; group 2 (G2: green) clustered around @TRNSMTFest; and group 3 (G3: blue) clustered around performing artist @AdamLambert.

Key Themes
Prior to the Festival.The five most frequently used words in the content of the tweets were  Furthermore, the key theme of the content of the tweets was extracted and processed for each group in order to identify the commonly used words and phrases.These data were combined into themes and is presented in Table 2. Cluster groups are marked in brackets.

Hashtags
The most used hashtags used between the different dates agrees with the conversation found in the clusters.However, the host city, Glasgow, is most often used as a hashtag for marking the location of festival rather than being a topic of conversation.The most used hashtags in tweets mentioning TRN-SMT are shown in the Table 3.

Data Analysis
The following section analyzes the results of the SNA and tests the three hypotheses identified.

Brand Experience
One of the research questions asks how social media can influence a brand experience and its dimensions (Xie et al., 2017) and how it can lead to increased BRQ between the music festival TRN-SMT and its attendees.To understand this, H1 will be tested.
"Glasgow" (83).Other frequently used words were "YouTube" (49), "golden" (37), "amazing" (33), and "love" (22)."Golden" was mentioned in 37 tweets, which appears to have a positive association.However, a higher proportion of tweets referred to the golden circle and ticket problems.In contrast, "amazing" was mentioned in 22 tweets, most of them referring to having a great time and a positive experience at the festival.

Themes in the Clusters
In the SNA process, determining node centrality is vital.Jarman ( 2018) outlined four measures of centrality: degree, eigenvector, beta, and betweenness.Node betweenness measures how often a node appears on shortest path between nodes in a network.An important point in this regard is that nodes "with higher betweenness centrality have the potential to act as gatekeepers" (Jarman, 2019, p. 342).The key users within the clusters were identified based on the betweenness centrality within the modularity class (Williams et al., 2015).The top 10 to 20 users of the three largest clusters were identified and are listed in appendix.The clusters/groups (G1-G3) were identified as follows: Prior to the festival: G1, TRNSMTFest; G2, Mir-renMcKinlay, and G3, the1975_Tour The recognition of these clusters and their utility can help the brand image transfer of the festival toward other services represented at the festival.Brand image transfer is looked at further later in this article.
Affective Behavioral Response.Affective response refers to the attendees' feelings, sentiments, and emotions (Xie et al., 2017)-that is, using social media can increase or decrease the relationship between customers and the festival regarding these three aspects.Social media provides an opportunity to share feelings about the events with other likeminded people and allows for a collective reminiscing of the event.Two examples of this from the data are shown below: Actual feel like a child again can't sleep the night before from excitement like it's fucking Christmas but it's not, it's even better, it's finally fucking TRNSMT day.
Behavioral Response.Behavioral response refers to physical actions, bodily experiences, and behaviors (Xie at al., 2017).Social media can influence the customers' behavior in many ways, including buying tickets, going to festivals, repeat purchases, and influencing fans to listen to the performing acts prior to, during, and after the festival.One of the key themes prior to and during the festival was ticketing, H1: Social media positively influences brand experience (sensory, affective, behavioral, and intellectual dimensions) and positively influences the customer's relationship with music festivals.
The behavioral responses that influence brand relationships are investigated.
Sensory Behavioral Response.Sensory response refers to the physical experience (Xie et al., 2017).In this, social media can influence the sensory behavior when attendees share their experiences.One of the main clusters found in the TRNSMT social network is YouTube, through which attendees can share their sensory experiences (what they hear and see) with others (i.e., fan accounts).With YouTube, fans can see and listen to the performance online.An example of this is shown here: I'm not a festival type, so I would like to thank the maker of this video, YouTube and #Wolfalice for bringing me this great live performance in my living room.It's like being there.Wolf Alice -2018-06-30 TRNSMT, Glasgow, Scotland https://you tu.be/0CRrzPBkqxE via @YouTube.
This example (as well as others) indicates how YouTube has a greater significance as a social media platform than has been identified in related literature.Other articles included in the study did not identify the same clusters.
Another example of sharing sensory experiences is when attendees share their experience of the festival, food, and/or services on social media.Two examples of this are shown below: A big thanks to Box Office Pizza at @TRNSMTfest for making me a cheese free pizza.#TRNSMT#pizza #trnsmtpizza . . .https://t.co/dTQCc11FFc.

STRAND AND ROBERTSON
image of the festival towards their own brand through social media (Smith, 2004).To investigate this, H2 was created and tested.
H2: Social media integration with sponsor brands positively influences attendance and BRQ.
Brand image transfer (BIT) occurs when the brand awareness and brand image of the festival transfer to the sponsor, or vice versa, to increase mutual benefits (e.g., reaching a mutual target market) (Scott, 2007).Scott (2007) suggested that the two major influences on how consumers assess the linking together of brands this way is fit and quality.In an effort to understand how social media can positively influence attendance and BRQ through sponsor brands, the fit and quality of the brands Twitter conversations will be investigated utilizing Scott's (2007) framework (see Fig. 6).
The initial thought by the researchers was to understand how the TRNSMT Festival could transfer their brand image to Glasgow as a destination.However, the Twitter conversations about Glasgow related mostly to the location of the festival and not to how TRNSMT made them feel about Glasgow.The more dominant festival sponsors mentioned in the tweets were BBC Scotland (@BBCScotland), which formed communities/clusters both during and after the festival.Hence, it is interesting to explore how the fit and quality between the two brands (TRNSMT and BBC Scotland) links together (see Fig. 6).
To summarize, TRNSMT is a music festival brand, with performing acts from the rock/pop genre.The attendees span a wide age range, with the youngest recorded age being 16.BBC Scotland is a sponsor of the festival because the performances are recorded and highlighted on their iPlayer.Thus, fans and audience members can watch or rewatch their favorite performances.For people at home, it is possible to feel as if they are taking part in the festival when accompanied by external (social media) communication.BBC Scotland (@BBC-Scotland) actively tweeted about the TRNSMT Festival during and after the festival; they shared the live performances and highlights, and linked these to the iPlayer output.By doing this, the sponsor brand created mutual benefits: TRNSMT received more attention and recognition, and the fans of the as one can see in the G1: TRNSMTFest cluster prior to the festival as well as the G3: BBCScotland cluster during the festival.A lot of the content in the tweets were concentrated on looking for, buying, or selling spare tickets through social media, for example: Looking for a ticket for TRNSMT today.If anyone hears of one going let me know.Cheers.#TRNSMTfest.
Two weeks after the festival, @TRNSMTFest tweeted that they had sold out the 2019 early bird tickets, which indicated that the 2018 attendees bought tickets again because of their experience at the festival and from information found online or on social media: @TRNSMTfest: Early bird tickets for TRNSMT 2019 are now SOLD OUT! Sign up for email alerts ~ https://t.co/r8pxgmzvaWhttps://t.co/XFGm. . .

Intellectual Behavioral Response.
Intellectual response refers to the stimulation of curiosity and problem solving (Xie et al., 2017).By using social media, consumers could have their curiosity stimulated and, thereby, be tempted to buy tickets because of their exposure to the festival's online presence.The consumers might then start looking for tickets online to satisfy their curiosity, which can be seen in the content of the tweets in cluster G1: TRNS-MTFest prior to festival and the G3: BBCSCotland cluster during the festival.
In conclusion, these forms of behavioral responses and interactions through social media let the customers (attendees) communicate and give positive or negative feedback to the festival organizers directly by using @TRNSMTFest.The organizers then get the opportunity to respond directly to their target market, which could positively impact the customers' experience and enhance their relationship to the festival brand.Therefore, the findings concerning brand experience supports H1, as social media can positively influence the consumers' experience and relationship with the music festival brand.

Brand Image Transfer
Another research question concerned whether corporate sponsorships could transfer the brand influences attendance and BRQ if the sponsors fit with brand and quality.

BRQ
The third research question is whether social media has a positive or negative impact on BRQ between the festival attendees and the music festival.
H3: Social media has a positive impact on BRQ between the attendees and the music festival.
Commitment/Connection measures the extension of the consumer-brand relationship and the degree of fit between them (Zhou, 2007;Zhou et al., 2012;Zhou et al., 2016).Based on the SNA and the clusters identified one can see that there are distinct clusters of communities connected to the music festival, e.g., @TRNSMTFest (the official festival or performers got an opportunity to respond by retweeting, mentioning the sponsor and festival, and clicking through to the BBC iPlayer.Thus, sponsors were able to reach an expanded target market.Fans watching online or on TV may be tempted to buy tickets to the festival the next weekend or even for the following year.Two examples of interaction with @BBCScotland are shown below: RT @BBCScotland: "Can I drum?"The amazing moment when @thekillers picked a cheeky Scots drummer to join them on stage at @ TRNSMTfest# . . .@BBCScotland Oy! What's the problem with the TRNSMT highlights on BBC2?Why am I seeing nothing from Friday night . . .https://t.co/w7gVaI4yY0 On reflection, the findings of BIT do support H2: social media presence of sponsor brands positively Association/Recognition measures the consumer's impression of the brand's logo or symbol (Jarman, 2018;Zhou, 2007).Concerning the festival tweets, more positive words are used in conversations about the festival than negative ones.Such words as "amazing" and "great" are used at a much more than negative ones, such as "sh**e" (an expletive common to the Scottish vernacular) and "disappointed/disappoint." To summarize, the research shows that the factors Commitment/Connection, Attachment/Attention, Familiarity/Awareness, and Association/Recognition are weighted higher for TRNSMT, whereas Trust/Respect is weighted less, with chances of negative word of mouth spreading (Fig. 7).However, in summary, the use of social media could lead to a stronger BRQ if the factors are taken into consideration prior to and during negative conversations and if preparation is made as how to respond to these conversations.

Conclusion
With the explosion in both the numbers of SNSs and the practice of building brands with SNSsparticularly through online brand communities twitter account of the festival), which is a distinct cluster prior, during, and after the festival.
Attachment/Attention measures the kindness between consumers of the brand as well as the consumers' intention to pay attention to the promoter and the brand (Brakus et al., 2009;Xie et al., 2017;Zhou, 2007).The distinct clusters identified within the social network of the festival shows that some of the clusters were made up of individual attendees speaking about their excitement of the festival (e.g., Cluster G2: @mirrenmckinlay).The individual attendees were also retweeting and helping each other by sharing content, videos, and information with likeminded people.
Familiarity/Awareness measures the consumer's knowledge of the product as well as marketer's important attributes (Zhou, 2007).The consumer's awareness and familiarity can be shown through the increase of social media usage concerning the event through time.On May 29th, 2018, a month prior to the event, 181 TRNSMT tweets are recorded.This increases to 5,555 TRNSMT tweets on June 29th, 2018.
Trust/Respect measures the consumer's trust of the marketer and the respect the consumer receives (Brakus et al., 2009;Llopis-Amorós et al., 2018;Zhou, 2007).The tweets mentioning the TRNSMT Festival in a negative way were relatively few.However, there were complaints on Twitter concerning the ticketing policy.For example, tickets respect among some of its attendees.This factor is incredibly important for fan engagement in outdoor music festivals (Robertson et al., 2018a) and could be improved by focusing more on customer service through social media.For example, the Glastonbury Festival in 2017 had four festival accounts on Twitter of which @Glastoinfo, was one dedicated to customer service alone; it was a helpline account sharing on-site updates and providing support to people at the festival (GlastoInfo, 2017).Such an approach has the capacity to help music festival experience and also support a greater understanding of the psychosocial needs of young people (Brown & Hutton, 2013;Hutton et al., 2020;Hutton, Zeitz, Brown, & Arbon, 2011).The authors of this work propose that at a time of increasing social pressure, and social and economic turbulence (Robertson, Ong, Lockstone, & Ali-Knight, 2018;Shapiro & Gross, 2013), this is particularly important.
The SNA here focused on one brand and one social media platform.The Twitter platform alone is not able to give deeper understanding of specific generations.Accordingly, there are limitations in generalizing the findings.Future research should build on more brands and utilize other social media platforms.
(e.g., online forums)-a body of research has emerged.There is an indication that SNSs can build brands, most particularly through brand communication.Furthermore, research has shown the viability of brand communication management (Hudson, 2015;Muniz & O'Guinn, 2001) for music festivals.The results of this analysis also indicate that when the social media tools are utilized efficiently, social media can create a stronger consumer-brand relationship.The hypotheses tested and the findings are summarized in Table 4, indicating whether the respective hypotheses were supported or partially supported.
The findings show that using social media to enhance the brand experience dimensions of the festival could positively influence the customer-brand relationship for the attendees of the TRNSMT Festival.Consequently, the findings support H1, whereas H2 was partially supported, as BIT using social media can positively influence how the consumers feel about the brand if the sponsors' fit and quality are equal to the brand.However, it can be difficult to measure the attendance based on sponsor-follower communications.In looking at the Twitter conversation, H2 is only partially supported.
The research literature proposes that five factors of BRQ can be applied through social media.The present research provides positive signals that the conversation around TRNSMT supported four of these factors.However, the trust/relationship factor has a less positive weighting, verifying the real potential of negative word of mouth spreading.However, overall, the use of social media could lead to a stronger BRQ if these factors are taken into consideration prior to and during negative conversations and if preparation is made with regards to appropriate responses to these conversations.
This analysis indicates that a challenge for the TRNSMT Festival is a feeling of lack of trust/ Supported H2: Social media interaction with sponsor brands positively influences attendance and BRQ.Partially supported H3: Social Media have a positive impact on the brand relationship quality between the attendees and the music festival.
Partially supported
) and/or figure(s) cannot be used for resale.Please use proper citation format when citing this article including the DOI, publisher reference, volume number and page location.

Table 1
Demographics and Location Compared Delivered by Ingenta IP: 146.176.247.82On: Sun, 29 Nov 2020 19:02:14 Article(s) and/or figure(s) cannot be used for resale.Please use proper citation format when citing this article including the DOI, publisher reference, volume number and page location.

Table 3
Most Used Hashtags in Tweets Mentioning "TRNSMT" Delivered by Ingenta IP: 146.176.247.82On: Sun, 29 Nov 2020 19:02:14 Article(s) and/or figure(s) cannot be used for resale.Please use proper citation format when citing this article including the DOI, publisher reference, volume number and page location.

Table 4
Hypotheses and Findings Summary Social Media positively influences Brand Experience (sensory, affective, behavioral, and intellectual dimensions) and positively influences customer relationship with music festivals.