Brand positioning in music : factors changing the market performance of popular albums

! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !

BRAND POSITIONING IN MUSIC:

FACTORS CHANGING THE MARKET

PERFORMANCE OF POPULAR ALBUMS

!

!!!!!!

08!

Fall$

Jack Roland Brugger

MSc in Business Administration

Entrepreneurship and Management in the Creative Industries

(10826602)

Amsterdam, 31.08.2015

Supervisor: Prof. Dr. Nachoem Wijnberg !

!

Table of Contents

4.1! Database and research design!...!22!

4.2! Model!specification!and!Variables!...!22! Table!I.!Overview!of!the!Variables!...!27! 5! Results!...!27! 5.1! Endogeneity!of!a!sequel!release!...!27! Table!II.!Probit!regression!results!...!29! 5.2! Descriptives!and!Correlation!matrix!...!29! Table!III.!Correlation!matrix!...!30! 5.3! Model!Estimation!and!Hypothesis!Tests!...!31! Table!IV.!TwoEStage!Regression!Estimation!...!32! 5.3! Additional!Tests!...!33! Table!V.!TwoEStage!Regression!Estimation!–!Additional!Test!...!34! 6! Discussion!...!35! 6.1! Theoretical!foundations!and!hypotheses!...!35! 6.2! Discussion!of!the!findings!...!36! Limitations,!and!Future!Research!...!39! 7! Conclusion!...!41! Bibliography!...!42! !

$

Abstract

Brand positioning in music is found to have differentiated effects on the market performance of music albums. This study integrates theory on brand positioning into a theoretical framework for album survivability. In addition, two indicators of change in brand positioning are explored. These are: changing record labels and entering into a new categorical identity. Changing record labels is the action of entering new market segments by leveraging secondary brand associations - generating awards to market performance. Entering into a new categorical identity is a sociological concept that captures audience expectations, idiosyncratic authenticity, and genre fuzziness- generating penalties to market performance. These concepts are of importance, because they provide insight into how music artists can establish a stable consumer base through successive releases on the charts. The current study extends our knowledge on the effects of changing record labels and categorical identity. In contrast to previous studies, this study conceptualizes established music artist’s albums as sequels within a series. We test whether sequels will perform better based on changes in record labels and changes in categorical identity. A database consisting of market performance information for 272 albums from the Billboards allows us to test the effects of both changing record labels and categorical identity on sales. As expected, we find that changing record labels has a positive effect on market performance, whereas changing categorical identity, measured as changing sub-genre, has a negative effect on market performance. The findings of the current study indicate a need for future research to further investigate this topic.

! !

1

Introduction

There are many music releases that reach the Billboard charts, but do not propagate future success for the artist (e.g., Kimbra's 'Somebody That I Used to Know', Nena's '99 Red Balloons', and Carl Douglas' classic, 'Kung Fu Fighting' ). In contrast, other successful artists make reappearances on the charts (e.g. The Pet Shop Boys, KMFDM, LCD Soundsystem). The difference between the two is that the latter managed to accomplish three goals. Survive early competition, establish a consumer base, and subsequently maintain this consumer base. But how does one propagate the sale of later albums? By applying branding, artists not only achieve an established position on the charts, but also maintain a stable consumer base. More specifically, artists can position their brand in the minds of consumers, because consumers of music use brands to create an authentic self and build their identity (Napoli et. al, 2014).

A key aim of this research is to integrate theory on brand positioning into the theoretical framework of album survivability. To explain and test the outcomes of brand positioning, we focus on multi-charting Billboard artists. Furthermore, we apply a stream of literature on product series to conceptualize albums as belonging to a sequel series (see Situmeang et al., 2014; Sood and Dreze, 2010; Karniouchina, 2011). The research objective is to test whether particular actions of brand positioning will reward or penalize an artist’s sequel market performance on the music charts. Thus this leads us to the research question, which the hypotheses will seek to answer:

Do changes to an established artist’s brand positioning change the stability of their consumer base?

To answer this research question, we regress possible co-determinants of market performance for sequel albums in order to set controls for the study. These determinants are - whether the album was contractually signed to a major or indie label, the artist’s career age, the artist’s experience, the number of the release, the album’s sub-genre(s), and the average market performance of past albums. In the literature review, these will be explained in detail. Following this, we test whether two changes in brand positioning have an affect on market performance alongside the previous co-determinants. These two are changing record labels and entering into a new sub-genre.

Changing record labels is when an artist signs a contract with a new record label to release their sequel album. This paper argues that this action of brand positioning is beneficial for the artist, helping maintain a stable consumer base. The second effector can be entering into a new sub-genre they were not previously members of (Hannan, 2010; Mattson et al., 2010). If artists make a ‘genre-deviating’ entry into a category, it has been conceptually found to penalize an artist (Mattson et al., 2010). We test and measure whether an entry into a new category results in a penalty in market performance for an established music artist or not.

To illustrate, for established artists or those who have survived the early stage of competition in the billboards (e.g., by charting at least three albums), one would expect that the determinants of an album’s market performance would be similar to the determinants of the next album - unless something changed their consumer base. A change in the consumer base is regarded as the change in the market performance between a previous album and the focal (sequel) album of the same artist. What causes

manner in which artists are placed within the minds of consumers; Keller, 2008). According to branding literature, by changing their brand positioning, artists alter their target market and enter a new competitive arena (Keller, 2008).

While music artists release content through multiple types of products – albums are the only music product used in this study for three reasons. First, there were limitations in available data. Second, an album is a representation of the artist’s identity. Third, compared to other types of releases, albums can be conceptualized as sequels. In marketing literature, previous research on sequels has been focused primarily within the film and video game industries (see Situmeang et al., 2014; Karniouchina, 2011; Sood and Drèze, 2010; Moon et al., 2010). However, established artists’ albums can function in a similar fashion to television, film, book, or video game editions within a series. This is in the sense that each music release from the artist that charts is subsequently a sequel or brand extension to the artist’s brand identity.

As for a general outline for the remainder of the paper, my approach in answering the research question involves an empirical analysis of music artists with three or more charted albums on the electronic/dance album Billboard charts. The paper will begin with a literature review on the most important and relevant articles to this topic, followed by a conceptual framework, which discusses the core variables in the study and hypotheses. This is followed by the fourth part that explains the methodology. Then the results of the analysis will be presented and the paper will finish with a discussion and conclusion.!

2

Literature Review

In this section, an overview will be given of existing literature surrounding album market performance. The first sub-section will consider other factors that may have an impact on how the determinants for market performance are weighted and ultimately affect the success of the focal album. It will become clear why the market performance of an album may be different from one album to the next. This is in the sense that controls will be set and relations will be defined in order for the core variables to be better examined later in the analysis.

2.1 Primary factors of chart performance

The music industry is arguably the most unstable and prominent industry in the creative industries (Rothenbuhler and McCourt, 2004). There is a considerable oversupply of products, product turnover is rapid, and the options for consumers and diversity of products are extensive (Hitters and Kamp, 2010). However, the demand of music products such as albums are difficult to predict as there are many factors governing the market performance of a product. According to previous research on factors affecting popular music album chart performance, there are three main factors of success that have been referenced the most and are regarded as the most prominent. These are: promotion, advertising and the presence of record labels (see Bhattacharjee, 2007; Hendricks and Sorenson, 2009; Ordanini, 2006; Spellman, 2006; Strobl and Tucker, 2000), superstardom (see Rosen, 1981; Hamlen, 1991; MacDonald, 1988; Gopal et. al, 2006; Crain and Tollison, 2002; Strobl and Tucker, 2000), and the artist’s reputation (see Elliot & Simmons, 2011; Hennig-Thurrau et al., 2009; Shao, 2012). If we look at each album both individually and collectively amongst an established artist’s entire series of albums,

these factors can act as determinants of both surviving the early stage of competition on the charts (e.g., getting a second release charted), and subsequently maintaining an established position on the charts, respectively (e.g., getting n sequel releases).

2.1.1 Promotion and advertising

Promotion and advertising, or marketability, can play a major role in meeting consumer demand (Elberse & Eliashberg, 2003). Marketability is a key predictor for opening week performance and securing a large opening audience (Elberse & Eliashberg, 2003). The level of marketability, however, varies album-to-album depending on if the distributor is a major or indie record label. Albums released by major labels are promoted more, have a wider audience exposure, and consequently, tend to last longer on the charts (Strobl and Tucker, 2000). The findings of Bhattacharjee et al. (2007) also confirmed these results, indicating that major labels are marked with higher levels of marketability and promotion. This is because the major record labels exert significant control in the recording, distributing, and promoting of music albums and possess the financial resources to gain access to large consumer bases (Strobl and Tucker, 2000). In contrast, there are thousands of independent labels, which together reportedly account for nearly 35% of market-share in the United States, while their market share is less than 15% in the world-market (Statista, 2015; Nielson Company Report, 2012). These independent labels, shackled by the lack of resources to reach wider audiences, tend to operate in niche segments (Spellman 2006). In which case, you would expect that sequels released by Major record labels to be more successful than those released by Indies. Thus they act as a co-determinant and are controlled for in the analysis of our variables.

2.1.2 Superstardom

While promotion and advertising have strong weighting effects on establishing an artist on the charts, there are some artists who become ‘superstars’ in the process. The phenomenon of superstardom in the music industry and its relationship with album success plays a significant role on the market performance of an album (Fox and Kochanowski, 2007; Strobl and Tucker, 2000). This occurs when “relatively small numbers of people earn enormous amounts of money and dominate the activities in which they engage” (Rosen, 1981). As a superstar dominates the charts, their market performance is a signal to consumers that other consumers are satisfied. This is better known as the ‘superstar effect’ in which these artists have a superior advantage in maintaining a stable consumer base compared to non-superstars (see Cox and Chung, 1994; Hamlen, 1991). A superstar artist’s consumer base is susceptible to change only when something highly dramatic occurs, such as a misspoken political statement, legal issue, or otherwise. Because of these reasons, a superstar is generally immune to considerable changes in their consumer base. For the progression of this review, superstar artists are controlled for. This is in the sense that, for superstars, there are ‘forward spillovers’ of past performance and past information onto the sequel album (Hendrick and Sorenson, 2009). For superstars, these spillover effects likely have a more significant weighting role on market performance than a change in brand positioning does, because the strong reputation of the existing product increases the demand for the product under the same brand (see Hendricks and Sorenson, 2009; Elliot & Simmons, 2011). However, even for established artists, this can too carry a weighting role on a sequel’s market

performance, just to a lessor magnitude. This leads us to the next section, which discusses information theory and forward-spillover effects of past market performance.

2.1.3 Performance from previous albums

The term “forward spillovers” refers to the impact of a new album on the sales of future albums (see Hendrick and Sorenson, 2004). This theory stems from previous literature that developed theoretical models that study the impact of information spillovers on firms’ decisions about whether to release new products under existing brand names (Wernerfelt, 1988; Choi, 1998; Cabral, 2000; see Hendrick and Sorenson, 2009). Within the stream of this literature they discuss the information hypothesis. According to the information hypothesis, a larger stock of informed consumers could increase demand for an artist’s future albums (Hendrick and Sorenson, 2004). The information hypothesis suggests that the past success of an album gives rise to a greater amount of album information in the public domain, contributing to further album success and sales (Elliot & Simmons, 2011; Hendrick and Sorenson, 2009). While this theory is integral in superstar theory, it is applicable to non-superstar artists charting multiple releases.

Hendrick and Sorenson (2009) argued that more popular albums are more widely promoted, so more consumers are aware of them, thus popularity becomes self-reinforcing. They also noted that the fact that artists release multiple albums serves only to amplify the sale of a subsequent album. Chang and Dhar (2009) also agreed, noting that you would expect artists who had hits in the past to have a greater likelihood of producing more hits based off the performance of previous albums. These arguments all point towards a similar theory known as the ‘carry-over mechanism’, which focuses on a previous product’s quality carrying over to later editions within a series. This is the

phenomenon that the image of earlier products can color the way for product extensions, or later editions in a series (Hennig-Thurau et al., 2009; Sood and Dreze, 2006; Situmeang et al., 2014). The reasoning is that the original product must have created a positive image and this image is carried over to the sequel, making the sequel more attractive and also creating awareness, excitement and anticipation, which will help subsequent sales (Situmeang et al., 2014). To reinforce this claim, a stream of literature generally shows that sequels perform better in the marketplace (Ho et al., 2009; Karniouchina, 2011; Sunde and Brodie, 1993).

In this regard, due to the fact that consumers are known to correlate past performance with future outcomes (MacDonald, 1988), we would expect that the value created by the market performance of a previous album would carry-over to the subsequent sequel album. Thus, this is used as a co-determinant of sequel market performance, in which we observe the average market performance of past albums on the success of a sequel album. This is in order to control for the fact that a sequel’s performance may be determined by past information or knowledge that is exogenous to the decision of an artist changing their brand positioning.

2.1.4 Artist’s Reputation

In the previous sub-section, forward spillovers and the carry-over effects of previous albums were discussed. When consumers are uncertain about album qualities of an established artist, the strong reputation of an existing album increases the demand for products sold and creates a forward spillover (Hendrick and Sorenson, 2004). In which case the artist’s reputation is a co-determinant of the sequel’s market performance. An

charts in the past should have a higher probability of attaining a sequel success. Elberse (2010) found that artists with a stronger reputation are associated with higher revenues. This same study however looked at the effect of reputation on product bundling sales. These results implied that sales for the album portion of the bundle are decreasing as the digital music buying rate increases, but particularly so for artists with a weaker reputation (Elberse, 2010). Furthermore, artists with a reputation can carry advantages in combination with the effects of word-of-mouth, as it helps artists attract traffic, which, in turn, is influenced by their content (Dhar & Chang, 2009). For these reasons, for each album we control for the artists career age on the charts.

3

Conceptual framework

3.1 Branding Theory

! So far, the promotion and advertising of an album, its distinction of being major or indie, information carry-over effects, and the artist’s reputation have been discussed as the key determinants of causing a change in the consumer base. What has yet to be discussed is how the artist’s branding decisions affect their albums’ performance. Musicians in electronic music have been increasingly known to integrate brands within their marketing strategies, using logos, alias names, costumes, and strange alter egos to establish a brand image (McLeod, 2001). However, the level of research focused on the connection between these artists, brands, and market performance at the artist level is sparse (see however Muñiz et al., 2011; de Chernatony, 2006; Schroeder, 2005). Thus in this section, the relationship between artists and brands is discussed. First, we integrate theory on brand extensions to better support the previous argument how albums from an

established artist reflect similar market performance dynamics to sequels. Second, brand positioning is brought into the discussion, setting the theoretical foundation for the hypotheses.

3.1.1 Albums as brand extensions

The connection between artists and emerging conceptualizations of brands is an interesting research area. Successful artists can be thought of as brand managers, actively engaged in developing, nurturing and promoting their albums under a complex, multifaceted public identity in the competitive cultural sphere (Schroeder, 2005, Muñiz et al., 2011). Scholars studying the careers of artists have asserted that artists and their works share important characteristics with corporate brands, luxury brands and cultural/iconic brands (Muñiz et al., 2011). Furthermore, brand theories suggest that cultural items such as product sequels can be conceptualized as brand extensions (Sood & Drèze, 2006; Chang & Ki, 2005). In branding literature, a brand extension is when a firm uses an established brand to introduce a new product (Chang & Ki, 2005). Brand managers do so in order to captivate consumers and reduce expenditures on promotion during introduction (Chang & Ki, 2005). Furthermore, in reference to the previously mentioned “forward spillovers” theory by Hendrick and Sorenson (2004), Backhouse et. al. (2015) found a similar reciprocal spillover effect between the extension’s success and the parent brand’s success (i.e., parent brand: the artist). In this regard, the brand extension will help consumers form a positive product-quality perception of a new line extension because of its similarities to the established brand (Basuroy et al., 2006). Within the framework of this study, an established artist’s first charted album functions as the introductory album reflective of the artist’s brand identity. The subsequent album

thus can be conceptualized as a brand extension from the original. The method of how a music artist can market their ‘brand extensions’ is discussed by introducing the process of brand positioning.

3.1.2 Brand Positioning

Brand positioning can be described as the act of designing and specifying an artist’s attributes, offer, and image so that it occupies a distinct and valued place in the target audience’s minds (see Keller, 2013; Kotler, 1997; Sujan and Bettman, 1989). For a popular electronic music artist, deciding on their brand’s position involves determining a frame of reference; which is built into two steps, identifying the target market and the nature of competition (Keller, 2013). In order for artists to survive the early stage of competition, we assume that these artists have successfully completed both these steps. However, once an artist begins to release multiple albums, they may be pressured to re-identify their competitive landscape due to market forces or pursue a new artistic direction or goal. If an artist does change their brand positioning, they can do so by either changing record labels or categorical identities. For both of these actions, the authenticity of the brand plays as a strong positioning device. The effects these two decisions have on an established artist’s sequel performance will be examined in the next section.

3.2 Changing Record Labels

Record labels are known as a trademark or brand associated with the marketing of music recordings. Record labels typically contract artists who have a specific focus on a group of aesthetics, styles and sub-genres in their music, curating works from artists with similar oeuvres. Under a standard contract, labels provide marketing and distribution advantages for the artist (Bryne, 2007). When an established artist changes to a new

record label, they enter into a new distribution channel. However, if the record label belongs to a Major label the distribution channel is much larger. In both cases, an artist would change record labels for two reasons: the artist sees the record label’s distribution channel as a new and fashionable, positive secondary brand association (Keller, 1993), and/or secondly, to simply expand their consumer base.

3.2.1 Secondary brand associations

Production and artistic content are clearly linked (Lena, 2013). In the music industry, record labels’ distribution channels can act as secondary brand associations (Muñiz et al., 2011; Keller, 1998). This is because consumers can form "brand" images of retailers (Jacoby and Mazursky, 1984) on the basis of their product assortment, pricing, quality of service, and so on. This section suggests that established artists who enter new record labels leverage the secondary associations of the record label could improve the market performance of their albums (Keller, 2013; Keller, 1993).

Muñiz et al. (2011) conducted a case study on visual artist Pablo Picasso, noting how aware he and others were of their distribution channels, and how it affected their artistic brand image. To illustrate within the context of this study, record labels in the electronic music industry are known to produce additional content for consumers other than just music recordings (e.g., T-shirts, bags, ‘insider’ magazine issues, special events, etc.; Rush Hour, 2015). The images and content these products (i.e., brand elements; Keller, 2013) purvey are secondary associations of music recordings, which include each individual artist’s work. With this said, a record labels’ secondary associations act as components to market the artist’s albums to consumers (see Bottomley and Doyle, 2006).

Furthermore, the music of other artists on the label are also secondary associations to an artist’s recording. When artists contract with a new record label they conform to a palette of artistic content offered in the label’s entire selection of music. They do so in order to seek a position, or status, within that group (Lena & Patchuki, 2013). Therefore, it is mutually beneficial for artists, other suppliers, and record labels to form a pre-established relationship and understanding between one another (de Chernatony, 2006, p. 192; Muñiz et al., 2011). In doing such, artists invest in the growth and maintenance of their brand community by activating a wide set of mutually reinforcing benefits for both producers and consumers (Marzocchi et al., 2013). Due to the fact that established artists are commercially successful artists, we can assume that artists reaching the charts leverage their secondary brand associations accordingly when they move to a new one. In this case, when established artists do leverage the secondary associations upon changing record labels, they may be able to draw in new consumers.

3.2.2. New Market Segments

In more avant-garde independent scenes of the electronic music industry, the artists’ priorities are to achieve the greatest possible in-scene network reputation (Lange and Bürkner, 2013). In contrast, artists who are established on the Billboard charts have greater interests in increasing their public reputation and consumer base (Hesmondhalgh, 1999; Lange and Bürkner, 2013). Entering into new distribution channels can be seen as a beneficial action as it provides access to new market segments and access to untapped consumer demand (Keller, 2013). Such expansion provides access to various new populations of consumers and stakeholders, cultivating an appealing and vivid identity that translates into purchases (Muñiz, 2011). Due to the fact that the artist has a prior

presence on the Billboards, the reputation of the artist will likely reduce consumer uncertainty.

Importantly, the extent of this consumer base expansion depends on whether the newly contracted record label is major or indie. Some independent artists who chart a successful album may get an offer from the Majors. In this case, they will likely experience an increase in market performance, due to the promotional advantages the Major label provides (Bhattacharjee et al., 2007; Strobl and Tucker, 2000). However, in doing so, artists sacrifice their brand image amongst consumers who may view the change as a ‘sell-out’ action (Hesmondhalgh, 1999). For this reason, we control for artists who do not ‘upgrade’ their contracts to majors, and ‘downgrade’ their contracts to Indies. In summary, artists who do switch record labels will increase the level of exposure to new audiences and replenish any lost consumers, helping achieve sequel success and the maintenance of a stable consumer base.

Hypothesis 1: Changing record labels between the former album and sequel (focal) album will have a positive effect on the market performance of a sequel album.

3.3 Entry into a new category

When established artists on the charts enter new categories, they enter into new competitive landscapes. But many commercial artists in the music industry do not sometimes realize the magnitude categorical entry has on their market performance. However, some of these artists enter into a new category under two circumstances. First, in the most high-risk scenario they may have a change in artistic direction or intension to establish a newly defined consumer base (i.e., a new target market; Keller, 2014), or

secondly, there are forces of competition that pressure these artists to move with trends of genre popularity in order to maintain their position as a top grossing artist. Without a strong prior-reputation supporting their new categorical entries, artists trying to sustain their career suffer penalties.

According to research on categories and categorical systems, the entry into another category or deviation from a pre-existing one has suggested that diversification brings with it certain penalties: less attention and legitimacy and lower chances of survival or success (Hsu et al., 2009; Mattson et. al, 2010; Negro, 2013). A seminal article to this argument is Mattson et al. (2010), in which they studied genre-deviating penalties upon the artist’s entry. However, my research intends to expand this literature by applying the authors’ propositions to an established artist’s entry into a new category through the release of a sequel album, using reputation as a co-determinant. The authors indicated that there are two types of genre-deviations at artist entry, however we will only use one.

The entry into a new category is defined as an artist entering into a new category, but in the process, deviates from one or more of the core feature values associated with the collective schema of the underlying form (Mattson et al., 2010). An example of this would be an established artist or group who claims to belong to the Trance sub-genre on their first album but includes a style from a completely different sub-genre, such as Acid Jazz, on the subsequent album. Thus the artist deviates from the Trance sub-genre by combining it with some other feature that is not characteristic to the underlying sub-genre (Mattson et al., 2010).

It is important to note that music genres are socially constructed categories, which allow audience members to effectively schematize and label a music album or other type of artistic work (Hsu and Hannon, 2005; Hannon et al., 2007; Hannan, 2010). However, music genres (i.e., including sub-genres and styles) tend to have fuzzy and blurred boundaries (see Hannon et al., 2007; Mattson et al, 2010). This makes it more difficult for audience members to ‘judge’ how well an artist fits into a category.

Mattson et al. (2010) made three propositions regarding an artist’s genre-deviating entry. We will argue for these propositions and then empirically test the effect entering into a category (i.e., sub-genre) has on the market performance of sequel albums.

Proposition 1: Deviations from existing forms upon entry are associated with stronger and more frequent penalties by audience members.

Proposition 2: Artists with high idiosyncratic authenticity are associated with lower and less frequent penalties by audience members.

Proposition 3: Categories with high genre fuzziness would have fewer penalizations on genre-deviating artists.

First, when Mattson et al. (2010) refer to audience members, they refer to consumers, stakeholders, critics, and other evaluators of an artists’ membership within a category (Hannan et. al., 2007). Second, in reference to the second proposition, idiosyncratic authenticity is defined as the features and behavior of an actor that represent sincere choices, i.e. being true to some underlying form of music. With this said, I argue that established artists on the billboards would be penalized more (i.e., experience lower market performance on sequel albums) if they do make a new entry into a sub-genre,

which they were not previously members of. Using the three propositions in combination with the aforementioned literature on artists’ brands, we must further reason why established electronic artists can be penalized and have a lower market performance on their sequel albums.

First, for deviations from existing forms upon entry - it can be argued that artists are penalized because audiences build two types of expectations, collective expectations and artist expectations. Established artists’ first albums are indicators that audience members (i.e., those who made this artist’s billboard entry possible) legitimated them. In the process, these audience members, who consist of primarily a mainstream consumer base, have developed collective expectations and norms for the genre (Mattson et al., 2010; Zuckerman, 1999). Furthermore, as artists establish connections with audience members, audience members build expectations for the artist’s work. An action such as a entering into a new category thus likely increases uncertainty in the artist’s work, and it has been found that consumers generally do not like uncertainty (Situmeang et al., 2014).

To argue for the second proposition, genre charts are reflective of an institutionalized industry-based genre constructed by mass-media gatekeepers rather than a culturally defined one (Hibbett, 2005; Lena & Peterson, 2008; Hirsch, 1972). For example, a majority of artists who reach the Electronic/Dance billboards can be placed into the category ‘EDM’. This can be described as a ‘marketing tool’ constructed by major institutions beginning in the late 1990s - rather than an actual genre (see Hibbett, 2005; Pigeons, 2012). Artists in this institutionalized genre include artists such David Guetta, Deadmau5, Skrillex, Hardwell, Swedish House Mafia, LMFAO, and so on – all culprits who are repeating what the Indies did in the 1990s – which was creating a

tendency towards classical pop aesthetics, and ‘arm’s length’ institutional ties with the corporations (Hesmondhalgh, 1999). The situation of EDM can be compared to the institutional structuration of ‘Indie Rock’ and ‘Hip/Hop & RnB’ (see Hibbett, 2006; Hesmondhalgh, 1999; Lena, 2013). For this reason, there is a strong argument that the artistic ambitions of those who reach charts are more driven by financial incentives and public fame rather than seeking idiosyncratic authenticity from audience members (see Lange and Bürkner, 2013; Hesmondhalgh, 1999).

Finally to add the third proposition into this discussion, this chart does not entirely represent the vast field of sub-genres and styles in the electronic music industry (see McLeod & Kembrew, 2001). In actuality, the chart is primarily composed of a small range of ‘contemporary’ sub-genres and styles (AllMusic, 2015; Billboards, 2015). Lena (2013) examined the lyrical content of rappers on the charts – finding that production and content are clearly linked. She found that that there were strong similarities across styles and artists. Thus, one could argue that due to the similarities in style amongst artists, genre charts are marked with higher contrast than fuzziness (Hannon et al. 2007; Mattson et. al, 2012).

In summary, we can make a strong assumption that entrants in new sub-genres will have more penalties on the charts due to the two types of audience expectations, institutional structure of the genre, and crisply defined nature of popular music charts (Hannan et al., 2007; Hesmondhalgh, 1999; Hibbett, 2006; Mattson et al., 2010). Thus, in combination with the three arguments for the propositions, the entry into a new category will result in a penalization in market performance.

Hypothesis 2: Entering into a new sub-genre between the former album and sequel (focal) album will have a negative effect on the market performance of a sequel album.

4

Methodology

!

4.1 Database and research design

!

The empirical material for this study consists of music albums belonging to various artists that charted on the dance/electronic album billboard charts between 2001 and 2014. The data (n=272) consists of solely albums from artists with two or more sequels (excluding the first release). The average number of sequel albums charted by an artist in this sample is 4, and the largest is 22.

It is also important to note that the chart reflects the most popular selling electronic albums in the United States (Billboard, 2001). This chart excludes singles but includes EPs, which are musical recordings that contain more music than a single, but are too short to qualify as a full length LP. These were excluded in the study, because LPs tend to be ‘larger’ works of an artist and thus may likely get promoted more. This was set as a control in order to avoid any difference in promotion between an artist’s sequels.

4.2 Model specification and Variables

The hypotheses were tested by estimating a comprehensive model regarding the antecedents of a sequel album’s market performance (PERFi). An overview of the variables is presented in Table II. Most importantly, the model takes the action of changing labels and entering into a new sub-genre into account, including the size of the

consumer base, career age of an artist, and sequel edition number as control variables. Additionally, I set seven of the sub-genres as dummy variables. The variable (Prob(Sequeli)) is estimated by means of a first-stage probit model described later in

Section 5.1.

PERFi = β 0 + β1.PERFAi + β 2.CAEAi + β 3.ALBNi + β 4.LABCi + β 5.ENCi + β 6.MAJi + β 7.GENRE1i + β 8.GENRE2i + β 9.GENRE3i + β 10.GENRE4i

+ β11.GENRE5i + β 12.GENRE6i + β 13.GENRE7i + β 14.Prob(Sequel i ) + εi

Market performance of the sequel album is the dependent variable of the comprehensive

model, and was determined via the billboard data as the total number of weeks the album was present on the charts. This is better known as the album’s survival. Survival is defined as the length of time that an album remains on the charts before dropping off, or its duration on the charts (Bhattacharjee et al., 2007). While the two terms can be used interchangeably, some studies have used survival as a measure for an artist’s life span on the charts (Ordanani, 2006), while others have used it as the lifespan of an album (Chon, 2006). The survival of an album is regarded as the standard measure of album market performance, as it best represents the amount of sales earned per album and an artist’s non-hardcore consumer base (Bhattacharjee, 2007). The minimum value of survival was 1 and the maximum 79 weeks, with the average sequel surviving for 11 weeks.

All figures of an album’s performance on the charts are based off sales figures and do not include radio airplay, online streams, and social media activity as some newer charts on the Billboards do. The variables for how an album performed were determined strictly from the information provided from this dataset, however sales figures were not used. Other information such as the date of chart, weekly position, debut position, and record label titles were given.

For additional tests and comparison, initial debut rank was used. The initial debut

rank is defined as the rank of the album on the first week it hits the charts. Initial debut

rank is related to an album’s acceptance by early adopters, which can create further demand from remaining consumers. These early adopters can be described as loyal fans who have often completed their purchase by the time the album has appeared on the chart (Bhattacharjee et al., 2007).

Estimated size of consumer base prior to sequel was used as another market performance

indicator in the study and is operationalized as the average number of weeks the album survived on the charts across all albums in the series prior to the focal albumi. While

there is no unit count of each album sold, the duration of weeks the album remains on the charts is reflective of sale figures and therefore provides a relative estimate of the size of their consumer base and the level of demand in the past.

Career age of the established artist is a control variable for the comprehensive model.

This is operationalized as the number of years the artist had been established on the charts after the first charting album up until the focal album. The average career age of an artist in the sample was nearly 6 years.

The years of prior activity was used as another control variable in the comprehensive

model. This is operationalized as the number of years since the artist first became active on the charts, i.e. years of experience. This is to control for the fact that the artist’s level of experience may affect their decision to develop a subsequent sequel album. For each album, this is measured by calculating backwards in time from the focal album.

Major or Indie Label distinctions of albums were also taken into account and used as a

billboard data. However, in order to determine its major/indie distinction, data was gathered by visiting the labels’ websites or finding a secondary source confirming its link to either a major conglomerate, sub-label of a major record label, or whether it was completely independent.

Changing Record label of the sequel album was used as a predictor in the comprehensive

model for the first hypothesis. This was determined by observing whether the artist made a change to a new record label from the previous album to the sequel album. If there were a difference in record labels between the previous album and the sequel (focal) album, it would be classified as changing labels. We controlled for the fact that artists changing from independent labels to major labels would experience an increase in performance due to more promotion.

Changing record label during the early stage of competition was calculated by observing

the change in record labels between the first and second chart of the artist. This was used as a dummy variable for sequel albums that belonged in a series where the artist changed labels between the first and second album in the series. This was used in the first stage probit model in order to control for the differing effects between changing record labels early in their career vs. changing record labels later in their career.

An artist’s entry into a new category was a predictor used in the second hypothesis. This

variable is operationalized as an artist’s entry into a new sub-genre on the sequel album. I control for the differences in categories between the categories in the previous span and those in the next. Lastly, due to data constraints - it is assumed that the artist had a high degree of typicality with sub-genres on the first album (see Mattson et al., 2010; Hannan

et al., 2007), thus any change from this would count as a deviance from their original form (see Mattson et. al., 2010).

The data for this variable was obtained from AllMusic, in which data regarding an album’s sub-genre were obtained. For some albums, typicality scores were available, but a majority of the artists in the used sample did not have typicality scores and were thus entirely discarded from the study. AllMusic provides a classification hierarchy of all categories, so this was used as well to maintain consistency of the classification schemes. All styles were recoded into sub-genres, because some album entries contained styles but excluded the sub-genres in which they belonged. There were a total of 83 styles that were recoded into 7 electronic sub-genres. These sub-genres were Downtempo, Electronica, Experimental Electronic, Jungle, House, Techno and Trance. Regarding the categorical placement of each album, the majority of artists within this chart belong to styles and sub-genres of the electronic genre. However there are other artists associated with Pop-oriented dance music and electronic leaning Hip-Hop and R&B. The sub-genres that were classified as being within the electronic genre, however, were used to determine an artist’s entry into a new category.

Table I. Overview of the Variables Variable Name Description

PERF i Market Performance of sequel i, measured as the number of weeks on the charts

PERFA i Estimated size of general consumer base prior to sequel i, measured as the average market performance of all albums prior to the sequel CAEA i Career age of the artist, measured as the number of years since the

artist became established ALBN i Edition number of sequel i

LABC i Record label change of sequel i, Dummy variable

LABCE i Record label change in early stage of competition prior to sequel _{Dummy variable for probit} i, ENC i Entry into new category of sequel i, Dummy variable

YAPS i Years of prior activity outside the charts as artist prior to sequel i, measured as the number of years of experience prior to focal album - for probit model

MAJ i Major or Independent Record Label for sequel i, Dummy variable GENRE1 i Downtempo sub-genre, Dummy variable

GENRE2 i Electronica sub-genre, Dummy variable

GENRE3 i Experimental Electronic sub-genre, Dummy variable GENRE4 i Jungle & Drum n Bass sub-genre, Dummy variable GENRE5 i Trance sub-genre, Dummy variable

GENRE6 i House sub-genre, Dummy variable GENRE7 i Techno sub-genre, Dummy variable

5

Results

5.1 Endogeneity of a sequel release

Before we estimate the comprehensive model, we need to control for the fact that not all albums are succeeded by a new sequel album on the charts. In other words, our estimation procedure must address the complex nature of the endogenous decision of a sequel being released by the same artist or not. This is approached by setting up a two-stage equation system (see Ho et al., 2009; Situmeang et al., 2014). First, you estimate a probit model to approximate the probability of a sequel reaching the charts using a set of exogenous variables. In this study, the probability of a sequel being released by the artist

is estimated based off whether there are changes in the artist’s consumer base from one album to the next, in the sense that a stable consumer base would give the artist the means to produce a sequel album. In the second stage, the estimated probabilistic scores are included in the main model. This variable captures endogenous factors, which would otherwise be embedded in the error term of the main model (Situmeang et al, 2014).

The exogenous variables used in the first stage of the probit model relate to the edition preceding the focal sequel only. The model is specified as follows:

Probability (sequel album i)

= φ ( α0 + α1 PERF i-1+ α2 ALBN i-1+ α3 YAPS i-1 + α4 GENRE1 i-1 + α5 GENRE2 i-1 + α6 GENRE3 i-1 + α7 GENRE4 i-1 + α8 GENRE5 i-1 + α9 GENRE6 i-1 + α10 GENRE7 i-1 + α11 LABCE _{+ εi )}

Table II displays the probit regression results; φ(.) is the cumulative distribution function of the standard normal distribution. The Probit estimation shows that three determinants are significant at p < 0.05, namely, PERF i-1, GENRE1 i-1, and GENRE3 i-1. This indicates that an artist with a successful album in terms of number of weeks on the charts has a higher probability of maintaining a stable consumer base for the sequel album (as expected) and albums in the Downtempo sub-genre (GENRE1) are likely to get more sequels than albums in the experimental electronic sub-genre (GENRE3). The estimation for label changing in the early stage of competition indicates that there is a lower probability of maintaining a stable consumer base when the artist had changed labels early in their career. However the result for this estimation was not statistically significant (B= -0.270, p=0.202). Most importantly, the probit estimation provides estimated values of the probability of charting a sequel album, and this additional variable is included in the main model as a control.

Table II. Probit regression results

5.2 Descriptives and Correlation matrix

Table III presents the correlation matrix and the descriptive statistics of the study’s key variables. The correlation analysis provides preliminary evidence that changing record labels and the size of the consumer base have a positive and significant correlation with the market performance of a sequel album (r= 0.106, p<0.05 and r= 0.427, p<0.01, respectively). The electronica sub-genre was found to be positive and significantly correlated to the performance of a sequel album (r= 0.222, p<0.01), whereas experimental electronic was found to be negative and significantly correlated (r= -0.162, p<0.01). Furthermore, artists with albums signed to a major record label have a positive and significant correlation to sequel performance and average market performance of previous albums (r= 0.369, p<0.01 and r= 0.433, p<0.01, respectively). Additionally, the

Variables Parameter Estimates (B)

YAPS i-1 -0.012 PERF i-1 0.094*** ALBN i-1 GENRE1 i-1 GENRE2 i-1 GENRE3 i-1 GENRE4 i-1 GENRE5 i-1 GENRE6 i-1 GENRE7 i-1 LABCE 0.016 0.436* 0.314 -0.651* 0.059 0.225 0.026 0.013 -0.270 Pseudo R2 0.264

Notes: Dependent variable: sequel (1) or no sequel (0); i-1 relates to the edition before the

Table III. Correlation matrix Variable Name Mean SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 PERF i 11.25 16.296 PERFA i 14.13 14.904 .427** LABC i 0.49 0.501 .106* 0.076 LABCE i 0.35 0.477 -0.093 -0.049 0.07 ENC i 0.29 0.456 -0.09 .197** -0.086 -0.016 MAJ i 0.43 0.496 .369** .433** 0.099 -0.023 -0.028 CAEA i 4.88 3.057 -.147** -0.015 0.044 .236** 0.051 -0.085 ALBN i 3.92 2.068 -0.066 0.032 -0.098 .176** 0.022 0.09 .672** YAPS i 10.80 6.628 -.137** -0.092 -0.074 .125* 0.093 0.017 .516** .486** GENRE1 i 0.25 0.436 0.052 0.096 -0.042 -0.084 0.038 0.009 .173** 0.068 0.047 GENRE2 i 0.62 0.486 .222** .205** 0.077 0.052 0.009 .298** 0.012 -0.03 0.03 -0.021 GENRE3 i 0.16 0.371 -.162** -.295** -.211** -0.045 -0.058 -.313** 0.076 -0.062 -0.036 -0.022 -.185** GENRE4 i 0.14 0.347 -0.048 -0.074 -.109* 0.037 -0.034 -.177** -.163** -0.092 -.132* -0.036 -0.009 -0.029 GENRE5 i 0.06 0.238 -0.039 -.126* -0.063 -0.04 -0.037 -0.035 -0.062 -0.029 0.098 -0.069 -.228** -0.019 -0.102 GENRE6 i 0.33 0.469 0.046 0.019 0.000 0.02 0.014 .159** -0.024 0.03 0.068 -.205** -0.008 -.198** -0.044 .266** GENRE7 i 0.21 0.407 -0.040 -.104* -0.041 -0.006 -0.019 -.108* -0.013 -0.01 0.032 .134* -.154** .373** -0.012 .125* -0.026

** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed).

sub-genres electronica and house are positively correlated with major record labels, whereas techno, experimental electronic, and jungle & drum n bass were negatively correlated. There is a significant negative correlation between performance of the sequel and the career age of the established artist (r= -0.147, p<0.01), and performance of the sequel and years since first activity as an artist (r= -0.137, p<0.01). Changing record labels in the early stage of competition was found to be significant and positively correlated with the career age of the artist and the years of prior experience from the focal sequel (r=0.236, p<0.01 and r=0.125, p<0.01, respectively). Perhaps most interestingly, the average market performance of previous albums is positively correlated to the sequel’s entry into a new category (r= 0.197, p<0.01).!

5.3 Model Estimation and Hypothesis Tests

The results of the two-stage regression estimation are presented in Table IV. Our results show that there is a significant improvement in explained variance between our simple Model 1 and the comprehensive Model 2 (ΔR2 = 0.029; ΔF = 5.404; p < 0.01). The ΔR2 shows that there is some change in variation in the model. The variance inflation factor values show that multicollinearity is not a major concern here. We use the standardized results of the comprehensive model (Model 2) to test the hypotheses. The β compares the magnitude of the different variables by standardizing the variables.

The first hypothesis stated that changing record labels between the former album

and sequel album will have a positive effect on the market performance of a sequel album. As shown in Table IV changing labels on the sequel album has a positive and

Table IV. Two-Stage Regression Estimation

Model 1 Model 2

Variable Name β VIF β VIF

Market Performance PERFA i 0.452*** 2.303 0.330*** 2.320 Artist Characteristics CARA i -0.178** 1.751 -0.195** 1.842 Album Characteristics ALBN i 0.005 1.572 0.031 1.656 LABC i N/A 0.111* 1.169 MAJ i ENC i Sub Genres GENRE1 i GENRE2 i GENRE3 i GENRE4 i GENRE5 i GENRE6 i GENRE7 i 0.171** N/A 0.077 0.086 0.073 -0.002 0.047 0.053 -0.012 1.551 1.219 1.281 1.571 1.143 1.219 1.195 1.262 0.157* -0.137** 0.078 0.073 -0.017 0.010 0.054 0.064 -0.014 1.562 1.162 1.240 1.291 1.692 1.157 1.222 1.203 1.273 n 272 272 R2 0.282 0.311 Adj R2 0.248 0.273 F Change 8.458*** 5.404**

Notes: *Significant at p < .05; **Significant at p < .01; ***Significant at p < .001; Note 1: The dependent variable is weeks on the charts; We estimated the model with the first stage parameter [Prob (Sequeli)]. In Model 1: 0.032; Model 2: β= 0.107

entry into a new genre. The second hypothesis states that artists entering a new

sub-genre between the former album and sequel album will have a negative effect on the success of a sequel album. According to the results there was a significant and negative

relationship with sequel market performance (β = -0.137; p < 0.01), supporting hypothesis 2. Other variables such as being signed to a major on a sequel had a significant and positive relationship with sequel album sales, as would be expected (β = 0.157; p < 0.05). The average market performance of previous albums had a strongly significant and positive relationship with sequel market performance (β = 0.330; p < 0.001). Interestingly, the career age of an artist had a significant and negative relationship with a sequel’s performance (β = -0.195; p < 0.01). The first stage parameter for a sequels probability however was found to be statistically insignificant, but had a positive relationship with a sequel’s market performance in our Comprehensive model 2 (β = 0.107; p = 0.216). However, none of the sub-genres had significant relationships with sequel album market performance, and were all quite weak.

5.3 Additional Tests

Additional tests were conducted in which the measures for market performance in the main model were swapped with initial debut rankings. This was done in hope that it might provide more interesting results, which could be compared to the main model. These however did not end up with any significant β estimates amongst any of the variables, and the comprehensive model 2 was found statistically insignificant (ΔR2 = 0.013; ΔF = 1.824; p = 0.163).

Table V. Two-Stage Regression Estimation – Additional Test

Model 1 Model 2

Variable Name β VIF β VIF

Market Performance 1_{PERFA2 i 0.026 1.505 0.025 1.609} Artist Characteristics CARA i -0.052 1.231 -0.036 1.341 Album Characteristics ALBN i -0.001 1.616 -0.016 1.431 MAJ i ENC i LABC i Sub Genres GENRE1 GENRE2 GENRE3 GENRE4 GENRE5 GENRE6 GENRE7 -0.067 N/A N/A -0.031 0.130 -0.034 -0.020 -0.063 -0.077 -0.037 1.441 1.505 1.268 1.492 1.155 1.225 1.391 1.261 -0.061 -0.008 -0.074 -0.040 0.134 -0.050 -0.035 -0.061 -0.057 -0.026 1.527 1.213 1.239 1.461 1.274 1.541 1.235 1.295 1.461 1.373 n 272 272 R2 0.041 0.054 Adj R2 -0.004 0.002 F Change 0.991 1.824

Notes: *Significant at p < .05; **Significant at p < .01; ***Significant at p < .001; Note: The dependent variable is weeks on the charts; We estimated the model with the first stage parameter [Prob(Sequeli)]. In Model 1: 0.017; Model 2: β= 0.025;

6

Discussion

In this section the results of the regression analyses presented in the previous section are discussed and possible explanations are presented. The section starts with a synopsis of the theoretical foundations of this study. Secondly, the results of the hypotheses are discussed. Third, limitations of this study, directions for future research and contributions to the literature are provided. Lastly, practical implications are presented.

6.1 Theoretical foundations and hypotheses

This study explored the effects of changes to brand positioning on artist survivability on the charts. This was approached by examining two common actions made by artists during their careers, changing record labels and entering into a new sub-genre. Furthermore, we conceptualized multi-charting artists’ albums as belonging to a sequel series to best explain the effects of our hypotheses. Although previous studies (Karniouchina, 2011; Moon et al., 2010; Situmeang et. al, 2014; Sood and Drèze, 2010) have researched the concepts of sequel series, they mostly investigated the film and video game industries. To our knowledge, no study has investigated these concepts in a way that converges theory on music branding and sequels to assess the survivability of established music artists. Through the combination of literature from different fields - album market performance (Bhattacharjee, 2007; Hendricks and Sorenson, 2009; Elliot & Simmons, 2011), sequel market performance (Situmeang et. al, 2014; Hennig-Thurau et

al., 2009; Sood and Drèze, 2010), branding (Keller, 2013; Muñiz et al., 2011), and social

labels to have a positive effect on an album’s market performance, and changing categorical identities to have a negative effect on an album’s market performance. Both hypotheses were supported with significant results. First, possible explanations for the results of the hypotheses concerning the main effects are presented. Then, we confirm similar findings in prior research.

6.2 Discussion of the findings

The first hypothesis suggested that changing record labels has a positive effect on sequel market performance. We found evidence to confirm this, showing a significant positive coefficient for this hypothesis, suggesting that changing record labels has a positive effect on sequel market performance. In other words, the results suggest that contracting to a new record label for the release of a sequel album would lead to an increase in market performance of that sequel album. There may be several reasons for these findings.

First, there are information effects impacting firms’ decisions on whether to release new products under existing brand names (Wernerfelt, 1988; Choi, 1998; Cabral, 2000). The size of their consumer base is attractive to consumers in the new market segment, and gives rise to a greater amount of public information about the album (Elliot and Simmons, 2011). A larger stock of informed consumers could increase demand for future albums. On the other hand, if consumers have a taste for diversity (i.e. albums from different artists), then the new release are likely to reduce demand for their future albums.

The second hypothesis suggested that entering into a new sub-genre has a negative effect on sequel market performance. We also found confirmatory evidence,

showing a significant negative coefficient for this hypothesis, suggesting that entering into a new category has a negative effect on sequel market performance. In other words, the results suggest that this leads to a decrease in market performance of that sequel album. There may be several reasons for these findings.

First, the genre may have higher contrast (Hannon et al., 2007). One finding is that the number of new categorical entries is low, which shows that genre-deviation is atypical in this chart. Categories marked with higher fuzziness have fewer penalizations on genre-deviating entries (Mattson et al. 2012). However, it is important to note that these results may be biased. The data for sub-genres were gathered from a site that is editable by anyone. This means that the artist’s membership to some categories may be invalid and ambiguous. Regardless, we did have significant results.

Second, mainstream consumers have strong expectations for artists in regards to avoiding uncertainty. Uncertainty has been found to have a less favorable weighting of past qualities in the series (Situmeang et. al, 2012). If there is brand consistency for a product, the consumer can expect the taste from subsequent purchases of the brand to remain consistent. In this regard, consumers engage in uncertainty avoidance – in which consumers go with the established brand because it lowers the risk of uncertainty and information costs (Lam et al., 2012). If an artist deviates from their original genre, this can result in inconsistency and thus lower market performance.

Thirdly, there is significant evidence that electronic artists do not have idiosyncratic features. However, such factors are likely to affect the outcomes of later-stage deviations from existing genres. This study did not control for early or later later-stage

Thus in this case, this points to future research, in that there may be differences between the two stages.

While sub-genres did not have weighting factors on sequel performance there were a few interesting correlations. The electronica genre was positively correlated to major record labels, whereas the experimental sub-genre was correlated to the Indies. It is important to understand that there are strong contextual contrasts between these two genres. Electronica is a sub-genre that consists of influences that are more straightforward, whereas experimental is typically associated with a wide-mix of influences (McLeod, 2001). This confirms prior results that majors have been found to stick to proven styles whereas independents are more active in the creation of new genres (Gander and Rieple, 2004; Mezias and Mezias, 2000; Mattson, 2010).

As for the additional tests for robustness, these results were likely due to the nature of the variable. This variable was a ranking only for the first week rather than for the entire lifecycle of the album - such as what was used in the main model. However, this result is somewhat surprising because the performance measure used in this test was to measure the artist’s loyal audience. One would expect that the results would be at least partially different from the main performance indicators, album survival (see Ordanini, 2006), however the results indicated that all the variables were non-significant and weak.

! ! ! ! ! ! ! ! !

Implications! !

There are a few practical implications that can be taken from this study. Apart from the substantive findings on sequels on the music charts, we show the applicability of brand positioning as a framework in the music industry. Second, this study provides implications for record labels and artists as they are making strategic choices regarding the artist’s identity and relationship to the prevailing structure of genres. Furthermore, the empirical support of information effects highlight a needs for record labels to time promotion and advertising accordingly. Artists can best focus on building strong brand identities prior to entry, because all albums are determinants of future success (Situmeang et. al, 2012). Furthermore, established artists should be inclined to release music through different record labels to improve market performance. However, it is important that artists act as brand managers and leverage secondary associations (Muñiz et al., 2011; Keller, 2013). These results may be beneficial to artists who are uncertain of signing with a new record label. While loyalty with a single entity can establish stronger connections between the artist and distributor, having more transactional ties across the industry can provide greater opportunities.

Limitations, and Future Research

This study has several limitations that at the same time point the way toward further research. First this study was limited to successful artists in the electronic music industry. More research is needed to validate the findings and analyze other types of music products. This is due to the declining nature of album sales in general, and thus its significance as a unit of analysis may become obsolete in future research. By including other forms of performance indicators such as singles, ticket sales for live events,

merchandise, and social media or online traffic, one can create a more comprehensive metric for an artist’s consumer base. Second, for measuring artist-deviating entry, it was only measured at the artist level. Due to limitations in data, it was not possible to calculate grade-of-membership of the actors within the category, and thus the fuzziness and contrast of the categories. Therefore, future research should strongly consider examining artist deviating entry at both the artist and genre level.

Another limitation of this study was that it did not take into account the changes in the popularity of styles. This would have had an impact on the interpretation of results for a few reasons. First, in the electronic music industry the popularity of sub-genres change on an annual basis (Billboards, 2015). Knowing whether the artist belonged to a less popular genre may be an explanatory factor for a change in an artist’s consumer base. Second, the popularity of a particular good reflects its fashionable state. If an artist does not move with fad changes or does so in an unfashionable way, it may penalize the artist (Hendricks and Sorenson, 2004). Thus, if an artist’s former album was a member of a popular sub-genre, but the sequel album was not, the consumer base could change.

Furthermore, a key determinant of a music album’s demand is change in consumer taste. Prior research has indicated that forecasting consumer tastes is difficult (Burke, 1996). However, changes in consumer tastes and preferences for particular genres of music have been found to have an impact on the market performance of music (Hendricks and Sorenson, 2004). Thus future research should consider controlling for such.

7

Conclusion

The first aim of this research was to test whether the market performances of two albums resemble the market performances of two editions within a series. A second aim of this study was to test the effects of changing record labels and new categorical entry in the empirical setting of the music charts. Based on previous studies, we hypothesized that changing record labels would have a positive effect on sequel market performance, whereas entering a new category would have a negative effect on sequel market performance.

Both hypotheses were supported. Alongside expectations, changing record labels was found to positively affect sequel performance, whereas changing categorical identities was found to negatively affect sequel performance. In conclusion, this study contributes to the literature by showing that the application of branding in music is important by making a first attempt to study albums as a sequel series. The findings of this study indicate a need for future research to further investigate this topic.