The “self made artist”. A real opportunity or by large a necessity?
Collaborations in the production of music and their effects on performance
MSc Business Administration:
Entrepreneurship and Management in the Creative Industries
Author: Anna Gatti 11680377
Supervisor: Prof. Dr. Michele Piazzai
Word count: 13058 Date: 22nd June 2018
Statement of Originality
This document is written by Anna Gatti who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Table of Contents
Abstract ...4
1. Introduction ...5
2. Literature Review ...7
2.1 Cultural Goods and Innovation ...7
2.2 Innovation Determinants ...10
2.3 Music Industry ...11
2.4 Meso Factors and Innovation ...14
2.5 Meso Factors and Performance ...16
3. Data and Method ...22
3.1 Data Collection ...22 3.2 Sample ...24 3.3 Variables ...25 3.3.1 Independent Variables ...25 3.3.2 Dependent Variables ...26 3.3.3 Control Variables ...26 4. Results ...28 4.1 Descriptive Statistics ...28 4.2 Hypothesis Testing ...33
4.2.1 Hypothesis H1a, H2a ...34
4.2.2 Hypothesis H1b, H2b ...39 5. Discussion ...43 6. Implication ...46 7. Conclusion ...49 8. Limitations ...50 9. References ...52
Abstract
This study investigates how the decision of working independently or with an indie or major record label affects the performance of artists in the music industry. Artists in the music industry can release their works (0) autonomously or sign a contract with a record label, (1) independent or (2) major. The present research will look at the impact of this strategic decision on the twofold aspects of performance: sales volume and user ratings. On the basis of data collected from Billboard 200 charts, 184 top albums from January 2013 to March 2018 are considered in this research. Evidence is found that albums released with the support of major record labels enjoy better sales volume compared to independent artists and artist that released album with indie record labels. Additionally, non-statistical evidence is found with regard to users’ evaluations performance. This result might have been caused by a bias in the sample choice of only the most successful released albums. From an organizational perspective the objective of this research is to explore how individuals and organizations interact in the production of cultural goods, and the effect this collaboration has on the final product.
1. Introduction
The music industry is one of the most unpredictable markets. It is characterized by extreme uncertainty regarding the potential success of any specific output (Peltoniemi, 2015). Factors and determinants for an artist’s breakthrough in the music industry are ambiguous and mostly not related with the characteristics of the product/individual itself. Oversupply amplifies these aspects, generating heavy competition and persistent failures. Novelty plays a special role in this kind of industries; in fact, artistic products are consumed for enjoyment and pleasure, thus through innovation, artists and companies aim at constantly introducing different kinds of goods to please diverse and fickle tastes (Peltoniemi, 2015). An open debate is still going on whether smaller or bigger firms are more likely to introduce knockout products on the market and win the competition. In particular, the paper examines how these organizational variables influence the production of music. The type of organization artists release a product with is expected to influence their output, specifically, either working for an independent or major record label has potential effects on the musicians’ performance. Finally, the increased tendency for creative employees to embrace entrepreneurship, facilitated by the modern technologies (Deresiewicz, 2015), makes it relevant to include in the study also the possibility for artists to work independently. In conclusion, an artist in the music industry can choose three different paths: 1) Release a product with a major record label 2) Release a product with a minor record label or 1) Work autonomously. Little is known about which strategic choice is the most valuable and profitable. Thus leading to the formulation of the following research question:
How does the decision of releasing a product independently or with an indie or major record label affect the performance of an artist in the music industry?
Taking into consideration the different options that musicians can take, the present research will look at the impact of this strategic decision on the twofold aspects of performance: sales performance and users’ evaluation. From an organizational perspective the objective of this research is to explore how individuals and organizations interact in the production of cultural goods, and the effect this collaboration has on the final products. This study will potentially help entrepreneurs to take more appropriate decisions when choosing different employment or release options and give helpful guidance for building a thoughtful strategic plan. Additionally, this paper provides a clearer picture to executives in the cultural industries of the managerial challenges and opportunities when releasing artists on the market and the relative desirable conduct to embrace.
2. Literature Review
2.1 Cultural Goods and Innovation
The recorded music market is characterized by extreme uncertainty; only one out of ten albums released is profitable, with even lower rates if the artist is unknown (Ordanini, 2006). However, the lucky release can reach disproportionate success and cover all the previous failures, furthermore boosting the dreams of many candidates that didn’t succeed. This mechanism is known as the “winner-take-all” market system (Salganik, Doods and Watts, 2006). The model makes it very appealing for artists to undertake a career in the music industry; uncertainty is an attractor and it is reinforced especially by the undefined means that cause this extraordinary success. When the cause-effect relationship that governs the market is ambiguous, it is easier to assume the possession of those mysterious skills required to reach success. Therefore inducing many people to be extremely optimist for their time to become superstars. Indeed, there is a persistent oversupply of creative labour, which is independent of economic cycles; offering more creative people than the respective market can support (Peltoniemi, 2015). The saturated nature of the market is one of the main peculiarities of the creative sector and notably the music industry, it intensifies not only the imbalance of demand and supply but also the exploitation and underemployment of the creative employees. Which means long working hours, underpaid and most of the time with low job security. Even though this is one of the major problems for inequalities in the cultural sector, oversupply can still be beneficial. Indeed, it is one of the triggers of innovation: with such strong competition, people that do not succeed are motivated to strive for the better and bring novelties. If an artist is not successful, will be motivated to constantly improve and present unique ideas. The more competition, raises the rate of innovation or, at least, of differentiation between prototype-like works, in exploiting and stimulating consumer demand for novelty (Menger, 1999).
Before introducing the processes governing innovation in the Creative Industries and its origin from oversupply and heavy competition it is necessary to determine its construct. Not surprisingly, there is a lack of consensus on the definition of innovation due to the multiple facets it possesses. Some scholars prefer to refer to in in the terms of standardization and diversity, others instead agree with the adoption of the term non-conformity (Castaner, 2002). Nevertheless, some degree of consensus has been reached; innovation can be generally defined as the introduction in a field (or market) of something new (Cloake, 1997). With the new regarded as something that departs from the local or global existing conventions (Castaner, 2002).
The argument presented above, claiming a positive relationship between innovation and competition, has not only been proven by empirical studies: for instance Porter (1990) shows how industrial districts and clusters perform internationally better due to their strong level of rivalry that stimulates innovation. But also has found support in the classic formulations of human ecology literature, whose main principles develop from the concept of isomorphism. According to Hawley (1968), one of its main exponents, isomorphism is the tendency to conform organizational structures to environmental demands. In equilibrium, only one organizational form can optimally adapt the specific demands of each environment, so that the other entities will be forced to resemble to the leading unit in that specific environment (Hannan and Freeman, 1977). Looking at the recorded music scene with those lenses, failing producers that are motivated to stay in the market must conform their work to the optimally adapted individuals for that specific market. Losers are pushed to conform to the characteristics and behaviours of the winners in order to be competitive again. Therefore, the more innovation is demanded the more innovative artists need to be, thus stimulating continued creativity and originality. However, this prospective doesn’t take into account an interesting consideration: isomorphism can result not only because organizational decision makers learn optimal responses and adjust organizational behaviour accordingly (adaptation process) but also because optimal forms are selected out of a community of organizations (selection
process). Both processes are at work in most social environments, however the literature has barely considered the latter (Hannan and Freeman, 1977). In fact, often happens that the environment selects the optimal individuals and, functioning as Darwin’s natural selection process, only the “strong” manage to survive. In the context of cultural goods and in particular the music industry, the selection process is mirrored in the tendency of failing artists to be pushed out from the market, while innovative and successful artists are being selected to stay. In this sense competition does not motivate musicians to be more innovative but it is simply the result of a severe selection process. Identifying these two possible mechanisms makes it difficult to determine which one will dominate in each distinct situation. It is unclear whether is the environment that optimizes though the selection of “the best” or are the individuals that constantly adapt to the environment needs.
Yet, the diffusion of innovation is conditional to the relevant adopter who perceives a given idea, practice or object as innovative. The newness observed by the adopters can vary considerably, the same item might be regarded as innovative by one person but not by another, as well as what is perceived as innovative about an item may not be what another person see as new. Moreover, due to increasingly familiarity, innovations are perceived differently over time by given individuals. Indeed, the relevance of innovations changes over the different stages of adoption and diffusion processes. How entities recognize innovations is influenced by various physiological, psychological and cultural conditions that impact the perceptual process (Zaltman, Dubois, 1971). One of the most significant claims that emphasizes perceived newness by the adopter is found in the study of Zaltman and Lin (1971) where they define innovation as “Any idea, practice or material artefact perceived to be new by the relevant unit of adoption”. With regards to the creative industries the recognition of innovation is even more complex as a result of the perceptual process involved in the consumption and enjoyment of cultural goods. The individual rely on sensory perceptions in order to appreciate the “arts” and this is highly determined by the perception of one’s environment through various physiological, psychological and cultural conditions.
2.2 Innovation determinants
Taken into account the diverse possible interpretations of innovations, it is clear that delivering the right innovative breakthrough product is not an easy task. In order to survive in an unpredictable market it is vital to take into account distinctiveness and visibility features (Hsu, Negro, and Perretti, 2012). Namely, when innovating, organizations must find an optimal level of differentiation in order to offer a unique product, which is at the same time sufficiently similar to the current offer on the market in order to remain legitimate. Therefore firms face a trade-off between the need for exciting novelty and the consumers’ pressure for familiarity to understand the cultural product (Peltoniemi, 2015). The study in the motion picture industry by Hsu, Negro, and Perretti (2012) demonstrates how seeking a safer conformist strategy lead organizations to perform better on average but those that are brave enough to pursue higher risk paths of innovation have greater chances of exceptional success. Thus, the optimal differentiation point (Askin and Mauskapf, 2017) should be balanced between being different and being the same. Movies are experience goods and they are released in a hyper competitive environment where differentiation is crucial, characteristics that likewise apply to albums and songs. Indeed, the argument presented above is relevant also to the music industry: records that conform to the standards have better performances on average, although when deviating form the usual the chances to release a big hit are much higher. The goal is to find a balance between novelty and familiarity in order to increase the likelihood of having a profitable product.
As previously and extensively illustrated, novelty is essential in order to stand out from the crowd of the cultural goods producers and it is difficult to realize. It is hence fundamental to understand the factors that enable innovation as well as the obstacles that prevent it (Castaner, 2002). Various researches describe the elements that influence innovation. Economic studies mainly focus on the effect of the macro environment: the availability of resources and the local market and institutional
characteristics. On the one hand, an investigation of the U.S. opera repertoire in the 1989 – 1994 conducted by Pierce (2000), finds that the average income of the population in which the opera house is located is not a statistically significant factor for the conformity index. Concluding that the resources of the population, reflected by the households’ earnings, were not relevant for the development of innovation in the opera theatres. On the other hand, Heilbrun (2000) reports that the greater size of the population the higher the percentage of modern work programmed. Affirming thus a correlation between characteristics of the residents with the rate of innovation in that area. Consequently, findings up until now generate ambiguous conclusions on the effects that the characteristics of a specific region might have on innovation. Sociologists instead, have mainly studied the effects of meso level variables, that is, the funding mix linking the local community and the organization. For instance, Martorella (1977) and Heibrun (1998) observe the decreasing diversity of opera houses as the result of increasing importance of corporate sponsorships and parallel decline of public funding. Supporting the argument in favour of a negative correlation between sponsorships and innovation. While more recent studies from Pierce (2000) argue that receiving capital from the public funds encourages opera houses to take more risks and increase the level of unconventional programming (Castaner, 2002). For the purpose of this study the main focus will be on the examination of the meso level explanatory factors for innovation. Specifically, will look at organizational variables such as size, age and internal structure. In such manner the focal point of the analysis will be constricted to the exploration of the organizational dimension and its impact on successful product delivery on the market. In the following sections these aspects will be considered, with a special attention for the recorded music industry.
2.3 Music Industry
For the creative industries and the music records in particular, innovations are essential. Given the fact that so many products are released every year and that they are judged on the basis of their
ability to offer fun, enjoyment and pleasure, novelty is crucial for constantly introducing different kinds of goods to please diverse and fickle tastes (Peltoniemi, 2015). Indeed, for record labels the market share is largely determined by “product differentiation innovation”, which is the ability to launch new releases of established or unknown artists (Burke, 1996). With so many albums released every year, according to a recent estimate there were more than 150,000 albums released in 2017 that sold at least one physical or digital copy (RIAA, 2018), it is extremely difficult to be recognized and attract the public attention with your work.
The functioning of the recording industry mainly follows models such as have been developed by Burke (1996) where artists start their career as amateur, they record their first songs independently and they carry out autonomously few marketing and selling activities. For many of them the ultimate ambition is to be spotted and picked up by existing record labels (independent or major). In fact, many artists are believed to initially release music on their own to demonstrate their commercial potential, in order to secure a satisfactory contract later on (Burke, 1996). Consequently, after their early releases, many artists face the choice of licensing their musical output to record companies or courageously build their own independent career. Indeed, conducting interviews and surveys with Irish composers, Burke observed that artists tend to prefer the safer choice: sign an agreement with an existing record company. However, in terms of performance, is it actually the desirable choice?
On the one hand, the desire to control the use of one’s own music is an important stimulus for an independent career. A desire to have ‘more control over music’ and to be ‘independent/your own boss’ appear to be significant ‘pull’ factors for entrepreneurship. Many artists find out to be disappointed with the management of their music by major firms and hence crave to leave the company and set up their own independent career (Burke, 1997). In fact, as highlighted by Ordanini (2006) major companies have lower scouting capabilities as they are usually managed by decision
makers without any artistic background. The “patrons” or managers are actually the ones that take the final decision on what is going to be financed for a potential appearance on the market. Managers must select or reject ideas to release on the market based on their predictions of how they expect them to be. Therefore, their judgement is strongly biased on domain expertise, academic background and preferences (Berg, 2016). If you are not in their graces and style, there is a high chance that your production won’t be handed out by that label. In this kind of situation being independent gives more freedom of choices on which project will be carried on. In addition, an overall trend in the cultural industries is supporting this direction. The cultural scene is now dominated by the triumph of the market and its values and the professional craft-man is giving the way to the entrepreneur or more precisely the “self-employed” (Deresiewicz, 2015). Employees are becoming freelancers and independent contractors. Artists that register their records independently are in every aspect entrepreneurs. They need to invest money to record their work, which involves high risk taking triggered by the aspiration to reach the desired and hardly reachable success. The new technologies certainly facilitate this trend, giving space for self-online promotions, sales and the possibility to have direct contact with the users in a way that they can compete face to face with the existing organizations. The Internet has given an alternative route to connect and sell directly to the fans which many times is much more convenient then the deals made with record labels that clearly favour the latter. However, is the “self made artist” a real opportunity or is by large a necessity?
On the other hand, record companies have also a favoured position on the market, in fact, they benefit from access to capital, education, role models, music industry experience, distribution channels and marketing competences. They take care of the logistics and practicalities of the job with rapidity and vast expertise, leaving the artist plenty of time to be creative. Being contracted with a record label means also to be part of a whole experience and have relationships that can help you along your path, supporting professionally your activity (Dredge, 2018). In a research
conducted by Burke (1997), he found that musicians who start-up their own company usually do so because of “push” factors, because they are obligated by external forces or actors. These individuals haven’t been spotted either by the major’s recruiters or appreciated by the independent’s managers, and thus have to find a way in the market by themselves. This scenario is reported by Bruke to be more frequent than the actual drive from “pull” factors that are related to inner motivation and desire. Thus, on average, artistic entrepreneurship represents the individuals that have been rejected by incumbent firms and need to purse their dreams autonomously. Therefore, it seems that the labour market for musicians tends to push less commercially productive ones into the self-employed sector so that new entrepreneur are more likely to represent the rejected rather than the future superstars.
The following hypothesis tests the assumption that artists that release products with record labels will enjoy easier access to the market and resources needed to deliver breakthrough products in the music industry. Artists’ contracted with a record label enjoy greater support on the humdrum inputs, the organization validated expertise and experience combined with the supporting capital will be likely to increase the artists’ sales volumes and users’ evaluations. Overall enabling superior performance than the self-employed artists.
H1a: Releasing a music product with a record label has positive effects on sales volume H1b: Releasing a music product with a record label has positive effects on users’ evaluations
2.4 Meso Factors and Innovation
As previously mentioned, the literature reports contradictory theoretical arguments and puzzled empirical findings on the variables that affect the release of innovations; in particular meso variables, such as firm size. Indeed, size captures different theoretical constructs (such as resource
availability and stability among others), making the interpretation of its effect ambiguous (Castaner, 2002). The literature on innovation has generally recognized the twofold aspects of size, namely the impact of being big or small (Castaner 2012).
Organizational size is not an isolated concept, first of all is linked to the market structure. Indeed, oligopolistic markets are defined by the dominance of few large firms, while highly competitive markets are characterized by the presence of copious smaller firms. An open debate is still going on whether oligopolistic markets are more likely to stimulate innovation or rather greater competition activate greater innovation. With regards to the music industry, Peterson & Berger (1975) have argued that increasing market concentration leads to a decrease in the diversity of products offered by the record companies, therefore concentration in the industry affects the diversity of product in a linear fashion. In contrast, Alexander (1975) claims that both low and high levels of market concentration are associated with decreased product diversity, and that maximum diversity results from a moderately concentrated market. Thus, when industry concentration is very high or very low, product diversity is reduced (Alexander, 1996). In his study of the Top 40 Hits on Billboard from 1955 to 1988, he presents findings that support his assessment. In fact, during the so-called “rock revolution” the top hits recorded were relatively homogeneous when compared to those of the following decades (1967 – 1977) when concentration in the industry was higher. These data contradict Peterson & Berger claim that increased concentration reduces product diversity. Nevertheless, from 1971 till 1988, diversity decreases while concentration increases, thus suggesting a nonlinear relationship between market concentration and product diversity in the popular music industry. Peterson and Berger (1996) reply to Alexander, in the same volume of the American Sociological Review, questioning his measure of concentration, musical diversity and the difference between diversity and innovation. Finally, supporting the Shumpeterian hypothesis (1942) that firms in competitive markets are more rapid to achieve technical progress, a recent research from Gayle (2003) finds empirical evidence that a more concentrated industry stimulates
innovation. The study support the proposition that major innovations require significant amount of resources that only large firms tend to have and that these large firms are most likely to be located in concentrated industries. The paper utilizes a novel and more accurate measure of innovation, a citation-weighted count, which accounts for the heterogeneity of technologies covered by patents. Evidently, researchers haven’t reached yet a point of agreement on the relationship between innovation and market concentration, and the music industry is certainly involved.
Supporting Alexander’s argumentations, a study made by Phillips and Owens (2004), clearly shows how incumbents, dominating the oligopolistic music market, were the first movers in launching a new genre: jazz. They examine the emergence of recorded jazz between 1920 and 1929, where dominant firms had substantial market success with the introduction of the new musical style; they invested in a radical innovation more rapidly and successfully than the small independents. In fact, large organizations have the benefits of greater resources and income stability that allow them to potentially start experimenting. In line also with transaction cost theory, dominant and larger firms were able to take advantage of transaction costs reduction and also trigger social contagion in the demand market thanks to their broader customers reach (Phillips and Owens, 2004).
2.5 Meso Factors and Performance
The latter argument is reinforced by consistent empirical findings. In the first place, Simmel (1957) verifies how market consumption, especially in the Creative Industries, is basically a way to share social experiences so people prefer to do what other people do. Furthermore, Salganik, Doods and Watts (2006), through an investigation of a “fabricated music market”, demonstrate not only the strong impact of social influence but also the effects on both inequality and unpredictability of success. In an attempt to explain how success in the cultural markets can be extremely distinct form the average performance and yet so hard to anticipate, they find two possible reasons for this odd
inequality. Firstly, they assert the correlation between quality and success as convex, that is, differences in quality correspond to larger differences in success (Salganik, Doods and Watts, 2006). Secondly, they incorporate social influence in the model, taking into consideration the fact that individuals do not make decisions independently, but rather are influenced by the behaviour of others. The experiment shows how social influence contributes to inequality of outcomes in the creative market and how individuals that are subjected to stronger forms of social influence, report more increasingly unequal collective outcomes (Salganik, Doods and Watts, 2006). Therefore, it is widely accepted that social influence has a strong impact on individual behaviour. Especially for the cultural markets, where it has effects on the way in which particular products turn out to be regarded as good or bad. The judgment is made according to the choices of others causing experts and producers to fail in the prediction of success. Thus, the ability to influence rapidly large shares of consumers is a key resource in achieving market power and success.
Nowadays three majors dominate the music industry: the French-owned Universal Music Group, the Japanese-owned Sony Music Entertainment and the US-owned Warner Music Group. They control between 65% and 70% of both the recorded and streaming market shares (MIDiA, 2017). These major record labels control the larger portion of resources, capabilities and connections thanks to their longer presence on the market, which enables them to gain a competitive advantage. In addition, they hold the ability to activate the social contagion described by Phillips and Owens (2004), Simmel (1957) and Salganik, Doods and Watts (2006). Artist can reach success easily and faster thanks to the major’s diffusion scale, which is many times broader compared to the small independent consumers reach. For those reasons, a superior performance, in terms of volume of sales, will be expected for artists that sign a contract with major record labels.
H2a: The positive effect of releasing a music product with a record label on sales volume is more positive when the record label is a major
Nonetheless, in many instances small companies present advantages. In fact, contrary to large organizations that suffer from inertia and high bureaucracy, which slow down every action toward change, small size businesses are fast and agile. They can quickly adapt to changes and conform to the needs of the market. Moreover, smaller businesses, due to their limited size and scope, can be related to new entrants and thus hold common benefits. In fact, Aron and Lazear’s (1990) affirm that new entrants are more likely to succeed in the release of new products because of benefits in high variance strategies; cannibalization that affect existing firms and newcomer asymmetrically and finally thanks to diseconomies of scope that influence incumbents performance negatively. Incumbent firms recognize new possibilities but find it too costly or risky to actually implement them (Aron and Lazear, 1990). Moreover, the new knowledge required by the innovation renders an incumbent obsolete and thus incompetent in its production, while favouring the young newcomers (Phillips, 2004).
In the music industry context, small Independents fully enjoy this favoured position. They are able to deliver the latest innovative artist on the market, rapidly and ahead of time. Furthermore, musicians can profit from more professional and closer working relationships. Indeed, niche record labels can offer specialized and tailored management because guided by experts in the music field who are able to deliver adequate advices to their employees (Ordanini, 2006). The small Indies are more competent than Majors in discovering artists potentials, and thus able to survive short-term fashions and trends, leveraging some specific selection capabilities (Kretschmer et al., 1999). This assumption is also confirmed by the study conducted by Ordanini on the Italian record market. He verified that artists working for the small Independents, even though the longer path to reach success have longer presence on chart, thus greater survival.
Moreover, when looking at contracts arranged between musicians and record companies, that is, between an individual and a firm, many social aspects and mechanisms involved can be compared
to agreements executed between companies. In both cases the negotiation involves the disposition of many terms and clauses that need to be approved by the two parties. Although, when two companies are negotiating, specific remarks need to be ascribed for these legal entities with regards to their specific rights and responsibilities. Firms are empowered with unique legal rights such as buying or selling trademarks and hiring employees, which are not conferred to single individuals. Yet, many of the issues implied in the agreements between two organizations are still very similar to the ones between record labels and artists. Post merger integration is a challenge for both the acquirer company and the firm acquired (Thomson, 2001) as it is for record labels and artists when building new alliances with musicians. In fact, acquirers (record labels) desire a successful integration of the new assets so that they can gain profit from them. Acquired (artists), on their side, wish to find a place in the new corporate culture but not loosing confidence and morale. The post acquisition process is a very delicate moment for both actors; many authors have tried to analyse this process and find the best way to deliver value to the subjects involved. Birkinshaw, Bresman and Hakanson (2000) focus on the importance of task integration combined with human integration. The former leads to a satisfying solution of transfers of capabilities and resources, the latter concerned primarily with generating satisfaction, and ultimately a shared identity, among the employees from both companies. These two processes are distinct but not independent, when used in combination they lead to greater interdependencies and benefits (Birkinshaw, 2000). In the creative sector this “human aspect” is extremely important as artistic quality can rise only if the creator feels supported, sustained and conform to the company ethics. Cultural integration becomes easier when occurring in small and independent record companies that acquire artist with similar characteristics, mind-set and goals. Creativity is enhanced and artists can gain self-confidence and explore innovative directions. In contrast, big international companies are mainly focused on the big masses demands and have more profit focus strategies. In this case artists could quickly feel alienated from their works and aspirations. Every move is directed to gain market shares, with no adequate attention to artistic quality and personality of the artist. The feeling of not being
appreciated and working only for money’s sake can surely emerge more intensively when working for big major companies, inducing less creative outcomes.
Therefore, superior agility in the delivery of new musical products, major attention to individual excellence and sensibility for artistic taste, typical of the indie record labels, is expected to enhance the musicians’ release quality. Thus, being contracted with independent record labels, increase the likelihood to deliver high quality products and, as a consequence, higher evaluations in the ratings.
H2b: The positive effect of releasing a music product with a record label on users’ evaluations is more positive when the record label is an independent
To summarize, artists contracted with record labels will enjoy easier access to the market and resources needed to deliver breakthrough products in the music industry. They enjoy greater support on the humdrum inputs thanks to the organization validated expertise and experience. Moreover, artist contracted with a major record label can reach success easily and faster thanks to the major’s diffusion scale, thus for those reasons, a superior volume sales performance for products released by major record label will be expected. However, superior agility in the delivery of new musical products and greater attention to individual excellence is expected to enhance the musicians’ quality when products are released with indie record labels. Thus, being contracted with an independent record label, increase the likability to deliver high quality products and, as a consequence, higher evaluations.
Figure 1. Conceptual Model H1a H2a H1b H2b Record Label Major
Independent Users’ Evaluation
Sales Volume (Highest position, Duration) Record Label Sales Volume (Highest position, Duration) Users’ Evaluation
3. Data and Method
3.1 Data Collection
An empirical quantitative research in the popular music industry is conducted to test the hypotheses. Similarly to all creative industries, the popular music industry is characterized by a “winner takes all” system (Salganik, Doods and Watts, 2006) and main indicators of success are popularity charts, such as Billboard Hot 100 and Billboard 200. In fact, Billboard magazine’s weekly performance charts are crucial for making decisions and signalling reputations for all those in the commercial music field (Arnand and Peterson, 2000). They provide highly accurate hierarchical data about market performance of the best performing songs or albums every week. Rankings are based on multi-metric consumption, blending traditional sales and track streaming of equivalent albums providing an accurate standard system for evaluating career success. The quickly interpretable chart information is used to coordinate and evaluate the performance of organizational subunits (Peterson and Berger, 1971) and is also used to make attributions about individual careers. Job tenure of artist in the music field is continually under review and individuals are regularly promoted, reassigned, or fired based on the success of the work they have been associated with (Karshner 1971, Peterson 1977).
Billboards Chart sales data are the most important information source in the recording industry (Ordanini, 2006), they have been used extensively to analyse the recording industry, but mainly for study interested in the level of concentration or differentiation in the market (Peterson and Berger, 1996; Alexander, 1996). Although chart performance is the most important indicator of success, it has never ben applied to artist release choices in the recording industries. This study will investigate the relationship between chart success and the collaboration with an organization, particularly the release of an album with the support or not of a record label. For this purpose the Billboard 200 is
used as secondary database to collect data on album’s sales performance. Indeed, albums are a key device for the purchase and consumption of music (Ordanini, 2006) and Billboard top 200 is the most appropriate indicator of market approval. In fact, it is based exclusively on sales, which is the unit of measure employed in the present study while other charts, for example the Billboard Hot 100, also factors other aspects such as streaming and radio airplay.
Not only chart sales but also users’ ratings will be analysed, in order to take into consideration the twofold aspects of performance. Public evaluations are one of the measures to assess the artistic quality of music products. Public ratings can have double nature: either pronounced by professional critics or popular users. Following Wanderer (1987) study on film ratings among professionals and lay audiences, he verifies the assumption that critics’ tastes are similar to consumers’ tastes. Indeed, his research demonstrates how the snobbism generally appointed to the experts’ judgement is in fact a mere representation of the upper-middle class taste they are member of. As part of a taste public, the critic is expected to share the taste of that public and find the same cultural products attractive (Wanderer, 1987). In addition, the role of the critic as predictor corroborates this expectation. In fact, experts are merely leading indicators of performance with no significant influence on box office revenues (Eliashberg and Shugan, 1997). From this perspective, professional critics are simply a representation of their audience, conveying beforehand the general public opinion.
On the one side the financial achievements and on the other side public’s evaluations. The two measures are used in order to offer a more coherent and complete picture of the artists’ performance form the audience prospective, in fact, sales and ratings both referred to consumers’ appeal. All Music.com is used to extract user ratings data, indeed the site features extensive information on album’s users evaluations. Furthermore, Musicbrainz.com offers comprehensive information on the artist discography and more precisely the record label of each album released. Musicbrainz.com is a rich database containing extremely accurate reports on artists and their related works. Thus,
information about the record label in charge of the release of the album, namely major, independent or no record label, are collected here.
3.2 Sample
To assess potential performance differences between artists contracted with independent, major or no record label, the research focuses on 200 artists with the highest presence on the Billboard chart during the five years from 2013 to 2018. The number of artists is fixed to 200 following previous studies also looking at discrepancies between musicians’ performances (Ordanini, 2006). A sample reflecting the composition of the market is created. The market share for each category is used to build a sample that mirrors the proportion of revenues for each group, namely, 3% of independent artists, 27% of artists contracted with an independent record label and 68% of artists contracted with a major record label (MIDiA, 2018). Quota sampling method is chosen because more suitable to study characteristics of particular subgroups. These subgroups are selected with respect to known features and traits; in this research quota sample based on market share revenues is applied in order to create an appropriate representation of the population that count for sales, the unit of measure of this research. However, it relies on the researcher’s judgement in giving the right weightages to each subgroup (distribution of revenues) and thus might have some bias.
Data are retrieved form Billboard.com, using a procedure that takes into consideration the end of month chart for each month between the years 2013 and 2018. A total of 1158 albums in 63 months are collected. Subsequently, a random extraction is computed until the quotas for each category are filled. The choice of analysing only the most successful artists and relative albums is due to the fact that the level of success of an artist does not only depend on the contract they have, but it is also heavily dependent on the strategies and resources invested for that individual (Ordanini, 2006). Thus, collecting data for the highest positioned on chart avoids the problems related to potential
asymmetric strategy efforts. Further, the reason for choosing this particular timespan is that previous studies suggest that differences in performance are already noticeable in a timespan of five years (Cappetta, Cillo & Ponti, 2006; Peterson, 1996). Moreover, in order to execute a more modern and up-to-date analysis the last five years were taken into consideration.
3.3 Variables
For the examination of the hypothesis the variables record label ownership, highest album position, number of weeks on chart, users evaluation, critics evaluation, month of entrance on chart and genre are used. Further explanations are found below. In this paper only albums that contain original music are analyzed.
3.3.1 Independent Variables
The contracting condition for each individual, namely having released the peak album with a major, indie or no record label, is used as independent variable. For the analysis of the hypothesis two dummy variables are created: independent, with values “independent” (1) and “major” (0) and major, with values “independent” (0) and “major” (1). The baseline group is the “no record label” released albums and it has values of (0) for each dummy variable. Information are collected from Musicbrainz.org looking at the record label ownership for each peak album master release. The choice of analysing only the most successful album is due to the fact that the level of success of an artist does not only depend on the contract they have, but it is also heavily dependent on the strategies and resources invested for that individual (Ordanini, 2006). Thus, collecting data for the highest positioned album on chart avoids the problems related to potential asymmetric strategy efforts.
3.3.2 Dependent Variables
Sales volume performance is measured using two different types of information, in order to reflect its twofold nature: the highest position and the number of weeks of presence on Billboard 200 are considered. Data regarding the highest album position on chart are retrieved from Billborad.com and reflect the highest place for that album during the period January 2013 – March 2018. The highest position corresponds to the 1st and the lowest to the 200th. This variable is a scale variable, in fact its values represent ordered categories with a meaningful metric. Data regarding the number of weeks of presence on chart are also retrieved from Billboard.com and reflect the longevity of an album that was released between January 2013 and March 2018. This variable is a scale variable, which reports a meaningful metric that uses a week as unit of analysis.
Evaluation performance is measured by looking at users’ ratings for each album. Ratings are based on a five star linkert scale, captured from AllMusic.com database. Score of 1 is given to the less attractive and 5 to the most attractive. In this case the variable taken into consideration is an ordinal variable, in fact it reflects categories with intrinsic ranges.
3.3.3 Control Variables
A control variable instrumental to detect potential differences in musicians’ artistic quality is added to the study. Critics’ evaluations are used to assess the possibility of dissimilar artistic abilities. This information is collected form Allmusic.com that records experts’ evaluations for each album with scores on a likert scale ranging from 1 (for the less praised) to 5 (for the most praised). Just as users’ evaluations, also critics’ evaluations are ordinal variables, reflecting categories with intrinsic ranges.
In addition, month of entrance on chart is used as control variable to minimize the negative effects that a later release can have on the dependent variable number of weeks on chart. Namely, album the were released at the beginning of the period of study have more chances to have longer presence on charts then the album released at the end of the period. Thus, a number is assigned to each month of album release, in particular from values of (0) to releases before January 2013, values of (1) to releases in January 2013 till values of (63) to releases in March 2018.
Finally, control for genre is integrated in the research. Data are collected from AllMusic.com database, which reports accurately each album genre classification. In addition, for the album considered in this study, AllMusic.com reports only one main genre, and this is the sole reported. This variable is a categorical dummy variable. For the study three dummy variables are created: Pop/Rock, with values “pop/rock” (1), “rap” (0) and “rnb” (0), Rap, with values “pop/rock” (0), “rap” (1) and “rnb” (0) and RnB, with values “pop/rock” (0), “rap” (0) and “rnb” (1). The baseline group are the albums with Other Genre with values of (0) for each dummy variable.
4. Results
4.1 Descriptive statistics
Firstly, missing values are spotted and following a listwise approach cases with no values recorded are excluded from the research. From a database of 200 cases only 12 were missing data, thus reducing the total data set to 184 cases. Secondly, frequencies analysis is run to check for errors in the data. No errors are found. Thirdly, descriptive statistics are run for both categorical and ordinal variables. For the categorical variables record label ownership and genre, frequencies and percentage are derived. The results are reported in Table 1. For the ordinal variables highest album position, number of weeks on chart, users’ evaluation, month of entrance on chart and critics’ evaluation mean, minimum, maximum and standard deviation values are derived. Results are depicted in Table 2. Bar graphs for the dependent variables, highest album position, number of weeks on chart and users’ evaluations are included for a better visual representation of the distribution of the data.
Table 1. Frequencies of Record Label and Genre
Record Label Frequency Percent No Record Label 8 4.3 Independent 51 27.7 Major 125 67.9 Total 184 100
Genre Frequency Percent Pop/Rock 100 54.3 Rap 22 12.0 RnB 17 9.2 Others 45 24.5 Total 184 100
Table 2. Means, Maximum, Minimum, Standard Deviations
Mean Maximum Minimum SD
Highest Album Position 28.2 175 1 36.25 N. Weeks on Chart 21.99 203 1 38.16 Month of Entrance on Chart 25.34 63 0 17.45 Users’ Evaluation 3.84 5 .5 .58 Critics’ Evaluation 3.54 5 2 .61
Bar graph 1. Highest Album Position on Chart
Bar graph 3. Users’ Evaluations
Moreover, correlation between the variables is computed with SPSS with the correlation matrix function for all the combinations of variables. The results are reported in Table 3. From the correlation analysis, critics evaluation is the stronger predictor of users evaluation with a Pearson correlation coefficient of r= 0.36 and the significance value less than 0.01. This finding support Wanderer (1987) assumption that critics’ tastes are similar to consumers’ tastes. Indeed, being correlated means that the two evaluations are moving on the same direction and with similar magnitude, indicating that they are comparable measures. Further, the number of weeks on chart is a moderately strong negative predictor of highest album position with a Pearson correlation coefficient of r= -0.32 and the significance value less than 0.01. This correlation is logical, in fact, when an album reach high positions on chart (smaller figures) it should also indicate better quality and acceptance, and thus longer presence on chart.
Table 3. Correlation matrix 1 2 3 4 5 6 7 1. Record Label . 2. Highest Album Position -.22** . 3. N. Weeks on Chart .19* -.32** . 4. Month of Entrance on Chart -.10 -.05 -.07 . 5. Users’ Evaluation .04 .02 .11 .11 . 6. Critics’ Evaluation -.12 .07 .09 .08 .36** . 7. Genre -.03 -.06 .01 -.05 -.13 .07 .
** Correlation is significant at the 0.01 level (2-tailed) * Correlation is significant at the 0.05 level (2-tailed)
Finally, skewness and kurtosis measures are computed to determine the normal distribution. The variable highest album position is not normally distributed. Its Skewness is above 1 with a value of 1.96, which indicates that the tail on the right side is likely to be longer that the left side. The Kurtosis value over 1 with a value of 3.74, this is called Leptokurtic distribution and as the value is high it indicates that the distribution might have slightly flatter tails. The item users’ evaluation is not normally distributed. Its Skewness is below -1 with a value of -1.72, which indicates that the tail on the left side of the probability function is longer that the right side. The Kurtosis value over 1
with a value of 7.85, this is called Leptokurtic distribution and as the value is high it indicates that the distribution might have very flatter tails. However, “with reasonably large samples, Skewness will not make a substantive difference in the analysis” (Tabachnick & Fidell, 2001, p. 74). “Kurtosis can result in an underestimate of the variance, but this risk is also reduced with a large sample” (approximately 200 cases: Tabachnick & Fidell, 2001, p. 75). In this research there are 184 cases, therefore the risk is reduced and Skewness and Kurtosis would not make a substantive difference in the analysis. Thus, in this case, the study will be conducted assuming the variables highest album position on chart and users’ evaluations as normally distributed.
The variable number of weeks on chart is not normally distributed. Its Skewness is above 1 with a value of 2.59, which indicates that the tail on the right side is longer that the left side. The Kurtosis is over 1 with a value of 6.79, this is called Leptokurtic distribution and as the value is high it indicates that the distribution might have very flatter tails. The variable month of entrance on chart is normally distributed. Its Skewness is close to zero with a value of .402 and the Kurtosis is close to -1 with a value of -.931, that might indicate a slightly Platykurtic distribution with the left tail slightly flatter than the right one. Finally, the variable critics’ evaluation is normally distributed. Its Skewness is close to zero with a value of -.30 and the Kurtosis value slightly below 1 with a value of .36.
4.2 Hypothesis Testing
Three regressions are run in order to test the hypothesis of this research. Firstly, a hierarchical multiple regression and a Poisson regression are run in order to test H1a and H2a, respectively in regard to the relation between highest position and duration on chart and record label ownership. Secondly, a hierarchical multiple regression is run to test H1b and H2b, specifically to determine the relation between record label ownership and users’ evaluations. Below the empirical analysis.
4.2.1 Hypothesis H1a, H2a
Two regressions are run to test Hypothesis H1a and H2a. Respectively they test if signing a contract with a record label has positive effects on sales volume and if having a contract with a major record label enhances this effect. Sales volume performance is measured first as the highest position reached on chart and secondly with number of weeks of presence on chart.
Firstly, a hierarchical multiple regression is run to examine the relation between the independent variable record label ownership and the dependent variable highest album position on chart. In the first place, the regression is performed for the control variables: critics’ evaluation, month of entrance on chart and genre. This was done for these three variables so that the variable highest album position on chart can be controlled. In step 1 the model was not statistically significant F(5, 178) = 1.836; p= .108 and explained 4.9 % variance in highest album position on chart. So it can be concluded that most variance of the variable of highest album position on chart is explained by other factors. In step 2 record label ownership was entered as a predictor and the total variance explained by the model as a whole was 12.3% F(2, 176) = 7.456. Thus the introduction of record label explained an additional 7.4% in highest album position, after controlling for critics’ evaluation, genre and month of entrance on chart. These results are statistically significant with a p value of p= .001. Table 4 gives an overview of the results from the regression analysis.
The negative sign of the beta for the dummy variable of the group of major record label ownership reveals a decrease of respectively -.171 on the dependent variable when comparing this category with the baseline group of artists that released without any record label. On the contrary, a positive sign of the beta for the dummy variable independent record label ownership reveals an increase of respectively .110 on the dependent variable when comparing this category with the baseline group of artists that released without any record label. These findings reflect better positions on chart
(smaller figures) for artists that are contracted with a major record labels in respect to the group of independent artists and relative worst positions on chart (bigger figures) for indie record label artists compared to the independent artists. Therefore these results partially support H1a and support for H2a with respect to performance on position on chart. Indeed, only artists that released an album with a Major record label enjoy higher rankings on chart (represented by lower figures), followed by independent artists, which reveal higher positions on chart compared to artists contracted with Independent record labels. For a better visualization of the results a Boxplot graph has been created.
Table 4. Hierarchical Regression Model of Highest album position
R R2 R2 Change B SE ß T Step 1 .221 .049 .049 Critics’ Evaluation 5.916 4.421 .099 1.338 Month of Entrance -.074 .154 -.036 -.484 Dummy Pop/Rock 6.148 6.412 .085 .959 Dummy Rap -16.343 8.764 -.157 -1.865 Dummy RnB .254 10.183 .002 .025 Step 2 .351 .123 .074 Critics’ Evaluation 4.273 4.292 .071 .996 Month of Entrance -.099 .149 -.048 -.663 Dummy Pop/Rock 7.454 6.201 .103 1.202 Dummy Rap -14.423 8.641 -.138 -1.669 Dummy RnB 1.647 9.845 .013 .167 Dummy Independent 8.909 13.606 .110 .655 Dummy Major -13.256 13.049 -.171 -1.016 Statistical significance: *p <.05; **p <.01; ***p <.001
Boxplot 1. Record Label and Highest Album Position on Chart
Secondly, a Poisson regression is run to examine the relation between the independent variable record label ownership and the dependent variable number of weeks on chart. Firstly, the regression is performed for the control variables: critics’ evaluation, month of entrance on chart and genre. A high number for value divided by degrees of freedom is observed Value/df = 39.182, which indicates an important over dispersion. Moreover, the Omnibus test reveals a statistically significant value of p= .000 for the model and the Tests of Model Effects shows statistically significant values for each variable, namely, genre p= .000, critics evaluation p= .000 and month of entrance on chart p= .000. These statistically significant results are reported in the next table where, the exponentiated beta values, which explain the contribution of each control variable (critics evaluation, month of entrance on chart and genre) on the dependent variable (number of weeks on chart), are determined. Data are collected in Table 5.
Afterwards, the independent variable (record label ownership) is added to the model. A high number for value divided by degrees of freedom is observed Value/df = 35.2, which indicates an important over dispersion. Moreover, both the Omnibus test reveals a statistically significant value of p= .000 and the Tests of Model Effects shows statistically significant values for each variable, namely, genre p= .000, critics evaluation p= .000, month of entrance on chart p= .000 and record label ownership p= .000. These statistically significant results are interpreted below. The exponentiated beta values are determined, and specifically explain the contribution of the independent variables on the dependent variable. They show a 50.5% decrease in the number of weeks on chart for independent artists compared to artists contracted with a major record label and a 70.4 % decrease on number of weeks on chart for artist contracted with an independent record label compared to major record labels. Results are presented in Table 5. Therefore, these findings partially support H1a and support H2a with respect to performance on position on chart. Indeed, artists that released an album with a major record label enjoy longer periods on chart (represented by higher number of weeks on chart), however they are followed by independent artists which revel higher numbers of weeks on chart compared to artists contracted with Independent record labels. For a better visualization of the results a Boxplot graph has been created.
Table 5. Parameter Estimates. Number of Weeks on Chart
Model 1 B Std. Error Exp (B)
Pop/Rock .111 .044 1.118
Rap 1.132 .049 3.103
RnB .528 .059 1.695
Others . . 1
Month of Entrance -.011 .001 .989 Model 2 Pop/Rock .078 .044 1.081 Rap 1.085 .049 2.958 RnB .519 .059 1.680 Others . . 1 Critics’ Evaluation .250 .025 1.284 Month of Entrance -.010 .001 .990 No Record Label -.705 .083 .494 Independent -1.216 .053 .296 Major . . 1
Overall, the above findings show partial support for H1a and support for H2a. Similar results for both regressions confirming the fact that highest position on chart and duration on chart are related between each other and together reflect sales volume performance. In fact, reaching high chart positions is a sign of great quality that definitely enables longer and stable presence on chart. Both regressions reveal that the best performances on sales volume are associated with artists that release musical products with major record labels (supporting H2a), however, results show also that artists releasing musical works independently enjoy better performances on sales then the ones working with indie record labels (only partially supporting H1a).
4.2.2 Hypothesis H1b, H2b
A hierarchical regression analysis is used to examine the relation between the independent variable (record label ownership) and the dependent variable (users’ evaluations). The hierarchical regression tests Hypothesis H1b and H2b, specifically that having a contract with a record label has positive effects on users’ evaluation and this effect is more positive when signing a contract with an Independent record label.
A hierarchical multiple regression is performed to research how having a contract with a record label predicts users’ evaluation, after controlling for critics’ evaluation, musical genre and month of entrance on chart. As a first step of the hierarchical multiple regression, three predictors, namely the control variables were entered: critics’ evaluation, musical genre and month of entrance on chart. This was done for these three variables so that the variable users’ evaluation can be controlled, thus the observed effect of being contracted with an independent, major or having no contract with a record label is independent of the effect of the control variables. In step 1 the model was statistically significant F(5, 178) = 7.74; p < .001 and explained 17.9 % variance in users’ evaluation. So it can be concluded that most variance of the variable of users’ evaluation is
explained by other factors. In step 2 record label ownership was entered as a predictor and the total variance explained by the model as a whole was 19.7% F(2, 176) = 1.991. Thus, the introduction of record label explained an additional 1.8% in users’ evaluation. Table 6 gives an overview of the results from the regression analysis.
The negative sign of the beta for the dummy variables reflecting the groups of independent and major record label ownership would reveal a decrease of respectively -.195 and -.068 on the dependent variable when comparing with the baseline group of artists that released the without any record label. This would reflect a greater decrease in users rating for artist that are contracted with an indie record label in respect to the group of independent artists and a relative lower decrease for artists contracted with a major record label compared to the group of independent artists. It would therefore indicate that releasing a musical product with a record label do not have positive effects on users’ evaluation, instead, artists that would release their album independently would be better off than their colleague contracted with record labels. However, this finding is not statistically significant with p values of p= .229 and p= .673 each. This means that the estimated betas are not statistically different from zero and that there is no evidence of any effect, whether positive or negative. Therefore, no evidence was found against the hypothesis. Not providing any support for H1b and H2b.
Additionally, a Boxplot has been created in order to enable us to better study and visualize the distributional characteristics of each study group and their users’ evaluation level. The plot shows how the values for each independent variable are distributed in respect to users’ evaluation. Independent artist display relatively higher users’ rating compared to the other two groups with a third quartile above 4.0. While Major record label ownership has slightly greater values than independent record label ownership. Interesting to observe is that the three groups all present a
median value of 4.0, which gives us an idea of the typical value for each group, thus suggesting very high similarities in users’ evaluation between the three.
Table 6. Hierarchical Regression Model of Users’ rating
R R2 R2 Change B SE ß T Step 1 .423 .179*** .179 Critics’ Evaluation .363 .066 .377*** 5.483 Month of Entrance .003 .002 .090 1.309 Dummy Pop/Rock .236 .096 .201* 2.454 Dummy Rap -.015 .131 -.009 -.114 Dummy RnB 0.221 0.153 0.110 1.451 Step 2 .444 .197 .018 Critics’ Evaluation .373 .066 .387*** 5.627 Month of Entrance .003 .002 .089 1.301 Dummy Pop/Rock .226 .096 .193* 2.359 Dummy Rap -.052 .133 -.031 -.393 Dummy RnB .217 .152 .108 1.425 Dummy Independent -.254 .210 -.195 -1.208 Dummy Major -.085 .201 -.068 -.423 Statistical significance: *p <.05; **p <.01; ***p <.001
5. Discussion
In this study three different alternatives when releasing an album are studied, namely releasing a musical product with the support of a record label (major or indie) or independently. These options are analysed in relation to their effects on performance, on the one side volume sales (highest position on chart and number of weeks on chart) and on the other side users’ evaluation. Different elements can affect the production of musical products, innovation, rate of adoption and selection, market concentration, social contagion abilities and finally integration issues. This paper tries to offer an overall perspective on the mechanisms that influence the relation between creators and producers of musical products and the implications on the final output.
In the first place, looking at findings on sales performance and how they are influenced by the collaboration with record labels, hypothesis H1a and H2a are only partially supported. Releasing a musical product in collaboration with a record label does contribute to reach better positions and longer presence on chart. However, this is valid only for major record labels and not for independent ones. Indeed, major record labels, thanks to their greater resources, enjoy larger marketing and advertising abilities to influence the particular tastes of the consumers and thus reach better sales volume and approval for the albums released with their support. This abundance of resources designated to commercial activities has also positive effects in the long period, with sustained presence of albums on chart. Moreover, major record labels have the advantage, enabled by their greater capital and income stability, to deliver breakthrough products more frequently, in fact they can take more risks, and more often benefit from incredible successes and profits. In addition, large organizations are able to trigger social contagion: especially for the cultural markets, people prefer to do what other people do, due to the unknown outcome and consequences of cultural products, it is common habit to follow the herd when choosing what experience good to consume. Thus, major record labels, thanks to the influence they have on bigger audiences, are able
