The effect
of
user-generated
content on
charts
in
the music
industry
Tom Jelte Heij
1457608
MSc
Strategic
2
Abstract
3 Table of Contents 1. Introduction ... 4 Research Questions ... 5 1.2 Thesis Outline... 5 2. Theoretical Framework ... 6
2.1 Market Information Regimes ... 6
2.1.1 Scope ... 6
2.1.2 Methodology ... 7
2.1.3 Political tone ... 7
2.2 Consequences of Market Information Regimes ... 8
2.3 Selection systems ... 9
2.3.2 Expert Selection ... 10
2.3.3 Selection systems and music ... 10
2.4 Conceptual Model ... 13
3. User-generated content ... 14
3.1 Volume and valence of UGC ... 15
3.2 Volume and valence in the music industry ... 16
4. Methodology and data ... 18
4.1 Settings & data collection ... 18
4.2 Measures ... 19 4.2.1 Dependent variables ... 19 4.2.2 Independent Variables ... 20 UGC Volume ... 21 UGC Valence ... 21 4.2.3 Control variables ... 21
Critic reviews volume & valence... 21
Album Type ... 22
5. Analysis and results ... 22
5.1 Chart appearance ... 22
5.2 Cumulative success ... 23
5.3 Opening success ... 25
6. Conclusions & Discussion ... 26
6.1 Managerial implications ... 28
4
1. Introduction
The rise of digital technologies has had a radical impact on the world we live in today. Especially with regard to information the emergence of the internet and search engines such as Google provides consumers with the possibility to gather information on virtually every subject imaginable. Research as to how this type of information is used by consumers and what influence it might have on retail sales is abundant (see Floyd, Freling, Alhoqail, Cho and Freling, 2014). Music is a particular field of interest as it is a market that is frequently said to be moving entirely to the online realm.
Success in the music industry has historically been measured by the performance in ―the charts‖. Charts are ordinal ranked lists based on sales and airplay data which reflect the most popular music. Charts therefore function as the market information regime for the music industry. A market information regime as defined by Anand and Peterson (2000) is ―the medium through which producers observe each other and market participants and market participants make sense of their world‖. For publishers this means observing how their products fare against the competition and for consumers it is a tool to see what the current most popular products are.
Technological changes have always affected the way sales professionals deal with their customers varying from the invention of the printed press to the telephone and the internet and modern communication channels available through the internet. In addition to ―old‖ media such as television, radio and print the internet now offers multiple channels through which consumers can experience, rate and purchase products such as music. In theory this alters the selection systems within the industry, specifically because consumers can now communicate directly with each other about the perceived quality of products through so-called ―user-generated content‖ and thus alter the outcomes in the marketplace. Research into online content has uncovered volume and valence as possible indicators of quality (Dellarocas, Zhang and Awad. 2007, Duan, Gu and Whinston. 2008.) meaning that the amount of content generated as well as a rating given by consumers can have an effect on the performance of certain products.
5
industry it is vital to understand how new technology shapes and affects the competitive process in order to adapt to an industry in transition from physical to almost total online distribution and promotion.
Research Questions
This leads to the following research question:
How does online user-generated content affect chart position and performance in the music industry?
Considering the main research question the following sub-questions are formulated: - What is a market information regime?
- What are the selection systems within the music industry? And how do they determine the credibility of the source?
- How does user-generated content affect the music market and consumer decision-making?
1.2 Thesis Outline
6
2.
Theoretical Framework
2.1 Market Information Regimes
Sense-making in the music industry is important as music is a cultural and information good of which the value can only be truly determined after consumption (Hirsch, 1972). In such a competitive marketplace where it is difficult to determine the quality or potential for success a market information regime provides insight for both publishers and consumers into what is actually going on in the market, and what songs are popular (success).
In the music industry the typical form of market information regime has been the ―chart‖. A musical chart generally consists of a collection of musical products ranked ordinally whereby the most popular product is ranked number one, the second most popular as number two and so forth in a descending order. The traditional chart system has been the dominant market information regime for consumers alongside radio and other sources such as television and magazines. The methodology of the chart system has historically been based upon sales and radio-airplay but recent technological developments have forced chart makers to consider and integrate online influences. For example, the leading American chart maker billboard now has a ―Social 50‖ chart where an artists' popularity is determined by a formula blending their weekly additions of friends/fans/followers on online platforms along with artist page views and weekly song plays.
This research follows the definition for a market information regime as given by Anand and Peterson (2000) which is ―the medium through which producers observe each other and market participants make sense of their world‖. Charts function as the market information regime for the music industry as producers, publishers and consumers measure success for music products by their performance and appearance on the charts. Therefore, in order to explain the effect of charts on the music industry a theoretical understanding is needed of the role charts play as a market information regime. Market information regimes as, defined by Anand and Peterson (2000), have three important characteristics: scope, methodology and political tone.
2.1.1 Scope
7
chart or on a broader and more diversified, specific set of genres such as Latin, R&B, Dance etc. where each genre has its own specific chart to give information about that specific genre. Since the conception of the chart system the number of charts has diversified in order to reflect the performance of not only all music (Top40, Billboard 100, etc.) but also of specific genres (R&B/Hip-Hop chart, Latin chart, Dance chart etc.)
2.1.2 Methodology
Research by Anand and Peterson (2000) has shown that a change in methodology can reveal significant differences in performance. Therefore, the methodology of a market information regime is perhaps the most important aspect as it dictates the outcome of the market information regime and thus the actual perceived performance of the product. Changes in technology can therefore be of influence on how performance is measured. In the music industry concerning the methodology of the chart system there is a continuous debate on how to accurately report the most popular music. For example, in the U.S. the chart composer billboard has started a chart that tracks the most popular songs on the social media twitter, ranking songs based upon acceleration in shares: total shares over the past hour compared to the song‘s hourly average over the past day.
2.1.3 Political tone
8 2.2 Consequences of Market Information Regimes
Research into market information regimes has shown that they have an effect on actual performance in the market (Anand and Peterson, 2000. Andrews and Napoli. 2006). Anand and Peterson (2000) revealed that a change in the market information regime can severely alter the understanding of the market and thus the actual performance of a product. Similarly, Andrews and Napoli (2006) researched the effect of a change in market information regime in the book publishing industry. They found that publishers resisted a new, more objective market information regime as the old system served as a promotional tool which not only reflected consumer demand but was also an important component in driving consumer demand. Giles (2007) found that there was a significant change in time spent at the top of the chart from the moment ‗album cuts‘ were included in the compilation of the Billboard chart and not only traditionally released ‗singles‘ were included. This revealed that other individual songs not released as commercial singles were sometimes more popular amongst consumers than commercial singles thus revealing new information about the market.
The selected market information regime for the music industry is the chart system in which the most popular products are ranked ordinally, where the success of products is ranked. According to Hakanen (1998) charts function as a marketing tool for the music industry, while serving audiences as an information tool. A high ranking in such a chart can potentially lead to bandwagon effects, in which consumers strongly base their purchasing decisions on the information derived from others (Leibenstein, 1950). Also, if a market information regime serves as an information tool for consumers then it can also have an impact on consumers as an information cascade.
9
choose the best products. Similarly, in a market for music files where there is high quality uncertainty Salganik, Dodds and Watts (2006) found that an increase in the strength of social influence leads to an increase in unequal outcomes in cultural markets. In relation to the market information regime for the music industry this would imply that a product appearing on a chart would serve as a signal of quality towards potential other consumers.
Furthermore, market information regimes can be used as a communication tool whereby companies actively use the inclusion by a market information regime to communicate the relative success of a product because ‗charts success will act as publicity that can increase sales of current and future work‘(Strobl and Tucker, 2000).
Concluding, market information regimes serve three important functions in the music industry. First, they can function as a quality signal towards consumers. Being adopted by the market information regime, in the music industry appearing in the chart, may indicate a relative performance vis-à-vis competing products. Second, a market information regime can function as a structuring device by ranking the most popular products from the ‗most‘ popular to the ‗least‘ popular among products eligible for ranking and thus drive demand for certain products. Third, a market information regime can function as a communication tool for companies in order to promote their products over others because they are adopted by the market information regime.
2.3 Selection systems
Market information regimes play a significant role within the music industry both as a measuring stick and possible influencer of success. As a producer of commercial musical products it is therefore crucial to be adopted by and achieve success reflected by the market information regime. In order to obtain this goal, products generally must go through a selection system which separates products from each other based on tangible or intangible reasons. As consumers now become less reliant on traditional (offline) sources of information this might disrupt traditional selection systems within industries such as music.
10
select products: market, expert and peer. These three ideal types of selection system relate to the type of information consumers rely on when making decisions. For this research we will focus mainly on market and expert selection. Peer selection is excluded from this paper because it is not a common selection system within the music industry. There are no charts or awards where musicians select fellow musicians.
2.3.1 Market Selection
Market selection is a system in which consumers select from the available offerings solely based on their own instincts and judgments. This simply means that consumers select from the marketplace based on their own judgments of quality. With the growth of the internet consumers have seen an increase of information sources through which product quality can be indicated. Consequently, consumers can also seek out other consumers opinions regarding product quality which can possibly affect their purchasing decision. User-generated content is a form of online word-of-mouth through which consumers can see how others evaluate and rate products. Consumers can thereby be made aware of the perceived quality of a product as judged by other consumers in the marketplace. This form of market selection has gained greater potential through the internet. Consumers can easily look up evaluations of products of any kind by other consumers and possibly use this information in their purchasing decision. 2.3.2 Expert Selection
An expert selection system is a system whereby independent experts who are accredited certain knowledge and who are not producers are consulted to select products. The most evident example of an industry that uses expert selection is the art world where experts such as art critics and gallery owners usually have an important influence on consumers‘ buying decisions. The use of expert selection systems is also frequently used in markets for experience goods whose value is difficult to ascertain beforehand and most of the research done has focused on the influence of expert reviews on the consumer demand for experience goods.
2.3.3 Selection systems and music
11
lead to an influence on consumer behavior and influence future success. According to Gemser and Wijnberg (2000) a selection system consists of the selected, actors that are in competition with each other, and the selectors, actors whose decisions determine the results of the selection process.
The music industry is characterized by an abundance of supply that because of online technologies is only expected to increase. Even for the most avid music fan it is impossible to fully take in all available products. New music is omnipresent therefore some kind of pre-selection must exist in order to narrow down choice and serve as a potential signal of quality. Professional reviews in magazines or newspapers by assigned experts have historically been a selection system within the music industry.
Gemser, Leenders and Wijnberg (2008) used selection systems theory in relation to salience and fit to determine what types of awards in the movie industry were the most effective as a quality signal in different contexts. Their research into movie awards revealed that the strength or impact of different types of awards varies across industry segments. Similarly, today‘s music industry has seen a vast increase in information sources through which consumers can select what music to consume. There is now a greater variety of sources through which consumers can select music, with the bulk of these sources coming from the internet.
Due to the enormous supply of songs in the music industry, selection systems are inevitable in order to create some clarity and provide consumers with a pre-selection of what is available.
12
Expert-based information has historically been an influential force within the music industry. Previously, expert-based media such as DJs/VJs, magazines and radio stations have always played a defining role in determining the success of music. For example, new singles would predominantly be sampled via the radio or television (music videos). In the 1950s and 1960s it was common practice for record labels to pay radio DJs to play and promote certain songs over others resulting in the ―payola‖ scandal resulting in this practice becoming illegal. Mol, Chiu and Wijnberg (2012) conducted a study into the determinants of new entry in popular music. They concluded that one of the defining factors of new entry success was the access to dominant selectors within the market, in this case radio DJs, thereby immediately increasing the audience to which the product was exposed.
Also, for cultural goods such as music, critic reviews as an expert-based selection system have been an important factor in predicting success. Desai and Basuroy (2005) found that positive (negative) reviews had more impact on performance for alternative than for mainstream movies. Reinstein and Snyder (2005) found that positive reviews had a significant influence on the demand for narrowly-released movies. Joshi, Das, Gimpel and Smith (2010) used the text of pre-release critic reviews to predict opening weekend revenue of movies and concluded that they are a good indicator for success.
In the online world expert-based selection has also partially shifted to market-based selection through the creation of among others digital radio stations, music websites and websites that specialize in selecting a particular genre of music. Furthermore, a lot of online sources allow consumers to express their own opinions and share those with others. Consumers can now post their own personal reviews, assign their own scores to products and express their sentiments and opinions about cultural products such as music.
13 2.4 Conceptual Model
This research aims to investigate the effect of online (market-based) communication technologies on performance in the music industry. Specifically the effect on the performance on the music charts will be investigated as charts are a widely accepted tool to evaluate the performance of products in the music industry. This research hypothesizes that the emergence of modern online communication technologies (i.e. user-generated content (UGC)) allows users to communicate amongst each other about musical products and in doing so they can bypass traditional selection systems such as critic reviews, directly influence consumers and theoretically alter purchase behavior and chart performance. Furthermore, because of the internet and UGC, consumers also have access to a larger quantity of music from both independent and mainstream sources. This – in theory - creates opportunities for independent labels and artists who generally do not have the resources to compete with mainstream labels and artists for the attention of traditional selection systems due to smaller promotion and marketing budgets. The impact of UGC is established in terms of volume and valence. Volume relates to the number of messages shared about a product and valence relates to the score (positive or negative) assigned to products by individual consumers.
14
3.
User-generated content
Market-based information has long been a subject of interest and since Katz and Lazersfled (1955) much attention has been paid to the phenomenon of consumer interaction, its influence on consumers and how firms can use this potential to their advantage (Godes and Mayzlin 2004). Consumers are driven to give their opinion to others by their desire for social interaction and economic incentives (Hennig-Thurau, Gwinner, Walsh and Gremler, 2004). Technological development has enabled market-based information to reach a larger audience and its impact on sales and performance has been of significant interest for cultural fields such as movies (Liu, 2006. Dellarocas, Gao and Narayan, 2010.) and books (Chevalier and Mayzlin, 2006).
Research in the online realm thus far has focused mainly on the impact of online consumer reviews. Consumer reviews are said to be the truest form of market-based information as consumers are not being paid to criticize a product and are therefore thought to objectively give their review. Significant research has been conducted on the influence of consumer-to-consumer information sources on the web, mainly through the use of online consumer reviews. Zhu and Zhang (2010) examined how product and consumer characteristics moderated the influence of online consumer reviews on product sales finding that online reviews had more influence on the sales of less popular games, suggesting that producers of niche products should be more concerned with online reviews. Similarly, Park, Lee and Han (2007) found that both quality and quantity have an effect when determining the effectiveness of online reviews, finding that the quality of a review had a positive effect on purchase-decisions and that this intent to purchase increased with the quantity of reviews. Clemons, Gao and Hitt (2006) focus on online reviews in the context of hyper-differentiation and resonance marketing theories suggesting that better informed consumers will demand more differentiated products and that the dispersion of user-ratings is positively correlated with sales growth. Moretti (2011) examined the effect of social learning on movie sales finding that this effect was greater for narrowly released movies than movies that were released on a larger scale.
15
communication becomes an important factor as information is increasingly diffused via such networks. Bakshy and Rosenn (2012) examined the role of social networks in online information diffusion finding that those exposed to friends‘ information sharing are significantly more likely to spread information than those not exposed and that weaker social ties play an important role in the spread of new information. Similarly, Aral and Walker (2011) in a research on how to create word-of-mouth peer influence and social contagion found that passive messages (e.g. a status update, a general message sent to all connections) created more peer influence and social contagion than messages that were personalized and active (e.g. messages sent directly to other users specifically about a certain product). Online market-based information can thus be passed on through larger groups of consumers who do not necessarily have strong social ties therefore increasing its potential reach. Duan, Gu and Whinston (2008) showed that the effect of online WOM can act both as a forerunner of and reason for an increase in retail sales by creating a positive feedback mechanism. This research makes use of data gathered from the social networking site Twitter which consists of predominantly passive messages sent out to multiple users instead of personalized direct messages.
3.1 Volume and valence of UGC
16
Table 1. Volume/Valence studies
Valence (Persuasive effect) Volume (Awareness effect) Article
Duan, Gu and Whinston (2008)
Eliashberg and Shugan (1997)
Reinstein and Snyder (2005)
Chevalier and Mayzlin (2006)
Finding
Rating of online user reviews had no significant influence
Critics are more predictors of box office success than influencers of it
Positive expert reviews have siginificant influence on movies released on a smaller scale
Improvement in book reviews led to more sales.
Article
Duan, Gu and Whinston (2008)
Godin and Mayzlin (2004)
Davis and Kazanchi (2008)
Li and Hitt (2008)
Liu (2006)
Finding
Volume of online user reviews significantly influenced box office sales High dispersion of online word-of-mouth
siginificantly correlated with early performance of a TV-show
Interaction between product views and product category that are statistically significant in explaining changes in unit product sales
Firms should encourage early positive reports in order to let products self-select into the market
The volume of online word-of-mouth provides significant explanatory power about a movie‘s performance during its early stages after release
(Floyd, Freling, Alhoqail, Cho and Freling, 2014)
3.2 Volume and valence in the music industry
17
pre-selected by expert-based sources such as critic reviews and rely more heavily on content generated by other users.
Examples of user-generated content sites are blogs and social networking sites where content can be suggested to and shared amongst users. Preliminary research into music blog buzz and predictive sales by Dewan and Ramprasad (2009) indicated that buzz created by blogs had a significant and positive relationship with album sales and that this influence was the strongest for independently released music. Dhar and Chang (2009) found that the volume of blog posts about an album is positively correlated with album sales. Frick, Tsekouras and Li (2013) found that online user interactions had substantial predictive power on both physical and digital album sales. Furthermore they found that an increased volume of online user interactions had a significant and positive relationship on the album sales of independent artists. Preliminary research thus suggests that online user-generated sources such as blogs play an important role for niche products such as independent artists because of the quality uncertainty of these products vis-à-vis mainstream artists (Dewan and Ramprasad, 2012). Therefore the volume and valence of UGC theoretically can be important factors influencing the performance of a music album.
Considering all of the above this research comes to the following hypotheses:
H1: The volume of UGC has a positive influence on performance H2: The valence of UGC has a positive influence on performance
This research also considers the interaction between valence and volume and the possible effects of this interaction between the two. First, in this research we expect that valence and volume can have a possible synergetic effect on each other whereby volume can enhance the effect of a negative or positive valence and vice-versa. Therefore:
H3: Volume and valence interact with each other in such a way that they strengthen the effect
of the other on performance
18
of mainstream music than by experts (greater similarity in evaluative criteria). Furthermore, buyers of mainstream music are more likely to select these mainstream sources to make their purchasing decision. Conversely, independent buyers are more likely to select based on expert sources. In both cases there is a high salience between consumer and selection systems. Information becomes more effective as a means of certification if it is adjusted to the dominant selection system within the industry segment: mainstream albums are more influenced by UGC and independent music buyers are more influenced by expert reviews.
H4a: Album type moderates the relationship between UGC and performance, such that the
effect of UGC is stronger for mainstream songs
H4b: Album type moderates the relationship between critics and performance, such that the
effect of critics is stronger for independent songs
4. Methodology and data
4.1 Settings & data collectionTo investigate the influence of UGC – relative to that of critics – we use the messages from the social media service twitter. Data was gathered from the social media website Twitter which enables users to publicly share short messages on almost any subject to count as the volume of messages being shared about a particular album. For the purpose of this research we shall use the social media service twitter as an example of user-generated content that has a possible influence on consumer behavior. With 284 million monthly active users and over 500 million short messages (‗tweets‘) sent between users each day the potential reach and influence of this medium is evident and it serves as the primary example of a social media thought to have great influence. Twitter has already gained much scholarly interest from the field of data analysis (Broniatowski, Paul and Dredze. 2014.), sentiment analysis (Go, Bhayani and Huang. 2009) and the influence on other users (Bakshy, Hofman and Mason. 2011).
19
traditional sources such as music critics were collected to accompany online sources as to study their relative significance on chart performance in the music industry.
Data collection
For this research data was gathered on 93 albums all released in March and April 2014. As a condition for selection all albums had to have been reviewed by critics and given a score. Data on review scores was acquired from the review site www.metacritic.com which collects and aggregates scores from professional critics. All data was gathered from publicly available websites and manually recorded. The list of albums and their release dates were obtained from the website www.officialcharts.com , a website that compiles the official music charts as used by the BBC and other media outlets in the United Kingdom. Chart positions were therefore also obtained from the same website for each album. Only new releases were considered therefore excluding old material such as reissues and albums such as soundtracks, compilations as well as albums that had not been reviewed by music critics.
The final sample then consisted of 93 albums. All further data were collected 21, 14 and 7 days prior to the release of an album. All data were gathered on Tuesday of each week as albums are generally released on Tuesday so that they can possibly be included in the chart published at the end of the week.
4.2 Measures
4.2.1 Dependent variables
To test whether volume and valence of online sources have any impact on the chart success of music albums a linear regression model is used on three dependent variables ―chart‖, ―cumulative success‖, ―opening success‖ and several independent variables.
Chart appearance
20
Therefore, the binary dependent variable ―Chart appearance‖ is created and albums that appear on the chart up to six months after their release date are given a value of 1 and albums that do not appear on the chart are given a value of 0.
Cumulative success
To measure the overall performance of a music album, including the length of time spent in the charts and the relative position in the chart the dependent variable ―Chart performance‖ was created. Obviously an album that spends 20 weeks in the charts at different positions in the chart can be considered more successful than an album that spends one week in a relatively low position in the chart. Therefore, the weekly chart positions of each album were tracked during a period of six months after initial release whereby the position in the chart was converted by assigning a score from 1-100 (1st postion=100, 2nd position=99, 3rd position=98, etc.) and totaling the score over a six-month period for each album.
Opening success
As charts in their role as market information regimes can have a cascading effect on the market the performance of an album in its first week is often crucial to the further performance. Therefore for each album in this research the performance in the chart was noted in the first week after its release by checking whether or not the album charted in the first week after its release. Each album that charted in its first week was assigned the value of 1 and each album that failed to chart in the first week was given the value 0.
4.2.2 Independent Variables
21 UGC Volume
For this research data was gathered on each album and the amount of tweets each album generated three (t-3), two (t-2) and one (t-1) week prior to release. Data was collected by using Twitter‘s own search function for the amount of tweets between specific dates. For each album a search was conducted for tweets containing the name of the artist and the name of the album together in the same tweet. As such the volume of tweets for each album was found and recorded.
UGC Valence
In order to reflect ‗valence‘ or the score users themselves give to a certain album this study uses data also obtained from the website www.metacritic.com where users can give each album a score on a 1-10 scale. For this study the average user score on a scale of 1-10 generated on metacritic.com for each album was taken and used as the user score and thus used as the ‗valence‘ score for each album similar to Dhar and Zhang (2009).
4.2.3 Control variables
Critic reviews volume & valence
Critic reviews have traditionally been a part of the evaluation process of music and other art forms. Research into the predictive effect of critic reviews on the box office success of movies has shown that critic reviews can be an indicator of box office success (Basuroy, Chatterjee and Ravid. 2003.) partly depending on what type of movie (Gemser, van Oostrum and Leenders. 2007). This study takes into account critic reviews for music albums as a traditional source of information for consumers. Data on critic reviews was collected from the website www.metacritic.com which collects critic reviews on amongst other things music albums and assigns a score to each reviewed album based on a weighted mean score of each review. For this study, only reviews were taken into account three 3), two 2) and one (t-1) week before the release of the album. The weekly as well as the cumulative score of each week was taken into account thereby reflecting the information available to consumers.
22 Album Type
Although there are a lot of different music labels most of them generally fall under the umbrella of a major record label. Generally, major label artists enjoy bigger budgets with regards to marketing and promotion and can therefore create attention for their album releases more easily than independent artists. To correct for this effect a dummy variable was introduced whereby each mainstream album was given the value 0 and each independent album was given the value 1.
5. Analysis and results
5.1 Chart appearance
In order to analyze the gathered data multiple statistical analyses were performed. First, a logistic regression was performed to analyze the relationship between the gathered data whether or not a particular album charted (a dichotomous outcome). As success in the music industry is measured by appearing in the chart this means that simply appearing in the charts can be considered a form of success. Using the binary dependent variable ―chart appearance‖ and the independent variables this leads to the following results.
Table 1.
Variables Model 1 Model 2
AlbumType -.85* -.87 UGCVolume t-3 .00 UGCVolume t-2 .00 UGCVolume t-1 .00 UGCValence .12 CriticVolume t-3 -.13 CriticVolume t-2 .10 CriticVolume t-1 .13* CriticValence t-3 -.00 CriticValence t-2 .00 CriticValence t-1 .00 R-squared (adjusted) .06 .44 DV=chart appearance ***p<.001, **p<.01, *p<.05, +p<0.1
23
Specifically, albums that are labeled mainstream have a higher probability of appearing in the charts. When including all the independent variables only critic reviews that appear one week before the release of an album show a statistically significant influence on the likelihood of albums making it onto the charts.
Table 2.
***p<.001, **p<.01, *p<.05, +p<0.1
When including the aggregated variables and interaction effects between UGC volume and valence critic volume shows a statistically significant influence on the likelihood of appearing in models two and three. When running the regression with just the interaction effects the interaction between critic volume and album type for independent albums is significant but only in isolation. When other factors are included there is no explained variance above and beyond the independent variables.
These results indicate that album type, critic reviews one week before release, critic volume and the interaction between critic volume and album type in some cases are significant predictors of an album reaching the chart or not and that mainstream albums have a higher probability of reaching the charts than independent albums.
5.2 Cumulative success
In order to ascertain whether the gathered data have an influence on the cumulative success of an album, (i.e. the position of the chart and duration in the chart) a multiple linear regression was ran with ―cumulative success‖ as the dependent variable.
Variables Model 1 Model 2 Model 3 Model 4 Model 5
24 Table 3.
Variables Model 1 Model 2
AlbumType -268.90*** -295.67*** UGCVolume t-3 -.25 UGCVolume t-2 .11 UGCVolume t-1 -.11 UGCValence 19.82 CriticVolume t-3 13.73 CriticVolume t-2 -24.15 CriticVolume t-1 9.75 CriticValence t-3 -2.17 CriticValence t-2 2.48 CriticValence t-1 9.75 R-squared (adjusted) .10 .08 DV=cumulative success ***p<.001, **p<.01, *p<.05, +p<0.1
In both regression models only album type shows statistical significance in relation to the cumulative success of an album. This means that mainstream albums are more successful than independent albums in relation to cumulative success.
Table 4.
Variables Model 1 Model 2 Model 3 Model 4 Model
5 AlbumType -268.90*** -294.54*** -302.01*** -328.11** UGCVolume -.07 .25 .19 UGCValence 14.2 19.21 18.35 CriticVolume 6.80 7.48 -37.58 CriticValence -.26 -.04 -.05 Valence*Volume UGC -.04 -.04 Valence*VolumeCritic -.19 -.15 UGCVolume*AlbumType .09 -.18 CriticVolume*AlbumType 44.88 9.13+ R-squared (adjusted) .10 .10 .08 .01 DV=cumulative success ***p<.001, **p<.01, *p<.05, +p<0.1
25 5.3 Opening success
To test whether the variables have any effect on the opening success again a linear regression was performed including the independent and control variables.
Table 5.
Variables Model 1 Model 2
AlbumType -.18+ -.13 UGCVolume t-3 .00 UGCVolume t-2 .00 UGCVolume t-1 .00 UGCValence .01 CriticVolume t-3 -.03 CriticVolume t-2 -.00 CriticVolume t-1 .00 CriticValence t-3 -.00 CriticValence t-2 .00 CriticValence t-1 .00 R-squared (adjusted) .03 .09 DV=opening success ***p<.001, **p<.01, *p<.05, +p<0.1
In this model album type was only statistically significant at the 10% level when inserted as the sole variable indicating that mainstream albums enjoy higher opening success than independent albums. When including the other independent variables no statistical significance was shown.
Table 6.
Variables Model 1 Model 2 Model 3 Model 4 Model 5
AlbumType -.18+ .17 .19 .64 UGCVolume .59 .94 .65 UGCValence .27 .37 .34 CriticVolume .27 .28 .18 CriticValence .54 .52 .55 Valence*Volume UGC .83 .83 Valence*VolumeCritic .83 .71 UGCVolume*AlbumType .11 .00* CriticVolume*AlbumType .13 .07** R-squared (adjusted) .03 .08 .06 .08 .09 DV=opening success ***p<.001, **p<.01, *p<.05, +p<0.1
26
mainstream albums. Also, the interaction effect in isolation between album type and UGC volume for opening success is stronger for independent albums than for mainstream albums.
The positive interaction effect between UGC volume and independent albums can be explained by the poor signaling qualities of independent releases such that any signal has a strong impact on opening success.
6. Conclusions & Discussion
The research conducted tested but could not find any significant evidence that user-generated content such as Twitter has an influence on the charts and chart performance in the music industry. Furthermore, there was no significant interaction effect between volume and valence.
In isolation, the interaction effect between album type and UGC showed significance in relation to opening success, such that the impact of UGC is stronger for independent albums. Also, in isolation the interaction effect between critic volume and album type showed positive significance in relation to opening success for independent albums for opening success, chart appearance and cumulative success.
This research also shows that signals given by external parties are important factors in determining success in the music industry and that these signals work more strongly for independent releases. Previous research (Gemser et al. 2008) has shown that quality signals such as awards are stronger for independent releases, this research also shows that critic reviews as quality signals are stronger for independent releases. This study also shows signs that UGC as an externally given signal helps in the diffusion of the signal at an early stage thus aiding opening success, possibly by creating an awareness effect about the product similar to the findings of Duan, Gu and Whinston (2008).
27
Although it is a relevant subject and the influence of online sources such as twitter may change in the future, this research did not find conclusive evidence that it can break down and replace traditional selection systems and power structures within the music industry. This is in line with conclusions drawn by previous research carried out by Mol, Wijnberg and Carroll (2005) and Broekhuizen, Lampel and Rietveld (2013) that showed that the emergence of the internet has not yet significantly impacted traditional structures.
Considering the still relative novelty of phenomena such as user-generated content, the field of user-generated content remains a field of significant interest for further research. One can for example imagine that in the future an artist‘s presence and following on social media can serve as a proxy for measuring popularity and subsequent success. Also, as can already be seen now, charts may start incorporating the popularity of a song on social media as part of the formula upon which the chart is constructed, which may lead to different make-ups of the chart and lead to different dynamics within the music industry with regard to promotion and the marketing mix.
Results from the performed regression showed no impact of UGC volume and valence, neither on the individual or aggregated level. The findings suggest that user-generated content has no effect on chart listing and chart performance. This research does find that artists that are signed to a major label are generally more likely to be successful than independent artists. This difference in success is most likely linked to higher marketing and promotion budgets that are required to gain critical mass amongst consumers and become visible or that they can attract higher quality artists. Independent labels generally lack these resources and generally struggle to gain enough attention from consumers to achieve chart success. Therefore, theories that predict that user-generated content would negate this competitive advantage of major labels are premature at this point in time.
Furthermore, this research found that the control variable critic review was significant one week before release for reaching the chart and that the volume of critics was statistically significant in isolation in two models for cumulative success and therefore indicative of the predictive validity and/or influencer effect of critic reviews. This would indicate that critic reviews are still of value to those operating within the music industry. Also, despite the fact that information about a product is easily disseminated amongst consumers themselves this would indicate that expert opinions are still valued and can have an effect on performance.
28
charts provides an accurate reflection of what is popular within the music market. Considering the novelty of user-generated content it is fair to ask if the modern-day methodology has caught up to new trends within the music industry.
6.1 Managerial implications
The results did not yet show any significant effect of user-generated content on charts. Therefore it can be concluded that at least for the time being the traditional structures of the music industry have not yet been fundamentally altered. If any artist wishes to be commercially successful (i.e. appear in the charts) then being released by a major label is still the most likely way to achieve this goal. The emergence of the internet as a means to break down traditional power structures of the major labels versus independent labels does not yet seem to be a reality as of yet as major labels continue to be more (commercially) successful in the market place than independent labels. However, this could also be explained by the fact that the current methodologies used to compile the charts fail to correctly incorporate modern technologies and that future charts will be better suited to accurately reflect the marketplace.
Furthermore, managers active in the independent segment of the industry should focus on creating any type of quality signal for their artists. Results showed that despite the fact that the industry segment did not fit with the dominant selection system UGC volume still had a strong impact on opening success of independent albums. This would indicate that for independent albums any type of signal has an impact and that having any type of signal is even more important than salience or fit with the dominant selection system.
6.2 Limitations and future research
29
media on different genres e.g. dance music and folk. Also, further research is necessary to see if the findings hold
30
Literature
Anand, N., & Peterson, R. A. (2000). When market information constitutes fields: Sensemaking of markets in the commercial music industry.Organization
Science, 11(3), 270-284.
Andrews, K., & Napoli, P. M. (2006). Changing market information regimes: A case study of the transition to the BookScan audience measurement system in the US book publishing industry. Journal of Media Economics, 19(1), 33-54. Aral, S., & Walker, D. (2011). Creating social contagion through viral product design: A randomized trial of peer influence in networks. Management Science, 57(9), 1623-1639.
Bakshy, E., Rosenn, I., Marlow, C., & Adamic, L. (2012, April). The role of social networks in information diffusion. In Proceedings of the 21st international conference on World Wide Web (pp. 519-528). ACM.
Bakshy, E., Hofman, J. M., Mason, W. A., & Watts, D. J. (2011, February). Everyone's an influencer: quantifying influence on twitter. In Proceedings of the fourth ACM international conference on Web search and data mining (pp. 65-74). ACM.
Bikhchandani, S., Hirshleifer, D., & Welch, I. (1992). A theory of fads, fashion, custom, and cultural change as informational cascades. Journal of political Economy, 992-1026.
Broekhuizen, T. L., Lampel, J., & Rietveld, J. (2013). New horizons or a strategic mirage? Artist-led-distribution versus alliance strategy in the video game industry. Research Policy, 42(4), 954-964.
Broniatowski, D. A., Paul, M. J., & Dredze, M. (2014). Twitter: Big data opportunities. Science, 345(6193), 148.
Chevalier, J. A., & Mayzlin, D. (2006). The effect of word of mouth on sales: Online book reviews. Journal of marketing research, 43(3), 345-354.
Clemons, E. K., Gao, G. G., & Hitt, L. M. (2006). When online reviews meet hyperdifferentiation: A study of the craft beer industry. Journal of Management Information Systems, 23(2), 149-171.
Davis, A., & Khazanchi, D. (2008). An empirical study of online word of mouth as a predictor for multi‐product category e‐commerce sales. Electronic Markets, 18(2), 130-141.
Dellarocas, C., Zhang, X. M., & Awad, N. F. (2007). Exploring the value of online product reviews in forecasting sales: The case of motion pictures.Journal
of Interactive marketing, 21(4), 23-45.
31
contribute online reviews for hit or niche products?. Journal of Management Information Systems, 27(2), 127-158.
Dertouzos, J. N. (2008). Radio Airplay and the Record Industry: An Economic Analysis. National Association of Broadcasters, USA.
Desai, K. K., & Basuroy, S. (2005). Interactive influence of genre familiarity, star power, and critics' reviews in the cultural goods industry: The case of motion pictures. Psychology & Marketing, 22(3), 203-223.
De Vany, A., & Lee, C. (2001). Quality signals in information cascades and the dynamics of the distribution of motion picture box office revenues. Journal of Economic Dynamics and Control, 25(3), 593-614.
Dewan, S., & Ramprasad, J. (2009). Chicken and egg? Interplay between music blog buzz and album sales. PACIS 2009 proceedings, 87.
Dhar, V., & Chang, E. A. (2009). Does chatter matter? The impact of user-generated content on music sales. Journal of Interactive Marketing, 23(4), 300-307.
Duan, W., Gu, B., & Whinston, A. B. (2008). Do online reviews matter?—An empirical investigation of panel data. Decision support systems, 45(4), 1007-1016.
Elliott, C., & Simmons, R. (2008). Determinants of UK box office success: the impact of quality signals. Review of Industrial Organization, 33(2), 93-111. Eliashberg, J., & Shugan, S. M. (1997). Film critics: Influencers or predictors?. The Journal of Marketing, 68-78.
Floyd, K., Freling, R., Alhoqail, S., Cho, H. Y., & Freling, T. (2014). How Online Product Reviews Affect Retail Sales: A Meta-analysis. Journal of
Retailing.
Frick, T., Tsekouras, D., & Li, T. The Times They Are A-Changin: Examining the Impact of Social Media on Music Album Sales and Piracy.
Gemser, G., Leenders, M. A., & Wijnberg, N. M. (2008). Why Some Awards Are More Effective Signals of Quality Than Others: A Study of Movie Awards†. Journal of Management, 34(1), 25-54.
Go, A., Bhayani, R., & Huang, L. (2009). Twitter sentiment classification using distant supervision. CS224N Project Report, Stanford, 1-12.
32
Gopal, R. D., Sanders, G. L., Bhattacharjee, S., Agrawal, M., & Wagner, S. C. (2004). A behavioral model of digital music piracy. Journal of Organizational Computing and Electronic Commerce, 14(2), 89-105.
Hakanen, E. A. (1998). Counting down to number one: the evolution of the meaning of popular music charts. Popular Music, 17(01), 95-111.
Hennig‐Thurau, T., Gwinner, K. P., Walsh, G., & Gremler, D. D. (2004). Electronic word‐of‐mouth via consumer‐opinion platforms: what motivates consumers to articulate themselves on the internet?. Journal of interactive marketing, 18(1), 38-52.
Hirsch P. M. Processing fads and fashions: An organization-set analysis of cultural industry systems. Amer. J. Soc. (1972) 77:639–659
Huang, P., Lurie, N. H., & Mitra, S. (2009). Searching for experience on the web: an empirical examination of consumer behavior for search and experience goods. Journal of Marketing, 73(2), 55-69.
Joshi, M., Das, D., Gimpel, K., & Smith, N. A. (2010, June). Movie reviews and revenues: An experiment in text regression. In Human Language Technologies: The 2010 Annual Conference of the North American Chapter of the Association for Computational Linguistics (pp. 293-296). Association for Computational Linguistics.
Leibenstein, H. (1950). Bandwagon, snob, and Veblen effects in the theory of consumers' demand. The Quarterly Journal of Economics, 183-207.
Lazarsfeld, P., & Katz, E. (1955). Personal influence. New York, 174.
Li, X., & Hitt, L. M. (2008). Self-selection and information role of online product reviews. Information Systems Research, 19(4), 456-474.
Liu, Y. (2006). Word of mouth for movies: Its dynamics and impact on box office revenue. Journal of marketing, 70(3), 74-89.
Mol, J., Ming Chiu, M., & Wijnberg, N. (2012). Love Me Tender: new entry in popular music. Journal of Organizational Change Management, 25(1), 88-120.
Mol, J. M., Wijnberg, N. M., & Carroll, C. (2005). Value Chain Envy: Explaining New Entry and Vertical Integration in Popular Music*. Journal of
Management Studies, 42(2), 251-276.
Moretti, E. (2011). Social learning and peer effects in consumption: Evidence from movie sales. The Review of Economic Studies, 78(1), 356-393.
33
radio airplay. Working paper, Marketing Dept., The Wharton School, University of Pennsylvania.
Reinstein, D. A., & Snyder, C. M. (2005). The influence of expert reviews on consumer demand for experience goods: A case study of movie critics*. The journal of industrial economics, 53(1), 27-51.
Salganik, M. J., Dodds, P. S., & Watts, D. J. (2006). Experimental study of inequality and unpredictability in an artificial cultural market. science, 311(5762), 854-856.
Strobl, E. A., & Tucker, C. (2000). The dynamics of chart success in the UK pre-recorded popular music industry. Journal of Cultural Economics, 24(2), 113-134.
Wijnberg, N. M. (2004). Innovation and organization: Value and competition in selection systems. Organization studies, 25(8), 1413-1433.
