dean.ero.com

Introduction

The music industry’s challenges in the face of digital media have been widely publicized. Over the last decade, US music sales have declined by 42% from the $32.8 billion in revenues seen in 2000 (Mulligan 2010: 4). During these ten years the recording industry’s approach to digital media has moved from willful ignorance, to active suppression, to a fumbling scramble for workable business models.

We’ve seen similar patterns in the past when the industry has been confronted with new technology. This paper aims to analyze the particular technological and business conditions that have left the music industry so flat footed in the face of digital media proliferation, and to further examine where the business could be headed. There are similarities between the recent disruptions and both the advent of radio as a listening medium and the the rise of independent labels during the rock and roll era; I hope to deepen the analysis by looking at the current state of the industry in light of historical disruption. It’s only with a clear understanding of both historical patterns and current technological trends that we can try to extrapolate into the future.

A cluster of relatively recent technologies led to a tipping point where illegitimate services became more effective at distributing music than the system that the labels had build over the course of more than forty years. In fact, these technologies have lead to entirely different consumption models. In light of this, the labels reacted using the same suppression patterns seen historically during periods of technological disruption in the music industry: by attempting to exercise control over distribution with copyright. The mixed success of this campaign has helped lead to the current Balkanized state of the digital music distribution.

In the longer term, present technological trends look to be swinging the pendulum back in favor of rights holders. Our perspective is informed in part the historical cycles of consolidation which have followed periods of disruption in incumbent music distribution systems. But it is also based on the move away from general purpose computing devices toward a more tightly controlled special purpose platforms and the emergence of Internet “hyper-giants” who increasingly play a mediating role between users and content. We’re likely to see increased control by rights holders in siloed environments; the advances in rights management and user convenience will come at the risk of oligarchical control over content by those who are increasing in control of its distribution.

The Present: Disruption, Suppression and Balkanization

The mainstream adoption of digital music started in the 1980s with adoption of the compact disc. Oddly enough, a decision made by the inventors of the compact disc set the stage for the digital disruption which came nearly twenty years later. The compact disc was not designed with any sort of  digital rights management (DRM) scheme. This was a choice made by the designers of the medium: at the time, the equipment required to produce a compact disc was specialized and expensive and DRM was judged unnecessary. In retrospect, this assumption would prove fatal as the computer industry developed CD reproduction techniques within 10 years for data applications. The duplication of compact discs, however, was just a thorn in the side of the industry when compared to the extreme disruption induced by later digital formats.

MP3 stands for MPEG-1 Layer 3, and is a compressed audio format. The format was established as an ISO/IEC standard in 1991 with the intention of creating a compressed audio format for digital film. The technology developed over the course of several decades of research, employing psychoacoustic masking developed in the early 1980s (“Data Compression”). Psychoacoustic techniques enabled a dramatic reduction in file size; by masking out large swaths of the audio signal which are largely inaudible to human hearing, MP3s are able to reach a compression ratio of nearly 10:1. The extreme compression ratio was vital to the early adoption of the format on the Internet–the first implementation of an encoder was in 1994, when most end-user internet connectivity happened via modem. Over a 14.4 kbps connection, it would take about five minutes to download a three-and-a-half minute pop song. Although this seems like an eternity by todays standards, an uncompressed, CD quality version of the same song could take nearly an hour to download over a modem.

As the network became faster and more ubiquitous, the pieces fell into place for the development of an alternate distribution chain for recorded music to take shape: most folks had moved their music collections to CD during the early 1990s, providing vast amounts of DRM-free digital source material; MP3 encoders were widely available for download either under shareware or freeware licenses; and distribution outlets were no further than USENET groups, IRC chat rooms, anonymous FTP servers, and, eventually, hosted links on web sites. Traffic agnostic networks connecting general purpose computing devices made it possible for most any node to fulfill any role the reproduction and distribution chains.

Early Services

For the most part, record labels ignored online activity until the end of the 1990s, reaping record profits from the compact disc until the early 2000s (Tschmuck 2006: 161). New releases were lavished with seven figure marketing budgets, and massive promotional campaigns were mounted to brand artists and push lead singles. Since the industry had basically killed the single in its move to the CD format, hit songs were only available on full-length compact discs, which are high margin goods (Knopper 2009: 40-65). In this environment, capital-intensive investment in new technology had little appeal to incumbents.

The technology and business models of legitimate online music providers have changed amazingly little since the late 1990s, mostly just shifting under the weight of litigation exercising control via copyright. In 1997, MP3.com launched with an open model, allowing independent artists (read: non-litigious rights holders who maintained their own copyright), to make MP3s of their recordings freely and widely available. The business model was based on ad revenue and was built around a click-through model used by its sister company, the FTP search site filez.com. At MP3.com’s peak, their server infrastructure delivered millions of MP3s per day (“MP3.com,” 2010). Ultimately, the site would run into problems from the incumbent rights holders. In 2000, My.MP3.com was launched, allowing users to upload their CDs in MP3 format to a server, allowing them to access streams of their music from any computer. In UMG v. MP3.com, the court ruled that this was not fair use, and MP3.com eventually settle with four of the five majors for $80 million (Tschmuck 2006: 172). Financially weakened from the suit, MP3.com was not able to survive the bursting of the dot com bubble in the early 2000s. It’s worth noting that this is nearly identical to the business models of newer, “cloud-based” services such as Spotify, BlueTunes and LaLa.

Peer-to-Peer

The service that most famously drew the ire of incumbent rights holders was Napster. Founded by Shawn Fanning in 1999, Napster represented a new paradigm in digital distribution, connecting network endpoints for file exchange without storing the files on a server. Although not truly a peer-to-peer network (the indexing of users’ files was mediated by a server, which would ultimately enable the rights holders to shut down the service out-of-hand), Napster was the first service to capitalize on one of the most fundamentally powerful ideas of the Internet Age: that interconnected general purpose computing devices can function as true peers on an open Internet, sidestepping the standard a client-server relationship.

The original Napster business model was to build a large user base with the file sharing service and then sell music related merchandise through the site (Knopper 2009: 125).  By making its profit elsewhere, Napster executives expected to be held harmless for the reproduction and distribution of music. The service became a phenomenon before any monetization was to occur. By July 2000, Napster had nearly 20 million users and began to stir up the ire major recording artists when pre-release material was made widely available. This also attracted the attention of the deep pocketed and litigious major labels. Napster had avoided paying rights holders to this point, relying on the fact the music was never on their servers and holding that the exchange between users was fair use. Their position was supported by public opinion; music fans were tired of buying full length CDs for one or two songs, and Napster allowed them to explore new music without emptying their wallets. Furthermore, the major labels had an image problem due to corporate consolidation and several high profile spats with artists. The courts, however, supported the labels’ claims. In A&M Records v. Napster (2001), Napster was effectively declared illegal when it was found that they were knowingly exploiting the exchange of copyrighted material to build their business. After a failed appeal, they were required to shut their service down, and mediated peer-to-peer was no longer viable in the United States. Similar legal action was taken against the true peer-to-peer networks (Kazaa, LimeWire and Grokster) in the years that followed. The courts again sided with the rights holders, citing an “inducement to infringe” on the part of the peer-to-peer networks, and the technology was effectively relegated to dark net status (MGM v. Grokster, 2005).

The peer-to-peer paradigm is, perhaps, one of the best examples of the power of networked general purpose computing devices, and the P2P heyday may have been a high-water mark for the open Internet. With every node in the network acting as both a client and a server there is effectively no third party mediation. As Lawrence Lessig argues in Code 2.0 (2006), it’s through this mediation by software that user activity on the Internet is controlled. In the series of rulings against the peer-to-peer networks, the courts effectively disallowed any business model built around the connection of peer nodes without an intermediate layer providing copyright protection. While illegal peer-to-peer file sharing remains commonplace, the suppression of business models attempting to use unlicensed content has been effective, giving the labels what amounts to veto power over new music distribution services hoping to use their content.  Successful companies licensing content directly from unsigned artists, such as MySpace, have generally served artists as promotional tools. While MySpace generates huge amounts of ad revenue, its model for monetizing music is shaky at best; it only pays artists or rights holders who have the wherewithal to legally oblige them to do so.

The Current Digital Marketplace

Considering the legal environment around digital service providers for music, it’s of little wonder that the industry has spent the better part of ten years floundering through business models. Some of the models currently seen:

• All-you-can-eat, on-demand streaming. This model is used by Rhapsody, the revamped and legal Napster, and is very close to the model of European upstart Spotify. You basically pay a fixed monthly rate for streaming access to a vast catalog of music. Spotify operates on a “freemium” model; the free service is ad supported and the premium service allows you access to the catalog, sans ads, and via mobile applications. Spotify has not yet been licensed the major label catalogs in the United States.
• Streaming radio. This model is used all over the place, but most successfully by Pandora, which uses an algorithmic recommendation system to program personalized radio. Since the music is not “on-demand” streaming radio is able to use existing compulsory licensing and pay performance royalties (as per terrestrial radio). The rate paid by webcasters has been a huge issue, resulting in a long line of congressional hearings, settlements, and legal contortions. Again, we see rights holders (in this case publishers and songwriters who hold rights to a different set of royalties) exerting their influence over an emergent industry.
• “Cloud”-based libraries. Lala is the prototypical example of this type of service, although several similar services exists and Spotify overlaps into this space as well. Basically, these services take your existing music collection and move them on to servers which provide you with streaming access. This should sound familiar, as it’s exactly the model that caused MP3.com to be sued out of business. Ironically, Lala was briefly owned in majority by the Warner Music Group (Kafka 2009). It has since been sold to Apple outright (Kincaid 2009).
• Download services. Most successfully adopted by Apple with their iTunes Music Store, the download model simply charges for digital downloads that you keep on your computer and/or use on compatible devices. While MP3 is the most popular format for listening, not all services use MP3 due to the inability to apply DRM to the format. This led to a proliferation of proprietary digital formats, including WMA (Microsoft), AAC (Apple), and ATRAC (Sony) among others. Each format is tied to the proprietary hardware and software which is able to play the format. The only format to have any success among these (in terms of adoption) has been AAC, which piggybacked on the success of the iTunes Music Store and Apple’s hardware. More recently, MP3 sales have been more widely adopted, with Apple providing high quality MP3, Amazon entering the market to some success and Emusic finally reaching terms with the major labels.
• Direct-to-Fan and fan pyramid. The Fan Pyramid is built around the idea that fans with different levels dedication will purchase different products: a few ravenous fans will purchase deluxe packages with high quality FLAC, vinyl, tee-shirt, etc. where as a greater number of casual fans may be willing to inexpensive MP3s. These product tiers are then generally marketed to fans directly, allowing artists to take a large cut of a (usually) smaller number of sales. An example is TopSpin, who provides marketing and sales widgets for bands. Their pricing model incentivizes direct-to-fan marketing of high end packages.

The only approach that has been significantly profitable for labelsis the download model. In fact, iTunes has been the brightest spot inwhat was generally regarded as dismal decade for the industry,becoming the largest single music retailer and having just passed 10billion track downloads (Apple 2010). Unfortunately for the labels,the shift from full-length CDs to single track sales has resulted in asmaller per unit margin. And along the way, sales of physical product havesteadily declined, although the compact disc is still the best selling format.

An interesting footnote is that the independent labels have not suffered quite the same fate as the majors. In fact, one could argue that independents have mostly benefited from the adoption of digital media.  Because their marketing budgets were small to begin with the indies have actually gained attention share as the public’s attention center has shifted from broadcast media to online outlets. Aggressive distribution of free MP3s, and early adoption of MySpace streaming and YouTube for promotional distribution have provided indies access to a broader audience, and increased consumer participation in niche markets has brought the labels new fans. Over the past several years the larger indies have at times nearly caught up with the smaller majors: in 2007, The Shins’ Wincing the Night Away (Sub Pop), and The Arcade Fire’s Neon Bible (Merge) both debuted at #2 on the Billboard chart; most recently Vampire Weekend’s Contra (XL) debuted at #1. While these records’ sales numbers don’t compare with the largest records of the majors, they’re often produced and marketed at a fraction of the cost, becoming profitable much sooner. We can see parallels between the current renaissance of independents in the state of the music industry during the 1950s. At that time, the new distribution methods enabled by the adoption of “unbreakable” vinyl and access to local FM radio led to a dramatic shift in market share between the majors and the indies. Today, we’re also seeing majors get smaller, and larger indies get bigger; market forces seem to be leading us toward ideally sized labels somewhere in between.

The Future: Long Stagnation, Consolidation and Walled Gardens

So what does the future hold? In this case, the best historical parallel may be the advent of radio. At that time, new methods of listening and production combined with a radical price restructuring as records were suddenly forced to compete with freely broadcast competition. This led to a long era of stagnation in the recorded music business. The forces at play now are fundamentally similar to those in the early 1920s: digital media has changed both the listening paradigm and the price structure of music; massively capitalized competitors with deep expertise in the new medium exist outside the music industry; and the hangover from the compact disc era has challenged the cultural relevance of an A&R process accustomed to a top down approach to marketing (Knopper 2009: 81-104). When the majors controlled promotion and distribution channels they were almost able to push trends into culture by sheer force of will. Now that effective distribution and production channels exist outside of industry control, however, it’s more necessary than ever to engage in organic trends that may be outside of their control. Digital media technology has shortened the timescale under which culture operates, by the time a new trend bubbles up through the traditional marketing apparatus, it may be too late.

Consolidation is happening, but in the new medium rather than in the music industry proper. The University of Michigan and Arbor Networks just completed the most extensive study of Internet usage to date, analyzing 264 Exabytes of traffic at Internet exchange points (IXPs). In 2007, thousands of autonomous systems numbers (ASNs) contributed half the content served on the Internet, but by just 2009, half the Internet’s content was served from only 150 ASNs. During this short two year span, the largest content providers (Google, Microsoft, Yahoo!, Facebook, etc.) have also become some of the largest transit providers, and the bulk of Internet traffic now stays within their networks.

What does this mean? Broadly, it indicates mediation by these “hyper-giants” the majority of point to point traffic on the Internet. Facebook’s application layer, for instance, is increasingly mediating the day to day services that make up the average person’s activities on the Internet. This not only means that Facebook is party to my behavior, but that it’s also providing data to a broader class of third party applications which are also able to track my identity on the network. The usage of the web is increasingly shifting from from unmediated access of distinct sites through unique URLs to mediated access through “apps” and widgets which keep users on a parent site where users can be exposed to targeted advertising (Messina 2009).

Silos are similarly being built on the hardware side. We’re in the midst of a coming of age of networked digital devices. The days of people using cumbersome general purpose computers  as their primary means of accessing the network are numbered. This is generally a good thing–the details of networks and computation need to become an afterthought for their real human benefits to be realized. Ubiquitous access to information is the new norm and it shouldn’t require specialized expertise or awkward user interaction. But as technology companies vie to be dominant in the new space of these devices, they’re injecting more and more mediation on developers and users. iPhone, Droid and WebOS (Palm Pre) apps are all built on different technology stacks each allow users and developers varying degrees of access to the hardware and to new applications. While this has been relatively benign to date, it looks remarkably similar to the layers of mediation being introduced in the web layer. In both cases users trade convenience for a degree of control over how they interact on the network.

While this may make the future of “open” sound ominous, it should bode well for rights holders. The intermediary layer is where distribution can be controlled and where profit can be made. It’s already the case that you cannot save an arbitrary MP3 on the iPhone, when that type of access becomes normal, the easiest way to consume music will be through approved applications which are bound to be royalty bearing. Whether or not these royalties will be sufficient to sustain large labels, however, is anyone’s guess. The safe bet is that more and more of the business will be controlled the new intermediaries–the hyper-giants like Google, Apple, Microsoft and Facebook–while the influence of the old middle-men continues to wane. To continue radio parallel, these hyper-giants are the new RCA and CBS, and with their mastery of the new medium they will be the only ones able to enact enough control over their users to enforce copyright.

Conclusion

The chaos of the early Internet era is likely behind us. We’re beyond the main disruptive phase of this generation of technology and heading into the consolidation phase. Not media consolidation, but a structural consolidation of the platforms through which we access that media. So the next distribution channels for music are likely to be controlled by the same parties who are mediating our access to the Internet.

I haven’t discussed what this means for the creators and consumers of music. Broadly speaking, the Internet has been good to both. Access to music has never been so widespread or convenient. Artist reach has grown and the cycles of creation, production and distribution have been shortened.  While there is room for discussion about the effects of a  such short feedback loops on the integrity of art, the more immediate connection between artists and fans seems to have benefitted both.

A lot has been made of this artist-fan interaction and the role of the artist in marketing a new age. While digital literacy is necessary, the role of the marketer is not something that I want to thrust in the laps of artists. Using hyper-connectivity to get artists to spend their time on marketing misses an important point: good music is, and always has been viral. We’re more than one hundred years past the first million selling record, and during that time the tools of promotion and distribution have advanced exponentially. In a world of relentless self-promotion and instant exposure, I’d like to think that quality is what is most likely to set an artist apart. As the world becomes ubiquitously hyper connected, there really aren’t too many places for good art to hide.

References

A&M Records, Inc. v. Napster, Inc., 239 F.3d 1004 (9th Cir. 2001)

Apple, Inc. (2010). iTunes Celebrates 10 Billion Songs Downloaded [Press Release]. Retrieved March 6, 2010 from http://www.apple.com/itunes/10-billion-song-countdown/

Arcade Fire. (2010, March 3). In Wikipedia, The Free Encyclopedia. Retrieved March 5, 2010, from http://en.wikipedia.org/w/index.php?title=Arcade_Fire&oldid=347535650

Compact Disc. (2010). In Encyclopædia Britannica. Retrieved March 17, 2010, from Encyclopædia Britannica Online: http://www.search.eb.com.offcampus.lib.washington.edu/eb/article-92845

Data Compression. (2007). In Encyclopedia of Measurement and Statistics, (Vol. 1). (pp. 225-231). Retrieved March 20, 2010, from Gale Virtual Reference Library via Gale: http://go.galegroup.com.offcampus.lib.washington.edu/ps/start.do?p=GVRL&u=wash_main

Kafka, P. (2009, May 7). Warner Music Group Walks Away From Digital Start-Ups Lala and Imeem, Loses $33 Million. Retrieved March 5, 2010, from http://mediamemo.allthingsd.com/20090507/warner-music-group-walks-away-from-digital-startups-lala-imeem-and-loses-33-million/

Kelly, K. (2008, March 4). 1,000 True Fans. Retrieved March 5, 2010 from http://www.kk.org/thetechnium/archives/2008/03/1000_true_fans.php

Kincaid, J. (2009 December 4). Apple Has Acquired Lala. Retrieved March 5, 2010, from http://techcrunch.com/2009/12/04/apple-acquires-lala/

Knopper, S. (2009). Appetite for Self-Destruction: The Spectacular Crash of the Record Industry in the Digital Age. New York: Free Press.

L3enc. (2010, March 3). In Wikipedia, The Free Encyclopedia. Retrieved 03:34, March 5, 2010, from http://en.wikipedia.org/w/index.php?title=L3enc&oldid=347409794

Lessig, L. (2006). Code Version 2.0. New York: Basic Books.

Labovitz, C., et al. (2009, October 19). Atlas Internet Observatory 2009 Annual Report. Presented at NANOG 2009. Retrieved 2010, March 5, from http://www.arbornetworks.com/en/arbor-networks-the-university-of-michigan-and-merit-network-to-present-two-year-study-of-global-int-2.html

Messina, C. (2009, November 16). The Death of the URL. Retrieved March 5, 2010 from http://74.125.155.132/search?q=cache:tkGtolvIuKMJ:factoryjoe.com/blog/2009/11/16/the-death-of-the-url/+the+death+of+the+url

Metro-Goldwyn-Mayer Studios Inc. v. Grokster, Ltd. (04-480) 545 U.S. 913 (2005)

MP3.com. (2010, March 2). In Wikipedia, The Free Encyclopedia. Retrieved 06:43, March 5, 2010, from http://en.wikipedia.org/w/index.php?title=MP3.com&oldid=347324074

Napster. (2010, March 2). In Wikipedia, The Free Encyclopedia. Retrieved 18:29, March 6, 2010, from http://en.wikipedia.org/w/index.php?title=Napster&oldid=347273962

Tschmuck, P. (2006). Creativity and Innovation in the Music Industry. Dordrecht, The Netherlands: Springer.

UMG Recordings, Inc. v. MP3.com, Inc., 92 F. Supp. 2d 349 (S.D.N.Y. 2000)