It is surprising that, to date, most of the conversation around the music licensing challenges faced by new music services has centered on cost rather than friction — the specific prices of royalty rates and the dollar value of advances, instead of the complexity of the process. When that complexity is addressed, it is spoken of in very general or anecdotal terms: music licensing is “complicated,” requiring a “long time,” because negotiations take place among “countless copyright owners with different perspectives.” Furthermore, it is often argued only broad changes to copyright law — regardless of the impact of other factors — would be sufficient to overcome this friction.
In May of 2012, Sean Parker, an investor and board member of Spotify, claimed the firm’s US licensing efforts required upwards of two-and-a-half years before completion.1 Reuters had reported it spent at least eighteen months attempting to license the service in the US2, while Forbes wrote it had taken two years in Europe.3
Turntable.fm launched a beta version in late May of 2011. While still an invite-only platform, it announced a license from Sound Exchange covering performances of sound recordings, as well as blanket licenses from ASCAP and BMI for the broadcasting of musical works.4 Eight months later, the firm would announce direct licenses from EMI, Sony, Universal, and Warner music.5
In August of 2011, the Financial Times reported that the negotiations between the record labels and Beyond Oblivion, while nearing conclusion, had been ongoing for upwards of eighteen months.6 Four months later, the FT reported that Beyond Oblivion would close shop before launching its Boinc music service. Licensing agreements had yet to be completed.7
These accounts suggest different problems. Spotify was rumored to have spent nearly four years seeking music licenses. Instead, Turntable.fm, launched without them, then obtained the requisite compulsory and blanket licenses within sixty days, and, within ten months of the service becoming available, negotiated directly with the record labels. Finally, Beyond Oblivion was a non-starter after apparently two years of discussions.
Research and Disclosures
This brief was prompted by the concern that there may be a far wider range of factors beyond the law that contribute to the current friction in music licensing. We need a better understanding of how technology firms and copyright holders navigate the intersection at which these two parties meet. For this reason, attention will be given to licensing timelines, process maps, and logics — indirect measures of the transaction costs and the frames within which these negotiations take place.
An expanded description of this research can be found in a report from the same author, released jointly by the National Association of Recording Merchandisers and DigitalMusic.org (see http://digitalmusic.org/blog/our-new-white-paper-gives-music-startups-the-info-they-need-to-negotiate-licensing-deals/). No third party other than Washington and Lee University, through a Lenfest Grant, provided funding for this research. DigitalMusic.org provided the ideal partner for the release of the larger report — not only is their charter “to advocate, educate, and organize on behalf of the entire digital music ecosystem,” but also their various advisory boards include representatives of both copyright owners and technology companies.8
A combination of publicly available and privately obtained data were gathered from more than twenty interactive music services launched or attempted to be launched in the United States, between 2000 and 2012. Published news stories and company statements were inputs for the data. Also, semi-structured interviews were conducted with more than twenty individuals who had been directly involved in music service licensing efforts over the period. In addition, and in order to gain as holistic an understanding of the licensing process as possible, private data was gathered from individuals who had represented both so-called “sides” in these negotiations: technology companies and copyright owners.
Importantly, individual sources for this research have been and shall be kept confidential. The direct mention within this article of any particular service does not mean that anyone affiliated with that service provided private data for this project. This writer was not given access to private contracts or any deal terms, and any privately collected data is addressed in the aggregate and/or without attribution.
Any service that makes musical recordings available to the public, whether as downloads, internet streams, or some combination of the two, likely needs to secure licenses from two types of copyright owners: the owners of the sound recording and the owners of the musical work. Sound recording copyrights exist in the music you hear when you hit “play.” Copyrights in the musical work exist in the words and notes (i.e., the lyrics and composition) expressed in what you hear. Record labels, featured artists, producers, and performing musicians have their stake in the sound recording. Publishers, administrators, composers, and lyricists, have their interest in the music work. A single individual might be the performing artist, label, songwriter, and self-publisher. Alternatively, different individuals or entities might play each of these roles.
Most interactive services will need to pursue licenses that cover some combination of reproduction, distribution (or publication), and performance rights granted to copyright owners. Two types of entities predominantly license these rights for the use of sound recording and musical copyrights in the US: copyright holders themselves and rights collectives (or aggregators). Rights collectives — such as ASCAP, BMI, SESAC, and SoundExchange — represent specific rights or sets of rights on behalf of their members or stakeholders. Copyright holders, or their appointed representatives/agents, can directly negotiate license for the rights they hold.
First, and most importantly, the licensing process unfolds as it does for reasons far more complex than one party ‘gets it’ while another party does not.
For example, startup founders and investors may express different attitudes toward uncertainty and failure than those expressed by incumbent firms or copyright holders. These differences in attitudes are reflected in clear differences of opinion and outcomes in the negotiation. While startups are encouraged to defray the cost of uncertainty as long as possible, copyright holders have an incentive, if not a fiduciary responsibility, to address this cost upfront. As a result, the parties end up pricing particular uses of music before the value of those uses may be fully understood. Furthermore, for some of the parties involved, failure is seen as a normal part of the innovation equation — a somewhat positive badge, earned pushing the edge of what is possible — while for others failure presents a legitimate threat to professional advance — a negative brand, signaling the lack of procedure or credible analysis.
Second, the directly negotiated licensing activities of interactive music services have required, at the median, roughly eighteen months of effort before service launch.
The majority of licensing time is spent completing deals with major record companies who are also major publishers, while the remainder of time is spent completing deals with rights aggregators and collectives. Between ten and fifteen sound recording deals, across major owners and aggregators of these rights, are believed to be necessary for initial service launch. After initial launch, however, the number of sound recording owners with whom negotiations unfold has expanded from between 20-50 to greater than 500.
For example, one firm spent upwards of six months discussing potential features and pricing with rights holders before even beginning to build some version of the service. While some individuals expressed the belief that ‘familiar’ services might be licensed within six months, almost all services were self-described as ‘groundbreaking’,” or ‘never been done before’. And so, while some standardization of service characteristics might hasten the licensing process, most operators are looking to innovate in some way, leading to an expansion of the time spent both coming to an agreement over service features and pricing those features for a license.
Over the past decade, the length of time spent negotiating prior to launch does not appear to have significantly changed (the decrease appeared to be no greater than three months). What has changed is the number of tracks services license before launch. As examples: both Pressplay and MusicNet launched in December of 2001 with approximately (or less than) 100,000 tracks.9 In May of 2005, Yahoo! Music Unlimited, powered by MusicNet, launched with roughly one million tracks.10 Rdio launched in August of 2010, claiming upwards of seven million available tracks.11
Third, as noted earlier, interactive music services obtain licenses that cover both the reproduction and the performance rights in musical recordings and works. The sound recording licenses are obtained directly from the appropriate copyright owners. The musical work licenses often engage a combination of both collective and direct (or voluntary) licensing, involving discussions with not only performance rights organizations, but also the copyright owners themselves (or their appointed publisher/representative).
The point is that consent decrees obviously and significantly hasten the amount of time it takes for a new service to be licensed by the core performance rights organizations in the US. When rates cannot be eventually agreed upon, however, the resulting rate-setting proceedings can and have extended for years. Simply stated, due to the terms found in government-established agreements that cover the licensing activities of ASCAP and BMI, a service can operate as licensed (for public performance rights) by simply requesting a price for a license with characteristics not covered by the standard license terms published by these collective rights organizations. The result is a situation wherein a service can operate as licensed, yet the rate for that license has yet to be agreed upon.
Fourth, the amount of time it has taken to obtain a sufficient collection of licenses covering what are called the mechanical (or reproduction) rights in the interactive use of musical works has decreased over the last decade — from what once was years to less than 90 days in some cases.
This newfound expediency can be had only as long as the service’s features fit within the categories outlined by a 2008 agreement covering interactive services, and the service chooses to license via what is known as the ‘notice of intent’ process. Essentially, this agreement requires that a service give written notice to the appropriate copyright holder, or their representative, of the intent to operate under established terms before the right to operate under those license terms can be enjoyed. If the service characteristics are outside the bounds of those prescribed terms, the service has had to directly negotiate with each copyright owner, or their representative, whose work might be used within the service.
This intent process is not necessarily efficient or affordable, however. Estimates of the number of points of contact for direct licensing or ‘noticing’ musical work copyrights vary substantially: from as few as 500 points, to as many as 6,000, to in excess of 30,000 potential points of contact. This variety depends upon how these rights holders are aggregated and how large the catalog of licenses is. That said, it may be no coincidence that a new cohort of music services were licensed and launched after 2008.
Fifth, the pathway through which innovation unfolds is largely similar across the services studied. Whether there is a right way or a wrong way to travel through the licensing pipeline, there is little variation in the way in which services make this trip. Furthermore, at any point over the last decade, it appears that no greater than two or three law firms, or individual lawyers and their staff, were most central in brokering directly negotiated licensing transactions.
For example, most services began the licensing process in discussions with one or more of the major labels. This initial stage might best be described as a ‘getting to know you’ conversation, during which both personal connections and basic service ideas were discussed. After a few months, the conversations shifted to more specific discussions of service features, pricing, if not also more technical white papers. The final months involved discrete discussions of licensing agreements, which saw between four and nine revisions over a span of one to four months. Once these major agreements had been negotiated, services shifted their attention to other aggregators of independent labels and artists.
Finally, while new services face an expectation to be novel — from the perspective of consumers, investors, and even copyright owners — service characteristics among competing services after launch seem quite similar. The variety in service characteristics that arrive at the table to be licensed appears to be somewhat greater than the service characteristics offered by services that launch as licensed. It is difficult to tell what would happen and what could be learnt by the industry from a much-expanded range of service characteristics.
While technology has historically paved the way for new developments in the music industry, it has both disrupted old business models and transactions over copyrights. Negotiating over music rights has become very complex, and the cost is not just measured in the price paid for a license. All the parties spend too much time, effort, and resources clearing every legal aspect in trade of recorded music. The cost for music startups is arguably large, and music right holders pay a hefty price too measured by the opportunity cost of losing new business. Whether the party is a copyright owner or a technology company, the process for procuring a music license weighs heavily all round.
By David Touve
1. Isaac, M. 2012. Sean Parker: Why Did Spotify Take So Long to Get Stateside? It Could Have Been Apple. All Things D, May 30, 2012.
2. Reuters. 2011. Spotify to launch in US after long wait. Reuters, July 14, 2011.
3. Bertoni, S. 2012. Spotify’s Daniel Ek: The Most Important Man In Music. Forbes, January 4, 2012.
4. Popper, B. 2011. Turntable.fm hits 140,000 users in its first month. BetaBeat, June 22, 2011.
5. Van Buskirk, E. 2012. Turntable.fm goes ‘legit’ with licenses from all 4 major labels. Wired, March 13, 2012.
6. Bradshaw, T. 2011. Beyond Oblivion reveals its Boinc service. FinancialTimes, August 23, 2011.
7. Water, R. & Garahn, M. 2011. Beyond Oblivion crashes before launch. Financial Times, December 30, 2011.
8. For further information, see: http://digitalmusic.org/about/
9. See: Garrity, B. 2001. RealNetworks bows subscription service. Billboard, December 15, 2001. Also: Kusher, D. (2001). The digitabt beat: Musicnet doesn’t rock. RollingStone, December 12, 2001. Available from: http://www.rollingstone.com/music/news/the-digital-beat-musicnet-doesnt-rock-20011212
10. Yahoo!, Inc.(2005) Yahoo! premiers Yahoo! music unlimited. Yahoo!, Inc., May 10, 2005 (company press release). Available from: http://docs.yahoo.com/docs/pr/release1237.html
11. Rdio, Inc. (2010). Rdio takes the wraps off social music service. Rdio, Inc., August 3, 2010 (company press release). Available from: http://www.rdio.com/press/rdio-takes-the-wraps-off-social-music-service/