The current digital landscape is based on platform capitalism and on the cloud concept in which different services (music platforms, instant messaging services, live music companies, sellers, device companies, music major labels, radio players) try to control music/sound circulation. The new century started with the development of new devices — the iPod, smartphones, AI speakers — and distribution modes (podcasting, streaming) and the emergence of voice interaction to control devices. In this context, this paper develops a value chain defining the diverse key intermediaries in the management, accumulation, distribution, and access to information phases.
1. Introduction: GAFA and the sound industries
2. Value chain, sound circulation, and the data ecosystem
3. The production: Human activity as a resource, and ways of acquiring data
4. Circulation and compilation: The great invisible mediation
5. Conclusions: Access to data to close the value
1. Introduction: GAFA and the sound industries
On 5 March 2019, the start-up Instreamatic.Ai, headquartered in San Francisco (U.S.), published a blog entry with the title: “The voice era is here. No doubt about it” (Instreamatic, 2019). In May 2019, the first Voice Summit, an event about the “voice business”, took place in Madrid (Spain). The summit’s Web site refers to many of the topics of the moment: business and voice, virtual assistance and voice, artificial intelligence (AI) and natural language processing. This blog article and this event, centered around conversational marketing, serve as a starting point for the present work on the value of sound in the current media context.
The transformation of the sound industries (radio, music, OTT-audio messaging, audiobooks, podcasts) during the last two decades has been radical. We do not exaggerate when we talk about a completely new ecosystem where actors are systematically reformulating themselves. In recent years, and based on my research on radio, podcasting, and the music industry, I have observed how content is no longer the epicenter of the debate, and in the work of Arsenault (2017), boyd and Crawford (2011), Albornoz (2015), Yúdice (2017), Schrock (2017), and Morris (2015a, 2015b) one can observe how data and data control are framed as the great hope (or threat) for cultural industries.
In an environment of platform capitalism (Srnicek, 2016), where the so-called GAFA companies (Google, Apple, Facebook, Amazon) are assuming important positions in different audiovisual sectors (Miguel de Bustos, 2016), the sound industries have not been immune to their impact. The Finnish intellectual property rights management entity Teosto (Osimo, et al., 2019) reported on large digital sound companies’ acquisitions of several companies that work with data and information .
These large companies work with hardware and/or software — in many cases, in a complementary way. Devices, platforms, data centers, networks, content producers, digital stores: there is a whole system where sound plays a fundamental role. This system is increasingly relevant where new practices, devices, and business models emerge, as exemplified by the emergence of intelligent loudspeakers, the importance of voice messages in messaging platforms and the development of the podcasting sector in different countries.
We live in times of ubiquitous listening where we are surrounded by sound artifacts at every moment thanks to our continuous connection with the cloud and a soundtrack that accompanies many of our daily activities via connected devices and digital music services (Morris, 2015a). This transformation means that even our music consumption is more marked by our daily activity than by our musical tastes (Prey, 2016) or personalities (Greenberg, et al., 2016).
During the years 2017/2018, and while working on the new value chain of digital music, I was faced with a new dilemma. Within the new ubiquitous listening framework, it was difficult to establish a simple value chain from songs, the natural object of the commodification of music, and to observe how the players in this system established a value chain of data from interactions with sound content; in other words, how this data are produced and by whom, how they are stored and distributed, and who has access to the data.
Here we refer to different types of data — not just the big data that occupy a large part of the current debates. A sound file is a piece of data in itself, together with related metadata. These data must be standardized and stored for distribution in a decentralized storage system that seeks to respond to user reproductions in a precise way (hence the importance of data centers, which we will address later). Likewise, all consumption of sound information involves the generation of assigned information that allows, for example, the distribution of money generated by the platforms between the different intermediaries and creators (if available), the offer of geolocalized advertising adapted to each user or the establishment of increasingly accurate consumption profiles. This is an Internet of Things environment, a system that measures and controls the activity of objects and living organisms through sensors that collect, process, and articulate the data that circulate through the networks (Mosco, 2017).
In this way, all the GAFA companies participate in the commercialization of sound content: Google launched its Podcasts application in 2018 and has YouTube Music, whose premium version also appeared in 2018 after revealing itself as the main music-accessing platform for many Internet users. Apple has a music service, a global streaming radio service, a podcast application, and a smartphone (mainly sound hardware). Amazon has its Amazon Music service and its audiobook portal, Audible. The three companies have each also placed a key device in the center of digital consumption: the AI speakers connected to virtual assistants — Alexa (from Amazon), Siri (from Apple), and Ok Google. On the other hand, we have Facebook, a key player in the circulation of contents in its different platforms, whose messaging platform WhatsApp keeps enhancing the circulation of sound messages (Stokel-Walker, 2018). To the GAFA platforms we would have to add specific sound players: platforms like Spotify, Pandora, or Deezer; aggregators and multi-channel networks (MCNs); music producers; radio stations; and other actors that will be addressed in this work.
Starting from this framework, where we observe a systematic commodification of the sound reality (which connects with the commercialization of reality defined by Zuboff  as surveillance capitalism), this work’s main objective is to trace a value chain where the production of data begins with humans’ actions and data intermediation favors different actors who have little or nothing to do with the producers of information.
This paper aims at exploring these actors’ capacity to position themselves in each step under the theoretical frame of the critical political economy of the media and, as Mosco (2014) states, the political economy of the cloud. This concept analyzes and critiques how the political economy of the cloud advances informational capitalism. The following sections of the present study examine various ways of extracting data generated by human activities related to sound. For this examination, I have divided the datafication process into three steps: data production, circulation/compilation, and access. In this context, quoting Mejias and Couldry (2019), the production of data and datafication combines two processes: “the transformation of human life into data through processes of quantification, and the generation of different kinds of value from data”.
Kaplinsky and Morris (2000) defined the value chain methodology as a simple way of articulating complex relations, but it allows us to visualize political economic relations and, based on that visualization, analyze those relations. With this tool, I achieve the objective of giving “an illustrative representation of the identified chain actors and the related product flows” (Faße, et al., 2009).
To define the sound commodification value chain, I completed a scientific literature review and analyzed reports, Web sites, and publications from different companies and institutions. In 2018 and 2019, I visited several companies that are key actors in the value chain. During these visits, I gathered information via participant observation (Emerson, et al., 2001) and open interviews. The types of companies from which I gathered information included : Data Center (Madrid, Spain); Telecommunications Company (Madrid, Spain); Voice Biometrics Company (Madrid, Spain); Digital Music Aggregator (Madrid, Spain); Music Label Multinational Company (Toronto, Canada); and, Distribution Warehouse E-Commerce Platform (Toronto, Canada).
2. Value chain, sound circulation, and the data ecosystem
Before beginning to dissect the different descriptions related to the value chain we are dealing with, I would like to discuss the main concepts around which this work revolves and the reasons for their use.
The “value chain” concept was introduced by Porter (2001) as a tool to improve the competitiveness of companies through the unbundling of the activities they carry out. In addition, and this is why we are interested in the value chain as a tool, it frames companies’ activities in the value system, where companies position their work in relation to other agents looking to differentiate the activities and specifying where they create value in the general dynamics of the system. According to Curry (2016), a value chain is made up of a series of subsystems, with raw materials (inputs) and their transformation processes generating final products (outputs).
The value chains of “analog” cultural industries have been well analyzed in terms of their relationships between production, distribution, and consumption, and the European Commission updated its analysis of them in its 2017 report Mapping the creative value chains. In a study on the cultural economy in the digital age, these industries stand out as being beyond the changes in power relations that have transformed value chains; the actors who have dominated these analog industries in the pre-digital scenario continue to have a fundamental role as gatekeepers in the current economic organization (European Commission Directorate-General for Education, Youth, Sport, and Culture, 2017).
Organisation for Economic Co-operation and Development (OECD) (2016), Curry (2016), and Miller and Mork (2013) have published maps of the “big data ecosystem” or the “big data value chain”. These works serve as a starting point for the design of our chain, created from the circulation of sound content.
In the ecosystem presented by the OECD (Figure 1), data appear in all transactions; consumers send data — in addition to money, attention, and taxes — in all of their exchanges with different services. Platforms are at the center of relationships with other actors and it is also noteworthy how the authors leave infrastructure providers out of the relationship with consumers, and the devices do not appear in any case. We must also rethink the role of public services broadcasters in this whole system.
Figure 1: Big data ecosystem (Source: OECD, 2016).
This central space that the platforms occupy in information extraction and intermediation, according to Srnicek (2016), is their main source of political and economic power. But to reduce the power of the platforms to their data is to limit their real impact, which is highly connected with the financialization of the economy and the conditions for futures (Arvidsson and Colleoni, 2012). These futures reside more in a quasi-monopolistic system with billions of users satisfied by a panoply of specialized private global platforms.
On the other hand, the real value of data as commodity is yet to be glimpsed, especially for small players in different industries. The Teosto report (Osimo, et al., 2019) highlights how data tends to be overvalued, especially by actors who cannot monetize them. In other words, the value of the data is much easier to determine when we consider the monopoly of large actors in each sector. In the long data queue, as with cultural products, not all actors have the same strength. Only cooperation and the creation of networks between small and medium actors will let them compete in the short and medium term in managing and generating value from information.
As for sound circulation, we understand the mediated forms that connect content to listeners. When we speak of mediated forms, we are not only referring to formal and distribution-centered channels, whose origin is the first great decade of the phonographic industry — the 1920s — and its development with electrified recording and the global market (Denning, 2015), but also to informal channels (P2P, USBs, CDRs, free radios) beyond purely capitalist logics. We are talking about a ubiquitous listening, which Kassabian (2013) associates with the concept of ubiquitous computing developed in the 1980s by Marc Weiser of Xerox PARC — the continuous presence of music in our lives and in our daily spaces. In this text, we go one step further with that convergence between ubiquitous computing, confirmed by the presence of sensors and hardware in each space, and ubiquitous listening, and arrive at the hypermediated commercialization of sound spaces of our daily lives, in which the generators of the initial raw material are the listeners and speakers themselves, as we will see in the following sections.
3. The production: Human activity as a resource, and ways of acquiring data
Creating value from information and data is by no means new. The telephone book would be an informative source based on its circulation, and there are valuable sources more related to the creation of audiences, such as hit lists based on record sales or radio play. Radio audience figures would also be an example of what Ang (1991) called the institutionalization of audiences or what Smythe (2001) referred to as the exploitation of audiences from their work as listeners.
Data acquisition has been called the first element of the big data value chain (Curry, 2016). Section 3.1 of this paper focuses on human activities (see Figure 2) that are exposed to the circulation of sound and begin to create value. Section 3.2 defines how they would come into contact with different software and hardware actors that would be in charge of gathering and compiling data (see Figure 3).
Figure 2: Human activities creating value with sound.
3.1. Human activities: The basic resource
Our value chain does not include all aspects of musical production incorporated into the music value chain. Although the pure activities of creation, recording, and production are key in information generation, applications such as Endel — which has signed an editing agreement with Warner — generate information by building soundscapes and composing songs for different times of the day from an algorithm (Guardian music, 2019).
Based on sound content and from its use and consumption, we can derive the following activities that generate value independently and combined:
Listening. Exposure to sound has been a classic means of generating information, especially when paired with media that have developed measurement techniques such as those of the radio industry. The amount of time devoted to consuming different sound media has been published in reports such as Audio Time (Radio Joint Audience Research Limited (RAJAR), 2015) and, although listening does not generate information from broadcast media directly, there are an increasing number of ways to measure audiences through hardware and software. Current bitcaster platforms or radios measure listening in terms of reproductions or connections, though they have difficulty differentiating between the machine and the listener.
Regarding Spotify, there is news about the measurement of reproductions and the action of bots (Ingham, 2018), and an investigation confirmed that the reproductions of the platform could be altered (Eriksson, et al., 2019). Radio Joint Audience Research Limited (RAJAR), the institution in charge of measuring audiences in the U.K., tried to differentiate between people and machines when making their reports: “Normally a computer doesn’t know where the person is, and would think they’ve been listening all the time. The stream is not intelligent” (RAJAR, 2012).
Extracting value from listening, which has always been linked to audience studies for subsequent monetization with advertising, is now connected to reproductions or streaming logs that generate other kinds of direct benefits. It is the change of focus from a single consumer listener to multiple consumer listeners that generates value in ways other than listening and recognizes that the listening activity is related to other data generated in different activities.
Speaking/interacting with one’s voice. The voice is a vital element in human communication and in generating information and value. Phones and phonographs are examples of how voice communication was commodified. But the real change in recent times is how the hypermediation of voice interaction has been centered in value creation thanks to AI. An article in Fortune (Dumaine, 2018) confirmed this with its headline and subheading: “It might get loud: Inside Silicon Valley’s battle to own voice tech. Amazon, Apple, and Google are investing billions to make voice recognition the main way we communicate with the Internet. It will be the biggest technology shift since Steve Jobs launched the iPhone.”
As highlighted in the opening of this paper, voice is measured and translated into objective information in a systematic way for different uses. From the devices, machine learning tries to understand different aspects of the human voice by objectifying it and trying to convert its different qualities into data. Biometric analysis extracts certain qualities for the verification of user identity, but these types of companies also work on speech analysis to identify user moods. Currently, the ability to identify these states starts more from word spotting than from tonal variation. This line of systematization of feelings from voice or sound introduces applications such as Chatter Baby, which aims to decipher why a baby cries, or Beyond Verbal, “The Emotions Analytics Company”. Cheney-Lippold (2017) questions the conversion of emotions into data and their value: their complexity, their cultural contextualization, and their material incoherence technically surpasses the computable.
Researchers and companies are concerned not only with the human qualities of the voice but also with how the synthetic voices of intelligence systems present a personality to users from a psychological point of view (Pérez García, et al., 2018).
Moving/running/dancing. The multiplicity of sensors in our devices means that listening to music and other sound content does not take place independently. The addition of geolocation to audio has led to multiple developments ranging from geolocalized advertising to the integration of data with music recognition applications, such as Shazam, that can be used while dancing in a club or at a concert. In addition, experiments such as the one carried out at the Sónar Festival (Artigues, 2015) made it possible to track the wireless signals generated by the devices of those attending the festival and thus determining the flows of people between spaces, the most viewed artists, and the patterns of activities according to their role within the festival. Other devices, such as wristbands connected by RFID (radio frequency identification), allow the festival promoters to follow the movements and purchases of users in these events, cross-referencing valuable information about users’ scenarios (and therefore the music and artist to which they are exposed) with the type of drink consumed in the bars, and other activities around the festival.
The last notable activity related to movement and sound would be running. Platforms like Spotify have linked the creation of playlists with the tempo marked by the runner’s speed. When this function was withdrawn from the platform in 2018, Spotify offered it through external applications such as Runkeeper and Nike+Run, which add an enormous amount of biometric and geolocation information to music consumption data.
Manual/visual/play interaction. It is possible to systematize digital game play (a multitude of listeners are now generated in connection with the game, constructing playlists on demand, or in completely musical video games), visual interaction, augmented reality (Olmedo and Augusto, 2013), and manual interaction. Gazi and Bonini (2018) highlighted the commodification of digital radio audiences’ tactile actions. This also connects with the actions of playing, pausing, or moving forward or backward in a song. All this provides relevant information for the platforms about their audiences’ usage.
Participating. Digital audiences’ conversion into data and commodification from participatory culture (or their work) has been extensively studied (Fuchs, 2010; Terranova, 2000; Arvidsson and Bonini, 2015). In their relationship with sound, we find the audience performs work such as creating playlists on different platforms, labeling files generating metadata, likes, ratings, disseminating content on social networks that promote circulation, comments, and retweets — a number of actions that when linked to sound are multiplied by millions of actions per minute that generate data and metadata. In the context of this work, we are not talking about the extraction of value pointed out by the aforementioned authors; we are talking about a complementarity that transforms the actions of the users into a sum of value within the totality.
Buying. To the classic exchange value that is introduced in the purchase process is now added data linked to this purchase. When buying a vinyl or a CD from an e-commerce service, when paying for a download in Bandcamp, or when subscribing to a streaming service, we are generating new data (credit card, location, device). To any exchange process we add the data of the process itself and the search and evaluation offered by users — an intangible not-put-in value for consumers but with much value for intermediaries if they know how to work the information.
3.2. The first step of data gathering
Human actions, the object of the capture and keys to the generation of value, produce different intermediaries that begin to acquire the data. In this first contact, we refer to raw data, without entering into its subsequent value, and this is useful for the subsequent representations of the value chain.
Figure 3: Processes of data collection.
a) Hardware: Contact devices or infrastructures. Hardware and sensors mark the first level of information gathering. Without microphones, there is no sound data collection. Without loudspeakers or headphones that reproduce sound, we cannot interact with devices or generate other information about sound consumption. Digitization has meant that the audio flows wirelessly in a personalized way, and all that circulation generates a value. In this step of the chain, we identified four types of devices that can mediate that relation:
Intelligent devices. In this section, we include those devices that integrate hardware, software, and sensors; computers that enter spaces where these devices previously played a role. To telephones, clocks, and televisions — three of the key technologies in everyday life in the twentieth century — add the prefix “smart” and include sensors (depending on the device: GPS, gyroscope, accelerometer, ambient light sensor, heart rate sensor, gravity sensor, rotation vector, etc.) and at least one microphone. The main device manufacturers, like Samsung, Huawei, or Apple, are included in this space, and we’re not just talking about smartphones — we’re also talking about connected TVs, as when Samsung was questioned about the intrusive activity of the microphones built into their devices when recording users’ voices (Gibbs, 2015).
To these devices as physical intermediaries we must add those that turn conventional televisions into smart TVs, and here we find devices from three of the big companies (Amazon Fire Stick, Apple TV, and Google Chromecast — with its Chromecast Audio version) which, besides having microphones, also connect with and can control other intelligent devices.
AI speakers and headphones. We wanted to dedicate a section to “intelligent speakers” because of their sonorous specificity, and here we include headphones with similar characteristics. We’ve already discussed the launch of the three speakers from Google, Amazon, and Apple. There is a fundamental work by Crawford and Joler (2018) on this topic, focusing on the Amazon Echo and the extraction of resources such as natural, data, or human work.
With a penetration of 21 percent of the population over the age of 18 in the United States (National Public Radio (NPR) and Edison Research, 2018), these types of devices are considered the great transformers of the industries of sonorous contents. The reality is that mediators with great global power are introduced in industries such as radio, where previously mediation was local and broadcast networks were controlled. The first contact is controlled by large software/hardware companies that will have control of the information and, in many cases, begin to control the contents in a context where automatic learning is fundamental.
Wearable technology. Some devices in one way or another participate in the musical and sonorous circulation and are attached to the body, and we highlight three types of technologies that could fall into this category. As an example of the first type, Nielsen’s Portable People Meter (PPM) has been measuring radio and television audiences in some markets in the United States and Canada since 2008 based on the recording of sound or non-listenable frequencies. PPM is becoming more wearable, with a better microphone that better captures the inaudible frequencies (Inside Radio, 2019). It has significantly modified the measurement of audiences in large markets and, quoting the radio executive Kevin Weatherly, “is a way of measuring that, at least, listens” (Carney, 2011); for authors such as Stiernstedt (2008), it facilitates the monitoring of audiences.
The second type of technology would be the wristbands that are used in festivals and events and that, thanks to RFID technology, allow companies to follow users’ activity from their entry or exit from the festival to their purchases in the different spaces. A report addressed to industry (Gingrich, 2017) praised the possibilities of this type of device (improving security, enabling cashless payment, and obtaining data and better engagement with the public) without mentioning the problems regarding customers’ privacy. Finally, it is worth mentioning virtual reality glasses, such as Facebook’s Oculus system, and the augmented reality glasses launched by Google in 2012, which also have different sensors as well as microphones.
b) Software. After defining the contact devices and infrastructures, we have to identify the software-related tools that extract information from the sound circulation of operating systems, applications, AI systems, and other relevant services. To establish broad and clearly categorizable frameworks, we recognize that there are other actors based on software (digital social networks and platforms); due to their importance, we dedicate a specific category to them.
Operating and AI systems. This is where the big companies (Apple, Microsoft, and Google) take on special importance because of their position and the way they handle information. Their operating systems are installed as standard in the various fixed and mobile devices, and in all of them sound interaction is being enhanced through their respective AI systems. An article showed that Android devices with pre-installed applications were a source of systematic information extraction from users (Gamba, et al., 2020). According to an investigation by the Washington Post (Fowler, 2019), occurs with Apple’s iPhone and the applications that work on its operating system. If in principle it seems that operating systems do not accumulate information or transport it, according to their configurations they give third parties access to certain data and sensors.
The case of AI software, such as Siri or Alexa, whose main activity is the challenge of understanding natural language for interaction, is different. Here user interaction is fundamental and, as indicated by information from Bloomberg (Day, et al., 2019) , its use is supervised by a global team that, from the audios recorded by users, without prior notice, help the system to understand natural language and improve its ability to interact. Therefore, there is a purely automated function that goes through each recording, sending segmented audios to the company’s central servers and interpreting them into individual sounds from the previous audio base, and a function performed by human teams. The reality is that for this whole process, computational power and speed of response is basic, and this is where infrastructures become fundamental and are controlled by very few operators at a global level. These systems, such as Alexa Voice Service, also allow third parties to design conversational interfaces. Manufacturers of devices (loudspeakers, cars, or lamps) use Amazon’s tool to facilitate connection and interaction (Amazon Alexa, 2019) in various spaces of daily life.
In this way, the IA systems of large companies are trained based on foreign content and the vocal interaction of users, in addition to specialized workers who check the correct functioning of these processes. The sound and the human word act as great trainers of the machine learning.
Apps. Applications on mobile devices have become key intermediaries within the ecosystems generated by different devices connected to the network. Regarding sound circulation, we have to differentiate between those apps with access to the microphone and those that base their service on access offered by music streaming platforms. On the one hand, and in addition to the applications that may access microphones of devices to which we have referred previously, here we find identification applications, such as Shazam, that have become key in the acquisition of information for their geolocation capabilities. On the other hand are those applications that extract information from the users’ listening activities in streaming services such as Spotify or Deezer. Here we can highlight, for example, apps dedicated to exercise (reviewed in Section 3.1c) or those such as Tinder where privacy spaces are commodified and use music in a way that helps shape the design of digital identity.
Voice biometrics. We must highlight voice biometrics companies, such as Verbio or Nuance, which, based on machine learning, offer their services to brands based on the voices of their clients. The development of this type of service will be essential in the coming years of sound development.
Ad programmatic agents. In this first contact, it is important to highlight the companies dedicated to making programmatic ad sales by tracking the activity of listeners on different platforms. These companies have been identified by Vonderau (2019), and some also appear elsewhere in our chain. Vonderau identifies more than 30 companies that obtain data from that first-contact reproduction of Spotify content. This is why we label such companies “advertising-related agents”.
c) Platforms. It is evident that the platforms could be included in previous sections as applications or as specific software. But their role goes further, especially regarding their central position in audio circulation. We will also see them in other steps in our value chain. There are many examples, but we can establish three platform categories depending on the type of main sound content that is accessed from subscription or advertising exposure. Note, though, that they are continuously mutating, incorporating other types of content: musical content (Spotify or YouTube), audiobooks (Audible), and audio (radio or podcast) content (TuneIn, iVoox). These applications work by adding third-party content and, according to David Fernández Quijada (2019), they allow audiences to access content quickly and easily, but with the disadvantage that radio operators lose control of relationships with listeners by not controlling the interface.
d) Social networks. Networks such as Facebook, Twitter, Instagram, and WhatsApp give content a qualitative and quantitative dimension. Companies such as Social Blade and Chartmetric offer statistics on one’s followers in different networks; platforms as well as content producers obtain important information from the interaction of users. We are not only talking about figures but also about an ability to read the moods of a sound production.
e) E-commerce. In e-commerce, information is secured through the purchase process and subsequent valuation. Sound e-commerce companies include Amazon (sale of vinyl, CDs, and digital music) and Ticketmaster (ticket sales).
4. Circulation and compilation: The great invisible mediation
In this step in the value chain, which we illustrate in Figure 4, much of the real power is concentrated around information within digital capitalism. We are talking about the combination of infrastructure and software that, together with large amounts of invested capital, gives control to few global companies. The control of the physical environment has historically been the key, and, in many cases, states have had and still have a key role as regulatory agents (or deregulators). In its first two decades, the Internet has been presented as a sort of de-mediated, ethereal space, sublimated to concepts such as “the cloud”, where real and physical mediation was invisible (Mosco, 2014). But the distribution infrastructure is key when analyzing media technologies (Sandvig, 2015), and strong investment means that the infrastructuralization of digital platforms is increasingly clearly appreciated (Plantin and Punathambekar, 2018). At this level, we also deal with what Lahey (2016) calls “invisible actors”, Morris (2015b) calls “infomediaries”, and Braun (2013) calls “transparent intermediaries”. It is true that some actors are common to the distribution or storage of any type of data (especially with regard to infrastructures), but we will see how in the field of software or compilation we find some specific actors. These actors are those who have the capacity to accumulate information, which can be valued in different ways (in the logic of platforms), and/or accumulate data in a more physical way, which requires ways to circulate and places where they can be routed or stored.
Figure 4: Circulation.
Focusing on the circulation of audio, according to a study by Sandvine (2018), audio accounts for 1.05 percent of network traffic. To this percentage generated directly by large audio platforms (see Table 1) and by other services (such as Shoutcast, a platform specializing in streaming radio) should be added the percentage of audio in circulation via BitTorrent (which accounts for 31 percent of upstream traffic) and messaging and VOIP (voice over Internet protocol) applications (which account for 1.72 percent of upstream traffic and 8.12 percent of upstream traffic, respectively) .
Table 1: Global audio streaming traffic share. Source: Sandvine (2018). Rank Company Percentage downloading Percentage upstream 1 Spotify 33.07 24.21 2 Shoutcast 18.01 21.92 3 HTTP audio streaming 16.94 32.40 4 TikTok 15.66 8.66 5 Apple Music 8.91 6.24 6 SoundCloud 2.34 1.52 7 Deezer 1.65 0.73 8 Google Play Music 1.40 0.63 9 Pandora 0.92 1.13 10 TIDAL 0.70 0.23
For the distribution and compilation of sound-related data, we establish two categories within the value chain, along with their corresponding subdivisions:
a) Infrastructure. Infrastructure owners have been the major beneficiaries of the flow of information since the beginning of the commercial Internet. The value is generated from data, without consideration of the information that it contains. One of the most repeated claims in recent years is that access to cultural products was free and did not generate wealth (especially P2P), when the simple circulation of bytes based on sound generated a clear value for owners of infrastructures. The reality is that cables, antennas, and property spaces are mostly private. They, in turn, needed state involvement for their development, with different modes of regulation depending on the country and the continent, but which neoliberal decision-making from the governments of Thatcher in the U.K. and Reagan in the U.S. consolidated as deregulated (Plantin and Punathambekar, 2018) and privatized (Burkart and McCourt, 2004) global spaces in the western Internet. It is important to understand the political architecture of cyberspace because companies that own and operate infrastructures continue to press for control over them (Deibert, 2013). They exert influence through lobbies in different states and supranational institutions.
Internet service providers (ISPs). ISPs have or rent submarine cables, different types of cable to connect users, antennas, and LTS or 5G repeaters. We are talking about the base on which sound information circulates, where we can distinguish between network owners/operators and marketers. Most ISPs are private commercial agents , although there are interesting examples such as Guifi.net — a community, neutral and open network connecting to the network developed in Catalonia — or the public municipal networks in Nordic countries (European Commission, 2017). ISPs’ relationships with sound data can be seen in their systematic offer of free temporary access to services/music platforms  or offers for unlimited consumption of audio data at a flat rate.
Content delivery networks (CDNs). CDNs are defined as networks of servers distributed (and hosted in data centers) by different regions to facilitate access to content in an efficient manner (Held, 2010). The closer that a given user is to the CDN the faster they receive content. Behind the CDN we find companies specialized in this work (e.g., Akamai) and others dependent on GAFA (e.g., Amazon Web Services or Google). The Google Cloud Platform CDN, for example, has hosted all Spotify data since 2016 (Konrad, 2016).
Data centers and Internet exchange points (IXPs). The logistic hosting centers of the network infrastructures or data centers host servers offering very specific geographical, security, and energy conditions. Many are located in the centers of big cities to be close to most of the users. In addition, some of these centers are IXPs and facilitate the physical exchange of information between different network providers. Data centers are strategic points and, according to Deibert (2013), they are key to monitoring and controlling traffic on the Internet. In this service sector, which is highly atomized at a global level, companies such as Equinix, Digital Realty, China Telecom, and Interxion, with data center networks throughout five continents, stand out.
Decentralized networks (P2P or blockchain). In decentralized networks, in many cases, network computers act as servers. For example, in its beginnings Spotify used its users’ computers, using P2P technology as the basis of its network (Kreitz and Niemelä, 2011). In the case of blockchain, anyone can process information by becoming a node, but the processing capacity is linked to the power of the equipment and high energy consumption.
b) Data compilators. Data compilators accumulate and treat information that has been captured in first contact, originating from different sources. They are the agents who value the long tail of data.
Traditional metrics companies. These are all companies in charge of measuring audiences in classical or digital media as well as generating sales charts. This includes Nielsen and ComScore as major global players, the Spanish IACM, RAJAR in the U.K. for sound audiences, and Billboard (U.S.) and Promusicae (Spain) as music streaming sales/consumption compilers.
Aggregators and MCNs. Digital media uses the aggregator concept to refer to different actors. In this case, we are referring to digital distribution intermediaries that operate between producers/artists and digital platforms dedicated to musical products; they are key players in the business-to-business market (Galuszka, 2015). Aggregators include The Orchard, owned by Sony, and the Spanish independent Altafonte. The number of MCNs operating as specific intermediaries and providing management services on YouTube has grown dramatically in recent years. Specialists in music include the American Create Music Group and the Spanish Real Key Music.
Apps aggregators of audio content and podcasts. This subcategory includes those applications that intermediate in the distribution of sound content based on the aggregation of streaming flows or RSS channels from different creators and companies. These aggregators’ value generation travels through intermediaries, offering programmatic advertising or subscription services to audiences. These applications include specialized actors such as TuneIn (located in Silicon Valley, U.S.), the Spanish iVoox, and podcast applications managed by Google or Apple.
Data manager companies. These are companies dedicated to compiling information from the circulation of music. The Echo Nest, the Spanish BMAT, and Soundcharts obtain information from social networks, platforms—playlists and charts—radio, clubs and festivals. These infomediaries offer key information for decision-making in sound industries (Morris, 2015b).
Ad programmatic agents. These have already been mentioned, but their role as intermediaries in different platforms, portals, and applications makes them key players in the accumulation of data.
Rights management organizations. The managers of intellectual property rights have long had exclusive access to a large amount of cross-referenced data. Today — and increasingly, thanks to the new services and their global presence — they continue to be key players in information accumulation.
5. Conclusions: Access to data to close the value
This paper analyzed a large number of intermediation agents that have arisen from information generated by users in their relationships with sound and that are represented in the final value chain (see Figure 6). Their exploitation is not limited to purely informational and involves creating informational value from its weight in bytes. This value is increasingly difficult to establish since its construction and monetary transformation is often carried out using non-transparent algorithms adapted to each value creator (De Marchi, 2018). In a change from other previous mediation scenarios, individuals do not reap the benefits of this massive generation of information due to a clear absence of public service in data management as well as privacy. Furthermore, this process of dominating platforms, or platformization, has changed the operating logics of cultural industries in terms of their dependence on new infrastructures and a production marked by and for data collection (Nieborg and Poell, 2018). Our value chain demonstrates how GAFA can be placed in all the levels of the value chain with an evident concentration of global power. This reflection has also been made by Napoli (in Canada Department of Canadian Heritage and the Canadian Commission for UNESCO, 2019), who highlighted vertical integration between creators and distributors of content.
Figure 5: Access.
Through automatic learning, and in a sort of algorithmic subjectivation, users act as unconscious designers and allow platforms to accumulate information and data in those processes that are modeling algorithmic responses. In this way, value is generated for information and structural intermediaries, but access to this information (see Figure 5) is exclusive to large digital operators. Even large corporations in the pre-digital scenario (media or record labels) had limited information about the trail of information generated by data from sound.
Essentially, we talk about a commodification of all spaces, the tracing of our preconscious routines (Langlois and Elmer, 2019), and the sonorous reality in which creators and audiences are often left out of the distribution of benefits. In the debate on public policies, van Dijck, et al. (2018) go further in their proposals on the regulation of platforms based on the incorporation of values more related to the common good than to corporate interests. They cite privacy, accessibility, and the principles of equity, inclusiveness, and accountability as fundamental values to be introduced into platform ecosystems. Anti-concentration measures are also beginning to be suggested (Napoli, in Canada Department of Canadian Heritage and the Canadian Commission for UNESCO, 2019), acknowledging the difficulty of imposing this type of measure on companies with so many functionalities and services that operate globally.
Figure 6: Final value chain.
Amazon, for example, starts its sound information control and integration with its Echo loudspeaker and Alexa software, then this information circulates and is processed in several ways: its content delivery network AWS; submarine cables connecting it to its properties (Bay to Bay Express, JUPITER, and Hawaiki); the ad programmatic service Amazon DSP; the audiobooks and podcast platform Audible; its Amazon Music streaming service; and, of course, e-commerce and physical sales spaces.
Approaches to creating public and common infrastructures or platforms are outside any political debate in the European Union. Perhaps now is the time to consider the role of a public and common service that places information at the service of the citizenry and not only of large corporations in collusion with governments in power.
About the author
J. Ignacio Gallego is Professor of Audiovisual Communication at the Carlos III University of Madrid (Spain).
E-mail: juanignacio [dot] gallego [at] uc3m [dot] es
The author disclosed receipt of the following financial support for the research, authorship and/or publication of this article: In preparing this piece, the author gratefully acknowledges the support of the project ‘Audiovisual diversity and online platforms: Netflix as a case study’ (CSO2017-83539-R), financed by the State Research Agency (AEI) within the National RDI Program Aimed at the Challenges of Society of the Spanish Ministry of Science, Innovation and Universities and by the European Regional Development Fund of the European Union.
The author would like to thank First Monday’s referees for their valuable comments as well as the Chief Editor for his patience during these difficult times. Also I would like to thank the team of The Catalyst-FCAD at Ryerson University where I stayed as visiting scholar in 2019 researching for this publication. This visit was funded by the UC3M Own Research Program.
1. Spotify acquired Echo Nest (2014), Pandora — Next Bight Thing (2015), Apple — Semetric (2015), Nielsen — Gracenote (2017), Apple — Shazam (2018), and Warner Music — Sodatone (2018).
2. The lists of participants and company names are reserved to protect confidentiality.
4. These applications include Skype (owned by Microsoft), which accounts for 28.14 percent of download traffic, and WhatsApp (owned by Facebook), which accounts for 23.15 percent of total data circulating through messaging and VOIP applications.
5. It is worth highlighting the privatization processes carried out in the 1990s, such as that of Telefónica in Spain, which went from being completely publicly owned to being a public limited company. On the other hand, there are companies such as the Uruguayan Antel (100 percent public and with a monopoly on Internet access by cable in Uruguay) and the German Deutsche Telekom (of which the German state owns around 15 percent).
6. Spotify has offered deals on access to its service through Movistar in Mexico and Spain and with Vodafone in Ireland.
L. Albornoz, 2015. Power, media, culture: A critical view from the political economy of communication. London: Palgrave Macmillan.
doi: https://doi.org/10.1057/9781137540089, accessed 11 May 2021.
Amazon Alexa, 2019. “What are Alexa built-in devices?” (25 April), at https://developer.amazon.com/en-US/alexa/devices/alexa-built-in, accessed 11 May 2021.
I. Ang, 1991. Desperately seeking the audience. London: Routledge.
doi: https://doi.org/10.4324/9780203133347, accessed 11 May 2021.
A.H. Arsenault, 2017. “The datafication of media: Big data and the media industries,” International Journal of Media & Cultural Politics, volume 13, numbers 1–2, pp. 7–24.
doi: https://doi.org/10.1386/macp.13.1-2.7_1, accessed 11 May 2021.
A. Artigues, 2015. “We know what you did last Sónar: Realtime user tracking & Interactive data visualization,” Barcelona Supercomputing Center, at https://www.bsc.es/viz/whatyoudid/press/presspack_whatyoudid.pdf, accessed 11 May 2021.
A. Arvidsson and T. Bonini, 2015. “Valuing audience passions: From Smythe to Tarde,” European Journal of Cultural Studies, volume 18, number 2, pp. 158–173.
doi: https://doi.org/10.1177/1367549414563297, accessed 11 May 2021.
A. Arvidsson and E. Colleoni, 2012. “Value in informational capitalism and on the Internet,” Information Society, volume 28, number 3, pp. 135–150.
doi: https://doi.org/10.1080/01972243.2012.669449, accessed 11 May 2021.
d. boyd and K. Crawford, 2011. “Six provocations for big data,” version at http://softwarestudies.com/cultural_analytics/Six_Provocations_for_Big_Data.pdf, accessed 11 May 2021.
J. Braun, 2013. “Transparent intermediaries: Building the infrastructures of connected viewing,” In: J. Holt and K. Sanson (editors). Connected viewing: Selling, streaming, & sharing media in the digital age. New York: Routledge, pp. 124–143.
doi: https://doi.org/10.4324/9780203067994, accessed 11 May 2021.
P. Burkart and T. McCourt, 2004. “Infrastructure for the celestial jukebox,” Popular Music, volume 23, number 3, pp. 349–362.
doi: https://doi.org/10.1017/S0261143004000236, accessed 11 May 2021.
Canada Department of Canadian Heritage and the Canadian Commission for UNESCO, 2019. “Diversity of content in the digital age: discoverability of diverse local, regional and national content,” cited in “Report — International Meeting on Diversity of Content in the Digital Age,” at https://www.canada.ca/en/canadian-heritage/services/diversity-content-digital-age/international-engagement-strategy/report.html, accessed 11 May 2021.
S. Carney, 2011. “Don’t touch that radio dial — Arbitron is listening,” Los Angeles Times (24 August), at https://www.latimes.com/entertainment/la-xpm-2011-aug-24-la-et-radio-ratings-20110824-story.html, accessed 11 May 2021.
J. Cheney-Lippold, 2017. We are data: Algorithms and the making of our digital selves. New York: NYU Press.
K. Crawford and V. Joler, 2018. “Anatomy of an AI system: The Amazon Echo as an anatomical map of human labor, data and planetary resources,” at https://anatomyof.ai/, accessed 11 May 2021.
E. Curry, 2016. “The big data value chain: definitions, concepts, and theoretical approaches,” In: J.M. Cavanillas, E. Curry, and W. Wahlster (editors). New horizons for a data-driven economy: A roadmap for usage and exploitation of big data in Europe. Cham. Switzerland: Springer, pp. 29–37.
doi: https://doi.org/10.1007/978-3-319-21569-3_3, accessed 11 May 2021.
M. Day, G. Turner, and N. Drozdiak, 2019. “Amazon workers are listening to what you tell Alexa,” Bloomberg (10 April), at https://www.bloomberg.com/news/articles/2019-04-10/is-anyone-listening-to-you-on-alexa-a-global-team-reviews-audio, accessed 11 May 2021.
L. De Marchi, 2018. “Como os algoritmos do YouTube calculam valor? Uma análise da produção de valor para vídeos digitais de música através da lógica social de derivativo [How do YouTube algorithms calculate value? An analysis of the production of value for digital music videos using the social logic of the derivative],” Matrizes, volume 12, number 2, pp. 193–215.
doi: https://doi.org/10.11606/issn.1982-8160.v12i2p193-215, accessed 11 May 2021.
R.J. Deibert, 2013. Black code. Inside the battle for cyberspace. Toronto: Signal.
M. Denning, 2015. Noise uprising: The audiopolitics of a world musical revolution. London: Verso Books.
B. Dumaine, 2018. “It might get loud: Inside Silicon Valley’s battle to own voice tech,” Fortune (24 October), at http://fortune.com/longform/amazon-google-apple-voice-recognition/, accessed 11 May 2021.
R.M. Emerson, R.I. Fretz, and L.L. Shaw, 2001. “Participant observation and fieldnotes,” In: P. Atkinson, A. Coffey, S. Delamont, J. Lofland, and L. Lofland (editors). Handbook of ethnography. London: Sage, pp. 352–368.
doi: https://dx.doi.org/10.4135/9781848608337.n24, accessed 11 May 2021.
M. Eriksson, R. Fleischer, A. Johansson, P. Snickars, and P. Vonderau, 2019. Spotify teardown: Inside the black box of streaming music. Cambridge, Mass.: MIT Press.
doi: https://doi.org/10.7551/mitpress/10932.001.0001, accessed 11 May 2021.
European Commission, 2017. “Investment models,” at https://ec.europa.eu/digital-single-market/en/investment-models, accessed 11 May 2021.
European Commission Directorate-General for Education, Youth, Sport, and Culture, 2017. “Mapping the creative value chains: A study on the economy of culture in the digital age: Final report,” at https://publications.europa.eu/en/publication-detail/-/publication/4737f41d-45ac-11e7-aea8-01aa75ed71a1, accessed 11 May 2021.
A. Faße, U. Grote, and E. Winter, 2009. “Value chain analysis methodologies in the context of environment and trade research,” Hannover Economic Papers, Discussion Paper, number 429, at https://diskussionspapiere.wiwi.uni-hannover.de/pdf_bib/dp-429.pdf, accessed 11 May 2021.
D. Fernández Quijada, 2019. “FM is king, digital is queen,” Radio World (6 May), at https://www.radioworld.com/columns-and-views/digital-is-queen-fm-is-king, accessed 11 May 2021.
G.A. Fowler, 2019. “It’s the middle of the night. Do you know who your iPhone is talking to?” Washington Post (28 May), at https://www.washingtonpost.com/technology/2019/05/28/its-middle-night-do-you-know-who-your-iphone-is-talking/, accessed 11 May 2021.
C. Fuchs, 2010. “Labor in informational capitalism and on the Internet,” Information Society, volume 26, number 3, pp. 179–196.
doi: https://doi.org/10.1080/01972241003712215, accessed 11 May 2021.
P. Galuszka, 2015. “Music aggregators and intermediation of the digital music market,” International Journal of Communication, volume 9, pp. 254–273, and at https://ijoc.org/index.php/ijoc/article/view/3113, accessed 11 May 2021.
J. Gamba, M. Rashed, A. Razaghpanah, J. Tapiador, and N. Vallina-Rodriguez, 2020. “An analysis of pre-installed Android software,” 2020 IEEE Symposium on Security and Privacy.
doi: https://doi.org/10.1109/SP40000.2020.00013, accessed 11 May 2021.
A. Gazi and T. Bonini, 2018. “‘Haptically mediated’ radio listening and its commodification: The remediation of radio through digital mobile devices,” Journal of Radio & Audio Media, volume 25, number 1, pp. 109–125.
doi: https://doi.org/10.1080/19376529.2017.1377203, accessed 11 May 2021.
S. Gibbs, 2015. “Samsung’s voice-recording smart TVs breach privacy law, campaigners claim,” Guardian (27 February), at https://www.theguardian.com/technology/2015/feb/27/samsung-voice-recording-smart-tv-breach-privacy-law-campaigners-claim, accessed 11 May 2021.
R. Gingrich, 2017. “Wearables in the performing arts: An RFID primer for arts managers” (20 June), at https://amt-lab.org/blog/2017/5/wearables-in-the-performing-arts-an-rfid-primer-for-arts-managers, accessed 11 May 2021.
D.M. Greenberg, M. Kosinski, D.J. Stillwell, B.L. Monteiro, D.J. Levitin, and P.J. Rentfrow, 2016. “The song is you: Preferences for musical attribute dimensions reflect personality,” Social Psychological and Personality Science, volume 7, number 6, pp. 597–605.
doi: https://doi.org/10.1177/1948550616641473, accessed 11 May 2021.
Guardian music, 2019. “Warner Music signs first ever record deal with an algorithm,” Guardian (22 March), at https://www.theguardian.com/music/2019/mar/22/algorithm-endel-signs-warner-music-first-ever-record-deal, accessed 11 May 2021.
G. Held, 2010. A practical guide to content delivery networks. Second edition. Boca Raton, Fla.: CRC Press.
doi: https://doi.org/10.1201/EBK1439835883, accessed 11 May 2021.
T. Ingham, 2018. “The great big Spotify scam: Did a Bulgarian playlister swindle their way to a fortune on streaming service?” Music Business Worldwide (28 February), at https://www.musicbusinessworldwide.com/great-big-spotify-scam-bulgarian-playlister-swindle-way-fortune-streaming-service/, accessed 11 May 2021.
Inside Radio, 2019. “Testing, testing: Nielsen puts wearable PPM devices through their paces” (7 January), at http://www.insideradio.com/testing-testing-nielsen-puts-wearable-ppm-devices-through-their-paces/article_a4f3598c-124d-11e9-b9af-f7fe9e78d3e9.html, accessed 11 May 2021.
Instreamatic, 2019. “The voice era is here. No doubt about it,” Medium (26 March), at https://medium.com/@instreamatic/the-voice-era-is-here-no-doubts-about-it-283eebc084b3, accessed 11 May 2021.
R. Kaplinsky and M. Morris, 2000. “A handbook for value chain research,” at http://asiandrivers.open.ac.uk/documents/Value_chain_Handbook_RKMM_Nov_2001.pdf, accessed 11 May 2021.
A. Kassabian, 2013. Ubiquitous listening: Affect, attention, and distributed subjectivity. Berkeley: University of California Press.
doi: https://doi.org/10.1525/california/9780520275157.001.0001, accessed 11 May 2021.
A. Konrad, 2016. “Spotify moving onto Google Cloud is a big win For Google over Amazon and Microsoft,” Forbes (23 February), at https://www.forbes.com/sites/alexkonrad/2016/02/23/spotify-is-a-big-win-for-google-cloud/#fcb5eff74b9a, accessed 11 May 2021.
G. Kreitz and F. Niemelä, 2011. “Spotify — Large scale, low latency, P2P music-on-demand streaming,” 2010 IEEE Tenth International Conference on Peer-to-Peer Computing.
doi: https://doi.org/10.1109/P2P.2010.5569963, accessed 11 May 2021.
M. Lahey, 2016. “Invisible actors: Web application programming interfaces, television, and social media,” Convergence, volume 22, number 4, pp. 426–439.
doi: https://doi.org/10.1177/1354856516641915, accessed 11 May 2021.
G. Langlois and G. Elmer, 2019. “Impersonal subjectivation from platforms to infrastructures,” Media, Culture & Society, volume 41, number 2, pp. 236–251.
doi: https://doi.org/10.1177/0163443718818374, accessed 11 May 2021.
U.A. Mejias and N. Couldry, 2019. “Datafication,” Internet Policy Review, volume 8, number 4.
doi: https://doi.org/10.14763/2019.4.1428, accessed 11 May 2021.
J.C. Miguel De Bustos, 2016. “Big data y Big GAFA. Reflexiones sobre la economía de los datos [Big data and Big GAFA. Thoughts on the data economy],” Economia della Cultura, volume 26, number 4, pp. 507–525, and at https://www.rivisteweb.it/doi/10.1446/85783, accessed 11 May 2021.
H.G. Miller and P. Mork, 2013. “From data to decisions: A value chain for big data,” IT Professional, volume 15, number 1, pp/ 57–59.
doi: https://doi.org/10.1109/MITP.2013.11, accessed 11 May 2021.
J.W. Morris, 2015a. Selling digital music, formatting culture. Berkeley: University of California Press.
J.W. Morris, 2015b. “Curation by code: Infomediaries and the data mining of taste,” European Journal of Cultural Studies, volume 18, numbers 4–5, pp. 446–463.
doi: https://doi.org/10.1177/1367549415577387, accessed 11 May 2021.
V. Mosco, 2017. Becoming digital: Towards a post-Internet society. Bingley: Emerald Publishing.
V. Mosco, 2014. To the cloud: Big data in a turbulent world. London: Routledge.
National Public Radio (NPR) and Edison Research, 2018. “The smart audio report,” at https://www.nationalpublicmedia.com/uploads/2020/04/Smart-Audio-Report-Winter-2018-.pdf, accessed 11 May 2021.
D.B. Nieborg and T. Poell, 2018. “The platformization of cultural production: Theorizing the contingent cultural commodity,” New Media & Society, volume 20, number 11, pp. 4,275–4,292.
doi: https://doi.org/10.1177/1461444818769694, accessed 11 May 2021.
H. Olmedo and J. Augusto, 2013. “Towards the commodification of augmented reality: Tools and platforms,” In: V.M.R. Penichet, A. Peñalver, and J.A. Gallud (editors). New trends in interaction, virtual reality and modeling. London: Springer, pp. 63–72.
doi: https://doi.org/10.1007/978-1-4471-5445-7_5, accessed 11 May 2021.
Organisation for Economic Co-operation and Development (OECD), 2016. “Big data: bringing competition policy to the digital era,” at https://www.oecd.org/competition/big-data-bringing-competition-policy-to-the-digital-era.htm, accessed 11 May 2021.
D. Osimo, L. Pujol Priego, T., Pekari, and A. Sirppiniemi, 2019. A symphony, not a solo: How collective management organisations can embrace innovations and drive data sharing in the music industry. Helsinki: Teosto, at https://www.teosto.fi/app/uploads/2020/10/27134714/a-symphony-not-a-solo-policy-brief-final-09012019.pdf, accessed 11 May 2021.
K. Paul, 2019. “Google workers can listen to what people say to its AI home devices,” Guardian (11 July), at https://www.theguardian.com/technology/2019/jul/11/google-home-assistant-listen-recordings-users-privacy, accessed 11 May 2021.
M. Pérez García, S. Saffon Lopez, and H. Donis, 2018. “Voice activated virtual assistants personality perceptions and desires: Comparing personality evaluation frameworks,” Proceedings of the 32nd International BCS Human Computer Interaction Conference.
doi: http://dx.doi.org/10.14236/ewic/HCI2018.40, accessed 11 May 2021.
J.-C. Plantin and A. Punathambekar, 2018. “Digital media infrastructures: Pipes, platforms, and politics,” Media, Culture & Society, volume 41, number 2, pp. 163–174.
doi: https://doi.org/10.1177/0163443718818376, accessed 11 May 2021.
M. Porter, 2001. “The value chain and competitive advantage,” In: D. Barnes (editor). Understanding business: Processes. London: Routledge, pp. 50–66.
R. Prey, 2016. “Musica analytica: The datafication of listening,” In: R. Nowak and A. Whelan (editors). Networked music cultures: Contemporary approaches, emerging issues. London: Palgrave Macmillan, pp. 31–48.
doi: https://doi.org/10.1057/978-1-137-58290-4_3, accessed 11 May 2021.
Radio Joint Audience Research Limited (RAJAR), 2015. “Audio time: What the RAJAR Midas Audio survey tells us about listening in the digital age,” at https://www.rajar.co.uk/docs/news/Audio_Time%20_FINAL.pdf, accessed 11 May 2021.
Radio Joint Audience Research Limited (RAJAR), 2012. “Making sense of the differences between audio streaming and RAJAR listeners,” at https://www.rajar.co.uk/docs/news/Diffs_btwn_audio_stream_and_RAJAR_listeners.pdf, accessed 11 May 2021.
C. Sandvig, 2015. “The Internet as the anti-television: Distribution infrastructure as culture and power,” In: L. Parks and N. Starosielski (editors). Signal traffic: Critical studies of media infrastructures. Champaign: University of Illinois Press, pp. 225–245.
doi: https://doi.org/10.5406/illinois/9780252039362.003.0010, accessed 11 May 2021.
Sandvine, 2018. “The global Internet phenomena report” (October), at https://www.sandvine.com/hubfs/downloads/phenomena/2018-phenomena-report.pdf, accessed 11 May 2021.
A. Schrock, 2017. “What communication can contribute to data studies: Three lenses on communication and data,” International Journal of Communication, volume 11, pp. 701–709, at https://ijoc.org/index.php/ijoc/article/view/5798, accessed 11 May 2021.
D. Smythe, 2001. “On the audience commodity and its work,” In: M.G. Durham and D.M. Kellner (editors). Media and cultural studies: Keyworks. Oxford: Blackwell, pp. 253–279.
N. Srnicek, 2016. Platform capitalism. Cambridge: Polity Press.
F. Stiernstedt, 2008. “Maximizing the power of entertainment: The audience commodity in contemporary radio,” Radio Journal, volume 6, numbers 2–3, pp. 113–127.
doi: https://doi.org/10.1386/rajo.6.2-3.113/1, accessed 11 May 2021.
C. Stokel-Walker, 2018. “Voice messaging — Conversational gain or pain?” Guardian (2 December), at https://www.theguardian.com/technology/2018/dec/02/five-reasons-why-voice-messaging-is-the-next-big-thing, accessed 11 May 2021.
T. Terranova, 2000. “Free labor: Producing culture for the digital economy,” Social Text, volume 18, number 2, pp. 33–58.
doi: https://doi.org/10.1215/01642472-18-2_63-33, accessed 11 May 2021.
J. van Dijck, T. Poell, and M. de Waal, 2018. The platform society: Public values in a connective world. Oxford: Oxford University Press.
doi: https://doi.org/10.1093/oso/9780190889760.001.0001, accessed 11 May 2021.
P. Vonderau, 2019. “The Spotify effect: Digital distribution and financial growth,” Television & New Media, volume 20, number 1, pp. 3–19.
doi: https://doi.org/10.1177/1527476417741200, accessed 11 May 2021.
G. Yúdice, 2017. “La diversidad musical en la nube,” In: L.A. Albornoz and T. García Leiva, (editors). El audiovisual en la era digital: La diversidad musical en la nube. Madrid: Ediciones Cátedra.
S. Zuboff, 2015. “Big other: surveillance capitalism and the prospects of an information civilization,” Journal of Information Technology, volume 30, number 1, pp. 75–89.
doi: http://dx.doi.org/10.1057/jit.2015.5, accessed 11 May 2021.
Received 7 October 2019; revised 5 May 2021; accepted 6 May 2021.
This paper is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
The value of sound: Datafication of the sound industries in the age of surveillance and platform capitalism
by J. Ignacio Gallego.
First Monday, Volume 26, Number 6 - 7 June 2021