Day One at Amazon: An In-Depth Analysis
of the Platform’s Power
Master’s Thesis New Media & Digital Culture
By Victor Frederick Camiel Bouwmeester
Supervisor: dr. N.A.J.M. (Niels) van Doorn Second reader: D. (David) Gauthier MSc
University of Amsterdam Date: 23-06-2020
Table of Contents
Abstract ... 1
1. Introduction ... 2
2. Methodology: ... 7
3. Theoretical Analysis ... 9
3.1 Day One ... 9
3.2 It is All About the Long Term ... 18
3.3 Obsess Over Customers ... 24
3.4 Infrastructure ... 28
3.5 Our Employees ... 33
4. Discussion ... 38
5. Conclusion ... 49
1
Abstract
This thesis sets out to undergo an investigation into Amazon’s growing platform power, using a critical theoretical analysis which is firmly rooted in the theories of critical political
economy and platform studies. A document analysis using a letter from Amazon’s founder and CEO Jeff Bezos to his shareholders, sent at the height of the global coronavirus
pandemic, is used as the primary document to form the narrative structure based on the themes addressed in the letter. The first part of this thesis aims to establish an understanding of how Amazon has been able to increase its platform power over time by leveraging several core features of the platform business model, including: the crucial role of data; the
programmability of the platform; utilizing the legal regime and the degraded role of labor; performing the structure of financialization.
The second part attempts to bring forward some of the implications of Amazon’s increased platform power, by looking at the impacts of the company’s market dominance on three themes which were extracted from the letter to shareholders: its customers,
infrastructure and employees. The themes served as a rhetorical tool to highlight that the implications of Amazon’s platform power often extend beyond what is immediately clear. This analytical model provides potential for future studies to build on, critically interrogating the platform power of monolithic firms in the digital economy.
The final part of this thesis forms the starting point for a broader discussion that I believe society should engage in, aimed at understanding what options are available to us to curb the growing platform power of companies such as Amazon. Amazon is not impervious to sustained pressure and criticism from society, which has already led to a number of profound changes within the company, such as its commitment to a $15 minimum wage and the Climate Pledge. Despite that, I would argue that there is still a role for governments in curbing platform power, both through newly defined antitrust regulations as well as by solving the problems that lead to and resulted from financialization.
Keywords: Amazon, Platform Capitalism, Platform Power, Data, Financialization, Labor, Digital Infrastructure, Antitrust
2
1. Introduction
As the COVID-19 crisis became a worldwide phenomenon, with fewer than 2000 daily new cases on March 1st and close to 85.000 on May 1st, governments around the world took action to limit the spread of the virus and lighten the pressure on healthcare professionals. Numbers show that at the time of writing in early May, some 3.5 million people have been infected, and 243.000 deaths have been recorded as a result of the COVID-19 pandemic (WHO, 2020). Public institutions such as schools and universities closed, and all non-essential workers were ordered to work from home. As public life came to a grinding halt and economies shut down, the impact of the virus on daily life became clear to everyone, as millions of workers lost their jobs. Economists predict a record high unemployment rate of between sixteen and twenty-one percent in the United States for April, up from the 3.5% half-century low of just two months prior (Davidson, 2020). However, not everyone is affected equally by the virus, as consumer demands change, individuals work from home and social gatherings are prohibited.
On the one hand, major platform businesses (e.g. Amazon) are among the biggest profiters of the coronavirus crisis, as their role as crucial intermediary highlights their usefulness and society’s dependence on them. On the other hand, as society adapts its professional and private life in the pandemic, there are a number of specific platform types (e.g. Uber & Airbnb) that are struggling immensely. With public life coming to an abrupt halt, the demand for platform services in the ride- and homesharing sectors all but collapsed, causing mass firings and a sharp decline in revenue (Airbnb, 2020). Due to social isolation, people are spending more of their time in their homes and online than ever and many of their interactions are made possible by ‘Big Tech’ companies. Social media usage is seeing a sharp climb, as well as video streaming services such as Netflix, Twitch and YouTube, who have seen web traffic go up by more than fifteen percent (Koeze & Popper, 2020). As people search for new ways to connect to others, be it their friends and family or their boss, video chatting applications are becoming increasingly popular, with millions of downloads worldwide (Koeze & Popper, 2020). Additionally, companies like Amazon saw a sudden, sharp rise in demand because more products are being ordered online, as people adhere to social distancing measures and invest in things such as new hobbies, DIY projects or home offices. Moreover, cloud services (e.g. AWS, Google Cloud) are in high demand, as businesses and schools are forced to move to teaching online (Wakabayashi, Nicas, Lohr & Isaac, 2020).
These trends are reflected in the companies’ combined $4.5 trillion market capitalization which, beyond the initial pandemic-induced market instability, has remained
3
largely resilient to the economic damage of the coronavirus, considering the nature and variety of their services. Moreover, all of the ‘Big Five’ (Alphabet Google, Amazon, Facebook, Apple and Microsoft) published higher revenues numbers for the first quarter of 2020 than the prior year, which could be an indication of the limited impact of the virus on their business model (ycharts, 2020). In addition, their substantial cash reserves will prove helpful in weathering an economic recession.
The coronavirus crisis is arguably the single most significant crisis since the Great Depression, creating extensive threats to society concerning healthcare, economic stability and overall welfare (Gopinath, 2020). It has furthermore made it clear that some platform businesses (e.g. the ‘Big Five’) stand to gain a lot from the current crisis, whilst others (e.g. Uber, Airbnb) are struggling amidst a radically changing society. As such, the pandemic can be seen as an unexpected gauge for platform power and as a trigger to reevaluate the grip these businesses have on a multitude of levels, ranging from the individual to society at large. In this regard, a letter that Amazon CEO Jeff Bezos wrote to the company’s shareholders in response to the coronavirus pandemic and its effects on the corporation can be viewed as an example of the scale and scope of a single platform’s power (Bezos, 2020). Moreover, because the letter includes a copy of Amazon’s very first letter to its shareholders dating back to 1997, it allows this work to take a comparative perspective between then and now, structured around themes central in letter from 1997.
Bezos published the letter as a blog on the aboutamazon.com webpage on April 16th,
2020 and in it, Bezos highlights Amazon’s role as a critical intermediary by noting how import the business is for its customers, ranging from the consumer e-commerce side to its Whole Foods Market grocery division and its Amazon Web Services (AWS) branch. The latter being particularly useful for outlining the sheer size and influence of Amazon’s operations around the world, as a segment of the letter gives us an example of how embedded Amazon is in daily operations throughout society. According to Bezos (2020), as a supplier of cloud computing services, AWS provides services to a variety of public and private organizations during the COVID-19 pandemic: Hospital networks, pharmaceutical companies and research labs; academic institutions; the World Health Organization; numerous governments; the New York City COVID-19 Rapid Response Coalition; the Los Angeles Unified School District; the Centre for Disease Control; the National Health Service in the United Kingdom; OTN (one of the world’s largest virtual care networks) in Canada; the São Paulo State Government. However, in Bezos’ (2019) own words, these are just some of “AWS’s millions of customers
4
rang[ing] from startups to large enterprises, government entities to nonprofits” (para. 2). Customers include companies such as Netflix and Spotify, but also the CIA and the Department of Defense (AWS, n.d.).
According to the Synergy Research Group, the web services wing of Amazon is currently market leader in the approximately 100$ billion market for cloud computing, controlling a roughly 33% share, more than its three largest competitors (Microsoft Azure, Google Cloud, IBM Cloud) combined (Richter, 2020). This is significant because it raises the question whether a single company should control such a substantial part of the market and what the implications are, especially if one considers AWS’ infrastructural role on the web, facilitating a wide variety of customers. However, AWS’ grip on the cloud computing market is not a solitary instance of the scale of Amazon’s platform power—rather, I argue that it is symptomatic of Amazon as a whole, regarding its business practices, how it operates and with what goal. This will be demonstrated by taking into account the philosophy of Amazon that Bezos presents in the first letter to shareholders from 1997, which will form the core narrative structure for this work.
By all accounts, the scale of Amazon’s growth over the past two decades has been extraordinary, showing exponential growth in the company’s: market capitalization, stock price, revenue, cash flow, number of employees (Macrotrends, n.d.). In addition, the business has widened the scope of its operations through a significant number of mergers and acquisitions in the fields of robotics, cultural production, eHealth, Internet of things and grocery stores (Moynihan & Payo, 2019). Consequently, as Nobel prize winning economist Joseph Stiglitz (2017) argues, if we allow unrestricted growth by platform businesses to continue, the economy of the future will increasingly be controlled by a handful of monolithic companies—reminiscent of the trusts of the American gilded age—in a way that is inconsonant with our democratic ideals.
Critical media theorists (Srnicek, 2016) and software and platform scholars (Bechmann, 2013) have pointed out the inherent inequalities that form the political economy of digital platforms, and other more popular critics (e.g. Tim Wu, Laurence Lessig, Evgeny Morozov) have furthermore been critical of how internet platforms undermine public health, democracy, privacy and innovation in unprecedented ways. I attempt to contribute to these critiques by offering a critical theoretical analysis of Amazon’s increasing platform power. Despite the growing sentiment in recent years in favor of regulation and reform, the current pandemic and Bezos’ letter to shareholders are testament that change is improbable when it comes to digital platforms and their growing power, as their worth to society is reinforced during this time of
5
global crisis. In this regard, I would argue that as a democracy we should ask ourselves the question whether it is justifiable for a single corporation to be so extensively rooted in many aspects of society and essentially ‘winning capitalism’. The effects of the coronavirus pandemic have forced us to radically rethink our priorities as an individual as well as collectively, on how we work, socialize and give meaning to life. We are at a crossroads moving forwards, do we continue on the path we were going, or choose a different future. This thesis is meant as a starting point for such a discussion by answering the following research question:
(1) How has Amazon been able to increase its platform power over time, (2) what are the implications of that increased power and (3) what options are available to curb it?
In order to critically assess Amazon’s platform evolution and power and its increasingly dominant position in the digital economy and society at large, I draw on academic work in critical political economy and platform studies, with insights from business- and legal studies. By building on these approaches, this research offers the conceptual tools to critically evaluate the historical context and economic conditions in which Amazon could emerge, the material-technical elements which facilitate its growth and the impacts of its business practices and monopolistic tendencies on the socio-technical and political-economic levels. With that context in mind, the latter part of this work attempts to initiate a broader discussion, exploring options aimed at curbing platform power.
The contribution of this work to existing literature is to provide a conceptual critical theoretical analytical framework through which future case studies can be mapped, allowing for a more succinct understanding of a single platform’s power, especially in the political economic sense. Furthermore, in a context of (new) media studies and digital culture this work serves as a critical perspective on the transforming society of the past decades, in which numerous aspects of our daily lives are increasingly controlled by fewer large corporations. This work aims to bridge a gap in existing research in a number of ways: (1) it takes a distinct historical perspective in explaining the political economic conditions that ultimately lead to the emergence of the platform business model; (2) it explains the inner workings of digital platforms and pays particular attention to the consequences of those inherent platform features; (3) it problematizes the growing power of platforms by zooming in on a single platform and— through a document analysis—using examples of Amazon’s business practices to get a grip on the theory on platforms and vice versa; (4) it explores and offers concrete legal and regulatory options to restrain platform power. The research gap is bridged not necessarily by one of these
6
individual elements, but rather because it combines the four elements into a single, comprehensive approach to understanding and dealing with platform power.
The theoretical framework is grounded in critical economic theory by taking a normative and historical perspective on ‘platform capitalism’ (Srnicek, 2016). Understanding the economic conditions that allowed platform businesses to emerge, as well as the financialization and business model are essential to get a grasp on the contemporary platform economy (Langley & Leyshon, 2017; Rahman & Thelen, 2019). Moreover, a platform capitalism perspective places business economics firmly in a societal context where power relations are rooted in issues surrounding ownership, exclusion, differential inclusion, exploitation and expropriation. In this regard, it is more critical of state of affairs than regular business studies, which will nonetheless be useful to describe how platforms function as multi-sided markets, how ‘network effects’ fuel platform competition and scalability and how ‘platform envelopment’ can leverage entry into and bundling between markets (Rochet & Tirole, 2003; Eisenmann, Parker & Van Alstyne, 2011; Kenney & Zysman, 2016; Cusumano, Gawer & Yoffie, 2019). Business literature is drawn on to highlight some of the driving forces of growth in a platform business, and by extension find out what makes the business model so appealing and a justification for starting a platform company. The overarching approach for this research is firmly rooted in what Bogost & Montfort (2009) coined as ‘platform studies’, precisely because it combines approaches rooted in critical political economy, computer science, software studies and cultural studies, which are necessary to undergo a critical investigation into processes of ‘platformization’ and the power shifts these processes bring about (Helmond, 2015).
Platform scholars have been concerned with examining “the technological affordances of platforms in relation to their political, economic and social interests”, their programmability through APIs, their business model based on ‘datafication’ and ‘commodification’, and the platform’s roots in the concept of Web 2.0 (Helmond, 2015; van Dijck, Poell & de Waal, 2018; Gehl, 2010). Other work has looked at the role of platforms in shaping ‘networked sociality’ or the ‘platform society’, which highlight the influence of platforms in shaping our everyday life (van Dijck, 2013; van Dijck, Poell & de Waal, 2018). Most of this research take social media or advertising platforms such as Facebook and Google as focal points for their analysis, arguably because the impact of those platforms on society is most apparent. However, I would argue that, taking into account the increasingly integrated nature of the platform ecosystem, other types of platforms such as Amazon, with its significant infrastructural connections, deserve more attention. Platform ecosystems in this regard refers to viewing platforms in
7
relation to each other, across markets and across societal levels—taking into account “ownership relations in terms of power over data flows as well as technical and organizational control over the ecosystem as a whole” (van Dijck, 2013; van Dijck, Nieborg & Poell, 2019, p8). In that sense, as van Dijck, Nieborg & Poell (2019) argue, we should “reframe platform power” to broaden our understanding of how technology businesses operate in digital ecosystems. Moreover, Nieborg & Helmond (2019) suggest “to move away from analyzing platforms as either (web)sites or (individual) apps and instead to consider platforms as data infrastructures that own and operate a variety of platform instances that each perform particular data work” (p.211). Consequently, rather than understanding platform power through a single, particular lens, this research’s multi-disciplinary approach has a broader focus that considers a platform’s infrastructural expansion/integration into society as well as the wider platform ecosystem.
2. Methodology:
This thesis attempts to assess Amazon’s platform power by conducting a critical theoretical analysis that is structured around several key themes that are central to Amazon’s philosophy, as put forward in the company’s first letter to shareholders. In this regard, I am using theory on the history and features of platforms in order to immediately get a grip on the history and business practices of Amazon, and vice versa. The analytical framework is structured around several themes which are extracted from the letter Amazon founder and CEO Jeff Bezos (2020) wrote to the shareholders of the company. As such, the letter forms the primary document—or starting point for the analysis but will be supplemented by other types of documents throughout that address the themes from the letter in different ways. Bezos’ (2020) letter to shareholders is made up of a first part that addresses the business’ current practices during the coronavirus crisis, and a second part that consists of the first letter to shareholders the company published from 1997. This duality is useful because it allows for a comparative perspective between the early history of the company and its current position of critical intermediary in society. Moreover, the foundational elements that formed Amazon’s business philosophy in 1997 will provide the narrative structure for this work.
A document analysis is a form of qualitative research that allows researchers to elicit meaning, gain understanding, and develop empirical knowledge by reviewing or evaluating documents (Bowen, 2009; Corbin & Strauss, 2008). The rationale for using this type of analysis is that it is particularly useful for this case study of Amazon, to produce a rich description of a
8
single organization (Bowen, 2009). Moreover, by combining it with a critical political economic approach, it allows me to make normative claims about Amazon’s business practices and monopolistic behavior. This research borrows from the methodological apparatus put forward by Nieborg & Helmond’s (2019) critical evaluation of Facebook Messenger as a ‘platform instance’, who performed a document analysis of Facebook quarterly investor calls, to develop a better understanding of Facebook’s business strategy as a public company. Consequently, it is useful for this analysis of Amazon because, as they argue, their framework “allows for a critical inquiry into the dynamics of platform evolution as it considers a platform’s technical and economic dimensions in tandem” (Nieborg & Helmond, 2019, p.211).
In order to undergo a critical political economic interrogation of Amazon using Bezos’ letter to shareholders, several themes will be extracted from the letter that form the core narrative structure both for the original 1997 letter, as well as this research. By using the same themes as the letter, I aim to critically analyze Amazon’s platform power from the inside, using the identical topics to compare and contrast, using the company’s core philosophy as an argumentative structure for this critique. Moreover, the themes identified could serve as a starting point for further research that aims to critically evaluate Amazon or other major platforms. The first theme is mentioned at the start of the original letter: “this is Day 1 for the Internet and, if we execute well, for Amazon.com” (Bezos, 1998). It is also featured at the end of the letter from 2020: “Even in these circumstances, it remains Day 1”, and appears to be a creed that is central to the way Bezos (2020) thinks about and sees Amazon. It does not act as a specific focus for either letter, but rather as a philosophy for the entire business. Bezos (2017) says he has “been reminding people that it’s Day 1 for a couple of decades” (para.2) because, he argues, ‘Day 2’ is the first step into decline and ultimately death. Amazon’s goal appears to treat every day as Day 1 and keep Day 2 away for as long as possible. However, for this research Bezos’ meaning of the ‘Day 1’ concept is less relevant, rather I would like to see it as a starting point for my investigation into Amazon’s platform power. The rest of the themes are formulated as headings in the company’s 1997 letter to shareholders and will form the structure of this work accordingly:
• Day One
• It is All About the Long Term • Obsess Over Customers • Infrastructure
9
Despite the usefulness of this type of analysis for my research, there are some limitations that are important to address. Firstly, the analysis uses a primary document for structure and builds upon that with a variety of other documents ranging from academic literature to empirical accounts. Although Bowen (2009) argues that the quality of the document is more vital than the quantity of the documents, the limitations of using a single primary document should be acknowledged. However, because of the extraordinary nature of the situation created by the coronavirus crisis, and the mounting criticism on Amazon as a result, the letter poses a unique and unexpected starting point for an examination into Amazon’s platform power. This idea is reinforced by taking a comparative perspective between ‘then’ (1997) and now, using the themes mentioned above to make claims about how Amazon’s platform power has increased.
A second limitation is the issue of bias, both from the creator of the document, and the researcher (O’Leary, 2014). The original purpose with which the document was created must be evaluated, in addition to its target audience (Bowen, 2009). I would argue that Bezos’ letter to his shareholders acts formally as a newsletter, informing the company’s shareholders how Amazon is responding to the coronavirus crisis. However, when examined within the context of the mounting criticism on the company and the knowledge that the letter will not exclusively be read by the shareholders but by society at large, other—less overt purposes for its release could be argued. This is in line with another limitation of document analysis Bowen (2009) mentions: ‘biased selectivity’ which, in an organizational context, means that the selected documents “are likely to be aligned with corporate policies and procedures and with the agenda of the organization’s principals” (para.5). Nevertheless, considering the approach of this work, I would argue that rather than seeing these limitations as such, they also justify the use of the letter as a means for analysis. This is because, after distilling the key themes from the letter, we can contrast the statements presented by Bezos with critical empirical testimonials from other documents, including academic, congressional inquiries, (former) employees’ testimonies and the press.
3. Theoretical Analysis
3.1 Day One
“Day 2 is stasis. Followed by irrelevance. Followed by excruciating, painful decline. Followed by death. And that is why it is always Day 1.” – Jeff Bezos
10
In order to start an investigation into Amazon and its platform power, it is necessary to go back in time by a few decades, to the socio-technical-economic shift towards platformization. Starting from the early years of the World Wide Web in the 1990s, transitioning into Web 2.0 at the start of the new millennium and the history of Amazon in the context of that shift. This period in time is what I would dub ‘Day One’ and in contrast to the metaphorical business philosophy it represents for Amazon, this work’s application relates to the more literal sense of the words. Day One refers to the start and growth of Amazon as an online bookstore, but furthermore places that start within the context of a technical history of the platformization of the web including the emergence of Web 2.0 and the role of APIs and data in facilitating the growth of Amazon’s platform power. The political economic conditions that enabled Amazon’s platform business model to thrive will be explored in the next part, after an understanding of what constitutes a platform have been established.
Before we can understand how platforms became the dominant economic and infrastructural model of the web in the past decade, a brief historical context on the development of the platform as a business model and the recent Internet is required. Even though the Internet as a worldwide connection of networks is some 60 years old, it has undergone significant developments in its short history. Although one must acknowledge the crucial role of technology such as ARPANET and the Internet protocol suite TCP/IP in the development of the early Internet during the period 1960-1990, they are beyond the scope of this research.
The ‘World Wide Web’ as we now know it was invented in 1991 by British engineer and computer scientist Tim Berners-Lee, when he managed to introduce hypertext technology to the Internet (van Dijck, 2013). In its first decade when this new form of network communication was still in its infancy it allowed users to take parts of their offline lives and organize it online via email, weblogs and by consuming information that was published online. These were one-way interactions from the user, meaning they would have to find and connect to various nodes on the web, considering one service would not automatically connect you to others (van Dijck, 2013). The webpages of the early 1990s were largely static, running on basic HTML code and containing hypertext with hyperlinks that would redirect users to other resources. By the end of 1994 only several thousand webpages existed on the web from around the world, most of them personal pages or connected to research institutes (Armstrong, 2019). In 1995, commerce on the web or, E-commerce began emerging with the founding of the websites such as eBay and Amazon.
11
Before Jeff Bezos started Amazon, he was a hedge fund manager at D.E. Shaw & Co. on Wall Street, a company which used computers and mathematical formulas to identify and exploit patterns in the global financial market (Stone, 2013). As a financial analyst, he was able to see the opportunity of the upcoming Internet before most others did and as such was tasked by the company with finding a way to exploit it. Among the first ideas discussed between Bezos and the company’s CEO Shaw was ‘the everything store’: “an Internet company that served as an intermediary between customers and manufacturers and sold nearly every type of product, all over the world” (Stone, 2013, chapter 1). In his research on the potential of the Internet, Bezos came across a collection of performance data about the entire World Wide Web in 1993, published by author John Quarterman as a newsletter (Stone, 2013). The data showed that the WWW’s simple, user-friendly interface was appealing to a far broader audience than other Internet technologies had done until then (Stone, 2013). Based on the data, Bezos extrapolated that between January 1993 and January 1994, overall Web activity (e.g. number of bytes and packets sent) had increased by a factor of 2300, or 230.000 percent (Stone, 2013). This enormous, unusual growth prompted Bezos to think of a business plan that would make sense in the context of that growth, he settled on books as the best option, as pure commodities, identical from store to store. The goal was to create a bookstore with a near infinite selection of books, more than any brick and mortar store could stock.
After considering names such as Cadabra Inc., Awake.com and Relentless.com, the site went live on July 16, 1995 as Amazon.com, as a reference to earth’s largest river. The business was run from Seattle, because of its reputation as a technology hub (Microsoft’s headquarters was located in the nearby Seattle suburb Redmond) but more so because of the relatively small population of the state. This is significant because it meant that Amazon would only have to charge sales tax on a relatively small number of its customers, considering federal law dictates that retailers do not have to collect sales tax in states where they have no physical operations (e.g. California and New York) (Milchen, 2011). This meant that Amazon could charge lower prices than brick and mortar retailers, attracting more customers and not lose any profits in the process. Within the first month of the site being visible to all Web users, it became clear that the company’s formula was successful, selling books to customers in all fifty U.S. states and in forty-five countries worldwide. Roughly half a year later revenue was growing between 30 and 40 percent each month and the company had to find a larger office, hire more employees and upgrade its servers and software several times in its early history to keep up with demand.
The company used user-generated reviews to attract customers and hired editors to make the site a more authoritative source of information on books by trying to replicate the
12
environment of a real bookstore. A controversial feature called ‘1-Click ordering’ was implemented, allowing customers to purchase an item in one click, using their stored personal and payment information to complete the transaction by Amazon (Richtel, 2000). The idea was that by removing friction from online shopping, customers changed their behavior and would spend more on the site. A year its inception, the company made the first steps towards what would go on to become the seminal model for internet businesses: a platform. It introduced the Amazon Associates Program, which allowed other websites to link to Amazon and collect a commission when those referrals made a purchase (Stone, 2013). Although Amazon was not the first to use affiliate marketing, it was the most prominent, and it helped the business to extend its reach across the Web early on, entrenching itself ahead of the competition. As a way to make the website more personal to each unique visitor, data about a customer’s previous purchases was used to make recommendations for other books that match their taste but would not otherwise have found. In this regard, Amazon could understand its customers in a truly individualized way using their data, giving E-commerce a distinct advantage over traditional retailers (Stone, 2013). Seen in this way, Amazon was one of the first to understand and utilize data as a resource, forming the foundation for the platform business model (Langley & Leyshon, 2017).
Data is a defining and perhaps the single most important feature of what makes platform. It has been of central importance in transforming the digital platform into a viable and the most dominant business model in developed economies. It is the raw material of the platform economy, which fuels the growth of the companies within it. In his book Platform
Capitalism Srnicek (2016) argues that as a result of the decline in manufacturing profitability
during the twentieth century, capitalism of the twenty-first century turned to data in order to maintain economic growth. Leaps in technological progress made data extraction increasingly cheap and the Web 2.0-inspired jump towards digital communication made it exceedingly simple (Srnicek, 2016).
Massive amounts of data became available and new industries arose to make use of them. Data can be any type of information used by a computer, ranging from text, image or audio to personal information such as name, age and gender (van Dijck, 2013). Furthermore, every action of a user made on the web generates data which is automatically collected by the owner of the site, or through the use of APIs. This process of capturing data out of previously unquantified user interactions on the web—and increasingly, outside of it—is what Mayer-Schönberger & Cukier (2013) refer to as ‘datafication’ and it is a central part of what constitutes a platform.
13
Movement of the mouse and clicks, but also geolocations, bookmarks, liking, watching, retweeting, rating, posting, searching, paying and browsing behavior are all being recorded and analyzed (van Dijck, Poell, de Waal, 2018). In this regard, user activities form the natural source of the raw material data (Srnicek, 2016). Through our extensive use of the internet, we allow detailed personal information about our interests and preferences to be collected. However, the name is misleading because to view ‘raw data’ in the same way as crude oil or ore, naturally occurring in the world waiting to be found and extracted, is a mistake (Sadowski, 2019). In the words of Gitelman (2013): “Raw data is an oxymoron”; considering data is never unstructured or unprocessed to begin with. More accurately, data is already shaped by the gathering and recording mechanisms of a platform (2013).
Rather than merely collecting data, these platforms employ ways to ‘trigger’ and ‘mold’ particular (inter)actions of users (Kitchin, 2014). By continuously collecting and processing data from these interactions it enables them to connect users to services and advertisements more effectively. As such, the complex underlying technological features of these platforms play an important role in shaping how users interact and what kind of data can be extracted from those interactions as a result (van Dijck, Poell, de Waal, 2018). This raises questions about the nature of data and supports the idea that data should generally not be viewed as a commodity but rather as a form of capital (Sadowski, 2019).
Depending on the way value is being derived from its use, some forms of datafication can be more valuable than others, encouraging platform businesses to engage in as many forms of it as possible to strive for maximum returns. In addition to selling advertisement, both Srnicek (2016) and Sadowski (2019) mention a number of other ways in which platform businesses extract value out of data: as a way to profile and target people to give insight into consumer preferences or to coordinate and outsource workers; as a way to optimize production processes; by enabling the creation of new products or services (e.g. Amazon Web Services, Amazon Echo) and by analyzing data to generate new data in a cyclical manner. Currently, Amazon is using data towards all these ends, which is a testament of how important the feature is in fueling the company’s growing platform power.
However, van Dijck, Poell & de Waal (2018) argue that although datafication is an important aspect of a platform’s techno-commercial strategy it can paradoxically be seen as a user practice. In addition to user data being extracted, that same data is simultaneously circulated by the platform through graphical user interfaces and APIs to third parties and end users. The circulation of this data enables the user to make use of some of the important features these services offer: such as keeping track of friends and their activities, read reviews and
14
recommendations or plan public events (2018). This mutually beneficial and interdependent relationship between users and owners highlights the interconnectedness of the different sides of a platform and the complex anatomy of the business model.
As mentioned, the power of data and datafication became more significant and easier as the early Internet evolved into what has been dubbed as ‘Web 2.0’, which arose sometime after the turbulent years around the turn of the century. As key features of Web 2.0 became ubiquitous across the internet, more users started to migrate their everyday activities to online environments. As a result, the companies who facilitated these interactions began to recognize this opportunity and began supplying them as a service. Data thus became an essential part of the business model for these companies. Rather than a new and improved version of software, Web 2.0 relates to a different kind of internet, one that is ‘open’ and ‘participatory’. The term was popularized by a seminal figure of the early Web; Tim O’Reilly, who used it to describe ‘the web as platform’ in the way that it functions as a “robust development platform” upon which “websites become software components” (O’Reilly & Battelle, 2004).
However, O’Reilly’s computational definition of Web 2.0 was not used widely after it was first introduced but the term was embodied in a metaphorical sense (cf. Gillespie, 2010). Web 2.0 was used as a “rhetorical technology” by the industry, to change the way we think of the internet; as a software development platform, rather than a publishing channel (Allen, 2013). Moreover, it was seen as the revival of the industry after the dot-com bubble had burst and, especially among scholars, as a revolution that would reform the existing media landscape (Gehl, 2010). Web 2.0 is characterized by websites that emphasis user-generated content, ease of use, participatory culture and interoperability for its users. More generally speaking, it refers to a broad set of services that encourage participation and collaboration (Madden & Fox, 2006).
Facilitated by these features, users were fueling datafication efforts by websites as their casual, everyday social interactions became ‘formalized inscriptions’ as data, which take on a different value once they are released in the broader economy of the public domain (van Dijck, 2013). Because data was now becoming an important element of doing business on the Web, companies like Amazon looked for ways to collect data more efficiently, across the Web. It was already doing so through its Associates Program but would expands its efforts significantly by incorporating the view of Web 2.0 that sees ‘the Web as a platform’. In this regard, the idea is of a platform as a collection of software components, one that can be (re)developed and (re)programmed.
Following O’Reilly’s definition of Web 2.0, Andreessen—the founder of the internet browser Netscape Navigator, argues that a website becomes a platform only when you can
15
program it (Helmond, 2015). He notes that a platform that can be reprogrammed or customized by outside developers opens itself up to different uses which the original developers could not foresee nor allocate time towards developing (Andreessen, 2007). For platform scholars, the concept of programmability is a vital part of any platform and has been key in understanding the logic of new media (Chun, 2011; Manovich, 2001). In her work, Helmond (2015) argues that programmability is central to what she coins as ‘platformization’: “the rise of the platform as the dominant infrastructural and economic model” (p.5). In order for a platform to become programmable it must make its services available to other software programs through what are called ‘Application Programming Interfaces’ (APIs) (Evans, Hagiu & Schmalensee, 2006). The model for APIs emerged from the open source software community of the 1990s and can be seen as a way to facilitate innovation by enlisting the knowledge of a distributed community (Langley & Leyshon, 2017).
According to Murugesan (2007) an API is: “an interface provided by an application that lets users interact with or respond to data or service requests from another program, other applications, or Web sites. APIs facilitate data exchange between applications, allow the creation of new applications, and form the foundation for the ‘Web as a platform’ concept” (p. 36). Among the first to introduce APIs were e-commerce businesses eBay in 2001 and Amazon in 2002. These APIs were business-to-business solutions and were designed to exchange data in order to enable sales management and transactions between different business applications (Lane, 2012). For example, Amazon Web Services allowed developers to incorporate Amazon.com content and features into their own web sites. Furthermore, Simon (2011) argues that it was the adoption of API’s that enabled Amazon to transition from an e-commerce bookstore into a platform intermediated online marketplace after the dot-com bubble.
In today’s world APIs are widely used by all kinds of services for different ends; for example, AWS operates an intermediary API service that creates and maintains APIs for third parties to use independently or in conjunction with Amazon Cloud services. This API also allows external developers to build an application and ‘plug-in’ to an existing platform infrastructure, such as building an independent store and selling on Amazon. Therefore, the use of APIs is of vital importance for platform businesses because it enables them to extend their reach beyond the environment of their own website into the wider web. This is significant because it allows them to collect and format external web data, using it to add value to their own platform (Helmond, 2015). In this sense, platforms make use of their programmability to enable the decentralization of data production, and a recentralization of data collection (Gerlitz & Helmond, 2013).
16
Amazon’s expansion toward ‘the everything store’ continued at the end of the 1990s, when it expanded its offerings beyond books, to include music and DVD’s as well as through several acquisitions such as Junglee and Planet All (Hansell, 1998). The company subsequently tried to include electronics and toys into their selection, but considering those products are significantly different in nature from books, it proved difficult to do. This meant that Amazon was stuck with millions of dollars in unsold toy inventory, as it proved too difficult to predict which toy would be popular. During the same time, as the craze around the Internet started to reach new peaks, brick and mortar retailers such as Toys “R” Us were hurrying to get online. However, their inexperience in the field of E-commerce lead to major outages of their website and delays in shipping, resulting in negative publicity and a fine from the Federal Trade Commission (Stone, 2013). The two companies announced a ten-year partnership in which they combined their strengths, with Toys “R” Us using its knowledge of the industry to predict and stock the right toys, and Amazon’s knowledge and infrastructure as an E-commerce website to facilitate ordering and delivery.
This solution turned out to become Amazon’s first big step towards platformization, a step the company’s executive team had been trying to make for a couple of years, but had failed (Stone, 2013). Following the example of Microsoft, which at the time was the archetype for the platform strategy, Amazon wanted to augment the e-commerce efforts of other retailers on their platform. The first efforts were made with services such as Amazon Auctions and zShops, but the first-mover advantage eBay had built up in the sector proved too significant to compete with (Stone, 2013). Using the Toys “R” Us deal as a template, Amazon executives travelled around the country to pitch the idea to other large retailers. The success of this strategy prompted a renewed focus on Amazon’s failing Auctions feature, which at the time was isolated from the main part of Amazon’s product catalog. The company began to understand that in order to host third-party sellers on its platform, it would have to list the sellers’ items alongside their own product offering. This new feature known as Marketplace was launched in 2000 and according to Amazon’s CEO, sales by third-party sellers in 2018 accounted for 58% of physical gross merchandise sold on Amazon (Bezos, 2019). It was with Marketplace that Amazon was able to move from being an online bookstore to becoming a metaphorical ‘online mall’, controlling the sale of goods online.
Amidst the rapidly changing landscape of the internet of the early 2000s, Amazon saw the growth of its business slowing down. In line with its mantra to put consumers first, it ensured to offer the lowest prices available, continuously adapting its prices to match with those of other major retailers. The idea was that if the prices remained low, customers would
17
be won over by its significantly larger selection of products as well as by the convenience afforded to customers through home delivery. Amazon executives theorized that:
Lower prices led to more customer visits. More customers increased the volume of sales and attracted more commission-paying third-party sellers to the site. That allowed Amazon to get more out of fixed costs like the fulfillment centers and the servers needed to run the website. This greater efficiency then enabled it to lower prices further (Stone, 2013, chapter 4).
In this sense, if one part of the cycle increases, it affects the entire cycle. This is important because, according to a number of prominent platform business scholars, utilizing the connection between all sides of a platform for growth and competitive advantage is key to unlocking the strength of the platform business model (Boncheck & Choudary, 2013; Evans & Gawer, 2016; Choudary, Van Alstyne & Parker, 2016; Cusumano, Gawer & Yoffie, 2019).
The phenomenon that the Amazon executive team described is known as ‘network effects’ and relates to a key feature of a platform business wherein a platform becomes more valuable as more people use it (Gawer & Evans, 2016). In this sense, each new user of a platform makes the platform more valuable to everyone else which in turn will attract more users to the platform increasing the value even further (Choudary, Van Alstyne & Parker, 2016). Amazon’s ‘1-Click ordering’ was implemented with the same idea, when a customer spent more on the site, the overall volume of business would increase, further feeding the cycle as sketched by Amazon executives. Although network effects existed in other business models, the online platform is aided by the pervasive connectivity of the internet and can thus increase the network effects in a way that was impossible until now. Take the telephone network for example, which only becomes a valuable technology when enough people actually own and operate a telephone line (Gawer & Evans, 2016). Being the only person with a telephone renders it essentially worthless but as one, five or ten others also install a telephone, it becomes significantly more valuable to everyone using a telephone (Cusumano, Gawer, Yoffie, 2019). The same can be said for online platforms: as customers flock to Amazon, writing product reviews and feeding the product prediction algorithm, the service becomes more valuable for other consumers on the platform. Moreover, selling products on Amazon is only sensible when there are others who are buying those products, and the more people that are buying on Amazon, the more likely new sellers will be to join the platform and vice versa (Cusumano, Gawer & Yoffie, 2019). This example furthermore serves to show that there are two kinds of network effects: ‘direct’ or ‘same-side’ network effects, where users attract other users and ‘indirect’ or ‘cross-side’ network effects, where users on one side of the platform
18
attract users on another side of the platform (e.g. third-party sellers on Amazon attract consumers)(Gawer & Evans, 2016; Cusumano, Gawer & Yoffie, 2019).
However, critics have argued that the presence of network effects leads to these platforms having an inherent tendency towards monopolization (Srnicek, 2016; Kurz, 2017). Others, such as PayPal co-founder, entrepreneur and venture capitalist Peter Thiel argue that network effects and its subsequent monopolization are simply natural to this type of business model and enables them to maintain high levels of innovation (Morozov, 2015). However, economists such as Stiglitz (2017) argue that instead of increasing levels of innovation, these quasi-monopoly platforms have leveraged their market power to stifle innovation, pointing to a decline in new innovative firms, especially those headed by young entrepreneurs. Nevertheless, Choudary, Van Alstyne & Parker (2016) argue that although network effects may seem too good to be true, if a platform is mismanaged or fails to innovate, the cycle can quickly turn around and cause an exponential decline in the number of users.
3.2 It is All About the Long Term
In addition to the features of a platform such as data and network effects, critics of the platform economy argue that there are a number of important external conditions that have enabled today’s digital economy to arise (Rahman & Thelen, 2019; Langley & Leyshon, 2017; Srnicek, 2016). Several key moments in the relatively recent history of modern capitalism have been shaped the new digital economy. Broadly speaking, this relates to changes in these three areas: the way businesses were modeled, the nature of employment and monetary policy. Ultimately, those changes have helped shape the current climate in which digital platform businesses have been thriving for the past decade or so. A brief exploration of these enabling conditions is required to understand how the process of financialization has played an instrumental role in creating an environment in which these platforms could arise. For one, the development of the business model from the mid twentieth-century consolidated firm toward the ‘network of contracts’ firm of the 1980s and 90s and most recently toward a platform business model (Rahman & Thelen, 2019). Under each of these models the role and position of labor has shifted, and the power of labor and labor movements has deteriorated. Finally, in an attempt to stave off the bust following the dot-com boom around the end of the 1990s and save the global financial order from collapse in the wake of the 2008 crisis, several financial policy changes were implemented that are key enabling conditions for today’s digital economy to develop. The postwar period was characterized by large consolidated firms built along Fordist lines, inspired by the automobile industry as a paradigm. Especially American manufacturing
19
was in a globally dominant position, after the Second World War had left most of the rest of the world devastated. These firms operated on the basis of mass production, a top-down control from powerful managers and ownership in the hands of ‘patient’ capital, dispersed between a large number of passive investors in the United States, or banks in Europe (Srnicek, 2016; Rahman & Thelen, 2019). The business model was focused on cultivating long term relationships that ensure stable gains for all stakeholders, including labor. Labor in these factories was organized along Taylorist principles, in which the production process is broken down into smaller, less skillful parts and reorganized to improve efficiency. The workers were mostly unionized, which meant that they were provided with extensive benefits such as high wages, healthcare, pensions, job security and upward mobility within the firm (Srnicek, 2016; Rahman & Thelen, 2019). As a result, these firms were instrumental in advancing broader goals for the New Deal-era society by providing for the welfare state and through productive investments. In addition, as powerful corporate entities, these firms were subject to regulation such as antitrust laws and financial regulation, which made sure they did not grow too big or channel their wealth into private rents. In that sense, argue Rahman & Thelen (2019), “the consolidated firm became a key vehicle through which the United States “improvised” a robust social contract even in the absence of a European-style social democratic political economy” (p.182).
However, facilitated by low cost, high skilled labor and favorable exchange rates, manufacturing in countries such as Japan and Germany began to soar (Brenner, 2006). Because their economies were stimulated by the American Marshall Plan, investment levels grew and the export market expanded, opening up the world market and growing global demand (Brenner, 2006). As a result, American manufacturing firms were suddenly competing with other countries on a global market, which lead to higher supply and thus lower prices. Unable to compete with the lower prices of its competitors, the domestic firms went through a crisis of profitability between the 1950s and 1970s (Moseley, 2013; Brenner, 2006). Threatened by the prospects of declining profitability, the previously patient investors began demanding more influence in corporate governance in order to promote efficiency and growth. In order to remain competitive, American businesses began to model themselves after their counterparts: the Fordist model was replaced by the Japanese Toyotist model, streamlining the production process further by breaking it down in even smaller parts and ensuring as little downtime as possible (Rahman & Thelen, 2019).
During the ‘shareholder revolution’ of the 1970s and 1980s, investors and securities analysts became increasingly powerful within the ‘network of contracts’ (NOC) firm, as stock
20
price was the core metric for success and share value was strongly dependent on reaching profit projections of analysts (Rahman & Thelen, 2019). Under pressure from their shareholders, firms were told to focus on core competencies, leading to outsourcing, asset stripping, soaring CEO pay and an increasingly precarious position of labor. Nike became exemplary for this model which became known as ‘Nikefication’: outsourcing all parts of its production process to low cost countries except for the design and marketing elements, which remained in high-income countries (Davis, 2009). Rahman & Thelen (2019) argue that managers faced coordinated pressure from wealthy investor groups, which punished businesses for pursuing long-term investments at the expense of short-term returns. As a result, CEO pay increased dramatically during this time, as managers acted upon investors interests in coalition. As firms were becoming leaner in an attempt to revive profitability, the power of labor was attacked throughout Western economies, as unions faced legal roadblocks, their membership dropped and several industries were deregulated (Srnicek, 2016; Brenner, 2006). Businesses took advantage of this weaker position of organized labor to abandon the previous norm of standard employment—full-time jobs with benefits and upward mobility—and move towards employment that is increasingly flexible, low wage and subject to managerial pressures (Srnicek, 2016; Rahman & Thelen, 2019). Both the increased power of investors as well as the diminished power of labor and subsequent retreat from standard employment have been vital conditions for the current digital economy to emerge.
The profitability of manufacturing industries continued to decline throughout developed economies and economic and productivity growth had a downward trend, even after the lows of the 1970s (Srnicek, 2016). However, the 1990s saw a remarkable break from this negative trend, as the possibilities of the internet fueled an economic boom that has had a profound effect on how the digital economy developed. It was during this time that the internet evolved from a largely non-commercial project into a commercialized ‘new frontier’ for financial and technological opportunity. Excited by the possibilities of the internet and confronted with low profit margins in manufacturing industries, the information and
communications technology (ICT) sector became the preferred destination for financial capital.
Internet related businesses were subject to excessive financial speculation, stimulated by large amounts of venture capital (VC) and resulting in high stock valuations. Venture capitalist funds invested in early-stage internet businesses with the idea that these could become highly profitable, as they helped people and businesses to get online. The goal for these venture capitalist funds is to cash-out their equity stakes in a portfolio (e.g. a selection of promising start-ups) through so-called ‘liquidity events’—an IPO, acquisition, or the sale of shares to
21
other investors (Langley & Leyshon, 2017). Traditionally those liquidity events only took place after a start-up had demonstrated an ability to generate revenue, and thus provide the investors with a way to recover the cost of their investment.
Crucially, this requirement was largely ignored for the internet businesses of the dot-com boom (Feng et al., 2001). Most of these businesses had no source of revenue let alone any profits, but the hope was that through rapid growth they would be able to secure a share of the market early on and over time go on to dominate what was assumed to become a major new industry (Srnicek, 2016). This ‘growth before profits’ model became an important philosophy for these internet businesses, something that has remained true for the platform businesses today. For example, Uber continues to operate with a net loss after their IPO and Amazon has operated with a net income close to zero for the majority of its nearly 25-year history as a publicly traded company (Macrotrends, 2020). The venture capitalist funds operate on the basis of the governing ‘2:6:2 rule’, which states that two investments will be losses, six will break-even and only two will realize returns. However, the two that do realize a return will be so significant, that it will generate overall returns for the portfolio that trump the equity markets over the same period (Mason, 2009). This is caused in part because of inherent features of the businesses that succeed, which coordinate network effects to generate revenue and thus monopolize the market, leaving little for their competitors (Langley & Leyshon, 2017).
Besides the large amounts of venture capital, the stock market boom was further exacerbated by other equity markets looking to make profits on the promises of the ‘new economy’ of internet companies. At the height of the boom between 1997 and 2000, technology stocks increased 300% and had a market capitalization of $5 trillion (Perez, 2009). Srnicek (2016) argues that as a result of the craze surrounding the new industry, massive amounts of capital were injected into the fixed assets of the internet, as companies were spending large sums to modernize their computing infrastructure: “millions of miles of fiber-optic and submarine cables were laid out, major advances in software and network design were established, and large investments in databases and servers were made” (p.22). Consequently, as the internet moved into the mainstream in the first years of the new millennium, important infrastructure was put in place, which would provide platform businesses with the technological basis for unrestricted growth for years to come.
Fearful of the Asian financial crisis that was unfolding throughout parts of East and Southeast Asia in the late 1990s, the US Federal Reserve implemented a number of interest rate reductions in an attempt to stop a spread of the crisis to the United States (Srnicek, 2016). This trend of ultra-easy monetary policy has become the preferred method for governments to
22
try and stave off economic recessions. Considering the US government was trying to reduce its deficits, fiscal stimulus was not an option. The idea behind a lower interest rate is that because being in debt becomes cheaper, consumers and businesses continue to spend money, fueling economic growth. Rather than the public deficit resulting from traditional Keynesian economics, this form of stimulus created private deficits and inflation of assets and is thus labeled as ‘asset-price Keynesianism’ (Brenner, 2007). This form of stimulus had an effect, the dot-com boom continued, and the East Asian crisis did not spread to the US.
However, as equity prices continued to rise, profit rates dropped and large investments fueled more overcapacity, the 2001 crash became inevitable. The lowering of interest rates continued after the crash, and household borrowing increased drastically, causing house prices to skyrocket (Brenner, 2007). Economic growth in the period between 2001 and 2006 was largely fueled by consumers, as businesses were more focused on increasing profit rates by launching an offensive on workers. As Brenner (2007) argues:
“by holding down job creation, investment and wages, they have held down the growth of aggregate demand, undermining their own incentive to expand. Instead, exploiting the cheapness of credit, they have devoted a record share of their resources to buying back their own shares, financing mergers and acquisitions, and paying dividends to stockholders - rather than expanding investment and creating new jobs” (para. 7). Moreover, because returns on investments were low, new sources for profits were sought after, eventually landing on the high returns from subprime mortgages and thereby creating the conditions for the following crisis (Srnicek, 2016; Brenner, 2007).
The 2008 crisis had a number of effects that explain the current platform economy. For one, interest rates around the world dropped even further, in some cases down to zero, creating an environment in which riskier investments were pursued for higher returns, ending up at unprofitable and unproven tech companies (Srnicek, 2016; Langley & Leyshon, 2017). Secondly, because governments provided corporate bailouts (turning private debt into public debt) and emergency tax cuts, their deficits rose drastically which ushered in a period of austerity. This meant that rather than fiscal stimulus, governments were looking at more unconventional options to revive their economies. Besides low interest rates, one such way was through ‘quantitative easing’ whereby central banks buy up financial assets (e.g. government or corporate bonds, mortgages) in order to inject money into the economy and expand economic activity. Although quantitative easing did seem to have a positive effect on economic growth, critics argue that its effects primarily benefits those already engaged in the financial market, passing little along to the rest of the economy (Frank, 2012). According to one critic:
23
“QE [sic] cash ends up overwhelmingly in profits, thereby exacerbating already extreme income inequality” (Frank, 2012, para. 4). Moreover, these cash profits were often hoarded or funneled to tax havens to enable tax evasion, done primarily by technology companies (Srnicek, 2016). For platform businesses, because they rely on intellectual property rather than physical property, relocating to different tax jurisdictions is particularly easy (Srnicek, 2016). This has left US technology companies with large amounts of money to invest into ‘unicorn’ start-ups or corporate acquisitions and mergers. Furthermore, by keeping their cash offshore and engaging in tax evasion, by definition they are exacerbating government deficits and furthering austerity (Srnicek, 2016).
The platform business model developed alongside the changes in the economy as a result of several booms and crises as well as developments in digital technology and the internet itself. Much like the NOC firm which took elements of the Fordist model—especially its specialization and division of labor—the platform model builds on and intensifies features of previous business models, such as its labor practices (Rahman & Thelen, 2019). This is not to say that the platform model will come to replace the other models, rather, they will exist together with other types of firms. However, what makes the platform model different is that because of their propensity toward monopolization, they put pressure on the markets they enter by setting the terms. For example, Airbnb disrupted the market of traditional hotels, Uber for the taxi business and retailers are only able to compete by emulating core features of Amazon’s model (Rahman & Thelen, 2019). The NOC firm’s labor- and cost-cutting strategy is an important feature of many platform businesses, often taking the precarious and flexible work arrangements to a new level by depending on independent contracting and using data analysis to manage and optimize labor on the smallest scale (Rahman & Thelen, 2019).
However, Rahman & Thelen (2019) argue that despite these similarities with the NOC firm, there are three distinct characteristics that are unique to the platform business model. Firstly, unlike meeting the short-term demands of investors in the NOC model, platform corporations have access to a more patient form of capital, a precedent that was set by the early internet businesses of the dot-com boom discussed previously. The ‘break it up and sell it off’ mentality to meet profit margins is replaced by investors with a vision for the future, in which their patience will be rewarded. Second, rather than the outsourcing and labor shedding to increase share value, which was at the heart of the NOC firm, the platform model centers around the generation of network effects in order to secure market dominance (Rahman & Thelen, 2019; Langley & Leyshon, 2017). As a result, investors are more patient when it comes to getting a return on their investment because they understand that once a dominant market
24
position has been established, monopoly rents can be extracted “while also multiplying the number of stakeholders whose dependence on the platform makes them potential allies in efforts to defend it against unwelcome regulation” (Rahman & Thelen, 2019, p.180). The third and final difference has to do with the role of consumers. Although consumers were already benefiting from the lower prices due to low labor costs and thus instrumental to the success of the NOC firm, the platform business’ relationship to the consumer goes even further. Because of the direct link between platform and consumer, often connecting through devices that are used daily, consumers are vital not only to the market strategy of a platform firm but to their political strategy as well. As Rahman & Thelen (2019) explain: “the most successful of such firms have proved to be extraordinarily adept in leveraging their loyal consumer base into an active public narrative and political advocacy strategy in order to secure legislative and legal support for the platform business model” (p.180). In light of things such as the degrading power of labor, concerns over privacy and data ownership and tax evasion, the relationship between these platform firms and consumers deserves more attention.
3.3 Obsess Over Customers
For Jeff Bezos, what sets Amazon apart from other businesses is its obsessive focus on customers. Bezos claims that unlike its competitors, who would focus on the competition instead of the customer, Amazon always operated with the best interests of the customer in mind (Stone, 2013). The rise of the World Wide Web allowed Bezos to take his ‘customer ecstasy’ further than any company had previously been able to. With a selection that dwarfs the size of any brick and mortar retailer; a user-friendly store that is open 365 days a year, 24 hours a day; low prices and at home delivery, the company acts on the fundamental impulse of modern capitalist societies: to consume. Since then, the company has become synonymous for the e-commerce industry in the Western world. By leveraging the data and programmability of the platform, as well as by using the financialization model of the dotcom boom to ‘Get Big Fast’ and secure his long-term vision as market leader, Amazon has helped define what it means to be a platform business. Paradoxically, as the platform and its power grew beyond its E-commerce operations, so too did its concept of ‘customer’. Amazon now caters to a large variety of customers, ranging from individual buyers and sellers on the platform to large corporations and governments. This also means that the company’s relentless focus on customers might benefit one customer, whilst being a detriment to others.
Amazon’s obsession with customers has also allowed it to largely avoid scrutiny from lawmakers and regulators over its growing market dominance and influence. To
