Essays on Power in Digital Platforms by Martín Harracá Submitted for the Degree of Doctor of Philosophy Department of Digital Economy, Entrepreneurship, and Innovation Surrey Business School Faculty of Arts and Social Sciences University of Surrey Supervisors: Annabelle Gawer Carla Bonina Itziar Castelló ©Martín Harracá 2023 1 Essays on Power in Digital Platforms Declaration This thesis and the work to which it refers are the results of my own efforts. Any ideas, data, images or text resulting from the work of others (whether published or unpublished) are fully identified as such within the work and attributed to their originator in the text, bibliography or in footnotes. This thesis has not been submitted in whole or in part for any other academic degree or professional qualification. I agree that the University has the right to submit my work to the plagiarism detection service TurnitinUK for originality checks. Whether or not drafts have been so-assessed, the University reserves the right to require an electronic version of the final document (as submitted) for assessment as above. Signature: Martín Harracá Date: August 25. 2023 2 Essays on Power in Digital Platforms Summary This thesis investigates power in a new organisational form: the digital platform. Platforms have disrupted our societies by creating and coordinating novel, flexible, and shifting forms of social interaction, changing how we trade, work, and communicate. The profound and abrupt economic transformation brought by platform companies like Amazon, Google, and Meta, while carrying initial expectations of increased freedom of choice, higher social cohesion, and power distribution, was soon shadowed by the growing evidence of platform power concentration and abuse. This research critically examines the characteristics and dynamics of power within digital platforms, specifically focusing on platforms’ organising structures and influence. This problem is approached from three different dimensions. The first dimension pertains to the structure of social interactions, recognising the mechanisms through which platforms shape power relations and autonomy by influencing agents’ interactions and modulating participatory practices. The subsequent dimension is centred on the platform's governance, contrasting the principles guiding the governance and how it is enacted, and how this shapes power dynamics within the platform. Lastly, the third dimension focuses on the interplay between firms, assessing the extent of influence wielded by the platform's owner in directing shifts in the value creation and capture strategies of other participant firms. Each dimension of the research problem is addressed in one study in a dedicated chapter, after charting the foundational concepts that characterise digital platforms and summarising extant research on platform power. The first study is based exclusively on conceptual development, while the second and third are a qualitative case study of the Amazon Marketplace, focused on Amazon’s and the sellers’ practices and how these affect each other. Overall, this thesis contributes to advancing our knowledge of how we understand platforms and provides new ways to characterise platform power, its dynamics, and its consequences. 3 Essays on Power in Digital Platforms Acknowledgements This project started way back in time, and initially as an informal activity. In the last years of my licentiate degree, I discovered my passion about research and knew I wanted it to become an important part of my life. At the time, I rejected the idea of making it my main occupation: I thought about keeping it as a side activity, so I would be relieved from any formal strings and focus on what I really wanted to say and explore. For several years, other than some dispersed writing retreats and moments of inspiration, that strategy didn’t seem to be working very well. I find now, however, that many fundamental ideas were sown during that period, and eventually came to expression once I formally enrolled in the program - and probably there are still others, waiting. In a similar way, this journey was not at all what I expected. With the hindsight of backward looking, I recognize in it the same pattern of the things that deeply changed me: I started by rejecting and feeling alienated by it, but that estrangement soon turned organically into enthusiasm and joy. This work, like all, is the result of a collective effort and the net of affection and support beneath it. An intricated net composed of people inhabiting different worlds whom, while mostly don’t know about each other, have here a point of convergence. I divide these worlds into three, inspired by Julio Cortazar and his Rayuela [Hopscotch]. Del lado de allá [from the side beyond] acknowledges those who were involved in this journey in Europe. Del lado de acá [from this side] acknowledges those who were involved in Argentina. This part is written in Spanish, as it is the native language of all the people I thank and my own. Finally, De otros lados [from other sides] acknowledges other brilliant people that has inspired this work in multiple unsuspected ways. Del lado de allá [from the side beyond] I need to start acknowledging my supervisory team, Annabelle, Itziar, and Carla, for the fundamental importance they had in this experience. It was a dream team, each contributing in their own unique way, and constantly pushing me into new and higher grounds. I am much obliged; it was truly a privilege to learn and work with you. I want to thank all the colleagues from the Department of Digital Economy, Entrepreneurship, and Innovation, and especially Glenn Parry, for the insightful conversations and their effort in creating a welcoming and integrative space, particularly during the periods of lockdown. I also want to thank Prof. Alexander, Karen, Veena, Carolina, and other members of the Surrey Business School staff for their continuous and warm support during these years. A very special thank you to Lydia and Hannah, for being such great allies in navigating together parts of this PhD. I also want to thank Yiwen, Adam, Parul, and Ivone, for those brief but meaningful connections in the worlds apart that are academic conferences. To my Argentinian friends in London, Matías, Jimena, Joel, and Cecilia, thank you for making so homely a new city and its culture. To Maria, Tejash, and Lea, for all the sharing and teaching, and for building a true home together. To Thi, Fernando, and Julio, obrigado de coração. I can’t imagine having gone through this without you, thank you for making it a bright experience. 4 Essays on Power in Digital Platforms Del lado de acá [from this side] En primer lugar, quiero agradecer a quienes son los principales responsables de haber despertado mi curiosidad intelectual, y por introducirme a nuevos mundos de pensamiento. Gracias Pablo, Axel, Juan, Mercedes, y Marcelo. También agradezco por la profunda influencia que tuvieron en mí a Daniel Heymann, Enrique Kawamura, y Carlos Mora. Agradezco al grupo de investigaciones en plataformas, liderado por Pablo, y a Sonia, Facundo, Andrea, y Nicolás, por las sesiones compartidas y las charlas inspiradoras. A mis compañeros de discusión y militancia Itai, Pablo, Andrés, Igal, Facu, Caro, Ema, Fede, Dani, Bruno, Martín, Flor, Seba, Guido, Luis, Lean y Vale, por enseñarme a hacer carne la reflexión. A esa experiencia única de investigación colectiva que fue la topa y que me marcó profundamente. Gracias Pilar, Cata, Pablo, y Lean. A los miembros del Seminario de París, Bruno, Leandro y Beto, por ser una fuente permanente de estímulo intelectual. Gracias a Nay, por estar siempre que andamos sin buscarnos. A Eugx, por tirar abajo paredes. A Mariana, por acompañarnos sin importar la distancia. A mis vecinos para siempre, Seba, Aye, Mariana, Andrés y Beto. A Ro y Flor, por toda una vida de conexión y entendernos. A Ger y Gonza, por empujarme cuando lo necesitaba y ayudarme a llegar, por estar allá y acá, y por Richard. A Liliana, Jorge, y Lito, por lo que hicieron para que sea lo que soy. A Jose, Inti, Juan, Allen, Charly, Quillén, Nelson, Lolo, Aye, Rodri y Alicia, por enseñarme a agrandar la familia. A Rosa, que está más cerca de lo que piensa. Gracias Hui por compartir tus días conmigo, por tu alegría, y por enseñarme nuevos horizontes de lo que creía posible. Gracias Doris. De otros lados [from other sides] To Elena, Paul, Alexandre, Paula, Bernard, Stephen, Gillian, Phil, Sacha, Marlon, Sam, Noé, Ian, Thomas, GuyManuel, Santiago, Pantro, Niño, Chatrán, Doctora, Laura, Mark, Donald, Dave, Mike, Stephan, Brian, Karl, Rick, Robert, Robin, Will, Elizabeth, David, Milton, Mariana, Dan, David, Stephen, Ed, Ben, Giorgio, Gustavo, Javier, Josh, Tom, Gaspard, Xavier, Ray, Leonard, Natalia, Lali, Sara, Jaime, Gernot, Klaus, Michael, Peter, Bjorn, John, Jeff, Tom, Jonny, Colin, Ed, Philip, Karinn, Olof, Jason, Macu, Jean, Terry, Victoria, Lea, Henry, George, Sumner, Becki, Barzana, Heredia, Delta, and Garay. Thank you for the inspiration and the peace you brought me. 5 Essays on Power in Digital Platforms Table of Contents Declaration ...................................................................................................................................................................... 2 Summary ......................................................................................................................................................................... 3 Acknowledgements ....................................................................................................................................................... 4 Table of Contents ........................................................................................................................................................... 6 List of Tables and Figures ............................................................................................................................................. 8 1. Introduction ...................................................................................................................................................... 9 1.1. Purpose and objectives of this study ....................................................................................................................................... 9 1.1.1. 1.1.2. 1.1.3. 1.2. Central concepts and related research ................................................................................................................................. 13 1.2.1. 1.2.2. 1.3. 2. Motivation and problem statement .................................................................................................................................... 9 Objectives and methodological approach .................................................................................................................... 10 This research in context: The impact of the COVID-19 pandemic ...................................................................... 12 Conceptualizing digital platforms..................................................................................................................................... 13 Extant research on power in digital platforms ............................................................................................................ 19 Thesis overview .............................................................................................................................................................................. 25 How digital platforms organize immaturity: A sociosymbolic framework of platform power ..........28 2.1. Organized immaturity ................................................................................................................................................................. 29 2.2. Platforms and the accumulation of power ......................................................................................................................... 30 2.2.1. 2.2.2. 2.3. A socio-symbolic perspective on digital platforms ......................................................................................................... 33 2.3.1. 2.3.2. 2.3.3. 2.3.4. 2.4. Fields, capitals, and habitus in digital platforms......................................................................................................... 33 A socio-symbolic perspective of power accumulation and its consequences for organized immaturity 37 Forms of Platform Power ...................................................................................................................................................... 38 Platform Power Dynamics .................................................................................................................................................... 44 Discussion ......................................................................................................................................................................................... 46 2.4.1. 2.4.2. 2.4.3. 2.5. 3. Platforms as organizing agents ......................................................................................................................................... 30 Platforms’ accumulation of power and the consequences for individuals’ autonomy erosion .............. 31 Rethinking Organized Immaturity from a Sociosymbolic Perspective .............................................................. 47 Contribution to the Understanding of Platform Power Accumulation ............................................................. 48 Contributions to Practice and Avenues for Future Research ................................................................................ 49 Conclusion ........................................................................................................................................................................................ 50 Two-faced governance in platform ecosystems ........................................................................................51 3.1. Introduction ..................................................................................................................................................................................... 51 3.2. Literature - Related Research ................................................................................................................................................... 52 3.2.1. 3.2.2. 3.2.3. 3.2.4. 3.3. 3.3.1. 3.3.2. 3.3.3. 3.4. 3.4.1. 3.4.2. 3.4.3. 3.5. 3.5.1. Governance in platform ecosystems ............................................................................................................................... 53 The consensus on platform governance: Four central claims ............................................................................... 54 Minority views on platform governance ........................................................................................................................ 55 Research Question .................................................................................................................................................................. 56 Research methods ......................................................................................................................................................................... 56 Empirical setting: Amazon and its Marketplace .......................................................................................................... 56 Data collection .......................................................................................................................................................................... 57 Data analysis .............................................................................................................................................................................. 59 Findings ............................................................................................................................................................................................. 59 Amazon governance practices ........................................................................................................................................... 59 Sellers’ adaptive ecosystem participation practices .................................................................................................. 71 Two-faced governance .......................................................................................................................................................... 76 Discussion ......................................................................................................................................................................................... 77 Contribution to the platform strategy literature ........................................................................................................ 78 6 Essays on Power in Digital Platforms 3.5.2. 3.5.3. 3.5.4. 3.5.5. 3.6. 4. Enriching the construct of platform governance ....................................................................................................... 79 Contributions to the literature on platform power ................................................................................................... 80 Managerial and policy implications ................................................................................................................................. 81 Limitations and avenues for future research ................................................................................................................ 81 Conclusion ........................................................................................................................................................................................ 82 The evolutionary dynamics of platform orchestration .............................................................................84 4.1. Introduction ..................................................................................................................................................................................... 84 4.2. Literature - Related Research ................................................................................................................................................... 85 4.3. Research methods ......................................................................................................................................................................... 90 4.3.1. 4.3.2. 4.3.3. 4.4. Findings ............................................................................................................................................................................................. 95 4.4.1. 4.4.2. 4.4.3. 4.5. Why do sellers adopt new business models? .............................................................................................................. 95 Amazon’s role: orchestrating change .............................................................................................................................. 96 Evolutionary interplay between Amazon and the Sellers ..................................................................................... 104 Discussion and conclusions..................................................................................................................................................... 106 4.5.1. 4.5.2. 5. Empirical setting: The Amazon Marketplace and its sellers’ business models .............................................. 90 Data collection .......................................................................................................................................................................... 93 Data analysis .............................................................................................................................................................................. 94 Contributions to platform research................................................................................................................................ 106 Limitations and avenues for future research .............................................................................................................. 107 Discussion and general conclusions ......................................................................................................... 108 5.1. On the nature of platforms as organising agents ......................................................................................................... 110 5.2. On the characterisation of platform power ...................................................................................................................... 110 5.3. On platform power dynamics................................................................................................................................................. 111 5.4. On the consequences of platform power ......................................................................................................................... 112 5.5. Policy Implications ...................................................................................................................................................................... 113 5.6. Limitations and future research ............................................................................................................................................. 114 5.7. Final remarks: repurposing platforms for social change............................................................................................. 116 References .................................................................................................................................................................. 117 Appendix .................................................................................................................................................................... 130 A. Glossary of Amazon’s programs and policies ................................................................................................................. 130 7 Essays on Power in Digital Platforms List of Tables and Figures Table 1. Extant research on power in digital platforms ........................................................................................................................ 20 Table 2: Summary of publications’ chapters ............................................................................................................................................. 27 Table 3: Forms of Platform Power and the Organization of Immaturity ....................................................................................... 43 Table 4. Data Sources: Detail ........................................................................................................................................................................... 58 Table 5. Amazon’s declared governance: Examples, practices, and outcomes .......................................................................... 60 Table 6. Amazon’s undeclared governance: Examples, mechanisms, practices, and outcomes ......................................... 62 Table 7. Amazon’s undeclared governance practices (1): Create an unpredictable and precarious environment for sellers. Evidentiary support .............................................................................................................................................................................. 64 Table 8. Amazon’s undeclared governance practices (2): Profit from an asymmetric position. Evidentiary support 65 Table 9. Comparison between Amazon’s claims, legal responsibilities, and actual practices.............................................. 70 Table 10. Sellers’ adaptive participation: Examples, mechanisms, practices, and outcomes ............................................... 71 Table 11. Sellers’ adaptive ecosystem participation practices: Evidentiary support ................................................................ 72 Table 12. Detail of Sellers’ adaptive practices for bending, eluding, or hacking the platform’s rules ............................. 75 Table 13. Challenging evidence to the literature’s central claims on platform ecosystem governance and consequences for theory ................................................................................................................................................................................... 78 Table 14. Complementor orchestration: main findings and issues raised in the literature reviewed ............................... 89 Table 15. Types of sellers’ business models in the Amazon Marketplace .................................................................................... 92 Table 16. Detail of the data sources ............................................................................................................................................................. 93 Table 17. Amazon’s orchestration activities: Examples, practices, and outcomes .................................................................... 98 Table 18. Summary of Amazon’s orchestration activities over time: evolution of functionalities availability and key changes in seller competition ....................................................................................................................................................................... 102 Table 19. Detail of functionalities available for sellers in the Marketplace over time ........................................................... 103 Table 20. Contributions of this thesis to research on platforms and platform power .......................................................... 109 Figure 1: Platforms as market facilitators ................................................................................................................................................... 15 Figure 2: Platforms as technological architectures ................................................................................................................................ 17 Figure 3: Platforms as organisational forms.............................................................................................................................................. 19 Figure 4: Platform Power Dynamics over Time ........................................................................................................................................ 44 Figure 5. Two-faced platform governance: A process model ............................................................................................................ 77 Figure 6. How new sellers’ business models appear on the Marketplace .................................................................................... 96 Figure 7. Amazon’s orchestrating role in the appearance of new business models ................................................................ 99 Figure 8. The appearance of the Private Label business model in the Marketplace.............................................................. 100 Figure 9. The appearance of the Aggregator business model in the Marketplace ................................................................ 101 Figure 10. Interplay between the sellers’ business models evolution and Amazon’s value proposition, orchestration, and value capture strategies ......................................................................................................................................................................... 104 Figure 11. Changes in the sellers’ importance in the Marketplace and of Amazon’s sources for value capture ...... 105 8 Essays on Power in Digital Platforms 1. Introduction 1.1. Purpose and objectives of this study 1.1.1. Motivation and problem statement The advent of digital platforms such as those operated by Amazon, Google, Meta, or Uber has brought a profound economic transformation, challenging established forms of coordination and exchange (Cusumano et al., 2019; Parker et al., 2016). They have created new markets and disrupted existing ones, as platforms excel at business growth and value capture (Cusumano et al., 2019; Manyika et al., 2018) and play a major role in fostering innovation (Ringel et al., 2019). Platforms have reshaped entire industries, like Google and Facebook with advertising or Amazon with retail–, have expanded the horizons of product innovation efforts by articulating networks of developers around operating systems such as iOS and Windows, have created global labour marketplaces, such as Upwork, and networks with billions of interacting individuals, like Instagram, Twitter, or YouTube. Overall, digital platforms have changed how society trades, works, and communicates. The novelty of platforms and their impact has been accompanied by a growing scholarly interest, attaining a stable position in the economics, information systems, and management research agendas. To a large extent, this research has attributed the platforms’ importance to how they contribute to creating value in new ways. Cusumano et al. (2019) differentiate between transaction platforms, which create value by facilitating matchmaking and exchange between groups of agents subject to network effects, and innovation platforms, which provide a technological foundation or infrastructure upon which third parties generate complementary innovation in the form of new products, new services, or new technologies. Correspondingly, foundational platform studies have focused on how they enable market creation and impact competition (Armstrong, 2006; Eisenmann et al., 2009; Evans, 2003, 2010; Gawer & Henderson, 2007; Hagiu, 2013; Rochet & Tirole, 2003; Rysman, 2009) and which platforms’ configurations and features drive innovation and how (Adner & Kapoor, 2010; Baldwin & Woodard, 2009; Chesbrough, 2006; M. Cusumano & Gawer, 2002; Gawer, 2009). Underpinning these new ways to create value is a more fundamental fact: platforms act as new organisational forms that, using a shared technological infrastructure, are capable of coordinating groups of heterogeneous and formally independent actors – both individuals and organisations – towards system-level goals (Gawer, 2014; Kretschmer et al., 2022; McIntyre et al., 2021). Conceptualising platforms as organisational forms stresses the centrality of their constituents’ agency and the interplays among their roles, positions, and strategies (Gawer, 2014). Further, this approach highlights the flexible and continuously evolving nature of the relationships formed between platform participants (Gulati et al., 2012; Kretschmer et al., 2022). Along with platform firms – companies created based on a platform, such as Google or Uber – other longstanding companies are also adopting this type of organisational strategy in sectors as diverse as real estate, transportation, health services, or finance (Evans & Gawer, 2016). This process can be led by the transformation in some of its business units (like General Electric or Siemens with their industrial “Internet of Things” platforms – Predix and MindSphere), but also by drastic and structural organisational overhauls (as in Haier, a prominent Chinese appliances producer that “platformized” both its internal divisions and supply chain network [Hamel and Zanini 2018]). 9 Essays on Power in Digital Platforms As platforms developed and spread globally, the predominantly optimistic views of the '00s and part of the 10's decades regarding public opinion and the attitude of governments towards platforms gave place to a steep "change of mood" (Cusumano et al., 2019). In an increasingly unequal world, platforms have delivered some of the largest and fastest-earned fortunes in history, while their expansion into new and diverse areas of human experience has turned widespread enthusiasm and support into a growing concern for the social implications of this phenomenon (Eubanks, 2018; Lanier, 2018; Lehdonvirta, 2022; Srnicek, 2016; Zuboff, 2019). Involved in highprofile conflicts such as those related to Cambridge Analytica in the Brexit referendum or the United States Capitol attack of January 6 2021, platforms have been pointed out for bringing issues of surveillance, manipulation of choice, disinformation and censorship, market dominance leading to discriminatory clauses, restriction of access, or limitation in content creation, and the worsening of labour conditions (Beer, 2017; Gawer & Srnicek, 2021; Landini, 2016; Lyon, 2018; Stigler report, 2019; van Doorn & Chen, 2021). Along with this “change of mood”, research related to the so-called platforms’ “dark side” and their power has also become more prominent (Flyverbom et al., 2019; Kölbel et al., 2023; Trittin-Ulbrich et al., 2021; Verbeke & Hutzschenreuter, 2020). In contrast with the initial characterisation of platforms as simply matchmakers (Caillaud & Jullien, 2003; Evans, 2003; Rochet & Tirole, 2003), it is increasingly recognised that its owners are not neutral intermediaries in the transactions they orchestrate but interested parts (Cutolo & Kenney, 2021; Furman, 2019; Khan, 2018). Platforms shape the rights of access, use, and property of its participants and have a distinctive role in the instrumentalisation of exchanges, as they supervise, drive, inhibit and even prohibit them (Alaimo & Kallinikos, 2017; Curchod et al., 2019; Landini, 2016). This allows platforms to tightly organise and command resources they do not own, redefining what has traditionally been considered an organisation’s scope and reach. My1 inquiry is motivated by the transformations brought by platforms as an organising agent and its impact on how power is structured and produced in this context. Platforms are both the vehicles and the embodiment of society’s digitalisation: they bring novel ways of interaction by building over while transforming the relations among its participants; they provide the space for these interactions to happen and establish the rules governing them. As these changes conflict with pre-existing social and economic structures, platforms reconfigure power relations. This thesis explores how power is produced and exerted in the context of platforms and how its agents’ practices are transformed as a consequence. 1.1.2. Objectives and methodological approach This investigation aims to understand how power is reconfigured in platforms as new organisational forms. I will approach this problem from three dimensions: The first dimension is at the level of the social interactions’ structure. Platforms have become a space on which an increasing number and variety of social interactions take place. The objective is understanding how platforms shape power relations and autonomy by structuring agents' interactions and influencing their participation practices. This problem is examined in Chapter 2. 1 In this thesis I use the first person in singular except in Chapters 2 and 3, which are co-authored. 10 Essays on Power in Digital Platforms The second dimension concerns the platform’s rules for participation. One of the foundations on which platforms operate is a set of governance rules that facilitate coordination and ensure participants’ good behaviour. The second objective, discussed in Chapter 3, is to identify the premises under which this organisational form is supposed to work, evaluate to what extent it does, and how this affects power dynamics in the platform. The third dimension consists of changes in the relation between firms. Orchestration is a fundamental activity in platforms by which the platform owner influences the participants to attain desired platform-level outcomes. The third objective, considered in Chapter 4, is to assess the platform’s owner power to orchestrate changes in the participant firms’ way of creating and capturing value. As a thesis by publications, these objectives are addressed in separate chapters, each with its own research question and specific methods. Nevertheless, the research activities that produced each publication are intertwined and were guided by a shared research objective, design, and approach. This research’s object, the transformation of power in new organisational forms, is a complex, multifaceted, relational phenomenon and consequently demands an approach that can deal with this intricacy. Further, it has multiple expressions, requiring a way to rigorously integrate different types of sources and data types documenting the phenomenon. In addition, platforms evolve fast, so the approach needs to adapt and cope with the object developing as the research takes place. Grounded theory was identified as a suitable methodological stance to accommodate these requirements, as it can be regarded as a systematic but still highly flexible strategy oriented to theory building (Charmaz & Bryant, 2008; Thornberg & Charmaz, 2014). The method is best suited for exploring new elements or connections in the social reality and, hence, might adequately address power in new contexts. Among the three differentiated branches that evolved from Glaser & Strauss’ original framework (1967) – namely Classical/Glaserian (Glaser, 1987), Straussian (Corbin & Strauss, 2008), and Constructivist (Charmaz, 2000; Charmaz & Bryant, 2008) – this research is inspired by the Constructivist approach, as its higher flexibility and its gnoseological point of view better suit the research problem and context. Grounded research follows an iterative process, where decisions on data collection and analysis are taken vis a vis the progress in the theoretical development. From the point of view of logic, this approach implies a movement between induction and abduction, where data first inspires theory development, followed by interrogation of that theory and further data exploration, subsequently producing new provisional hypotheses (Thornberg & Charmaz, 2014). This iterative process facilitates data triangulation, enhancing the project's trustworthiness and increasing the potential for contribution. The cornerstone of data analysis and theory development in grounded approaches is the process of coding, starting with open coding (initial familiarisation and exploration of the data), followed by focused or selective coding (selection from the initial codes based on analytical relevance or interest), that is combined with theoretical coding (categories brought the researcher based on existing theories) (Charmaz & Bryant, 2008; Kenny & Fourie, 2015; Thornberg & Charmaz, 2014). Theoretical saturation is the methodological tool used to give closure to this iterative analysis process, meaning that data fills the properties of a theoretical category (Charmaz & Bryant, 2008). While Chapter 2 is based exclusively on conceptual development, Chapters 3 and 4 are based on an in-depth case study of the Amazon Marketplace. The case study is informed by semi-structured interviews, participatory 11 Essays on Power in Digital Platforms observation, and various types of archival sources and was conducted between November 2019 and December 2022. These sources were used for both chapters, except for Amazon’s financial statements, that were used only in Chapter 4. The Amazon Marketplace is the platform Amazon uses to run its retail business, where consumers can purchase products sold by Amazon and third-party sellers (“sellers” from now on). The Marketplace is a wellsuited empirical setting to conduct this research, being simultaneously one of the longest-standing and largest platforms worldwide. The study focused on Amazon’s and the sellers’ practices, and how these affected each other. This focus yielded rich information about the organisational and power dynamics emerging from the relationship between the platform owner and a group of highly heterogeneous organisations – the sellers –. 1.1.3. This research in context: The impact of the COVID-19 pandemic The COVID-19 pandemic had various effects on this research, including on me personally as a researcher, in the research activities, and in the object of study. I provide next a brief account of them to build additional context for this research. I moved from Buenos Aires (Argentina) to Guildford (UK) in September 2019 to start the research that produced this thesis. Only six months later, in March 2020, the outbreak of COVID-19 was declared a pandemic, and all levels of social life were heavily disrupted. Along with the rising uncertainty and generalised social anxiety, all the activity in the University went soon online, and the people I knew (mostly national and international students) left the area to stay with their families. Previous to the outbreak, I had a flight booked to Buenos Aires to meet my (then) soon-to-be newborn nephew, in late March. After discussing it with my supervisors, we agreed I’d travel and continue the research activities online for the moment. Argentina’s air space closed just a few days before the flight. I managed to change my flight to Rio de Janeiro (Brazil) - where travelling was still possible -and after 24 hrs hoping through airports and taxis, I entered Argentina via Puerto Iguazu’s triple border by land. I made it to Buenos Aires using an emergency government charter flight, transporting groups of strained citizens, on the same day that a total lockdown was imposed in the country. The following months were marked by constant tension between the evolution of the sanitary situation, the assessment of the possibilities of physically returning to the University - driven by changes in local and international regulations -and the varying social unrest. Due to this changing situation, my experience in that period was very different from that of most people, who felt stuck and that could not move, as I moved fifteen times in twelve months. These conditions, while demanding significant efforts, also imbued me with energy and inspiration for the research. During these challenging times, I received full support from my supervisors and the University for better adapting to these conditions. The pandemic also impacted the research as it altered the possibilities of conducting the planned activities. On the one hand, the data collection plan was fundamentally based on interviews. The pandemic significantly delayed the recruitment of informants (described below in sections 3.3.2 and 4.3.2). In addition, limitations for in-person meetings and travelling reduced the possibility of conference participation and other formal and informal spaces for discussing the progress of the research. Finally, the pandemic also impacted the object of this study — first, limitations to move and meet dramatically increased society’s reliance on digital technologies. Most large digital platforms played essential roles during the pandemic, facilitating connectivity, and supporting governments with activities such as data analysis and logistics. 12 Essays on Power in Digital Platforms This position naturally meant an increasing importance and power of these companies. Second, the generalised restrictions for people to leave their houses led to a sharp and fast increase in e-commerce, abruptly increasing the volume of Amazon’s and its marketplace’s operations. This, in turn, stressed the system and created new tensions to be addressed between the company, the sellers, and the logistics providers. 1.2. Central concepts and related research This section provides a contextualisation for the general problem and objectives discussed in this thesis and the specific objectives addressed in chapters 2, 3, and 4. As each chapter counts with a review of the literature relevant to the particular problem addressed, I here offer a broader and deeper discussion of the fundamental concepts structuring this research. First, I review the concept of platform. By revisiting the foundational work of three complementary perspectives on platforms, I contrast the different ways platforms are characterised and deemed novel research objects. Second, I summarise extant research on platform power. I identify three approaches from which platform power has been studied, each with its focus and conceptual background. 1.2.1. Conceptualizing digital platforms Previous efforts to systematise platform research concur that two complementary approaches have laid the foundations for the conceptualisation of platforms (Gawer, 2014; McIntyre & Srinivasan, 2017; Reuver et al., 2018; Rietveld & Schilling, 2020; Saadatmand et al., 2019): One is based on the industrial organisation economics literature, portraying platforms as market facilitators, and the other is based on the product development and information systems literature, characterising platforms through their technological architecture. The bridging of these research streams (Gawer, 2014) gave birth to the organisational view of platforms that conceptualise them both as agents and as organising actors endowed with a unique structure. In this section, I highlight the theoretical underpinnings of these approaches and their primary contribution to the understanding of platforms. 1.2.1.1. Platforms as market facilitators In the economics literature, platforms are conceptualised as “multi-sided markets”: intermediaries that bring together, coordinate (match), and facilitate exchanges between heterogenous groups of actors by aligning market incentives (Caillaud & Jullien, 2003; Rochet & Tirole, 2003). A typical example of a platform acting as a market facilitator is Uber: the platform provides a space where different types of agents looking to exchange (drivers and passengers) are matched based on parameters such as time and proximity. The economics literature on platforms is rooted in two streams of research, both closely connected to the diffusion of information technologies in business activities. The first one consists of the foundational economic studies on networks’ properties (Katz & Shapiro, 1985, 1986), which introduced the central concept of network externalities, a situation where “the utility that a user derives from consumption of the good increases with the number of other agents consuming the good” (Katz & Shapiro 1985; 424). This characteristic makes networked markets prone to converge to a single product or technology, as each consumer will be better off in the network with more consumers. This stream argues that a sponsor of a specific networking technology can strategically 13 Essays on Power in Digital Platforms promote specific asset investments and take advantage of network effects to tip the market (Katz & Shapiro, 1986). The second stream focuses on the problem of matchmaking and the coordination of heterogeneous groups of users. Through studies on Internet information brokers and card payment networks they introduce the concept of two-sided markets (Baye & Morgan, 2001; Rochet & Tirole, 2002; Schmalensee, 2002). Here, the network coordinator can both maximise its profits and consumer participation by adequately balancing the fee structure, charging different prices to different groups of consumers and cross-subsidizing them. From the economics perspective, a platform acts as a matchmaker between groups of actors previously impeded to exchange because of the existence of transaction costs. The “chicken-and-egg” problem of attracting and keeping the different sides of the market is solved by defining a split price structure to cross-subsidize between them, hence forming a two-sided market (Caillaud & Jullien, 2003; Rochet & Tirole, 2003). By enabling the transactions, platforms internalise the network externalities and consequently increase social welfare (Evans, 2003). Since the size of network effects typically grows more than proportionally to the cost of expanding the network, it is argued that concentration is likely to be the most efficient market configuration in this context (Caillaud & Jullien, 2003; Evans, 2003). Subsequent foundational studies enriched this framework by distinguishing direct network effects (among the same type of actors) from indirect network effects (across different types) (Parker & Van Alstyne, 2005) and demonstrating their importance for determining the price structure and consumer multihoming decisions (Armstrong, 2006). Other contributions explored additional critical decisions in platform markets, such as how to adequately identify the platform’s sides and scope (Hagiu, 2006), how to define the governance and ownership structure (Nocke et al., 2007), and how to build strategic competitive advantages in platform competition (Eisenmann et al., 2006). Figure 1 shows digital platforms as conceptualised by the economics literature, that is, as multi-sided markets. The focal point is “the space for interaction”, created by the platform owner. Its role is to provide the vision and manage the balance within the space, a service for which it profits from interactions. The “sides” are each composed of relatively homogenous agents that need to connect to the other sides, but this is hindered by the existence of transaction costs. As the common space makes interactions possible, value is created by internalising the latent network externalities. 14 Essays on Power in Digital Platforms Figure 1: Platforms as market facilitators Transaction costs Transactions Agent group A Agent group B Value creation Value capture Matchmaking Incentives design Safeguard transactions Platform Owner Platform owner’s roles Value creation activities Platform’s value capture 1.2.1.2. Platforms as technological architectures The product development and information systems literature has characterised platforms based on their particular hierarchically layered technological architecture (Baldwin & Woodard, 2009; Tiwana et al., 2010). This stream’s primary motivation has been to explore how the configuration of this architecture and the relationships of cooperation and competition among its participants have greatly facilitated innovation efforts (Adner & Kapoor, 2010; Gawer & Henderson, 2007). In the field of product innovation, a platform is defined in relation to the idea of modularity (Baldwin & Clark, 2000; Baldwin & Woodard, 2009). A modular design is characterised by a group of fundamental and stable functionalities that are redeployed in combination with other complementary functionalities to conform a family of products (Ulrich, 1995). Interfaces are the rules governing the interactions among parts (Baldwin & Clark, 2000) and contribute to reducing systems’ complexity by reducing the amount of information required to develop each functionality (Gawer, 2014). Considered together, these elements define a platform’s architecture: the stable functionalities are named the platform’s core, the variable elements are the platform’s periphery, and their interaction is facilitated by the interfaces (Baldwin & Woodard, 2009; Tiwana et al., 2010). A classic example in this approach are smartphone’s operating systems, such as Android and iOS: the operating system contains the phone’s core functionalities, while the apps, developed by third-party developers through standardized interfaces, provide additional functionalities and constitute the periphery. 15 Essays on Power in Digital Platforms Modular architectures lead to economies of scale and scope, which are strong incentives to use modular design strategies (Baldwin & Woodard, 2009; Gawer, 2014). Economies of scale can be achieved through the reutilization of core functionalities, the amortisation of fixed costs, and the recombinant innovation of modules, while economies of scope can be gained mainly through the reduction in the cost of new product development. Critically, the extent of the economies achieved, and the system's potential to foster innovation will depend on adequately designing the architecture and defining which element plays each role and how they interact. Since this research deals with platforms that provide the basis for interaction among formally independent organisations (“industry platforms” [Gawer & Cusumano, 2014a]), solving the architecture design problem involves a complex strategic interplay. The organisation that solves some essential problem in the industry and provides the vision of how a product, technology, or service can be at the centre of the network becomes the platform “hub”, “owner”, “leader”, or “orchestrator”, with design rights over the core functionalities and the interfaces (Cusumano & Gawer, 2002). In turn, the agents contributing by creating, expanding, and delivering additional value building on top of the platform’s core are called “complementors” (Gawer & Henderson, 2007). It is not unusual for the hub to hide the core’s design parameters to protect its resources and make its functionality reachable only through visible interfaces (Baldwin & Woodard, 2009). Nevertheless, this setting creates the space for cooperation with minimum coordination requirements: by complying with the design parameters established by the hub, each module can be developed, operated, and even priced independently (Jacobides et al., 2018). The existence of the centre-periphery structure and the division of roles motivated scholars to explore how the hubs’ behaviour influences the overall system’s innovation dynamics, particularly concerning two decisions: platform openness and governance rules. Since interfaces are how the platform enables its connection with other actors, its degree of openness – accessibility to build interoperability by third parties – defines its potential to attract complementors. However, achieving a virtuous innovation cycle requires a delicate balance (Boudreau, 2010; Gawer & Henderson, 2007): ease of integration maximises participation and, therefore, the potential to add on and tap into external resources. On the other hand, it also provides opportunities for competitors to thrive using the hub’s unique developments and resources (Benlian et al., 2015; Boudreau, 2017). The definition of the platform’s governance is also critical to achieving harmonious growth in complementors’ engagement: while stringent rules can repress innovation potential, being too lax will turn the platform into a “no man’s land”, deteriorating commonalities and eroding its value (Boudreau, 2012; Tiwana et al., 2010). In this context, scholars have focused on how to design rules that align incentives to enable both competition and cooperation, as well as to prevent abuses of “star” complementors that might create bottlenecks (Daymond et al., 2022; Wareham et al., 2014). Figure 2 represents a platform based on its technological architecture. Within this framework, interactions occur under specific premises: actors do not interact directly but through devices that communicate with interfaces specified for each type of user. Interfaces grant and regulate access to the core functionalities of the platform. They further enable innovation by complementors, who develop products or services in the periphery and expand the platform’s functionalities, increasing overall value creation. The platform owner is responsible for designing the architecture and defining the governance layers that regulate the users’ and complementors’ participation. 16 Essays on Power in Digital Platforms Figure 2: Platforms as technological architectures Complementors Innovation Periphery Core Users Users Design architecture Define governance Platform Owner Platform owner’s roles Value creation activities Platform owner’s value capture Governance layers Technological architecture Interfaces Devices 1.2.1.3. Platforms as organisational forms While both the economics and the information systems frameworks offer valuable answers to the specific problems each addressed (how platforms create and reshape markets and foster innovation), these frameworks were essentially unconnected. The lack of cross-pollination between these streams offered an opportunity to develop integrative frameworks and unify the platform theory. The organisational view of platforms was an answer to that vacancy, first introduced by Gawer (2014) and later developed and expanded through subsequent studies (Gawer, 2021; Jacobides et al., 2022; Kretschmer et al., 2022; McIntyre et al., 2021; Rachlitz, 2023) The organisational approach expands the previous characterisations because, while considering both the platforms’ role as interaction coordinators and the implications of their specific technological structure, it focuses on the platforms’ particular way of organising the multiple sets of relationships among its constituents (Gawer, 2014; Kretschmer et al., 2022; Rachlitz, 2023). They can therefore be considered “evolving organisations or metaorganisations”2 (Gawer, 2014; 1240), structures “between organisations and markets, providing a mixture of Gawer (2014)’s original characterisation includes three key features in her definition: “technological platforms can be usefully conceptualized as evolving organisations or metaorganisations that: (1) federate and coordinate constitutive agents who can innovate and compete; (2) create value by generating and harnessing economies of scope in supply or/and in demand; and (3) entail a modular technological architecture composed of a core and a periphery.” (p. 1240) 2 17 Essays on Power in Digital Platforms market-based and hierarchical power, and a mixture of market-based and hierarchical incentives” (Kretschmer et al., 2022; 407). Under this approach, platforms can be considered a “new kind of social ordering” (Rachlitz, 2023) or a “particular kind of technology of organising” that solves the problem of coordinating autonomous entities toward a systemlevel goal without hierarchical formal authority (Gulati, Puranam, and Tushman 2012). This form of organising can be found in very different organisational contexts (such as firms or industry ecosystems), in each of which its constitutive agents will likely vary as well (Gawer, 2014). This approach does not fix, based on pre-existing characteristics, the role that participants (which can be both individuals and organisations) adopt nor the nature of their interactions. Instead, their roles and interactions, and how they change, are explained by participants’ strategic behaviour, recovering the centrality of agency (Gawer, 2014). This conceptualisation implies that relationships between members can vary from cooperation to competition and back to collaboration, or even coexist simultaneously on different sides or features of the platform. As stated above, this approach recognises the platform’s structure as a defining aspect, but its importance is also framed in the context of agency: the primary source of authority in platforms regulating constituents’ actions emanates from the control of the technological core and the interfaces (Kretschmer et al., 2022). In this context, the platform owner has a decisive role. The centre-periphery structure defines a framework, the playing field within which the participants collaborate and compete with uneven success (Adner & Kapoor, 2010; Dhanaraj & Parkhe, 2006). The asymmetry crystallised in this relationship is the ability to set the rules and manage participation in the platform. Another dimension that distinguishes the organising in platforms is the nature of incentives and their role in coordination and governance. Since constituents are formally autonomous in most cases, their contribution in the platform is subject both to their individual success and to how they perceive the platform’s overall success influences it (Kretschmer et al., 2022). Consequently, motivation in the platform is shaped simultaneously by market mechanisms, by the governance rules set to oversee participation, and by boundary setting, “nudges”, and other tools guiding participants towards desirable contributions (Claussen et al., 2013; Ghazawneh & Henfridsson, 2013; Kretschmer et al., 2022) Adequately designing the rules of interaction while balancing the incentives constitutes a coordination problem specific to the platform. For example, Gawer & Cusumano (2014) compare it with how coordination works in two other industrial configurations: dominant design and supply-chain. The first refers to “the standard for what form and features users expect a particular product to take in the future” (Gawer & Cusumano, 2014; 420), and its defining characteristic is that it emerges within an industry without the need for central coordination to guide it. At the other end, a supply chain requires tight coordination and has a “master builder” who plans it in detail and controls its evolution. Platforms differ from both because while there is hierarchical coordination, it relies on how modularity is designed, and while the outcome can be induced, it cannot be tightly planned in advance. Figure 3 depicts platforms as organisational forms. From this perspective, the defining aspect of the platform is how it organises autonomous participants towards system-level goals of value creation, either by facilitating innovation or enabling transactions. Participants’ roles are not pre-defined, as they result from their strategies 18 Essays on Power in Digital Platforms within the conditions set by the platform owner: how it designs the platform and sets the governance and incentives. The differences in roles create a hierarchical structure within the platform that is presided over by the owner. Figure 3: Platforms as organisational forms Periphery Transactions Participants Innovation Participants Core Organizing participants through: - Platform design - Governance and incentives setting Platform Owner Platform owner’s roles Value creation activities Platform owner’s value capture Governance layers Technological architecture 1.2.2. Extant research on power in digital platforms As the platforms’ social and economic importance increased, accounts associating platforms with power abuses also became more prominent. Both academic researchers and journalists began documenting a new set of platform-related issues, such as consumer and business harm based on market dominance, widespread surveillance, social opinion manipulation, harmful psychological effects on users – such as addiction – and the worsening of labour conditions. In addition to documenting the abuses, researchers began to explore the sources of platform power and the specific mechanisms that lead to abuse. I distinguish three streams of research in platform power – market-based, sociomaterial, and techno-economic –, summarised in Table 1. Each stream has a different focus and conceptual approach, although we can trace some shared arguments in their explanations: one stream focuses on how platforms’ features create market power and affect market competition, the second on how the technical objects involved in society’s digitalisation reshape agents’ actions and the forms of control, and the third stream focuses on how the platforms’ specific means of resource extraction and production control creates new forms of power and dependence. 19 Essays on Power in Digital Platforms Table 1. Extant research on power in digital platforms Approaches to platform power Market-based Focus Identifying the mechanisms by which platforms can shape market structures and exchanges and influence competitive outcomes Characterisation of power Three primary sources driving platform power: - Gatekeeping, resulting from the specific intermediary role that platforms play, allowing its owner in concentrated markets to define the market’s rules - Data capture and control, used to both improve and develop services and create entry barriers to fend off competitors - Leveraging, consisting of the platform’s owner's effortlessness to achieve a dominant position in an adjacent market, based on access to large strategic assets (e.g., informational, financial, etc.) Sociomaterial Analysing norm production and control mechanisms by exploring how actors mobilise digital artefacts to reshape actions and choices. Power asymmetries are created by how technological objects are deployed in platforms - Algorithms automatically manage the digital interactions and objects that constitute platforms, concealing the rationalities and decisions behind their design - Encoding defines how offline objects and actions are translated into a digital language Consequences of power Harms to consumers: imposition of supra-monopoly prices, discriminatory clauses, or restricted access to the platform. Limitation of choice by the creation of lock-ins and behaviour manipulation Harm to businesses: preventing competitors from entering the platform’s market (competition “for” the market, e.g., use of data holdings as barriers to entry, creation of interoperability lock-ins) and preventing competition within the markets created by the platforms (competition “in” the market, e.g., effects of the platform owner entering its complementors’ market space) Notable examples in the literature Crémer et al. (2019); Furman (2019); Hagiu et al. (2020); Jacobides (2021); Khan (2016, 2018); Stigler report (2019); Zhu (2019) Digitalisation exerts a structuring effect on interactions, reshaping them. Choice is constrained and directed by the continuously restructured platform interfaces. Algorithms are socially embedded: shaped by visions and expected outcomes; they express and reinforce some realities at the expense of others. This creates “black boxes” that eschew questioning and abrogate responsibility. Alaimo & Kallinikos (2017); Beer (2017); Calo & Rosenblat (2017); Curchod et al. (2019); Eaton et al. (2015); Flyverbom et al. (2016); Kelkar (2018); Orlikowski & Scott (2015); Stark & Pais (2020) Control of crucial digital resources creates dependence as platforms can restrict consumer access, impose participation rules, expand dominance into new markets, and further deploy additional forms of control. Andreoni & Roberts (2022); Coveri et al. (2022); Cutolo & Kenney (2021); Rikap (2022b); Srnicek (2016, 2021); Vasudevan (2022); Zuboff (2019) - Interfaces and other boundary resources manage users’ visibility and access to the platform’s functionalities Techno-economic Exploring how the interrelation between the platform firm’s technological features and its capital accumulation logic creates new forms of power and dependence The platform firms’ infrastructural position grants them control over resources that became critical for trading, working, and communicating in the digital phase of capitalism. New accumulation logics are based on surveillance and data capturing, for which platforms have been optimised. Platform firms act as capitalist rentiers since their infrastructural position creates barriers to capital mobility, originating rents that the platform firm captures. Source: Based on the literature reviewed 20 Essays on Power in Digital Platforms 1.2.2.1. Market-based platform power One approach to platform power, developed by industrial organisation and competition law scholars, focuses on the specific mechanisms by which platforms can shape market structures and exchanges and influence competitive outcomes (Busch et al., 2021; Jacobides, 2021; Khan, 2016, 2018; Lianos & Carballa-Smichowski, 2022). As a topical issue, there is an ongoing and vigorous debate between regulatory agencies and scholars about what market power and its abuse means within digital platforms (Crémer et al., 2019; Cusumano et al., 2021; Furman, 2019; Hagiu et al., 2020; Peitz, 2023; Petit & Teece, 2021; Stigler report, 2019). In the heat of this debate, three primary sources that drive platform market power have been distinguished: gatekeeping, data capture and control, and asset leveraging. Gatekeeping results from platforms' specific intermediary role in concentrated markets (Khan, 2018). Driven by network effects, digital markets tend to create winner-takes-all-or-most outcomes (Cusumano et al., 2019), with one or a few companies controlling the market. In this context, a platform has a favourable position from where to define the market’s rules, including who can access it and how the transactions occur, as well as to charge supramonopoly prices (Busch et al., 2021; Crémer et al., 2019; Furman, 2019; Khan, 2018; Stigler report, 2019). Data capture and control is a fundamental part of platform business models (Crémer et al., 2019; Furman, 2019; Khan, 2018). User and data bases are some of the more valuable assets platform companies have, providing the source to improve existing services and develop new ones and create entry barriers to fend off competitors. By collecting unprecedented amounts of user data, platform owners can thoroughly understand their users’ needs and preferences. This knowledge can be used to improve targeted marketing and product development but can also be monetised through business intelligence services to third parties (Busch et al., 2021; Crémer et al., 2019; Furman, 2019). Leveraging consists of the platform’s owner's effortlessness to achieve a dominant position in a market that is both different but adjacent to the one in which it is currently positioned (Khan, 2018). This conduct is possible due to the platform’s owner's access to large informational, logistical, and financial assets and because digitalisation eases the creation of large economies of scale (such as the ability to manage additional operations and users with a negligible marginal cost) and scope (like the use of large databases from one market to understand and tip another one). This has led large platform firms to enter apparently unrelated markets and quickly achieve a dominant position in them (Crémer et al., 2019; Furman, 2019; Stigler report, 2019). The market-based approach to platform power contributes by identifying anti-competitive platform practices and how they can cause harm to individuals and businesses. First, effects on individuals are assessed considering their role as consumers. Consumer harms derived from gatekeeping power include, for example, facing supra-monopoly prices, imposing discriminatory clauses, or restricting access to the platform (Furman, 2019; Gawer & Srnicek, 2021; Stigler report, 2019). The capture and manipulation of data have also been identified as a potential source of consumer harm, such as the limitation of choice by the creation of lock-ins and behaviour manipulation due to the use of nudges and “dark patterns” in user interface design (Busch et al., 2021; Crémer et al., 2019; Furman, 2019; Gawer & Srnicek, 2021; Khan, 2018). 21 Essays on Power in Digital Platforms Next, anti-competitive practices affecting businesses can be divided between those done to prevent competitors from entering the platform’s market (competition “for” the market) and those done to avoid competition within the markets created by the platforms (competition “in” the market) (Furman, 2019). Examples of practices limiting competition for the market are the use of data holdings as barriers to entry, the creation of interoperability lock-in to prevent user multi-home to other platforms, or the use of predatory pricing to capture and tip the market (Crémer et al., 2019; Furman, 2019). Competition in the market has been approached by the literature of competition with complementors and dual marketplaces (Hagiu et al., 2020; Zhu, 2019), exploring the effects of the platform owner entering its complementors’ market and the strategies used by both them in preparation and response to that action. These contributions identified product spaces more likely to be entered (Zhu & Liu, 2018), how complementors’ entry decisions (Lan et al., 2019) and profitability (Hagiu et al., 2020) are impacted by the platform’s business model, and responses in complementors’ innovation and value-capture strategies when facing platform’s entry threats (Wen & Zhu, 2019). 1.2.2.2. Sociomaterial approaches to platform power Another stream of researchers has approached platform power by focusing on society’s digitalisation and how the specific technological objects and processes deployed in platforms create different power asymmetries. These scholars discuss how digital artefacts are mobilised by actors to reshape actions and choices, analysing the mechanisms of norm production and control in the digital realm (Alaimo & Kallinikos, 2017; Beer, 2017; Flyverbom et al., 2016; Orlikowski & Scott, 2015). Algorithms are one of the central objects in this stream of research since, as a sequence of self-contained actions, algorithms automatically manage the digital interactions and objects that constitute platforms. The basic algorithmic function of search and match creates a multi-layered problem of power asymmetry since the inspection and control of its computing processes pose technical challenges for users and regulators (Beer, 2017; Just & Latzer, 2017). The criteria for how results are displayed and sorted, the fairness in the treatment of data entities, and the limitation of results (filtered selections) are of social interest while also being carried out in camera (Faraj et al., 2018). Maintaining the inscrutability of algorithms is argued to be necessary for competitive reasons, as their secrecy is usually critical to retaining competitive advantage (Agrawal et al., 2019). However, studies identifying algorithmic bias and discrimination have raised concerns about this issue and call for mechanisms of social algorithmic audition (Danaher et al., 2017; Larsson & Heintz, 2020). Furthermore, bias might be found not only in the algorithm's rules but also as an emergent property of the data used as input (as was observed for racial biases), creating additional complexity for regulation (Noble, 2018). The challenge to unpacking the consequences of algorithms for power lies in the need to grasp their technical features in conjunction with a socially structured analysis of the device (Beer, 2017; Introna, 2016; Orlikowski & Scott, 2015). This issue highlights the social embeddedness of algorithms, as they are shaped by visions and expected outcomes. Beer (2017) argues the need to move beyond the “immediate consequences of code” and understand algorithms' different roles in social processes. In particular, he draws attention to how the notion of algorithm circulates in the social fabric and notes its symbolic power to determine how things are or should be done. In a similar vein, Orlikowski and Scott (2015) delve into the algorithms’ performative character, as the constituted-in-practice materialisation of a service. Furthermore, they highlight that algorithms express and 22 Essays on Power in Digital Platforms reinforce some realities at the expense of others. As “algorithms act; they do things” (Orlikowski & Scott, 2015), we are brought back to the problem of how they make decisions and also how they constitute the world in which we make our choices (Beer 2017). The “black box of algorithmic culture” (Orlikowski & Scott, 2015; Pasquale, 2015; Striphas, 2010) creates a world of secrecy that eschews questioning and abrogates responsibility (Introna, 2016). In digital platforms, algorithms create specific power asymmetries as they are used for “algorithmic management” of social interactions (Calo & Rosenblat, 2017; Curchod et al., 2019; Lee et al., 2015; Stark & Pais, 2020). Typically deployed for this purpose in digital marketplaces for service provision, such as Uber, TaskRabbit, eBay, or Upwork, algorithms offer an unprecedented capacity to monitor and correct specific tasks and reshape labour conditions (Calo & Rosenblat, 2017; Neyland & Möllers, 2017; van Doorn & Chen, 2021). Algorithms transform the typical managerial function of evaluation, dislocating both hierarchical power and the subordinates’ exercise of agency (Curchod et al., 2019): as managers, algorithms might prove to be inaccessible, unappealable, and tireless, constraining human agency and negating the possibility to adapt, respond, or deviate from control mechanisms (Curchod et al., 2019; Introna, 2016). Encoding is a crucial process when assessing digitalisation’s power dynamics, as it is the activity that defines how offline objects and actions are translated into a digital language (Alaimo & Kallinikos, 2017). Once codified, actions must be performed in accordance with the rules established by the programmer. Thus, the codification and the rules exert a structuring effect over the initial interactions, shaping them (Alaimo & Kallinikos, 2017; Curchod et al., 2019). It has even been claimed that encoding makes cyberspace an arena of potentially perfect control (Landini, 2016) since it creates the actions that users are invited to perform in a way that can be stored and processed. Finally, another set of studies has explored how the design and manipulation of platform architectural elements, such as interfaces and other boundary resources (Ghazawneh & Henfridsson, 2013), shape power dynamics in the platform. As stated above (section 1.2.1.2), interfaces manage access to the platform’s core functionalities, and consequently, their design plays a fundamental role in defining the relationship with the platforms’ constituents. Indeed, a defining feature of the changing scope of the category of user is the fact that they are limited and controlled by various kinds of interfaces (Kelkar, 2018). This means that the type of tools available to users, what is visible to them and what is not, and the amount of freedom they enjoy in the digital platform are decisions taken at the architecture design level. In this train of thought, Bradshaw and Howard (2018) highlight how choice is constrained and directed by the continuously restructured platform interfaces designed to satisfy people’s various desires for identification and social connection. In addition to the impact on individuals, scholars have explored how the evolution of the deployment and use of boundary resources has consequences regarding power relationships among businesses. For example, Eaton et al. (2015) describe the “paradoxical tensions of generativity and control” arising from the cycles of resistance and accommodation between software developers and Apple in the iOS ecosystem. These cycles, whose outcome is innovation in the form of new software applications, result from how the differences in actors’ resources affect their easiness of overcoming the limitations imposed by the boundary resources. 23 Essays on Power in Digital Platforms 1.2.2.3. Techno-economic platform power A third stream of research has focused on how the interrelation between the platform firm’s technological features and its capital accumulation logic creates new forms of power and dependence. This approach is developed mainly by political economy scholars that argue limitations in the explanations of platform power based exclusively on market relations, bringing forward the importance of resource extraction and production control for value appropriation (Cutolo & Kenney, 2021; Langley & Leyshon, 2017; Rikap, 2022b; Srnicek, 2016; Vasudevan, 2022; Zuboff, 2019). This stream associates the advent of platforms with a profound change in social and economic relations and how capitalism works. One of the most influential contributions from this perspective is Zuboff’s thesis of “surveillance capitalism” (2019). This theory argues that with the advancement of large technological companies into most aspects of social life, capitalism has entered a new phase of accumulation based on surveillance that claims human experience as free raw material for translation into behavioural data. While some of this data is applied to product development, most is appropriated as “behavioural surplus”: data is fed into machine intelligence to create prediction products - anticipating what a user will do – sold in specific advertising markets. Furthermore, tech companies can produce statistically relevant user behaviour modifications by nudging, tuning, herding and conditioning users through the computational infrastructure. Zuboff summarises these dynamics by arguing the appearance of a new “instrumentarian power”, consisting of capabilities for computational-based behaviour shaping toward specific ends to achieve “guaranteed outcomes”. In a similar vein, other scholars have proposed the hypothesis of “platform capitalism“ (Langley & Leyshon, 2017; Srnicek, 2016). Rather than being solely a manifestation of wider social transformations, these authors depict platforms as “a discrete mode of socio-technical intermediary and capitalist business arrangement”, driving a profound change in economic circulation (Langley & Leyshon, 2017; 15). By providing a digital space for others to interact, platforms are considered the vehicle optimised to extract and exploit a crucial and now abundant raw material: data (Srnicek, 2016). Data can now be captured from an ample variety of sources such as natural processes (e.g., weather conditions, crop cycles, etc.), industrial activities (e.g., assembly lines, continuous flow manufacturing, etc.), and other businesses and users’ behaviours (e.g., web tracking, usage data, etc.). Finally, data serves vital functions that explain the platform’s dominance: they create competitive advantages by educating algorithms; they enable the coordination and outsourcing of workers; they allow for the optimisation and flexibility of productive processes; they make possible the transformation of low-margin goods into high-margin services; and data analysis is itself generative of new data, in a virtuous cycle (Srnicek, 2016). A common theme in explaining platform power from this approach is how the platform firms’ infrastructural position creates dependence for consumers and for other businesses (Andreoni & Roberts, 2022; Coveri et al., 2022; Cutolo & Kenney, 2021; Li & Qi, 2022). Unlike previous forms of monopolisation, limited to specific markets or sectors, platform firms control resources critical to all digital economy participants (Coveri et al., 2022; Li & Qi, 2022). Control of these resources creates dependence because platforms can restrict consumer access, impose participation rules, and further deploy additional forms of control. Cutolo & Kenney (2021) explain the dynamics of dependence over time, describing how power asymmetries evolve with the participants’ involvement in the platform. Using the example of entrepreneurs, they show how while new entrants typically benefit from the 24 Essays on Power in Digital Platforms resources provided by the platform that enable a fast and cost-effective way to enter a market, more mature businesses become locked in and highly dependent on the platform’s decisions. Further, the creation of dependence relations can also be identified between platforms: Li & Qi (2022) highlight platforms’ heterogeneity regarding control over digital infrastructures, leading to the existence of a hierarchical power structure where while small platforms have some power over users, large platforms dominate both of them. Finally, one of the expressions of platform power emerging in this approach is the characterisation of platform firms as capitalist rentiers (Andreoni & Roberts, 2022; Rikap, 2022b; Srnicek, 2021; Vasudevan, 2022). As argued above, the platform’s infrastructural position produces dependence and further creates barriers to capital mobility, originating rents captured by the platform firm (Vasudevan, 2022). This stream has identified several different forms of rent derived from infrastructural asset control, such as advertising, monopoly, innovation, or political rents (Andreoni & Roberts, 2022; Srnicek, 2021). Other factors besides the platform’s infrastructural position are also associated with rent capturing. For example, Rikap (2022) describes how platforms appropriate knowledge and data and transform it into rent-bearing intangible assets such as the platform’s core algorithms. 1.3. Thesis overview This section describes this thesis’ structure and summarises each chapter’s central argument. The body of this thesis by publications consists of five chapters plus the references and appendices. The current chapter (Chapter 1) provides a general introduction to this research, chapters 2, 3, and 4 contain the publications, and chapter 5 discusses the thesis’ findings and contributions and offers some general conclusions. Table 2 summarises the publication chapters' authorship, research questions, methods, findings, and specific contributions. The current chapter is a general introduction to this study, presenting the problem statement and motivation, the general and specific objectives, the methodological approach, and the research context. The leading research problem is understanding how power is reconfigured in platforms as new organisational forms. To contextualise this problem, this chapter reviews the concept of platform by revisiting the foundational work that built three different perspectives on platforms: as market facilitators, as technological architectures, and as organisational forms. Next, it summarises extant research on platform power and identifies three main approaches: marketbased, sociotechnical, and techno-economic. To conclude, it provides an overview of the thesis. Chapter 2, “How digital platforms organise immaturity: a sociosymbolic framework of platform power”, conceptualises how the way platforms structure social interactions and practices shapes power relations and agents’ autonomy. A guiding construct in this chapter is that of organised immaturity, defined as the erosion of the individual’s capacity for public use of reason (Scherer & Neesham, 2020). Drawing on Bourdieu’s concepts of field, capitals, and habitus, this chapter provides a sociosymbolic perspective on platforms’ power dynamics. Digital platforms are characterised as organisations mediated by a digital infrastructure and a digital habitus in which agents accumulate capitals by operating in a field. Power dynamics in the platform are explained by the agents’ practices, positions, and strategies and how the digital infrastructure and the digital habitus shape these. This chapter builds a framework identifying five forms of platform owner power accumulation (constitutional, juridical, discursive, distinction, and crowd) and two forms of counterpower (crowd and hacking). It also explains the evolution of the power dynamics over time and proposes a three-phase model in which the forms of power operate. These phases are platform formation and field enclosure, platform domination within the original field, 25 Essays on Power in Digital Platforms and platform cross-field expansion. Overall, this chapter offers a holistic and dynamic understanding of platform power, where multiple coexisting forms of power are exerted and confronted by different types of agents. It also explains the organising dynamics of immaturity and how immaturity is constituted in practice. Chapter 3, “Two-faced governance in platform ecosystems”, explores the problem of the platform’s rules for participation and its compliance through an inductive grounded study of the Amazon Marketplace. Discussing established claims in the economics and management literature on platform governance, this chapter explores the existence of discrepancies between a platform owner’s declared ecosystem governance and its actual practices and their consequences. The chapter shows the existence of two sets of governance practices: on the one hand, Amazon’s officially declared set of rules, roles, and activities and, on the other hand, a set of practices and “unwritten rules” of participation that depart from the official governance. It documents these discrepancies and their impact on the Marketplace’s sellers. Further, the study also identifies a set of adaptive practices that sellers adopt, finding that they, too, resort to sets of declared and undeclared practices. The chapter consequently argues the existence of a “two-faced governance” strategy in the Marketplace, defined as the platform owner’s practice of pursuing simultaneously seemingly contradictory governance practices. This two-faced platform governance creates unpredictability for sellers and enhances, at least in the short term, Amazon’s bargaining power over them. Overall, this chapter enriches the construct of platform governance by identifying that there are two types of governance practices (declared and undeclared), showing how the two interact and their consequences for power dynamics. Chapter 4, “The evolutionary dynamics of platform orchestration”, examines the reach of the platform’s owner orchestration capabilities by assessing its influence on its complementors’ business models through an inductive long-term study of the Amazon Marketplace. Observing the existence of multiple and heterogeneous business models for sellers in the Marketplace, this chapter first identifies the process through which new sellers’ business models appear and are adopted in the Marketplace. It then shows how Amazon influences this process, identifying the specific orchestration activities used and how they operate. The chapter then illustrates this by describing Amazon’s influence on the appearance of new sellers’ business models in two different periods. Next, it describes how the appearance of new business models benefits Amazon by improving its value proposition and increasing its value capture. Finally, the chapter discovers a pattern of dynamic adjustments between Amazon and the sellers’ business models, resulting in an evolutionary interplay where both are transformed. Overall, the chapter illuminates new platform owner orchestration capabilities and identifies ways in which it can create forms of complementor dependence. Chapter 5 discusses the contributions of the previous chapters, showing how this thesis enriches our understanding of platforms and how we characterise platform power, its dynamics, and its consequences. Overall, it argues that this research illuminates new aspects of platforms as organising agents and provides a multifaced, multidirectional, and dynamic characterisation of platform power. The chapter highlights the identification of new strategies of platform power, how these strategies change vis a vis the platform, and further the recognition of the consequences of platform power over autonomy and decision-making, both on individuals and organisations. This chapter also discusses implications for policy, the research’s limitations, and avenues for future research. It concludes with some final remarks about the potential social implications of this research for platforms. 26 Essays on Power in Digital Platforms Table 2: Summary of publications’ chapters Chapter 2: How digital platforms organise immaturity Chapter 3: Two-faced governance in platform ecosystems Chapter 4: The evolutionary dynamics of platform orchestration Authorship Martín Harracá, Itziar Castelló, Annabelle Gawer Martín Harracá, Annabelle Gawer Martín Harracá Publication / presentation Published in Business Ethics Quarterly (March 2023), doi:10.1017/beq.2022.40. Presented as a conference paper at the 2021 EGOS and the 2022 AoM conferences. Presented at the 2022 Platform Strategy Symposium, the 2022 RaDMA, 2023 EGOS, and the 2023 SASE conferences. It is currently under review for publication. Presented in the PhD workshop of the 2023 EU-DPRN conference and accepted to be presented at the 2023 ICIS Conference (acceptance rate ~25%). How do digital platforms organise immaturity? Are there sustained discrepancies between a platform owner’s declared ecosystem governance and its actual practices? What are the consequences of these discrepancies on the platform owner and its ecosystem members? How do platform owners orchestrate the appearance of new complementors’ business models? How does the platform owner benefit from this? Conceptual development Inductive grounded study of the Amazon Marketplace, based on interviews, participatory observation, and archival sources Discusses the platform’s role in organising immaturity, defined as the erosion of the individual’s capacity for public use of reason. Adopts a sociosymbolic perspective, based on Bourdieu’s concepts of fields, capitals, and habitus, to characterise digital platforms as organisations mediated by a digital infrastructure and a digital habitus in which agents accumulate capitals/power by operating in a field. Evidence of a “two-faced” platform governance, defined as the platform owner’s practice of pursuing simultaneously seemingly contradictory governance practices. It comprises the platform firm’s officially declared set of rules, roles, and activities and, on the other hand, a set of practices and “unwritten rules” of participation that depart from the official governance. Research question Methods Central argument and findings Specific contributions Develops an integrative and dynamic conceptualisation of platform power through a framework with five forms of power (constitutional, juridical, discursive, distinction, and crowd) and two forms of counterpower (crowd and hacking) accumulation along three-phases of field transformation (field enclosure, field domination, and cross-field expansion). Evidence of a set of adaptive practices that ecosystem participants adopt, also resorting to sets of declared and undeclared practices. This two-faced platform governance creates unpredictability for complementors and enhances, at least in the short term, the platform firm’s bargaining power over them. Explains the organising dynamics of immaturity based on the relations between the platform structure, the digital objects, and the agents’ strategies. Analyses how immaturity is constituted in practice and explains and nuances the possibility of emergence and the selfinfliction dimensions of immaturity. Contributes to the platform strategy literature by providing new evidence that runs contrary to the central claims of the platform governance literature. Builds a three-phase model of platform power dynamics over time. This model expands current views on platform power, providing a more holistic scheme in which power is accumulated and contested and highlighting how agents other than the platform owner play a role in producing and exercising forms of power. Enriches our understanding of how platform power can be identified and exploited. It provides evidence that platform power can be exerted in ways not previously specified by focusing on the strategic mismanagement of the rules and the creation of uncertainty. Enriches the construct of platform governance by identifying that there are two types of governance practices, declared and undeclared, and their interaction Identifies the process through which new sellers’ business models appear in the Marketplace. Evidence showing that Amazon successfully orchestrated the appearance of new sellers’ business models in its Marketplace. Identifies a pattern of dynamic adjustments between Amazon and the sellers’ business models, resulting in an evolutionary interplay where both are transformed. This interplay is triggered by Amazon’s orchestration activities, which, by providing functionalities, structuring seller competition, and increasing the competitive intensity sellers face, influence the sellers’ business model transformation. This transformation is instrumental in improving Amazon’s value proposition and increasing its value capture in the platform. Expands our understanding of the platform owner’s set of capabilities and responsibilities by showing that they can orchestrate changes in its complementors’ business models. Recognising these capabilities is relevant for research on platform power as it identifies ways in which the platform owner can create forms of complementor dependence. Contributes to the platform orchestration stream of literature by identifying new sets of the platform owner’s orchestration activities and explaining the mechanisms by which these activities transform the complementors’ business models. 27 Essays on Power in Digital Platforms 2. How digital platforms organize immaturity: A sociosymbolic framework of platform power3 Organized immaturity, defined as the erosion of the individual’s capacity for the public use of reason (Scherer et al., 2020), differs from other forms of control in that it is a self-inflicted and emergent (as opposed to orchestrated) collective phenomenon in which autonomy-eroding mechanisms mutually reinforce each other (Scherer et al., 2020; 9). The phenomenon of autonomy erosion and increasing user control has been discussed in the context of the dark side of digitalization (Flyverbom et al., 2019; Trittin-Ulbrich et al., 2021). Scholars have looked at how the automation of interactions through algorithms can lead to an emergent manipulation of choice and autonomy erosion (Alaimo & Kallinikos, 2017; Beer, 2017; Just & Latzer, 2017; Orlikowski & Scott, 2015), but there is still little exploration of the organizing role of platforms in this process. Digital platforms have been described as organizational forms that orchestrate activities between independent users through the use of digital interfaces (Constantinides et al., 2018; M. Cusumano et al., 2019; Gawer, 2014, 2021; McIntyre et al., 2021). Increasingly, scholars denounce the negative effects of power accumulation by digital platforms and platform owners. For example, studies of the structural constitution of markets criticize gatekeeping positions that impose discriminatory clauses or limit content access and creation, with consequences for users’ choices (Crémer et al., 2019; Jacobides, 2021; Khan, 2018). Other researchers, such as Kelkar (2018), Stark and Pais (2020), and Flyverbom et al. (2019), discuss sociomaterial perspectives on platforms and show how platform owners design the interfaces, prescribing what is accessible to users and what choices they may enjoy in the digital platform; this, again, restricts choice and creates negative psychological effects on users (Seymour, 2019; Wu et al., 2019). Lanier (2018) and Zuboff (2019) present systems of surveillance promoted by the power of digital platforms that explain how the datafication of human experience leads to increasing forms of domination. These studies provide valuable explanations of how the increasing power of platforms hinders freedom of choice and individual autonomy. However, their explanations are partial, focusing either on the market mechanisms that limit consumer choice or on the specific role of digital objects, such as algorithms, that constrain the platform users’ autonomy. The fundamental aspects of the organizing of immaturity, such as the tension between organizing and emergence, and the relationship between self-infliction and the power accumulation strategies of key agents, such as platform owners, remain unexplored though. These tensions are essential to explaining how organized immaturity is created and reproduced. We claim that there is a need to explain the power accumulation of the different agents of the platforms and its relation to the mechanisms that lead to the delegation of autonomous decision-making. Therefore, in this article, we ask, How do digital platforms organize immaturity? This chapter was published as a co-authored article with Itziar Castelló and Annabelle Gawer in Business Ethics Quarterly (2023), in the special issue “Socio-Technological Conditions of Organized Immaturity in the Twenty-First 3 Century”, doi:10.1017/beq.2022.40. It was also presented as a conference paper at the 2021 EGOS and the 2022 AoM conferences. 28 Essays on Power in Digital Platforms To tackle this issue, we build a sociosymbolic perspective of power accumulation in digital platforms inspired by Bourdieu’s writings (Bourdieu, 1977, 1979, 1984, 1987, 1989, 1990, 1991, 2005, 2011, 2014; Bourdieu & Wacquant, 2007). A sociosymbolic perspective supports building a dynamic conceptualization of power accumulation based on agents’ practices, positions, and strategies. The concepts of field evolution and habitus allow further explanation of the emergence of immaturity and the mechanisms of self-infliction. By situating the concepts of fields, capitals, and habitus in the context of digital platforms, we describe digital platforms as organizations mediated by a digital infrastructure and a digital habitus in which agents accumulate capitals by operating in a field. We explain the role of the digital habitus in organizing immaturity, complementing prior literature on materiality and affordances. We propose a framework of power accumulation in which the dynamics of platform owner power accumulation and counterpower accumulation coexist. The platform owner accumulates power in five forms: constitutional, juridical, discursive, distinction, and crowd. There are two forms of counterpower: crowd and hacking. We also explain the evolution over time of the power dynamics and propose a three-phase model in which the forms of power operate. These phases are platform formation, platform domination within the original field, and platform cross-field expansion. This framework makes two significant contributions. First, we build a theoretical apparatus that explains the organizing dynamics of immaturity by explaining the relations between the structure, the digital objects, and the platform owner’s power accumulation strategies. From these, we can explain the tension of emergence and selfinfliction. With this framework, we draw on sociological perspectives to expand the understanding of organized immaturity in digital spaces by focusing on describing the practices that constitute the webs of relations that configure the digital habitus and the processes of power accumulation. Second, we contribute to the platform literature by developing a three-phase model of platform power dynamics over time. This model expands current views on platform power, providing a more holistic scheme in which power is both accumulated and contested and highlighting how agents other than the platform owner play a role in producing and exercising forms of power. This article concludes by providing policy recommendations on how to understand and tackle organized immaturity and highlighting potential avenues for further research. 2.1. Organized immaturity Organized immaturity has been defined as a collective, albeit not necessarily orchestrated, phenomenon where independent reasoning is delegated to another’s guidance (Scherer & Neesham, 2020). It is inspired by the Kantian principle that humans should have intellectual maturity involving autonomy of judgment, choice, and decision-making without the guidance of an external authority. It also relates to the ability to use experience to reason and reflect critically and ethically on complex or problematic situations and to challenge norms and institutions (Scherer & Neesham, 2020). The concept of organized immaturity differs from other forms of control in two ways. First, it is a “self-inflicted” (Kant, as cited in Scherer & Neesham, 2020; 8) process, referring to harm done by humans to themselves, often in a nonconscious manner. From this perspective, “immaturity” is therefore a condition of the human being that arises when an individual defers or delegates their own autonomous reasoning to external authorities (Dewey, 1939). The second way in which organized immaturity differs from other 29 Essays on Power in Digital Platforms forms of control is that it is an emergent (as opposed to orchestrated) collective phenomenon in which autonomy-eroding mechanisms mutually reinforce each other (Scherer & Neesham, 2020; 9). According to Scherer and Neesham (2020), the study of immaturity relates also to its organizing elements. The perpetuation of modern forms of immaturity has been associated to organizations and institutions that create the conditions for the self-inflicted immaturity. Organized forms of immaturity have been addressed in the critical analysis of bureaucratic organizations, where the individual is subject to various forms of domination and control (Clegg, 1989; Hilferding, 2005). The Fourth Industrial Revolution (Philbeck & Davis, 2018; Schwab, 2017) has ushered in a consolidation of the globalized information and communication technologies that are driving the organization of economic life. However, the infrastructures and mechanisms behind these sociotechnological systems curb individual liberties and impact people’s autonomy (McCoy et al., 2018; O’Connor & Weatherall, 2019). The term organized immaturity is not explicitly used in most of the literature studying forms of control related to digitalization (with the exception of Scherer and Neesham [2020] and Scherer et al. [2020]), but scholars are increasingly analysing the “dark side of digitalization” (Flyverbom et al., 2019; Trittin-Ulbrich et al., 2021). In particular, attention has been directed to the use of big data and systems based on artificial intelligence and to how the automation of interactions through algorithms can lead to an emergent manipulation of choice. Even the basic algorithmic function of search and match creates power asymmetries, since the inspection or control of its guiding principles presents technical challenges for both users and regulators (Beer, 2017; Just & Latzer, 2017). Biases might be found in the criteria for how results are limited, displayed, and sorted (Faraj et al., 2018) and may even amplify properties of the data used as input, as has been observed in the context of racial biases (Noble, 2018). Researchers are increasingly pointing at the importance of unpacking the consequences of algorithms in conjunction with a socially structured analysis of the device (e.g., Beer, 2017; Introna, 2016; Orlikowski & Scott, 2015). Through this, they show how the “black box of algorithmic culture” (Orlikowski & Scott, 2015; Pasquale, 2015; Striphas, 2010) creates a world of secrecy that eschews questioning and abrogates responsibility (Introna, 2016), eroding autonomous decision-making. However, this emphasis on the artificial intelligence tools, algorithms, and coding processes that hinder autonomy in decision-making must be complemented by research into the organizing structures of immaturity, that is, the key organizing agents. Studying digital platforms can improve understanding about how organized immaturity happens, as these platforms organize social interactions and transform the power relations of the different agents who participate in the digital exchanges. 2.2. Platforms and the accumulation of power 2.2.1. Platforms as organizing agents In the platform literature, digital platforms have been described as new organizational forms that orchestrate activities between independent users through the use of digital interfaces (Gawer, 2014; Kretschmer et al., 2022; McIntyre et al., 2021). Platforms can be considered a “particular kind of technology of organizing” (Gulati et al., 30 Essays on Power in Digital Platforms 2012; 573) or “hybrid structures between organizations and markets” (Kretschmer et al., 2022; 4), as they use a mixture of market and hierarchical incentives to coordinate autonomous agents. Platform organizations are distinct from hierarchies, markets, and networks (Gawer, 2014) because, as Kornberger et al. (2017; 81) argued, “platform organizations question not only extant organization designs but also, quite fundamentally, the [Coasian] idea of the firm . . . and . . . of value creation processes.” Two fundamental characteristics define the digital platform as an organizing agent: how its digital architecture is structured and how it coordinates interactions. From an organizational perspective, platforms can be described by the common set of design rules that define their technological architecture. This system is characterized by a “core” functionality with low variety and a complementary set of “peripheral” functionalities with high variety (Tiwana et al., 2010). The rules governing interactions among the parts are the interfaces (Baldwin & Woodard, 2009). Interfaces contribute to reduce a system’s complexity by greatly simplifying the scope of information required to develop each functionality (Gawer, 2014). Together, the centre, the periphery, and the interfaces define a platform’s architecture (Baldwin & Woodard, 2009). The centre–periphery structure therefore defines an asymmetric framework in which the participants collaborate and compete (Adner & Kapoor, 2010), under conditions set by the platform owners on two elements: openness and governance rules (Boudreau, 2010; Gawer & Henderson, 2007). Platforms coordinate transactions by creating “multisided markets,” in which their owners act as intermediaries to bring together (match) and facilitate exchanges between different groups of users by aligning market incentives (Rochet & Tirole, 2003). Interactions occur in a networked structure, implying that the value derived from platform usage increases exponentially with each additional user (Katz & Shapiro, 1985). As the value for participants grows with the size of the platform, it is optimal for them to converge on the same platform, leading to the prediction that platforms will tend to create concentrated markets organized by increasingly powerful owners (Caillaud & Jullien, 2003; Evans, 2003). 2.2.2. Platforms’ accumulation of power and the consequences for individuals’ autonomy erosion The characteristics of platforms described in the preceding section have facilitated the accumulation of power by platform owners, leading to “new forms of domination and competition” (Fuchs, 2007; 7) that are increasingly eroding people’s capacity to make independent decisions. The consequences of the platforms’ power accumulation for manipulation of choice and autonomy delegation have been analysed from two perspectives: first, in relation to the structural constitution of markets and how this structure can lead to manipulation of users’ choices, and second, from a sociomaterial perspective that looks at the interaction of digital objects (e.g., algorithms) and the platform users. From the perspective of the structural constitution of markets, the accumulation of power and manipulation of choice is associated to the growing centrality of large platforms in the economy. Consumers and business partners can have their choices manipulated because of the specific intermediary role that platforms play. Once the market has been tipped, this role provides the platform owner with a position from which they can charge supramonopoly prices and define the rules of the market, including who can access it and how the transactions 31 Essays on Power in Digital Platforms occur (Busch et al., 2021; Jacobides, 2021; Khan, 2018). In this way, platforms are increasingly operating as gatekeepers, imposing discriminatory clauses or limiting content access and creation (Furman, 2019; Stigler report, 2019). Choice making can also be limited due to market concentration driven by platforms, in that a platform enhances its owner’s opportunities to leverage its assets (Khan, 2018). Thus the owner can entrench their (platform’s) position in a market and enter an adjacent one by creating economies of scale and scope occur (Jacobides, 2021; Khan, 2018). This brings the possibility of creating a dominant position in apparently unrelated markets through practices like vertical integrations, killer buys, predatory pricing, and self-preferencing (Crémer et al., 2019; Furman, 2019). In addition, the capture and control of transactional data may be used to improve platform services, while also enabling the creation of entry barriers that fend off competition (Khan, 2018). Market-based analyses provide a view of power accumulation based on asset control and market position. However, they have been criticized for overlooking the impact of other noneconomic dimensions and for portraying power as relatively unidirectional (Lynskey, 2017, 2019; Margetts et al., 2021). Such critiques recognize that the deep social impact of platform power cannot be tackled from a market perspective alone (Lianos & Carballa-Smichowski, 2022; Margetts et al., 2021). Sociomaterial perspectives place affordances and materiality of the digital objects at the centre of the platform interactions (Curchod et al., 2019; Fayard & Weeks, 2014; Kornberger et al., 2017). In this perspective, digital objects, such as code, interfaces, and algorithms, are described as central objects that can hinder autonomy. For example, when platform owners design the interfaces, they define the category of user, prescribing what is accessible to users and what choices they enjoy in the digital platform (Kelkar, 2018). Encoding, which comprises the rules for how offline objects and actions are translated into a digital language (Alaimo & Kallinikos, 2017), is also defined by platform owners. Once codified, actions must be performed in accordance with the rules established by the platform. Thus the affordances of technology shape and mold the interactions of the users with the platforms (Alaimo & Kallinikos, 2017). Furthermore, algorithms and codes have been denounced for their opacity (Etter & Albu, 2021). The inspection and control of a platform’s guiding principles present technical challenges for both users and regulators (Beer, 2017), which enables manipulation. For example, Seymour (2019) and Wu et al. (2019) describe how the manipulation design techniques employed by platform firms like Facebook and Twitter are worrying not only because they affect an individual’s freedom of choice but also because they can cause users to experience harmful psychological effects, such as addiction. Yet the aforementioned studies of affordances and materiality offer a limited understanding of how emergence and self-infliction of organized maturity are patterned by the strategic choices of platform owners and other agents. To further understand the organized immaturity of digital platforms, it is important to look at how practices are shaped and organized by the relations between the technological objects, the different users’ strategies, and structural elements that conform the power accumulation of the platform. Some scholars have begun to offer holistic models that explain the accumulation of power by platform firms and its consequences for the authority erosion of different agents. Lanier (2018) and Zuboff (2019) describe digital platforms’ datafication of human experience, which leads to increasing forms of domination in what they term “surveillance capitalism.” Surveillance is enabled by the asymmetric positions of platform owners and users, defined by technological architecture, and executed through monetization strategies based on user data. Zuboff 32 Essays on Power in Digital Platforms (2019) argues that despite the explicit narrative of platforms as both positive and objectively inevitable, their strategies and business models—based on voluntary data sharing—are fundamentally connected to the extraction of economic rents. Surveillance reduces human experience to free raw material for translation into behavioural data and prediction products (Zuboff, 2019), eroding individual autonomy and disrupting intellectual privacy (Richards, 2012). Surveillance has become a naturalized practice that we all—willingly or not—perform (Lyon, 2018). Surveillance theories therefore contribute to this debate by offering an understanding of the instrumental connection between the business model and technological objects that constitute the platform and the selfinfliction aspects of immaturity processes. Yet, we argue that further work is needed to understand not only the expansion of immaturity through a system of economic surveillance but also how the everyday practices of leading and participating in the platform relate to immaturity emergence. Moreover, we argue that these views should be enriched with a theory of how agency is constituted and transformed by platform power dynamics, how these dynamics have an organizing role in producing and reproducing the delegation of autonomous decision-making, and how the emergence of immaturity and the strategic power accumulation by platform owners are connected. 2.3. A socio-symbolic perspective on digital platforms To further explain how platforms organize immaturity, we draw on Bourdieu’s sociosymbolic theory and the concepts of field, capitals, and habitus. A sociosymbolic perspective situates the agents in a field and explores the power accumulation dynamics of each agent. It takes materiality into consideration, but, through the concept of habitus, it is able to explain how interactions are also mediated by previous history and the networks of relations in a way that complements the notion of affordances and its connotations for the perception of physical artifacts and technology (Fayard & Weeks, 2014). Furthermore, a sociosymbolic approach allows us to build an integrative conceptualization of power accumulation and its dynamics based on agents’ practices, positions, and strategies. It shows how multiple types of powers can coexist and accounts for how the relative positions of agents shape their motivations and actions, explaining the practices of immaturity and its relation to self-infliction. We explain this further, first by providing an overview of how a sociosymbolic perspective generally explains power and its dynamics through the concepts of field, capital, and habitus; we thus show how digital platforms can be understood through these lenses. Second, we describe the dynamics that lead to specific forms of power accumulation and explain how they can evolve over time. 2.3.1. Fields, capitals, and habitus in digital platforms Bourdieu’s sociosymbolic theory was developed to explain social stratification and dynamics in (offline) societies by focusing on how agents (people, groups, or institutions) produce, reproduce, and transform social structures through practice (i.e., what they do in everyday life). Through practice, agents produce particular social spaces with specific boundaries demarcated by shared interests and power relations; these social spaces are termed fields of practice (Bourdieu & Wacquant, 2007). 33 Essays on Power in Digital Platforms 2.3.1.1. Fields A field (champ) is a key spatial metaphor in Bourdieu’s work. It represents “a network, or a configuration, of objective relations between positions” (Bourdieu & Wacquant, 2007; 97). These positions are objectively defined “to field occupants, agents or institutions . . . by their present and potential position (situs) in the structure of the distribution of species of power (or capital)” (Bourdieu & Wacquant, 2007; 97). Individuals, groups, or organizations can be agents in a given field, and one individual may have different agencies (or “roles”), depending on their situation in the field. The concept of “field” can be related to digital platforms in the sense that the organization and production of practices situates the platform in relation to an existing field. This may be the field of cultural production (e.g., Facebook) or the field of goods exchange (e.g., Amazon). The fields have specific logics and structures that define them. Different agents can have multiple roles; for example, an Instagram user may be both a contributor and a consumer of content. The relational aspects of the fields are also very compatible with network-based perspectives (Portes, 1998) because the field in which the platform is embedded functions on the basis of relations created during the practice of exchanges that constitute the field. The technological infrastructure creates a centre–periphery structure, which provides the foundation on which the practices occur, both enabling and regulating them. This approach to platforms highlights the practice of the agent and its position but also simultaneously shows how the platform’s constitutive elements are deeply interconnected. Taking Twitter as an example, the extent to which a specific content generated by a user is reproduced depends on the user’s social position in the network but also on the priorities defined by the platform’s algorithms, which create the structure in which the content is shared. Multiple nested and overlapping fields can be found on any platform, just as they are in any (offline) social context. For example, YouTube constitutes a huge field of people broadly interested in sharing and viewing online video content. However, YouTube also hosts a variety of other, more focused subfields, for instance, a field cantered on cryptocurrency videos. At the same time, platforms do not necessarily constitute a field in its entirety, for while some online fields exist mostly in a single platform, like the field of video content sharing on YouTube, competing platforms have entered some subfields, such as gaming videos in Twitch. At the same time, other online fields are embedded in larger fields of practice. For example, job seekers would look at job opportunities in LinkedIn while engaging offline with the job-offering companies. Yet the creation of a digital platform can also be conceptualized as an attempt to “enclose” part of a field: an agent (the platform creator) designs a value creation model for users (the specific practices to be performed by them within the field) and develops the digital infrastructure that makes interactions possible. Digital platforms enclose the field because they attempt to create “exclusive control rights” (Boyle, 2003) over dimensions of practices that were previously in the public domain. Consider Google’s Street View, launched in 2007, which permits users to view the fronts of buildings from a pedestrian’s viewpoint. The service utilizes photographs taken by Google of objects that are not covered by intellectual property rights, albeit that the photographs were taken without the authorization or agreement of the communities, and their use is monetized (Zuboff, 2019). In this case, Google Street View becomes not only a new service for users but also a new way of exploiting value through dispossession of public goods and private data (Zuboff, 2019). 34 Essays on Power in Digital Platforms A field enclosure by a platform also includes encoding social interactions defined by more or less variable practices (e.g., hailing a taxi on the street) into a precisely defined process in a controlled space (using a ridehailing app). This appropriation is produced through the codification of social interactions, control over the digital space, and the data generated by these interactions. Moreover, by enclosing a field, digital platforms modify both the practices and the agents’ relative positions. For example, drivers and passengers are inscribed into a database owned by the platform owner and are organized into groups from which they are picked and matched. Furthermore, the creation of the platform can transform the scope of the field. Digitalized practices often involve connecting with deeply intimate aspects of users’ lives (Lupton, 2016), such as private data exemplified in photos, comments, or information about consumption habits. While typically regarded as private, the encoding of these portions of experience puts them into the potential reach of a field and exposes them to its specific field logic. Furthermore, because of the new ways of performing certain practices, platforms collide with the established scopes of the field, changing the agents and institutions involved in it. This is the so-called disruptive nature (Stigler report, 2019) of the platform. Examples can be found in conflicts around regulatory frameworks triggered by the introduction of platforms to some industries, such as Uber’s entry into the field of transportation and Airbnb’s into hospitality. 2.3.1.2. Capitals Fields are dynamic spaces defined by the relations of power between players that constitute the structure of the field (Bourdieu & Wacquant, 2007). These relations result from “the possession and activation of resources that are both materially and symbolically produced and perceived” (Bourdieu, 1989; 16). These resources are the capitals. The accumulation of capitals give access to “the specific profits that are at stake in the field, as well as by their objective relation to other positions (domination, subordination, homology, etc.)” (Bourdieu & Wacquant, 2007; 97). In each of the specific fields, the spaces of objective relations are the sites of a logic specific to those who regulate the fields. This logic does not need to follow purely economic rationalities to be described (Sandberg & Alvesson, 2011). For example, TikTok users who copy their nearest higher-status digital neighbours in a particular contest or “dance” might not be guided by economic rationality, but they do follow the logic of the platform. Capitals are therefore the resources—scarce and socially valued stocks of internalized abilities and externalized resources—that each agent has. Bourdieu defines three fundamental forms of capital through which power is accumulated: economic capital (money and other assets), cultural capital (knowledge and familiarity with accepted norms), and social capital (reflected in the actor’s creation of connections and social networks) (Bourdieu, 2011). To these, Bourdieu adds symbolic capital, “which is the form that one or another of these species takes when it is grasped through categories of perception that recognize its specific logic, . . . [that] misrecognize the arbitrariness of its possession and accumulation” (Bourdieu & Wacquant, 2007; 118)., that is, the reflection in the relations of the field of accumulated prestige, consecration, or honour (Bourdieu, 1993). For Bourdieu, power struggles are mainly symbolic, and agents who are willing to increase their power will ultimately exercise the symbolic capital that will help them to be “perceived and recognized as legitimate” (Bourdieu, 1989; 17) in what Bourdieu (1984) also calls “distinction.” 35 Essays on Power in Digital Platforms Social dynamics in fields are cantered on the generation of distinction(s) by agents, who “constantly work to differentiate themselves from their closest rivals” (Bourdieu & Wacquant, 2007; 100)., although the actors’ participations in these games are typically no more than “unconscious or semi-conscious strategies” (Bourdieu, 1969; 118). Distinction operates through the accumulation of capital that matters to the field. Thus, fields are spaces of conflict and competition in which the hierarchy is continually contested. However, agents can attempt to convert one form of capital into another or transfer it to a different space, depending on the specific logic of the field (Levina & Arriaga, 2014). The concept of distinction can be assimilated to the concept of “status” as it is used to explain the means of interaction on digital platforms (Levina & Arriaga, 2014). For example, on digital platforms like YouTube, a user’s social network position and cultural skills (e.g., their offline knowledge about a particular topic) combine with their taste and the time and money they invest into the field. Together, these shape which content gets noticed and which is ignored (Levina & Arriaga, 2014) and therefore which agents become “influencers” or agents with high status in the network. 2.3.1.3. Habitus Besides the description of how agents, through their collective actions, shape emergent field structures and the understanding of which capital matters and how, Bourdieu also looks at how structure shapes agency. Bourdieu uses the notion of habitus to describe the socially learned schemata of perception and inclinations to action (Bourdieu & Wacquant, 2007). Habitus is the internalization of the logic of the field. It is a set of historical relations incorporated within individual bodies in the form of mental and corporeal schemata (Ignatow & Robinson, 2017). These relations, or the “system of schemes of perception and appreciation of practices, cognitive and evaluative structures,” are “acquired through the lasting experience of a social position” (Bourdieu, 1989; 19); that is, they are acquired through interaction with other social agents. The habitus includes related comportment (posture and gait), aesthetic likes and dislikes, habitual linguistic practices, and ways of evaluating oneself and others via categories. It forges not only actions but also desires and aspirations (Ignatow & Robinson, 2017). While cognitively embedded, it is also embodied in gestures, postures, movements, and accents (Ignatow & Robinson, 2017). Its reproduction depends mainly on institutions like family and school. Mastery of the habitus tends to guarantee distinction and constancy of practice over time (Bourdieu, 1990). Crucially, the constitution of the habitus is recursive: while agents can reshape social distance and the ways it may be perceived, their own perception is likewise framed by their own position in the social structure. This recursive cycle is the process of constitution of the sociosymbolic space, where changes in position can be understood as the outcome of symbolic struggle. Habitus is therefore a way of conceptualizing how social structures influence practice without reifying those structures (Costa, 2006). In his studies of class, taste, and lifestyles, Bourdieu (1984) illustrates how habitus shapes taste in ways that make a virtue out of necessity. For example, working-class people develop a taste for sensible, plain food, furnishings, and clothes, and they shun fancy extravagances (Bourdieu, 1984). Hence habitus leads to the “choice of the necessary,” and in so doing, it tends to generate practices that ultimately reproduce the original objective conditions, through which it functions as structure (Costa, 2006). Thus, given a set of conditions, “habitus affords an actor some 36 Essays on Power in Digital Platforms thoughts and behaviours and not others, making those thoughts and behaviours seem more appropriate, attractive, and authentic than others” (Fayard & Weeks, 2014; 245). Ultimately, however, it is the actor who decides what to do. Often the decision occupies no conscious thought, but, as Bourdieu (1990; 53) argues, it is “never ruled out that the responses of the habitus may be accompanied by strategic calculation tending to perform in a conscious mode.” The concept of digital habitus has been used in the analysis of digital spaces (e.g., Ignatow & Robinson, 2017; Levina & Arriaga, 2014; Romele & Rodighiero, 2020) to explain the ways of acting, namely, the social and technologically ingrained habits, skills, and dispositions that define the practices in the digital field. Ignatow and Robinson (2017) argue that digital machines are not only the crystallized parts of habitus but also habitus producers and reproducers. This is because practices performed in digital platforms have technological and symbolic mediations: they are digitized—coded—and they are performed through a constant interaction with algorithms and the data that feed the learning of the algorithms. For algorithms to constitute the habitus, they need the platform to be able to extract increasingly large amounts of data and transform them into capital. In this context, the data work as the culture that informs the knowledge about the social space. The norms of the platform are constantly shaped by the interaction between the data, the algorithm, and the agents. The capital created by this interaction can be appropriated by certain agents who know how to use these results to their advantage. The mechanism of the digital habitus has two consequences. As socialization is increasingly done through digital platforms, the algorithmic logic becomes a norm that everyone needs to learn to play by or with (Beer, 2017), and thus it becomes part of the habitus. It becomes the representation of the current taste of a social class or group so that their decisions resemble each other. However, unlike the offline habitus, it derives from code as well as from action; thus it is somehow defined behind closed doors by the platform owners. Second, as Ignatow and Robinson (2017) argued, the digital habitus becomes a (re)generator of the social group because it is mediated by the property of the algorithmic practice that relates to aggregation for prediction. The singularities of social agents are reduced to aggregates of decisions, actions, desires, and tastes. This phenomenon has been called “personalization without personality” (Ignatow & Robinson, 2017; 100), personality being the principle that gives unique style to each human process of individualization. Having set the theoretical apparatus to explain how digital platforms can be understood from a sociosymbolic perspective, we turn now to defining how digital platforms accumulate power and how power accumulation increases the problem of organized immaturity. 2.3.2. A socio-symbolic perspective of power accumulation and its consequences for organized immaturity Building on Bourdieu’s later writings on the State and its forms of power (Bourdieu, 1989, 2014) and in light of the latest developments of digital platforms and their accumulation of power, we direct our analytic attention to the platform owner and its relations with the other platform agents and sociodigital objects. Thus, we go beyond the extant analysis of distinction in digital platforms done by scholars of digital sociology (e.g., Ignatow & Robinson, 2017; Julien, 2015) which focuses on users, to capture the mechanisms of field transformation led by platform 37 Essays on Power in Digital Platforms owners in their relationship with the other platform agents. We follow Bourdieu (2014) in terming these mechanisms “forms of power” and showing how these contribute to explaining organized immaturity. Drawing on Bourdieu’s writings (Bourdieu, 1984, 1989, 1991), we define the forms of power, distinguishing between two general dynamics. We first define five forms of power (constitutional, juridical, discursive, distinction, and crowd) that drive the accumulation of power within the platform. Second, inspired by recent literature on platforms (Bucher et al., 2021; Eaton et al., 2015; Krona, 2015; Ziccardi, 2013), we show how counterpower can also be performed by end users and other peripheral agents through crowd and hacking power. Crowd and hacking power are not concepts derived directly by Bourdieu’s theory but provide a more comprehensive view of power accumulation dynamics. We then articulate the platform power dynamics through three phases of platform evolution, which are derived from an interpretation of platform innovation research (Cutolo & Kenney, 2021; Kolagar et al., 2022; Rodon Modol & Eaton, 2021; Teece, 2017): formation, where the platform is launched and starts to be used by agents; domination, where the platform has been widely adopted and operates under a relatively stable design within the original field; and cross-field expansion, where the platform expands to other fields, leveraging their accumulation of power. Although we describe for each stage the dominant forms of power and counterpower accumulation that enable the transformation of the field, we acknowledge that several forms of power coexist in these phases, that the evolution of platforms is often nonlinear, and that not all platforms will become dominant. 2.3.3. Forms of Platform Power 2.3.3.1. Constitutional Power Constitutional power is the ability to “transform the objective principles of union and separation, . . . the power to conserve or to transform current classifications” (Bourdieu, 1989; 23). Within the platform, this power comprises both the architectural design (platform layers and modularity, design of user interfaces and experiences) and the capacity to define the rules, norms, categories, and languages that make up the digital interactions. Constitutional power shapes the digital medium for interactions and defines what may and may not be accessed by each type of agent within the platform. Constitutional power is exercised mainly by the platform owner. As the provider of the digital infrastructure upon which other agents collaborate, the owner defines the symbolic space through code. Code symbolically creates the objects that constitute the relations, being a neat, unified, and unambiguous language with no openings for interpretation (Lessig, 2009). In the digital realm, the actor who manages the code can increase its symbolic imposition and therefore its legitimization. As the legitimation process is unified, creation and transformation are delegated. This legitimation is “world-making” (Bourdieu, 1989), as it explicitly prescribes the possible realities and actions. The platform owner is therefore able to hold a monopoly over legitimate symbolic violence (Bourdieu, 1989), having a differential capacity to influence and settle symbolic struggle. The possibility of obtaining and activating this symbolic capital is associated with complex technological competences, which are scarce and highly concentrated (Srnicek, 2016; Zuboff, 2019). 38 Essays on Power in Digital Platforms The coherent body of code adopted by the symbolic space through constitutional power is not a neutral technical medium (Beer, 2017; Gillespie, 2010), and it can trigger autonomy eroding. Code is created and transformed in accordance with the objectives of the platform owner and correspondingly managed toward these goals. For example, Kitchens et al. (2020) show how the differences in platform design for Facebook, Twitter, and Reddit create a differentiated impact on the diversity of news and the type of content their users consume. Calo and Rosenblat (2017) and Walker et al. (2021) find that the algorithmic design in Uber reduces drivers’ insights about their working conditions and the competition they face, hindering their autonomy. Even without assuming strategic manipulation, the limited symbolic and repetitive action of users implies a delegation of users’ own independent reasoning and the emergent coordination of their actions by the platform. 2.3.3.2. Juridical Power Along with the architecture definition, a second feature that is critical to the thriving of the platform is its governance. While constitutional power has to do with the design of governance, juridical power is the capacity to sanction via the created rules and the authority to arbitrate in disputes (Bourdieu, 1987, 2005). Typically, it can take a variety of forms, such as sanctioning rule infringement, reporting abuses, or managing access to the platform (Adner & Kapoor, 2010). Digital technologies can enable increased participation and distribution of roles among agents, which is why studies of governance in these contexts have favoured the idea that digitalization processes are highly democratizing (von Hippel, 2006; Zittrain, 2009). However, the hierarchical structure of digital platforms facilitates the creation of governance layers, meaning that the importance of those decisions can be easily packaged, resulting in a limited distribution of power in the field. For example, transaction-oriented platforms like Amazon, eBay, and Uber rely on user-based rating systems to ensure good quality and sanction inadequate behaviour; however, the platform owner designs the rankings and retains control of other actions, such as account activation and suspension (Gawer & Srnicek, 2021). This role division effectively creates and redistributes power and therefore restricts the capacity of some agents to interact without the intervention of the digital platform owner. Hence the definition and distribution of roles will interact with (and eventually transform) the authority structure and the conflict management mechanisms that pre-exist in the field, including regulation. For example, Valdez (2022) explores how Uber uses what she calls “infrastructural” power to deploy a strategy of “contentious compliance,” both adapting to and challenging existing regulation. This strategy allows the company to exploit differences in regulation and regulatory scrutiny to reduce users’ access to information and acquired rights. 2.3.3.3. Discursive Power A third distinctive form of power that characterizes agents’ strategic interplay is discursive power. Discursive power is the power exercised in linguistic exchanges, which are embodied and learned but also generative of the habitus (Bourdieu, 1991). The way agents talk about platforms and the words they use to explain them—these discourses configure the collective narrative of what is possible on and valuable in a platform. 39 Essays on Power in Digital Platforms Platforms are narrated as part of a broader, already-institutionalized rational-technological narrative in which customer-centrism, effectiveness, and rationality of the exchanges are dominant values (Garud et al., 2022; Gillespie, 2010). Technological determinism discourses promoted by platform owners reinforce the idea that platforms’ algorithms are inscrutable and of a complexity unfathomable to the public or the regulator (Martin, 2021; Pasquale, 2015). These discourses have led to a broader narrative of a “Manifest Destiny” (Maddox & Malson, 2020) of digital platforms, where the user is explicitly asked to delegate their own reasoning to the platform. This, alongside user dispersion, is a fundamental element that enables prescribing actions. Critical to maintaining user dispersion is the narrative that users are directly connected through the platform, which is presented as an agora of exchanges. In actuality, platforms mediate that interaction, formatting it, regulating it, or even suspending it. 2.3.3.4. Distinction Power Distinction power is the creation of categories and the mechanisms of categorization that drive choice in the platform. It builds on the concept of distinction proposed by Bourdieu (1984). It defines the rules and practices that inhabit the habitus and designates which of them are legitimated and considered by society to be natural. The purpose of this type of power is to produce a behavioural response that serves some agents’ specific accumulation of capital. The platform owner can influence user behaviour by modifying the interfaces, the encoding, and the algorithms, thereby manipulating the user’s decision-making. At the same time, users can access and activate this power through their digital habitus, allowing them to influence and drive other users’ choices. On platforms, distinction power is often exercised through what Kornberger, Pflueger, and Mouritsen (2017) call evaluative infrastructures. Evaluative infrastructures are the different interactive devices, such as rankings, ratings, or reviews, that establish an order of worth among the users of the platform, driving the “attention” (Goldhaber, 1997) of other users. They relate agents and their contributions with each other, but they are also instruments of power. They define not only how agents are perceived and ranked in the community but also how the hierarchy is monetized by the platform’s owners (Kornberger et al., 2017). Status markers are examples of how distinction power is exercised. As they define how user activity and loyalty to the platform are rewarded, they become a fundamental element in guiding agents’ accumulation strategies. For example, YouTube and Wikipedia changed their strategy for recognizing content to stimulate newcomers (Kornberger et al., 2017). Ignatow and Robinson (2017) refer to this process as the “übercapital.” Übercapital emphasizes the position and trajectory of users according to the scoring, gradings, and rankings and is mobilized as an index of superiority that can have strong reactive or performative effects on behaviour (Ignatow & Robinson, 2017). A key feature of distinction power is that it is exercised heterogeneously over different users through differences created by constitutional and juridical power. Different types of users are granted different forms of agency, not only by the platform designers but also by their own intervention on the platform (Levina & Arriaga, 2014). For instance, passive users may be granted agency through technological features. For example, YouTube gives agency to passive users by displaying the number of views. Merely by viewing a piece of content, individuals cast a vote on its value, which has significant consequences for the content producers. Other users become judges or “raters” and producers of information at the same time. For example, retweeting on Twitter is both a contribution 40 Essays on Power in Digital Platforms to the platform and an act of evaluation. As well as users who are raters, there are often users who act also as “expert evaluators” (users who have accumulated significant cultural capital). One such example is the “superdonor” on crowdfunding platforms like Kickstarter, whose expert evaluations influence which projects are funded. Expert evaluators tend to form a tight-knit group within a field (Aral & Walker, 2012; Vaast et al., 2013). Other users might have what Bourdieu called “institutionalized consecration” (Levina & Arriaga, 2014), which is the formal authority to evaluate content given by the platform designers. These are typically site moderators and community managers, who have more power than others to judge contributions (Levina & Arriaga, 2014). In sum, these different types of agencies are designed by the platform owners to orient users’ actions and to promote and demote content (Ghosh & Hummel, 2014). They are typically linked to how the platform owner designs revenue models (Zuboff, 2019). The forms of power presented so far tend to reinforce the power position of the platform owner, but there are other forms of power that create the opposite tensions, that is counterpower accumulation. These are crowd and hacking power. 2.3.3.5. Crowd Power In the accumulation process, users are in a unique position in that they are the agents who produce the platform’s activity. Crowd power results from the influence that users can exert on the platform by the sheer mass of their actions, which may or may not be coordinated (Bennett et al., 2014; Culpepper & Thelen, 2020). These practices are, in essence, the exercise of the digital habitus. The exercise of the habitus can have a long-lasting effect on the platform’s structure. Practices can both inspire new functionalities and generate unexpected transformations to the value proposition, which the platform owner can recapture through redesigning the code. For example, this has been observed in the sharing and creator economies, in which, because the provider side of the platform is the main value creator—for example, graphic designers, programmers—the platform owner periodically changes the design to facilitate delivery of that value (Bhargava, 2021; Bucher et al., 2018). As Bourdieu (1990) argued, the agents ultimately decide what they do, and the digital habitus may be accompanied by strategic calculation, even if most of the practices are bound by parameters defined by the platform owners and managed automatically by algorithms. This creates the opportunity for practices not aligned with the value proposition to go viral, eventually posing challenges to the balance envisioned in the platform design. For example, Krona (2015) uses the notion of “sousveillance”— an inverted surveillance “from the bottom” or “from many to a few”—to describe the novel use of an audiovisual sharing platform by social movements during the Arab Spring uprising. This emergent use emphasizes the emancipatory potential of users to create collective capabilities and decision-making (Ziccardi, 2013), which we designate as crowd platform power– challenging forms of power. Yet, platform owners can attempt to use crowd power in their favour, in what we call the crowd platform power– enhancing forms of power, through constitutional power (architecture design, limiting the possibility of contact between users), juridical power (policing and sanctioning users), and distinction power (by shaping the evaluative infrastructure). For example, Thelen (2018) shows how Uber “weaponized” the volume of its users in a regulatory 41 Essays on Power in Digital Platforms dispute by introducing a button on its interface that would send a templated complaint email to local government on the user’s behalf. 2.3.3.6. Hacking Power Hacking power is the ability to identify the features and categories of digital spaces, such as overlooked programming errors and ungoverned areas, that may be used for a different purpose than the one originally intended (Hunsinger & Schrock, 2016; Jordan, 2009). There are numerous examples in the literature of expressions of this type of power in digital platforms. Eaton et al. (2015) have described the continuous cycles of resistance and accommodation performed by groups of hackers and Apple that surround the jailbreaking of each new release of the iOS. Bucher et al. (2021) and Calo and Rosenblat (2017) have shown how workers learn to anticipate patterns in algorithms that control their work processes and use this knowledge to defend themselves from abuses. Hacking power is the antithesis of individual immaturity, as it requires not only the exercise of independent reasoning but also a degree of understanding of the specific system in which the power is exercised. It is deliberate and purposeful, unlike crowd power, which is independent of users’ understanding because it stems from the combined volume of their actions. At the same time, hacking power necessarily operates in the margins or interstices of the platform. Furthermore, hacking power can be thought of as opposed to the constitutional and juridical powers; as such, it will be dispersed, under the radar, and is often considered illegal (Castells, 2011). This creates difficulties for creating and accumulating this power in the field and consequently for using it to challenge other forms of power. Table 3 summarizes the different forms of power and provides further examples. 42 Essays on Power in Digital Platforms Table 3: Forms of Platform Power and the Organization of Immaturity Form of power Definition Organization of immaturity Examples of the use of platform power Constitutional Design and control of the platform’s architecture (modules, interfaces, layers, and algorithms) and the capacity to define the rules, norms, categories, and languages that make up the digital interactions Explicitly prescribes realities and actions, granting the platform owner a monopoly over legitimate symbolic violence, leading to the delegation of users’ own independent reasoning and the emergent coordination of their actions by the platform. - Definition of user requirements to join (e.g., require proof of identity to join Airbnb) - Definition of possible interactions (e.g., the introduction of the “Like” button on Facebook) - Definition of allowed and forbidden user actions (e.g., the impossibility of editing a tweet on Twitter) Juridical Capacity to sanction via the created rules and the authority to arbitrate in disputes Disciplines users’ voice and participation and align them to the platform’s interests and values - Sanction rule infringement (e.g., suspension of a Lyft driver’s account for using an alternative route) - Report abuses (e.g., users flagging inappropriate content on Instagram) - Management of access to the platform (e.g., restrict blacklisted users from using Tinder) Discursive Power exercised in linguistic exchanges, which are embodied and learned but are also generative of the habitus Shapes collective discourse; asks users to delegate their own reasoning to the platform - Discourses of efficiency, technological determinism, or complexity promoted by the platform owners (e.g., accuracy and neutrality of Google Search results) - Narratives created within users’ communities (e.g., the superiority of PC/Windows gamers over Mac users) Distinction Creation of categories and the mechanisms of categorization that drive a user’s choice in the platform Enacts the platform owner’s capability to shape behaviour and the digital habitus - Definition of users’ performance, status, or visibility metrics (e.g., definition of a property’s valuation metrics in Booking) - Creation of differentiated tools to define hierarchies among users (e.g., ability to view profiles while remaining anonymous for LinkedIn premium users) Crowd Users’ influence on platforms by the shared mass of their actions, coordinated or not; can become manipulated by the platform owner Implicitly contests immaturity when used against platform power accumulation - User viralization of a message or practice (e.g., coordination of a protest through a Telegram channel) - Force a change in the platform’s functionalities (e.g., introduction of feedback options for service providers in UpWork) - Manipulation of users by covertly coordinating their actions (e.g., mobilization of Uber’s users to settle a regulatory dispute) Hacking Exploitation of a platform’s features and categories for a different purpose than the one originally intended Explicitly contests individual immaturity, as it requires the exercise of independent reasoning and the understanding of the platform’s rules - Evade the platform’s restrictions on functionality (e.g., jailbreaking of Apple’s iOS) - Performance of forbidden practices (e.g., get away with selling counterfeit products on Amazon) - Abuse the logic of the algorithm in users’ favour (e.g., disguise a user’s IP address to access additional content on Netflix) 43 Essays on Power in Digital Platforms 2.3.4. Platform Power Dynamics By discussing platforms in the context of fields, we have shown how the relations between the different key agents can be understood through dynamics of power accumulation. On one hand, users activate their capitals through the production of the practices that configure the digital habitus, which enhances their understanding of the ways of participating on the platform. However, it is mainly the platform owner who captures most of the value creation process through constitutional, juridical, discursive, and distinction power. This uneven distribution facilitates the creation of a leveled field upon which the relative positions can be consolidated while, at the same time, enlarging the distance between agents and therefore their capacity to decide in an autonomous way. We articulate these power dynamics through the phases of platform evolution to explain how platforms transform the agents’ relative positions over time and its impact on organizing immaturity. Figure 4 depicts the evolution in three phases. Figure 4: Platform Power Dynamics over Time Platform Evolution over Time Forms of Platform-related Power Platform Formation and Field Enclosure (Phase 1) Platform Domination within Original Field (Phase 2) Platform Cross-Field Expansion (Phase 3) Constitutional Power Juridical Power Forms of Platform Power Accumulation Discursive Power Distinction Power Crowd Power (Platform-Power-Enhancing) Forms of Platform Counter-Power Accumulation Crowd Power (Platform-Power-Challenging) Hacking Power 2.3.4.1. Phase 1: Platform Formation and Field Enclosure In platform formation, the primary objective for the platform owner is to get users to adopt the platform and regularly perform their practices on it. Constitutional, juridical, and discursive power are three of the forms of power through which platform owners attempt to enclose the field organizing the emergence of immaturity, for example, by designing a value creation model to create “exclusive control rights” (Boyle, 2003) over dimensions of practices that were previously in the public domain. At the same time, these forms of power organize immaturity. First, constitutional power (in the form of rules, norms, categories, and languages) defines how and under what conditions interactions are performed and how the different agents can express their preferences. Through juridical power, platform owners have the capacity to define the sanctions that will promote or restrict an agent’s capacity to operate on the platform, for example, who can exercise their voice on the platform and who cannot and what sanctions are going to be applied to misbehaviour. Finally, discursive power creates a common narrative 44 Essays on Power in Digital Platforms about the value of the platform, restricting the capacity of agents to think beyond discourses that are presented as truths. 2.3.4.2. Phase 2: Platform Domination within Original Field Platform adoption and sustained use create the conditions for it to increasingly occupy the field. The increasing participation of agents on the platform can change the predominant accumulation logics of the different agents in the field, shaping the digital habitus. The process of capital accumulation of different agents leveraged by distinction power defines further how immaturity is organized by promoting the processes of self-infliction of immaturity. Capital accumulation on platforms is expressed as more data and the levying of fees, and the influx of users is repurposed as capital to develop the platform. In turn, users invest different combinations of capitals (data about their practices, social networks, money, and other assets) with logics of both consumption (purchasing, sharing digital content) and profit and accumulation (influencer, merchant, driver). To thrive in the accumulation dynamics, agents must increasingly invest themselves in the platform, adapting their strategies so they align with those that are relevant to the platform and embedded in the digital habitus. Users adapt their practices to learn the specific logics. This brings user practices closer to their archetypical category, and because they are better rewarded, it further legitimizes the practices of the digital habitus. When users grasp the critical elements of the digital habitus that correspond to their type, their practices experience and enjoy a viral thrust that characterizes the platform logic (e.g., they may become social media superusers or influencers). This success in increasing the capital leveraged by the mechanism of distinction power, such as rating in the platform, calls for higher investment, increasing the users’ dependence on the platform and thus contributing to the self-inflictive process of immaturity. At the same time, the processes that reinforce platform power accumulation coexist with other processes that create tensions that call for change and adjustment. Misalignments between users’ practices and their expected behaviour can quickly accumulate, destabilizing the platform’s operation or posing challenges for its governance. In addition, platforms with massive user bases and innumerable interactions can become problematic for the platform owner to police, creating the space for agents to exercise their hacking power. These counterpower accumulation forces can therefore create an emergent enlightenment—as opposed to immaturity—for the agents. 2.3.4.3. Phase 3: Platform Cross-Field Expansion In a third phase, the platform’s domination over the field leads to the possibility of integrating new fields, further contributing to the accumulation of power and the organizing of immaturity. Once a platform has become the dominant agent, a position in which the structure itself acts on the owner’s behalf, it can expand the scope of the field to new geographies and users and even enter and integrate previously separated fields. For example, Uber’s launch of Uber Eats was deployed using the platform’s extant base of users (drivers and passengers, viewed now as commensals). From the domination position, the owner can operate in the various fields with great freedom, changing the exchange rate of the capitals at play and accumulating power. Highly dominant expressions of constitutional power include interoperability lock-ins, the use of dark patterns and biased information that impede sovereignty 45 Essays on Power in Digital Platforms of choice, and digital workplace design and control. Juridical power can be commanded from a position of gatekeeping, permitting arbitrary suspension of users’ accounts, biased arbitration in a dispute, the imposition of discriminatory clauses, restriction of access to the platform, or limits on freedom of speech. Abuses of power are typically supported by the discursive power that enacts the discourse of Manifest Destiny and uses opaque arguments to justify the increased accumulation of power and the need to enforce the juridical power measures of the platform owner. Also in this phase, a full deployment of distinction power relates to the platform owner’s ability to monopolize the capture and processing of data through control of the technological architecture. This can be used to drive user choice in multiple ways, such as information asymmetries about the activities of a market or participant and political influences on social media platforms. The activation of powers in the cross-field expansion phase depicts the dynamics within a field in a given moment, but it does not mean that the dominion of the platform owner is absolute or that the platform becomes a “total institution” (Bourdieu & Wacquant, 2007). What we highlight is how the field’s structural homology with the platform eases a fast concentration of powers and creates remarkable obstacles to modifying this situation, whether from within (due to users’ habituation) or outside of the platform (because of network and lock-in effects, barriers to entry, and technical complexity). In addition to this, the form of dominion that the platform’s specific logic enables is very effective because by creating multisided businesses, it invisibilizes the specific accumulation and struggle dynamics with respect to the core practices users perform. For example, Amazon is “a place to buy and sell online,” and the fact that the company accumulates capital from the capture of user data and the use of its sellers’ capitals is not evident to the platform’s users. Thus the platform’s “rules of the game” may appear to be somewhat objective and relatively neutral, but they are in fact part of the organization of immaturity. 2.4. Discussion In this article, we present a sociosymbolic approach to power dynamics in digital platforms and how they relate to organizing immaturity. A sociosymbolic approach explains the structural and agentic dynamics of power accumulation leading to organized immaturity. We contrast the power asymmetries between the platform owner, as the central coordinating agent, and the rest of the agents directly participating in the platform to present five main forms of power enacted by the platform owner: constitutional, juridical, discursive, distinction, and crowd. We also present two forms of power that explain how users counteract the platform owner’s power accumulation: crowd and hacking. We explain how these forms of power are fundamental for understanding the different ways in which immaturity is organized. We show that constitutional power limits the symbolic world of the users and therefore their capacity to influence new rules and vocabularies that orchestrate participation. We explain how through juridical power, the platform owners have the capacity to define the sanctions that restrict the voice and participation of users. We show how through the digital habitus, the logic of the field is constituted, explaining the emergence of immaturity and its self-infliction. However, we also argue that distinction power enacts the platform owner’s capability to shape behaviour through creating evaluative infrastructures that mediate the emergence of immaturity. Furthermore, we argue that the construction of a narrative of omniscience, through discursive power, explicitly asks users to delegate their own reasoning to the platform. We also highlight the existence of forms of 46 Essays on Power in Digital Platforms power (hacking and crowd) that help users to accumulate power and resist the central authority of the platform owners. Finally, we describe power dynamics and their relation to organized immaturity through three phases: first, platform formation, where forms of power—mainly constitutional, juridical, and discursive—operate to promote the field enclosure and set the basis for immaturity to occur; second, platform domination in the field, where distinction power promotes the field reproduction and processes of self-infliction of immaturity, while hacking and crowd power create resistance to the central authority; and third, platform cross-field expansion, in which power accumulation dynamics lead to the integration of new fields and increasing dynamics of immaturity. In defining the power accumulation dynamics, we explain the emergent character of immaturity and its relation to agents’ strategies. By focusing on the digital platform and its power dynamics, we contribute in two ways to the current literature. First, we build a framework that explains the organizing dynamics of immaturity, based on the relations between the platform structure, the digital objects, and the agents’ strategies. Through this, we expand the understanding of organized immaturity in the light of sociological perspectives. Our framework analyses how immaturity is constituted in practice and explains and nuances the possibility of emergence and the self-infliction dimensions of immaturity. Second, we provide a dynamic framework of platform power accumulation contributing to the platform literature. Finally, we also provide policy recommendations on how to tackle immaturity, and we highlight potential avenues for further research. 2.4.1. Rethinking Organized Immaturity from a Sociosymbolic Perspective A sociosymbolic perspective on digital platforms and its power dynamics can push the boundaries of current concepts of organized immaturity toward a post-Kantian and more sociologically grounded view (Scherer & Neesham, 2020). This contributes to the understanding of organized immaturity in three ways. First is by explaining the different components of the emergence of immaturity through power struggles. We show how struggles are the result of agents’ different strategies, heterogeneously shaped by their positions on the platform and their practices, but also by their discourses and the history of experiences of each individual that shape the digital habitus. By showing the dynamics in these struggles, we contribute to explaining the process through which immaturity emerges as a non-orchestrated phenomenon. Second, we explain self-infliction by moving away from the more political understandings of autonomy erosion. Political perspectives of immaturity look at the individual and its “(in)capacity for public use of reason” (Scherer & Neesham, 2020; 1) and consider the “delegation of decision making to impersonal authorities they cannot comprehend or control” (Scherer & Neesham, 2020; 4) as a condition of the individual. We, however, adopt a sociological view that focuses on the generation of practices and places the individual in a space of sociosymbolic power struggles. We complement previous literature exploring the symbolic aspects of technology and its impacts on society and, more concretely, on autonomy erosion (Fayard & Weeks, 2014; Stark & Pais, 2020; Zuboff, 2019) by providing a set of forms of power that articulate how self-infliction is embedded in the digital habitus and thus how immaturity is organized. Our sociosymbolic perspective explains how the conditions of agency are shaped by the specific structure of the platform and its power dynamics. 47 Essays on Power in Digital Platforms Last, looking at fields through the power dynamics between the different agents can shed explanatory light on the formation process of organized immaturity. The relationship between habitus and field operates in two ways: while the field structures the habitus as the embodiment of the immanent necessity of a field, the habitus makes the field a meaningful space in which the agents may invest their capitals and themselves (Bourdieu & Wacquant, 2007). By defining the stages through which this relationship unfolds, we contribute to showing the emergent, dynamic, and accumulative nature of organized immaturity. 2.4.2. Contribution to the Understanding of Platform Power Accumulation We have approached organized immaturity by analysing platforms as spaces of coordination and production of practices, shaped by relations engrained into a digital habitus and the logic of the field. By better understanding the forms of power and the role they play in field transformation, we have identified more clearly the different forms of power accumulation through which digital platforms can become vehicles for organized immaturity and its dynamics. This contributes to the literature of platforms in the following ways. First, our description of the structural process of power accumulation on the platform expands market and network approaches (Eaton et al., 2015; Jacobides, 2021; Khan, 2018) by showing the importance of the social, cultural, and symbolic dimensions of capital. This lays the foundations for fundamentally reconceptualizing platform power and further explaining how power is exercised by the platform owner (Cutolo & Kenney, 2021; Kenney et al., 2021). Second, we enrich structural approaches to platforms by presenting how fields can be transformed through dynamics of power accumulation that extend beyond the consequences of an asymmetric structure (Curchod et al., 2019; Hurni et al., 2021). Furthermore, our framework shows how platforms can be reshaped by the interaction of agents’ strategies and the reconfiguration of the fields. By introducing a field view, we provide a more holistic scheme in which power is both accumulated and contested. We also highlight how agents other than the platform owner play a role in producing and exercising forms of power. This nuances our understanding of field dynamics and agent interaction in the context of platform power dynamics. Third, our model complements sociomaterial studies on platform power (e.g., Beer, 2017; Kornberger et al., 2017; Stark & Pais, 2020) with the notion of the digital habitus and its relation to organized immaturity. Other authors have analysed technological affordances as social designations of a space and the social and cultural factors that signify a space and set a generative principle of governance (Jung & Lyytinen, 2014). Although these authors do not talk explicitly about habitus or social capital, they reflect on the generative reproduction of norms by individuals in contact with their social spaces; this is very similar to the Bourdieu definition of habitus in social spaces. We complement the sociomaterial view of platforms by showing how the digital habitus works and by emphasizing the role of the platform as an organizing agent with a privileged capacity of capital accumulation. We present the platform as a space of symbolically mediated power relationships in which the digital objects and structural elements interplay to conform the logic of the field. We provide an understanding of the multifaceted nature of power as a process resulting from agents’ practices and strategies, the habitus, and capital accumulation in a field. We argue that this conceptualization defines power in platforms not only as an “instrument” (Zuboff, 2019) at the service of the platform owners but as a web of relations utilized by agents who can better exploit the different forms of capital. We also contribute to the debate about the coordinating role of platforms and how they 48 Essays on Power in Digital Platforms create generative forms of distributed control while power remains centralized, in an interplay between hierarchical and heterarchical power relations (Kornberger et al., 2017). Bourdieu’s (1977, 1990) concepts of capitals, habitus, and distinction have been used before in the study of the social consequences of digitalization and platforms’ increase of power (e.g., Fayard & Weeks, 2014; Levina & Arriaga, 2014; Romele & Rodighiero, 2020). We complement that research with a view of platforms’ accumulation of power and its role in the organizing of immaturity. We go beyond the explanation of distinction power to define constitutional, juridical, discourse, crowd, and hacking forms of power, thereby offering a more complete view of how platforms accumulate power and organize immaturity. 2.4.3. Contributions to Practice and Avenues for Future Research Our article provides a conceptual framework to practitioners that can enable platform owners, users, and policy makers to fundamentally rethink how they might address the platforms’ negative consequences for society. First, it highlights immaturity as a relevant concept to address social issues in platforms. Our detailed understanding of the mechanisms leading to immaturity and the manipulation of individuals’ decisions can help policy makers to identify and set limits on these types of powers, especially in the light of platform domination. By explaining the organizing dynamics of immaturity, we direct attention to the more holistic assessments of the social consequences of platforms. Concretely, we emphasize how these are not just concerned with the concentration in specific industries (such as retailing or advertising) but also involve constraints on human rights (such as freedom of speech). Furthermore, we show how the consequences of organizing our practices through platforms are embedded in social structures and expressed in the transformation of fields. We believe that this line of thought is fundamental if we are to collectively rethink the social role of platforms. Our article has also limitations that open up avenues for further research. We have identified not only forms of platform power accumulation but also forms of platform counterpower accumulation. As our focus in this article has been on how platforms organize immaturity, we have devoted more attention to the forms of power accumulation. However, future work is needed to deepen our understanding of how platforms lose power. For example, in recent years, we have witnessed an increasing backlash against big tech platforms, fuelled by reputational scandals and vigorous societal complaints (Gawer & Srnicek, 2021; Joyce & Stuart, 2021). We have also observed a new wave of regulatory initiatives that intend to curb platforms’ power by forcing interoperability and limiting self-preferencing and acquisitions (Cusumano et al., 2021; Jacobides et al., 2020), even when the effectiveness of these policies is being debated (Rikap & Lundvall, 2021). For example, in Europe, the new legislation of the Digital Markets Act (European Commission, 2022b) and the Digital Services Act (European Commission, 2022a) are respectively intended to create more contestability in digital platform markets. In the United States, there has been intense debate around the possible revocation of Section 230, which has so far provided a shield for platforms’ activities in social networks (Stigler report, 2019), leading to abuses of power and increasing immaturity. In parallel to regulatory or external counterpower mechanisms, research into power dynamics could also analyse the flows of affects and affective intensification (Just, 2019) that happen with the abuse of the digital habitus. Incipient research (e.g., Castelló & Lopez-Berzosa, 2021; Just, 2019) has shown how these flows of affects not only shape collective meanings but can also lead to increasing forms of hate speech and the renaissance of populist politics. More should be researched about what forms of counterpower may emerge in 49 Essays on Power in Digital Platforms society to reduce populism and hate speech. We believe that our framework sets grounds for studying the more concrete practices of immaturity in platforms but also new forms of resistance. 2.5. Conclusion Building on the concepts of fields, capitals, and habitus, we propose a sociosymbolic framework to explain organized immaturity in digital platforms. We articulate six forms of power that characterize the different ways in which platforms organize immaturity. It is our suggestion that a more precise understanding of the digital platforms’ role in driving organized immaturity can become the basis for fundamentally rethinking the role of the digital platform in society. Can the processes that lead to organized immaturity be reoriented toward organized enlightenment? We argue that a first step in this direction is to better understand how power is performed in digital platforms, which is what our framework contributes to explaining. 50 Essays on Power in Digital Platforms 3. Two-faced governance in platform ecosystems4 3.1. Introduction Platform ecosystems have acquired enormous economic importance, disrupting and creating markets and challenging established forms of economic and social coordination (Cusumano et al., 2019; Rochet & Tirole, 2006; Parker et al., 2016). Platform firms orchestrate autonomous actors towards system-level goals without formal authority in platform-based ecosystems (Adner & Kapoor, 2010; Gawer, 2014; Jacobides et al., 2018). Thus, platforms must set governance mechanisms to govern their ecosystem members' participation (Gawer, 2014; Kretschmer et al., 2022). Platform governance, or platform ecosystem governance, refers to the policies and mechanisms the platform firm adopts to govern its operations on the different sides and maintain its ecosystem (Tiwana et al., 2010). It regulates access to the platform and the interactions among its different sides (Boudreau & Hagiu, 2008; Hagiu, 2014). In most of the largest platform ecosystems, the owner typically centralises governance design and enforcement (Gawer, 2021). A platform owner governs its ecosystem by defining the rules for ecosystem participation and encouraging its participants to engage in desirable behaviours (Cusumano et al., 2019). As such, platform ecosystem governance relies on designing, deploying, and enforcing incentives and control mechanisms to ensure collaboration and value-creating activities by the ecosystem participants. While this creates an asymmetric position between the participants and the owner, the literature tends to assume that participants have agency and decide whether to follow the rules or leave the platform (Kretschmer et al., 2022). A consensus in the economics and management literature has emerged on how platform governance should be designed and enforced. It claims that: (1) it is in the platform owners’ interest to define and enforce governance rules that aim to keep their ecosystem participants satisfied (Hagiu, 2013; Parker et al., 2016); (2) governance rules should be clear and precise enough to be followed by ecosystem members (Saadatmand et al., 2019; Wareham et al., 2014); (3) the platform owner should enforce these rules (Jacobides et al., 2018; Constantinides et al., 2018); and (4) sustained governance infringement would lead to platform failure (Boudreau, 2012; Huber et al., 2017). However, anecdotal evidence and investigative journalism have highlighted cases when some platforms do not seem to govern their ecosystem effectively. Examples include how Amazon includes thousands of banned, unsafe and illegal products in its Marketplace (Berzon et al., 2019) and how it experiences persistent problems with fake reviews (Nguyen, 2021). Other examples include Facebook’s failure to control what is generally framed as hate speech (The Guardian, 2022). These accounts are usually interpreted as operational failures from the platform in policing its ecosystem effectively. However, despite anecdotal evidence to the contrary, there has yet been a scant systematic empirical exploration of the extent to which platform owners may be deliberately engaging in practices inconsistent with their own declared governance rules. Furthermore, there has been limited empirical research on whether the platform owner may actually benefit from these discrepancies. This chapter was presented as a conference paper co-authored with Annabelle Gawer at the 2022 Platform Strategy Symposium (Boston University), and the 2022 RaDMA, 2023 EGOS, and the 2023 SASE conferences. It is currently under review for publication. 4 51 Essays on Power in Digital Platforms This article asks: Are there sustained discrepancies between a platform firm’s declared ecosystem governance and its actual practices? And if so, what are their consequences on the platform owner and its ecosystem members? To address these questions, we studied Amazon between 2019 and 2022. We analyse how Amazon has governed its Marketplace and examine the effect of its actual governance. We adopted an inductive, grounded approach based on semi-structured interviews, participatory observation, and secondary data sources. We find evidence of a sustained mismatch between the governance the platform declares to abide by and its actual practices. We identify what we call the platform’s “two-faced” governance, which we define as the platform owner’s practice of pursuing simultaneously seemingly contradictory governance practices. On the one hand, they consist of practices consistent with the declared governance rules, roles, and activities. On the other hand, we observe undeclared practices and “unwritten rules” of participation that depart from the declared governance. These undeclared practices include the lack of platform owners’ support for complementors and the strategic mismanagement of the declared governance rules, creating contradictory conditions for platform participation. We also identify a set of adaptive practices that ecosystem participants adopt and find that they, too, resort to sets of declared and undeclared practices. We present a process of two-faced platform governance that clarifies how these declared and undeclared practices interact. Our observations of the platform’s own contradictory rules, the discrepancies between official rules and actual implementation, the frequent and unpredictable changing of rules, and its discretionary enforcement are consistent with what might be a strategy used by the platform owner to further increase, at least in the short term, its bargaining power over complementors. Whether it is an emergent or a deliberate strategy is outside the scope of this paper. Our findings principally contribute to the platform strategy literature. First, we provide evidence that runs contrary to the four central claims of the platform strategy literature. Second, we enrich the construct of platform governance by claiming it needs to be expanded to include the declared and undeclared governance practices we identify and characterise. Third, we contribute to the growing literature on platform power (Crémer et al., 2019; Gawer & Srnicek, 2021). Fourth, we derive implications for managers and policymakers. In the next section, we review extant research on platform governance and summarise the key claims by which this literature describes how platform governance should be designed and implemented. We then outline the characteristics of the Amazon Marketplace and detail our data collection and analysis methods. Next, we present our findings, describing Amazon’s declared and undeclared governance practices and how sellers adapt and respond to them, and we present our process model of two-faced platform governance. Finally, we discuss the consequences of our results for platform theory and practice, present our study’s limitations, and offer some concluding remarks. 3.2. Literature - Related Research Platforms are novel organisational forms (Gawer, 2022; Kretschmer et al., 2022; McIntyre et al., 2021) that operate centrally within platform-based ecosystems (Adner & Kapoor, 2010; Gulati et al., 2012; Jacobides et al., 2018) and solve the problem of coordinating autonomous entities toward a system-level goal without formal hierarchical authority. They coordinate transactions by acting as intermediaries that bring together (match) and facilitate 52 Essays on Power in Digital Platforms exchanges between different groups of actors by creating market incentives (Caillaud & Jullien, 2003; Rochet & Tirole, 2003). The platform’s digital architecture is structured around a core of stable functionalities that are redeployed in combination with complementary functionalities through interfaces to generate a family of products (Tiwana et al., 2010; Gawer, 2014). This modular structure, designed by the platform owner, defines an asymmetric framework in which complementors collaborate and compete (Adner & Kapoor, 2010; Dhanaraj & Parkhe, 2006). The platform-based ecosystem consists of a constellation of agents connected via the platform. The ecosystem participants include individuals and organisations that can play various roles, including buyers, sellers, developers of complementary innovation, and end-users. 3.2.1. Governance in platform ecosystems The construct of governance has been deployed in the strategy, information systems, and organisation literatures with slightly different meanings and emphases in contexts as diverse as territories, corporations, or IT systems. A common trait is that governance entails “creating the conditions for ordered rule and collective action” (Stoker, 2018; 15) and that it consists of “the orchestrated use of mechanisms to encourage consistent behaviour” (Staub et al., 2023; 909, citing De Haes & Van Grembergen [2009] and Weill & Ross [2004]). In all streams of literature, in the context of platforms and ecosystems, platform governance is understood as aiming to encourage formally autonomous actors within the ecosystem to achieve the platform’s goals productively. Some of the literature in information systems has taken the construct of platform governance to mean the design and deployment of governance decisions (Tiwana, 2014) and has focused on how these decisions are distributed (Baldwin & Woodard, 2009; Tiwana et al., 2010). This literature has focused on how rules and values are used to design and implement governance decisions (Ghazawneh & Henfridsson, 2013; Huber et al., 2017; Kretschmer et al., 2022; Tiwana et al., 2010). Strategy scholars tend to focus on how platform governance reflects the platform owner’s strategies for creating and appropriating value (Chen et al., 2021a; Kretschmer et al., 2022). In this line of research, Cusumano, Gawer, and Yoffie (2019; 67) characterise platform governance as “decid[ing] what behaviours to encourage or discourage on the platform, and how to enforce the rules”. This definition summarises both platform governance’s practices (deciding and enforcing), tools (rules and incentives), and target outcomes (desirable behaviours). Consistent with this approach, Huber et al. (2017) define governance practices as the specific activities through which the decisions about behaviours and rules are implemented. In this study, we adopt Cusumano et al. (2019) characterisation of platform governance complemented by Huber et al. (2017) characterisation of governance practices because they summarise two critical aspects of platform governance. First, they are consistent with an approach that recognises the existence of different roles in the platform ecosystem (Adner, 2017; Cusumano & Gawer, 2002). Platform owners are uniquely responsible for defining and enforcing the rules. They are structurally central in controlling the technological architecture within which platform ecosystem participants interact (Gawer, 2014; Kretschmer et al., 2022; Tiwana et al., 2010). Complementors, in turn, have the role of expanding the core value proposition offered by the platform owner (Gawer & Henderson, 2007; Parker & Van Alstyne, 2017). Second, as the participants are formally independent, platforms must set mechanisms to encourage and discourage behaviours and enforce the rules (Gawer, 2014). These mechanisms can be designed and deployed as combinations of legally binding contracts and informal arrangements between a platform owner and its ecosystem participants (Tiwana et al., 2010; Uzunca et al., 2022). 53 Essays on Power in Digital Platforms 3.2.2. The consensus on platform governance: Four central claims A consensus has emerged in the economics and management literature around how platform governance should be designed and implemented. We identify and detail four central claims. First, it is in the platform owner’s best interest to design the governance to satisfy all the platform sides . The literature has emphasised the convenience for the platform of this confluence of interests, stating that “a successful ecosystem is one in which all actors are satisfied with their positions” (Adner, 2017; 42) or that “the platform owners’ priorities, then, are to protect their own interests and secure their competitive positions while also securing the interests of producers and consumers who contribute to value-creation on the platform.” (Constantinides et al., 2018; 384). This view of governance has been considered as not only essential to value creation but also to winning in platform competition (Teece, 2018). The underlying logic is that since participation is voluntary and all sides are needed to deliver the value proposition, strong enough incentives must be set in place (Hagiu, 2013; Parker et al., 2016). In addition, the literature claims that there are virtuous cycles, based on “network effects”, in attracting participants and creating additional value for them that has positive consequences for the platform owner. (Katz & Shapiro, 1985). However, satisfying all sides requires the platform owner to manage the conflicts between participants' divergent interests (Adner, 2017) and find the right balance between them. Platform governance research aims to illuminate these trade-offs and suggest ways to navigate them (Eaton et al., 2015; Ghazawneh & Henfridsson, 2013; Islam et al., 2023; Staub et al., 2023; Wareham et al., 2014). A second claim in the platform literature is that platform governance rules and roles should be clearly articulated and precisely defined. If, as Tiwana et al. (2010) suggest, platform governance should aim to improve coordination, then it needs to be clear and precise enough to be easily followed and enforced (Huber et al., 2017; Islam et al., 2023; Uzunca et al., 2022). This claim has been associated with the rules governing access to the platform’s core functionalities (Wareham et al., 2014) and the participants’ expected behaviours (Saadatmand et al., 2019). Overall, formulating clearly “who is allowed to do what” is a priority in platform governance (Cusumano et al., 2019; Jacobides et al., 2022) and has implications for both the platform owner and its ecosystem participants (Parker et al., 2016). A third claim in the platform literature is that platform owners should enforce the governance rules that they set. The platform owner’s role goes beyond simply formulating the rules of ecosystem members’ participation; it also requires it to enforce them to sustain the alignment of ecosystem members (Jacobides et al., 2018; Teece, 2018). As Jacobides et al. (2018) argue, “platform governance appears to require a firm hand” (Jacobides et al., 2022:20) to adequately coordinate operations and safeguard value creation. The behaviours included in enforcement include platform owners rewarding and punishing ecosystem members’ behaviours, resolving conflicts, and coordinating task execution (Constantinides et al., 2018; Uzunca et al., 2022). While a platform owner typically centralises governance-enforcing efforts (Gawer, 2021), some features of platform governance can be delegated to other actors through specific control mechanisms, often related to output or behavioural control (Levina & Arriaga, 2014). The fourth and last claim in the platform literature is that sustained violations of platform governance are bound to lead to platform failure. The violations mentioned refer to both the platform’s and its ecosystem members’ violations. If governance is set in place to align participation and overcome conflicts, infringements by the 54 Essays on Power in Digital Platforms ecosystem members will harm value creation (Boudreau, 2012; Huber et al., 2017). Consequently, “securing (…) complementors’ contribution of value-adding complementarities and their compliance with the platform’s rules and processes (…), is thus the most critical success factor of such organisations” (Saadatmand et al., 2019:1). For platforms operating within vast ecosystems, it can be challenging to prevent all rule violations, sanction all infringements adequately, or solve conflicts fairly and efficiently (Parker et al., 2016). Nevertheless, a platform owner’s systematic failure to address breaches will likely create a toxic user environment and hinder their participation (Boudreau, 2012; Huber et al., 2017). Parker et al. (2016) illustrate this effect in their account of how Myspace’s demise can be attributed to a failure in policing advertised content. Similarly, Zhang et al. (2022) argue that Apple’s governance failure after the iOS 7 jailbreak negatively impacted the developer community and led to a significant decline in knowledge sharing. Overall, extant research depicts a relatively unified view of platform governance and how it should be designed and enforced for a successful platform strategy. In that consensus view, platform governance is essential to a platform business design. It aims to create value for interconnected yet autonomous actors (the ecosystem participants), coordinate their interactions, and align their behaviours. Platform owners are responsible for setting and enforcing platform governance. 3.2.3. Minority views on platform governance Departing from the consensus described above, a few scholars have explored situations when the platform owner is not acting consistently with the governance they proclaim abiding by. These studies have tended to follow two approaches. In the first approach, the platform firm inconsistently applies its own rules in different situations, for example, across different complementors. For example, Huber et al. (2017) have explored how platform firms, in the context of the highly heterogeneous population of the enterprise software industry, do not consistently apply ecosystemwide governance rules and instead resort to specific dyadic arrangements. The authors observe and suggest that platform firms situationally departing from ecosystem-wide governance can be a viable alternative to orchestrating all complementors in an arm’s length way. Another example is Hurni et al. (2017)’s study, which examines heterogeneity in complementors’ “dedication” to the platform. They suggest that platform owners should either refrain entirely from practising rules with situational flexibility and benevolence, thereby achieving moderate complementor dedication, or, alternatively, practice rules with both flexibility and benevolence simultaneously, thereby maximising complementor dedication. In these studies, implementing specific rules for specific types of complementors in specific situations, departing from unified ecosystem-wide governance, is interpreted as benefitting both the platform owner and the complementors. While this line of research illuminates valuable strategies for platform governance, it does not depart significantly from the views of the platform owner’s roles and practices presented above. In the second approach, a small but growing stream of academic articles, books, and policy reports has begun to study platform owners’ abuses (Gawer & Srnicek, 2021; Stigler report, 2019; Cusumano et al., 2019). These studies have focused mainly on how platform owners treat their complementors (Cutolo & Kenney, 2021; Khan, 2018; Rietveld et al., 2020). This line of research describes the power asymmetries that enable abuse and has aimed to 55 Essays on Power in Digital Platforms identify its sources (Jacobides, 2021). It has begun to explore the impact of platforms’ abuses on complementors and has sought to assess the extent to which regulation and public agencies are equipped to deal with these harms (Busch et al., 2021). However, this literature tends to address existing regulation's limitations and suggest regulation changes rather than focusing on platform governance. An exception is Cusumano et al. (2021), who explicitly examine the relationship between platform governance and government regulation, asking whether selfregulation, essentially an act of platform governance, can save digital platforms from overly intrusive top-down government regulation. Furthermore, anecdotal evidence and investigative journalism have highlighted cases when some platforms do not seem to govern their ecosystem effectively. Examples include how Amazon includes thousands of banned, unsafe and illegal products in its Marketplace (Berzon et al., 2019) and how it experiences persistent problems with fake reviews (Nguyen, 2021). Other examples include Facebook’s failure to control what is generally framed as hate speech (The Guardian, 2022). These accounts are usually interpreted as operational failures from the platform in policing its ecosystem effectively. 3.2.4. Research Question Based on our literature review, and despite the importance of the topic, we argue that there has not been any systematic empirical exploration of the extent to which platform owners may be deliberately engaging in practices inconsistent with their own declared governance rules. Furthermore, there has been limited empirical research on whether the platform may actually benefit from these discrepancies. This article, therefore, asks: Are there sustained discrepancies between a platform owner’s declared ecosystem governance and its actual practices? And what are the consequences of these discrepancies on the platform owner and its ecosystem members? 3.3. Research methods To address the above questions, we studied how Amazon governed its Marketplace between 2019 and 2022. Amazon is a well-suited empirical setting to conduct this investigation, as Amazon’s ecosystem is one of the largest in the world, with 2 million sellers. We focus on the governance rules and practices of the platform owner – Amazon – and how sellers adapt and respond to them. We adopted an inductive grounded approach, best suited for problematising a field’s assumptions (Alvesson, 2011; Sandberg & Alvesson, 2011) and analysing a phenomenon in an exploratory manner (Charmaz & Bryant, 2008; Eisenhardt, 1989). It is oriented towards theory development rather than proving existing hypotheses (Thornberg & Charmaz, 2014). 3.3.1. Empirical setting: Amazon and its Marketplace Amazon is a global, US-headquartered technological company and one of the most successful platform firms worldwide. It is one of the top companies in the world by market capitalisation and revenue. As of February 2023, it had 310 million active users in twenty-eight countries and employed 1.6 million people. 56 Essays on Power in Digital Platforms While Amazon began in 1994 as an online bookstore retailer, it gradually diversified into selling many other categories of products and services, including toys, electronics, furniture, fresh food, and media entertainment. As of 2023, along with Alibaba.com, Amazon was the largest Internet-based retailer in the world. Amazon has separate retail websites (called stores) in the United States, the United Kingdom, and twenty-six other countries. It also offers international shipping globally for many of its products. Besides online retailing, the company produced consumer electronics and media content. As of February 2023, it operated the world's most extensive cloud infrastructure as a service provider, one of its largest online marketplaces, and a major global advertising service5. On Amazon's websites and apps, customers can purchase products sold by Amazon or sellers. The Amazon Marketplace is the digital platform through which sellers reach customers and manage operations. For having an account and listing products, Amazon charges a combination of five different fees: subscription, referral, closing, high-volume listing, and refund administration. The subscription includes analytics and automation services through the Amazon Marketplace Web Service. After receiving an order from a customer, the seller has three options to fulfil it: doing it by themselves, using a third-party logistics provider, or using the Fulfilled-by-Amazon (FBA) program for an additional fee. To gain visibility on the platform, sellers can use Amazon’s advertising services to bid for better placements in search results. Finally, Amazon offers other services to sellers (see Appendix A for a complete list of Amazon’s programs and policies): the seller online forum, the seller university (with educational materials), and the service provider network (a directory of partnered third-party service providers to sellers, such as accounting, taxes, marketing, etc.). As stated in the 2018 letter to Amazon’s shareholders, the Marketplace’s importance for Amazon has steadily grown. The sales volume provided by sellers surpassed that of Amazon in 2015 and kept growing. As of 2022, it represented 65% of the total sales volume. Furthermore, in 2022, the revenue from the Marketplace (combining seller and advertising services) amounted to 30% of Amazon’s total revenue, allegedly being its most profitable business unit (Amazon, 2019, 2020, 2021, 2022a). 3.3.2. Data collection We conducted this study between 2019 and 2022 and collected multiple data sources during that period (see Table 4). For a description of Amazon’s growth path and the interaction among its business units see Kenney et al. (2021), Coveri et al. (2022), and Rikap (2022). 5 57 Essays on Power in Digital Platforms Table 4. Data Sources: Detail Data Source Cases Hours 37 23 Sellers 20 12 Amazon 2 2 Service providers* 9 5 Industry experts & media** 6 4 17 109 Events 9 93 Forums & social media 8 16 - Understanding of sellers’ characteristics and general challenges - Triangulation to identify inconsistencies Documents Pages - 602 3395 Legal agreements 424 1734 Reports and studies Guides and educational materials 27 4 1060 95 Media articles 139 422 Legislation Other documents (campaigns against Amazon) 5 75 3 9 Interviews Participatory observation Archival Use in the analysis - Detailed understanding of the Marketplace’s characteristics and dynamics - Identification of Amazon’s and sellers’ practices - Understanding of the Marketplace’s operation - Understanding of sellers’ characteristics and general challenges - Triangulation to identify inconsistencies * Four of our service provider informants were formerly sellers ** Three of our industry experts & media informants were formerly sellers In a preliminary phase, between October 2019 and April 2020, we learned about how the Amazon Marketplace operates. We sought to understand its key functionalities and services, fee structure, main communication channels, and who the main actors are in its ecosystem. To achieve this, we gathered information from industry reports, media articles, and Amazon’s official outlets and through participatory observation in online forums and events. In this phase, we identified the sellers as the key informants for our research. In the next phase, between May 2020 and November 2020, we focused on understanding the sellers’ characteristics. We consulted specialised reports, surveys, guides, and educational materials designed for them. We outlined sellers’ demographics and composition and identified the most popular product categories and markets. Based on survey data (Jungle Scout, 2020, 2021), we also built a first outline of what sellers identify as their main challenges and goals regarding the platform. We then created lists of potential sellers to interview in the United Kingdom for reasons of proximity. Additional informants, including sellers in other European countries, the United States, and China, were later reached through snowballing (Parker et al., 2019). We conducted the first round of interviews between December 2020 and November 2021. We conducted and recorded 23 online interviews, averaging 40 minutes each. The COVID-19 pandemic caused specific problems, making it impossible to meet informants in person and causing industry-specific events to be cancelled. Consequently, we found recruiting participants for this first round of interviews very challenging. Nevertheless, once the interviews began, we were able to create rapport easily. The interviews followed a common 58 Essays on Power in Digital Platforms questionnaire but had an open-ended format. Our questions to our informants focused on sellers’ participation in Amazon Marketplace, their views on Amazon, and their relationship with the company. After the first round of interviews and thanks to the suspension of some pandemic-related restrictions, we collected more data by attending several large, in-person sellers’ conventions and meetups. This participation helped us to understand the Marketplace’s communities better. It also helped us validate preliminary findings via triangulation of the information obtained in the interviews. We conducted a second round of interviews between March 2022 and May 2022, which consisted of 14 interviews and averaged 40 minutes each. This second round maintained the semi-structured format, with questions now centred on specific practices, dynamics, and characteristics of the Marketplace. We stopped recruiting participants as we reached saturation (Rheinhardt et al., 2018) in the themes discussed. Finally, we did a closing round of data collection consisting of revising additional archive material (financial statements, surveys, market reports, and legislation) for triangulation (Rheinhardt et al., 2018) of our previous findings. Using this material proved fundamental to validate insights or nuance insights from the data provided by our informants when we noticed contradictions or discrepancies. 3.3.3. Data analysis The data analysis followed an inductive grounded approach (Charmaz & Bryant, 2008; Gioia et al., 2013; Rheinhardt et al., 2018; Thornberg & Charmaz, 2014). After the initial revision of archival resources for immersion in the field, we transcribed and analysed the interviews. At first, the interviews were open-coded (Corbin & Strauss, 2008) to identify key themes and conflicts. Based on these initial codes, we built data structures (Gioia et al., 2013), which allowed us to identify the relevant set of practices and start conceptualising the actors’ interactions and strategies towards the marketplace. We resorted to the literature on platform strategy and governance to enhance our code structure and gain precision. We did not follow a predefined sequence when analysing the practices. Instead, the iterations helped us identify relevant practices and clarify their connections. In this way, some of Amazon’s practices became clear as we coded sellers’ practices, and vice-versa. As we iterated in this analysis, both authors met periodically to discuss the emerging categories and our interpretations of the data. In the later stages of the study, we started conceptualising the interactions between the participants’ practices and building the process model. Once we agreed on it, we made minor revisions to the code structure and the categories used. Finally, we prepared the final version of the data structures connecting the practices with the supporting evidence (see Tables 7, 8, and 9). 3.4. Findings We organise our findings in two parts. In the first one, we document (1) Amazon’s governance practices and (2) sellers’ adaptive participation practices. In the second part, we identify and describe what we call the two-faced governance of Amazon’s platform ecosystem, and we present a process model of how it operates. 3.4.1. Amazon governance practices 59 Essays on Power in Digital Platforms We find that Amazon engages in two distinct sets of practices that seem, at first sight, to be inconsistent with each other. Amazon’s first set of practices consists of what we call Amazon’s declared governance: this includes the rules and roles that Amazon officially formulates and communicates externally and the activities through which it implements them. The second set of practices consists of Amazon’s undeclared practices that depart from its declared governance. 3.4.1.1. Amazon’s declared governance Amazon presents itself as striving to be “Earth’s most customer-centric company, where customers can find and discover anything they might want to buy online” (Amazon, 2022b). The company has declared that its commitment to delivering this vision led them to open its digital platform for other retailers (sellers) to compete with them through the Amazon Marketplace. In this way, Amazon offers sellers a space to reach a large audience of potential clients and a suite of services that facilitate showcasing and fulfilling the products. At the same time, it provides customers with an extensive range of products to choose from, secure means of exchange, and short delivery times. As the platform owner, Amazon provides this space for interaction and claims to safeguard its integrity, handling the communication with the customer and the payment and supporting sellers to complete the transactions. In official documents and brochures, Amazon highlights how it “depends on sellers’ success” and supports it by creating a trustworthy environment, preventing fraud and abuse, and offering the best services for sellers to choose from. Sellers, in turn, must act fairly, honestly, and following the established rules and policies. As specified in the Amazon Seller Code of Conduct, their primary responsibilities include providing accurate information to Amazon and customers, not attempting to damage or abuse other sellers, influencing ratings and reviews, and not circumventing the sales process. We identify three main outcomes of the governance Amazon claims to pursue (see Table 5): Table 5. Amazon’s declared governance: Examples, practices, and outcomes Declared Governance Practices: Examples (1st-Order Concepts) Declared Governance Practices (2nd-Order Themes) - Sales Rank and Advertising tools (better positioning in search results); Featured Merchant Status; Buy Box offer - - Fulfilment by Amazon; Worldwide Selling - Lower entry barriers to sellers - Amazon Accelerator; Brand Registry and Dashboard; A/B Experiments - Product Reviews; Seller and Product Ratings; Ato-Z Guarantee; Amazon Brand Registry - Addition of new segments and categories; Increasing product selection - Amazon Prime; Amazon Vine; Personalized Recommendations - Provide tools for seller development Increase consumer trust - Improve offering to consumers - Increase consumer loyalty and spending - Foment good behaviour - Ensure quality of interactions - Community Rules; Amazon Seller Code of Conduct - Communication Guidelines; Seller Performance Measurement Provide visibility and access to customers Declared Governance Outcome (Aggregate Dimensions) - Platform to attract and keep sellers by creating opportunities for profit - Platform to attract and keep consumers by assuring and improving its buying experience - Platform to ensure favourable conditions for transactions Source: Based on interviews, participatory observation, and multiple archival sources Attract and keep sellers on the platform by creating profit opportunities. 60 Essays on Power in Digital Platforms The main incentive of the Marketplace for sellers is how it provides visibility and access to a large customer base. As of 2021, 60% of US consumers declared to start their shopping searches on Amazon (Jungle Scout, 2022a). Amazon also lowers sellers’ barriers to reaching customers by providing additional services, such as fulfilment. In addition, the Marketplace includes various tools that facilitate the sellers’ business development, including data analytics, brand development and protection, and market research and experimentation tools. Attract and keep consumers on the platform by assuring and improving its buying experience. Amazon declares to keep a loyal customer base by building trust in the buying experience through initiatives such as encouraging customer product reviews, showcasing seller and product ratings, and offering a return guarantee. Amazon keeps consumers attracted by its ongoing improvement of the platform’s offering, for example by adding new product segments and reducing delivery times. Finally, Amazon aims to keep consumers on the platform and increase spending through loyalty memberships (Prime) and personalised recommendations. Ensure favourable conditions for transactions. Since Amazon does not supervise individual transactions between sellers and customers, the company must set mechanisms to ensure favourable transaction conditions. This involves specifying what types of behaviours are allowed and establishing the punishments for infringement. The Community Rules, the Amazon Seller Code of Conduct, and the typification of Offenses are some of the mechanisms used to ensure good behaviour. These rules establish that users’ infringements can lead to account suspension and financial penalties. Similarly, the quality of the interactions is safeguarded by overwatching seller performance and establishing communication guidelines. Typically, each outcome encompasses more than one practice and affects different types of participants differently. For example, the product review functionality contributes to all three outcomes. It provides customers with additional information about the product before the purchase, it can benefit the seller by showcasing the account of a satisfied customer, and it works as a quality control device through the ratio of positive and negative reviews. At the same time, while it is an incentive for customers, it works mainly as a control mechanism over sellers. We find that these governance practices are both contractually and materially embodied: on the one hand, responsibilities and rules are defined in formal documents, such as the terms and conditions for users and the legal agreement with sellers. But we also find that identifying how these practices are materially deployed is crucial for understanding how they work. Governance practices can be deployed through the technological infrastructure, for example, by the specific design of interfaces (e.g., how visible is the seller's performance in the product page) or functionalities (e.g., how the seller's performance affects the position in search results). Further, these practices are also visible in how digital and non-digital processes interact (e.g., how a delay in delivery is reflected in seller performance). We found that observing how the practices worked alongside platform participants illuminated aspects of governance that were not observable through the legal documents. 3.4.1.2. Amazon’s undeclared governance 61 Essays on Power in Digital Platforms We also find evidence of a wide range of Amazon’s practices inconsistent with its declared governance objectives, official rules, and declared practices. These practices are undeclared: although we observe their presence in the Marketplace’s operations, they are not part of any rules and roles explicitly set out. Furthermore, as these practices can constitute law violations, Amazon denies engaging in them. However, our ecosystem member informants openly and repeatedly mentioned them in events, forums, and social media, and they have been described in industry reports and media articles. Next, rather than being isolated cases, we recognise patterns in these governance practices, which systematically shape participants’ behaviours and change the value creation and appropriation dynamics. We find that the undeclared governance practices undermine the conditions for sellers’ activities on the platform. At least in the short term, they appear to benefit Amazon by increasing its bargaining power over its sellers. We first describe the undeclared governance practices (see Table 6) and identify their outcomes: (1) Create an unpredictable and precarious environment for sellers (see Table 7 for evidentiary support), and (2) Profit from an asymmetric position (see Table 8 for evidentiary support). Table 6. Amazon’s undeclared governance: Examples, mechanisms, practices, and outcomes Undeclared Governance Examples: Activities and Processes (1st-Order Concepts) - Automated service support; Representatives lacking authority or competence - Lack of internal communication and data sharing among Amazon’s teams Undeclared Governance Mechanisms (2nd-Order Themes) Undeclared Governance Outcomes (Aggregate Dimensions) - Limited seller service support - Siloed and highly competitive organisation - Sellers internalise costs for Amazon’s pitfalls due to the difficulty of claim procedures - Pass the error cost to the seller - Lack of details of how a policy would be implemented or how a critical parameter is measured - Policy ambiguity - Changes of forbidden and allowed practices or to terms of use of services and functionalities - Constant update of functionalities and rules - Changes to inventory allowance in warehouses; changes to seller performance measurement - Change of platform’s parameters - Unexpected and unexplained account or product suspensions; Warnings or suspensions based on allegations unrelated to the seller’s business - Unreliable enforcement of rules - Lack of policing of fraudulent sellers - Inconsistent policing of seller’s misbehaviour and legal obligations - Force sellers to use optional services - Penalize sellers for using nonAmazon shipping services; Force sellers to pay to receive a better support service Undeclared Governance Practices (3rd-Order Themes) - Lack of platform owner’s support for complementors - Platform to create an - Rules uncertainty unpredictable and precarious environment for sellers - Discretionary policing - Impose the use of services to sellers - Limit competitive alternatives to Amazon’s additional services - Lock-in sellers through the use of services - Identify successful sellers’ products and launch a similar - Copy sellers’ products - Platform to profit from asymmetric bargaining position with sellers - Unfair competition with sellers 62 Essays on Power in Digital Platforms Amazon version - Favouring Amazon’s products in search results and fulfilment times - Elude rules for Amazon's products - Obtain market insights; Get advantages by understanding sellers’ competitive position - Exploit sellers’ data Source: Based on interviews, participatory observation, and multiple archival sources 63 Essays on Power in Digital Platforms Table 7. Amazon’s undeclared governance practices (1): Create an unpredictable and precarious environment for sellers. Evidentiary support Practice Governance Mechanism Exemplary excerpts from interviews “Every time you have a problem is "that's not my department" or “can I speak to the relevant department?”, “No, they’re firewalled off” - Limited seller service support - Lack of platform support for sellers - Pass the error cost to the seller - Siloed and highly competitive organisation - Policy ambiguity - Rules uncertainty - Discretionar y policing - Constant update of functionalities and rules “But nonetheless, trying to get anyone to actually help you sorted out is another matter entirely.” “They are going so big that it's very, very difficult to get even the simplest things resolved with them.” "It's funny how many workarounds and systems you have to build just to be able to prove you overcharge me $1,000 on this item." "You get a black mark for dispatching it late. But it's not late. It's actually Amazon's delivery drivers collected it late." Other supporting evidence "Amazon’s Seller Support Central generally prompt automated, unhelpful responses, which may be entirely unrelated to the specific case, question, or concern raised by the seller.” (Nadler & Cicilline, 2022) Three surveys conducted between 2020 and 2022 show that between 58% and 66% of respondents have experienced poor seller support from Amazon" (ILSR, 2022; Jungle Scout, 2020, 2021) “A third-party bookseller told Subcommittee staff that Amazon delisted 99% of his business’s inventory in September 2019. The bookseller requested that Amazon return its products, which were stored in Amazon’s warehouses.1675 As of July 2020, Amazon had only returned a small fraction of the bookseller’s inventory and continued to charge him storage fees.” (Nadler & Cicilline, 2022) “The teams are very siloed… there's a lack of communication in a lot of cases… they're obviously very protective of the data, so there's a lot of things that certain teams can't see.” “Amazon is deliberately siloed, so you have these kinds of competing factions.” “Whenever Amazon put a new policy or procedure in place… they're not direct in their wording. So, they leave things kind of open ended” “That's one thing that's really frustrating: nothing's ever clear.” " I feel like that kind of makes it very difficult to excel in anything, because they're always keeping something just out of reach of you." "It's really, really difficult for sellers to keep up with all these constantly changing demands." "I do think that it's incredibly difficult, operating in Amazon… the rules are always changing." The constant change of the Amazon's policies and terms of service was identified as a concern for sellers in three surveys from 2021 and 2022, with answers ranging from 47% to 61% (ILSR, 2022; Jungle Scout, 2021, 2022b) - Change of platform’s parameters "Even if we're perfect in almost every single metric, we have bigger items, so they're penalizing us. So that's one area where we're not exactly meeting their standards and we're getting a slap on the wrist for somebody else who might not be performing as well but they have a smaller item." “Another form of retaliation that Amazon reportedly engaged in was showing publishers’ titles as out of stock or with delayed shipping times.” (Nadler & Cicilline, 2022) A 2022 seller survey reported that 67% of them are concerned about Amazon changing search results to favour paid results over organic results (Jungle Scout, 2022b) - Unreliable enforcing of rules “Amazon shut down that listing… they have a set percentage of what they'll allow, sometimes they don't even like at the same price. It just depends on how moody Amazon is that day.” “Among the most egregious examples of Amazon’s arbitrary treatment of sellers are its abrupt suspensions of their accounts, frequently made without explanation.” (Nadler & Cicilline, 2022) “I can't say who they pick and choose to pick on, you know, some people Abrupt account or listing suspension is a constant concern for sellers. In four 64 Essays on Power in Digital Platforms - Inconsistent policing of seller’s misbehaviour and legal obligations get away with things that other people don't.” surveys conducted between 2020 and 2022, between 41% and 76% of respondents shared this preoccupation (ILSR, 2022; Jungle Scout, 2020, 2021, 2022b) “Amazon put their head in the sand and ignored it. Because at the end of the day, they're still making money.” "These people are doing like really bad stuff, their account gets shut down for like, two weeks, and then they're back again, how do they manage this?" “Subcommittee staff uncovered evidence during the investigation that Amazon has used its ability to police counterfeits more or less aggressively by as leverage in contract negotiations with brands who attempt to resist Amazon pressure to sell on its platform.” (Nadler & Cicilline, 2022) “They make no efforts until the government kind of forced them to sort of go overseas sellers that weren't collecting VAT, which obviously, is a massive cost advantage.” A 2022 survey of sellers shows that 68% of respondents consider that Amazon does not do a good job policing unscrupulous and fraudulent Marketplace sellers (ILSR, 2022) Source: Based on interviews and archival sources Table 8. Amazon’s undeclared governance practices (2): Profit from an asymmetric position. Evidentiary support Practice Governance Mechanism Exemplary excerpts from interviews Other supporting evidence 65 Essays on Power in Digital Platforms - Impose the - Force sellers to use optional services use of services to sellers - Lock-in sellers through the use of services - Copy sellers’ products - Unfair competition with sellers - Elude rules for Amazon's products “We're having an increasingly difficult time finding appointments for our carriers… you can get an appointment very quickly if you use their shipping service. They're kind of funnelling you into using their shipping service, which is a little bit more expensive.” “Amazon has recently monetised the degradation of its seller services, rolling out a program where sellers can pay an extra fee for a dedicated account representative. Sellers are supposed to pay for representatives to help them solve the very problems that Amazon created in the first place.” (Nadler & Cicilline, 2022) “We understand that it's not going to be effective if we're not advertising.” “Advertising services have become “less of an option and more of a requirement for sellers to compete” on the platform, Amazon’s ads have also become more expensive.” (Nadler & Cicilline, 2022) “They offer me an account manager, it's $4,000 a month... it was pretty much nothing that they can do… It gives you one extra day to think about how you're going to answer Amazon's things and maybe give you some even indication of why you've been suspended, which is something that Amazon don't do.” "Italy's watchdog said in a statement that Amazon had leveraged its dominant position in the Italian market for intermediation services on marketplaces to favour the adoption of its own logistics service - Fulfilment by Amazon (FBA) - by sellers active on Amazon.it." (Nadler & Cicilline, 2022) “It starts like, oh, here's a new scheme, you can get first mover advantage. But actually, what you're doing is you're committing to it, you can't move to a non-prime model… but actually, I'm just a guinea pig. I'm dealing with all the system challenges and problems that go and get ironed out and rolled out and then the market becomes hooked on it like a drug and mounting that competitive advantage is eroded." “More than 73% of all Marketplace sellers worldwide reportedly rely on FBA Numerous thirdparty sellers told the Subcommittee that they feel they have no choice but to pay for FBA to maintain a favourable search result position, to reach Amazon's more than 112 million Prime members, and to win the Buy Box.” (Nadler & Cicilline, 2022) "They find really high selling items… we have our [product A], they started selling a very similar product, that they're always near the top. And, you know, they don't really have to advertise for their own products” “Armed with [the sellers'] information, it appears that Amazon would copy the product to create a competing private-label product.” (Nadler & Cicilline, 2022) "They can sell it for a lesser price because... they are not paying certain fees and commissions, and the margin itself." “In March 2020, Amazon announced that it would begin temporarily delaying shipments of all non-essential products from its warehouses, regardless of whether they were sold by Amazon or by competing third-party sellers. Amazon reportedly excepted this policy and continued to ship non-essential items sold by Amazon Retail from its warehouses.” (Nadler & Cicilline, 2022) "I know [M] actually email blasts them, due to fake reviews. He's called out a lot of like Amazon basics items. There's definitely not a fair standard." A 2021 investigation from Reuters, based on Amazon's internal documentation, shows the company ran a systematic campaign of creating knockoffs and manipulating search results to boost its own product lines in India. In 2022, the UK's CMA launched an investigation under similar allegations. In a 2022 seller survey, 74% of them considered that search results are manipulated to benefit Amazon's products over those of sellers (ILSR, 2022) In 2022, The US House of Representatives antitrust subcommittee asked for an investigation into Amazon’s preferencing of its own brands and exclusive products in search results. - Exploit sellers’ data “That does create conflicts, particularly when they start to bring in own-brand products where they've obviously got a rich data set to harvest. They obviously say "no, no, we don't do anything there." “European Commission's investigation “focuses on the use by Amazon of accumulated, competitively sensitive information about marketplace sellers, their products and transactions on the Amazon marketplace, which may inform Amazon's business decisions.” (Nadler & Cicilline, 2022) “One of the widely reported ways in which Amazon treats third-party sellers unfairly centres on 66 Essays on Power in Digital Platforms Amazon’s asymmetric access to and use of third-party seller data” (Nadler & Cicilline, 2022) Source: Based on interviews and archival sources 67 Create an unpredictable and precarious environment for sellers. We observe that Amazon has created an unpredictable and precarious environment for sellers, in line with the findings of previous research on entrepreneurs operating on digital platforms (Curchod et al., 2019; Cutolo & Kenney, 2021). Our interviewees describe that the main challenge they face is the high uncertainty surrounding fundamental aspects of their activities in the Marketplace. This uncertainty pertains to how the platform rules are applied, how they are changed, if they can continue to use their account, how unprotected they are from emerging issues, and how hard it is to solve these problems. This unpredictable environment sustains and increases the asymmetry of bargaining power between Amazon's and the sellers' positions, implies less platform management costs, and provides Amazon with opportunities to profit. We distinguish three primary governance practices in creating this environment: Lack of platform support for complementors Finding answers, solutions, or support as a seller in the Marketplace is very hard. The reasons behind warnings and account blocks are not always clear or even communicated, and seller support is largely automated, sometimes prompting responses unrelated to the specific query. Contacting a human representative is not necessarily better: Sellers describe that the service support is often uncooperative, claiming a lack of authority or suitability for addressing the issue. Further, Amazon's internal teams seem to be highly siloed and competing with each other. Lastly, this lack of clear communication channels and responsibilities favours Amazon in passing the error costs to sellers. Rules uncertainty While sellers are required to follow the Marketplace’s rules, this duty is hard to fulfil because rules are unclear and constantly changing. First, the terms of use of the different services and the characteristics of their functionalities are frequently updated, making it hard to keep up with and avoid infringement. Sellers also describe that the wording is not always direct and straightforward when a new policy or procedure is in place. Further, some policies’ aspects might not be declared (for example, a parameter of exclusion), with sellers finding out only when it affects them. This creates blind spots, making sellers unsure about their compliance with rules and leaving the decision to Amazon’s interpretation. Discretionary policing We observe that Amazon’s policing of the Marketplace is discretionary, which creates unfair differences in the treatment of sellers and how they compete. Sellers denounce that rule enforcement is unreliable and inconsistent. On the one hand, sudden account or product suspension is a shared concern. Appealing a suspension is tricky because the reasons for infringement are usually unclear, and Amazon’s seller service is not responsive. Sellers describe that it is not rare to receive suspensions with allegations entirely unrelated to their businesses. At the same time, sellers complain that Amazon does not do a good job policing fraudulent sellers. These might be performing attacks on other sellers through account or product page hacks, infringing the Marketplace’s rules to obtain a competitive advantage, or engaging in illegal activities. 68 Essays on Power in Digital Platforms Profit from an asymmetric position We find that Amazon uses its position as the platform owner to profit from sellers by imposing the use of optional services and unfairly competing with them. We identify two main governance practices that allow the platform to profit from its asymmetric position: Impose the use of services to sellers While all of Amazon’s services are legally optional, sellers describe several ways through which they find themselves forced to use them. For example, sellers that try to use non-Amazon shipping services have been penalised or put in abnormally long waiting queues to release their inventory. The use of paid advertising has become almost a requirement to keep a position in search results. More generally, we observe that Amazon has been able to monetise problems in the platform that it has created itself, such as the lack of seller support or late delivery pickups. In addition, sellers denounce that they are locked-in in using services once they start to use them, as they cannot move back without fundamentally compromising their competitive position. Unfair competition with sellers The fact that Amazon both manages the Marketplace and sells on it creates numerous possibilities for unfairly competing with sellers: Amazon has a distinctive competitive advantage by having access to all of the Marketplace’s data. Further, Amazon has been found to systematically identify successful sellers’ products and launch a similar version while favouring its positioning in search results. Amazon also competes unfairly with sellers by eluding the platform’s rules for its products, for example, favouring its positioning in search results. 3.4.1.3. Discrepancies between the platform’s declared and undeclared governance practices Overall, we observe sustained discrepancies between Amazon’s declared and undeclared governance. Table 9 shows examples of these discrepancies: we observe a stark contrast between what Amazon claims about its intent, role, and values, what they do, and how they try to limit its own liability. 69 Essays on Power in Digital Platforms Table 9. Comparison between Amazon’s claims, legal responsibilities, and actual practices What Amazon claims about its intent, role, and values How Amazon claims to support sellers “Every day, our global teams work to provide selling partners with the strategies, answers, guidance, programs, and solutions they need to succeed in our store… We have more than 24,000 Amazonians focused on supporting and inventing for selling partners”1 How Amazon claims to create a trustworthy environment for sellers “Amazon strives to be Earth’s most customer-centric company, offering vast selection, low prices, and fast delivery. We deliver on that vision by creating shopping experiences that customers, sellers, and brands can trust”2 How Amazon claims to support policing “In 2020, we invested over $700 million and employed more than 10,000 people to protect our store from fraud and the abuse of customers and sellers.”3 How Amazon claims to offer the best services for sellers to choose “We know sellers have lots of options, and we work hard to make Amazon the best partner to help drive their success.”4 How Amazon depends on sellers’ success “Our success depends on the success of our selling partners. Amazon’s work with third-party sellers is one of the greatest partnership stories in retail. Together, we make a great team.”5 Amazon’s legal responsibilities as defined by its contractual agreements Amazon’s practices What Amazon actually does Amazon does not warrant a service level. “We do not warrant that the functions contained in the Amazon sites Lack of platform owner’s support for complementors Finding answers and solutions is very hard because: - Amazon is a highly siloed organisation - Seller support is largely automated or operated by usually under-resourced offshore teams that lack authority. - The reasons behind warnings and account blocks are not always clear and usually are not explained or even communicated In addition, this lack of response facilitates passing the error costs to the sellers. Amazon reserves the right to change services and the platform Amazon reserves the right to modify the Agreement at any time with immediate effect. Continued use of the services constitutes acceptance Rules uncertainty When a new policy or procedure is in place, the wording is not always direct and straightforward. Further, some specific aspects are not declared (for example, a parameter of exclusion), with sellers finding out only when it affects them. The constant update of the terms of service and new functionalities makes it hard to keep up and don't infringe on them. and the services will meet your requirements or be available, timely, secure, uninterrupted, or error-free, and we will not be liable for any service interruptions”6 7 of the changes. Amazon reserves the right to change and suspend services, websites, and content (including seller’s products and listings). 8 Amazon has no responsibilities for policing seller abuse Sellers must comply with the Seller Code of Conduct, while no Amazon 9 Code of Conduct exists. Amazon is not responsible for policing counterfeit, tax evasion, or other illegal activities on the platform. 10 The agreement states that all the services are optional “Amazon Services Business Solutions, a suite of optional services for sellers including Selling on Amazon, Fulfilment by Amazon, Amazon Advertising, Transaction Processing Services, and the Selling Partner 11 API.” The agreement provides Amazon with leverage against sellers’ legal action All disputes will be settled by binding arbitration instead of court through a request letter sent to Amazon's registered agent. Disputes will not be conducted by class/consolidated/representative action. 12 Discretionary policing There is an unreliable and inconsistent enforcement of rules, creating differences between sellers. This difference is also created because of some sellers' legal obligations and misbehaviour overlooking. Impose the use of services to sellers Sellers have been penalised for not using optional services and forced to pay for advertising to keep a position in the search results. More generally, Amazon has used the problems it creates (like lack of seller support or late Amazon pickup) to sell additional services (like an account representative or the use of Amazon warehouses) Unfair competition with sellers Amazon has been found to exploit sellers’ data to obtain a competitive advantage, copy sellers’ products, and elude platform rules for Amazon products (like selfpreferencing in search results) In addition, it increases its margin by avoiding platform fees and commissions for its own products while defining price rules for the seller’s products. Sources: 1) 2021 Amazon Small Business Empowerment Report; 2) 2020 Amazon Brand protection report; 3) 2021 Amazon Small Business Empowerment Report; 4) 2021 Amazon Small Business Empowerment Report; 5) 2021 Amazon Small Business Empowerment Report; 6) Amazon Services Business Solutions Agreement, section 7.a.; 7) Amazon Services Business Solutions Agreement, section 15.a., 15. b & 15. c; 8) Amazon Services Business Solutions Agreement, section S-6; 9) Amazon Services Business Solutions Agreement, section 3 & Selling Policies and Seller Code of Conduct; 10) Amazon Anti-Counterfeiting Policy & Tax Policies; 11) Amazon Services Business Solutions Agreement, General Terms; 12) Amazon Services Business Solutions Agreement, section 18. 70 Essays on Power in Digital Platforms While Amazon displays at the core of its official mission a vision to create a trustworthy environment, the agreement allows them to change the services and the platform with short or no notice. We find that the legal agreement plays a critical role in setting up the tensions between the two governance faces: It dismisses Amazon from responsibilities related to its undeclared practices. For example, while Amazon declares how much effort and resources it puts into supporting sellers, the agreement states that a service level is not warranted. Amazon also declares to make significant investments to prevent fraud and abuse, but the legal agreement states that Amazon bears no responsibility for policing them. 3.4.2. Sellers’ adaptive ecosystem participation practices We found that sellers’ behaviour is shaped by both the declared and the undeclared set of Amazon’s governance strategies. We identify three patterns in how sellers respond to navigate the often-contradictory Marketplace’s governance: (1) Play by the rules of the platform’s declared governance; (2) Address and exploit the platform governance contradictions; (3) Fight platform dependence (see Table 10 for a summary of sellers’ adaptive participation and Table 11 for exemplary excerpts). Table 10. Sellers’ adaptive participation: Examples, mechanisms, practices, and outcomes Sellers’ Adaptive Participation: Examples of Activities and Processes (1st-Order Concepts) - Assign a job position exclusively to deal with policy updates - Deploy compliance verification activities in all Amazon-related processes - Invest in brand recognition; Launch new products Sellers’ Adaptive Mechanisms (2nd-Order Themes) - Keep updated on Amazon’s policies and rules changes - Assure compliance with Amazon’s rules - Product and brand development - Optimise sourcing; Develop demand forecasting; Improve packaging. - Optimise supply chain operations - Outsource marketing, shipment management, data analytics, administration, etc. - Bend description or picture rules to increase the position in search results and improve product display - Use fake or paid-for reviews, sabotage other sellers’ accounts and products, sell counterfeit products and manipulate review scores. - Develop procedures to revert unexpected account suspensions and blockings by Amazon - Learn about Black Hat tactics and possible remedies - Outsource activities - Drive sales through sellers’ websites; Use other transaction platforms; Partner with traditional distributors - Sell the account to specialised funds - Sales channel diversification - Attach notes inside the products encouraging direct communication with sellers - Drive buyers to the seller’s website to collect data - Create alternative communication channels with customers - Collect and own customer data Sellers’ Adaptive Practices (3rd-Order Themes) Sellers’ Adaptive Practices: Outcomes (Aggregate Dimensions) - Adapt to the platform’s rules and requirements - Play by the rules of the - Develop and profit platform’s declared governance within the platform’s declared rules - Use ‘Grey Hat’ tactics - Use ‘Black Hat’ tactics - Bend, elude, or hack the platform’s rules - Address and exploit the - Learn to deal with Amazon’s undeclared governance - Defend from other sellers' attacks - Sell the Amazon account platform governance contradictions - Defend from the ecosystem’s undeclared practices - Reduce exposition to the platform - Evade the - Fight platform dependence platform’s intermediary role Source: Based on interviews, participatory observation, and multiple archival sources 71 Essays on Power in Digital Platforms Table 11. Sellers’ adaptive ecosystem participation practices: Evidentiary support Participation practice Strategy - Adapt to the platform’s rules and requirements Play by the platform’s declared rules - Develop and profit within the platform’s declared rules - Bend, elude, or - Keep updated on Amazon’s policies and rules changes - Assure compliance with Amazon’s rules - Product and brand development - "My job revolves around making sure I'm keeping up on the new policy changes that Amazon has in the marketplace." - Optimise supply chain operations - “I built an FBA forecasting tool… not to sound like I'm bragging but I know our forecasting tool for FBA is pretty robust. - Outsource activities - - Use ‘Grey Hat’ tactics - - Use ‘Black Hat’ tactics - hack the platform’s rules Address and exploit governance’s contradictions - "We've made every mistake in the book, we've gotten hit with things that weren't really a mistake. And now we know how to address those problems." - "We're constantly looking to add new SKUs and branch off and define the brand" - - - Learn to deal with Amazon’s undeclared governance - Defend from the - Defend from other sellers' attacks - Reduce exposure to the platform - ecosystem’s undeclared practices Fight platform dependence Exemplary excerpts from interviews - Sales channel diversification - Sell the Amazon account - Like we put like five years of work into it.” "We manufacture our own items, you're able to control the cost a little bit more. And we were able to market it better that ties into PPC spend also, to be able to sell more and just make more money." "Administration is outsourced. All the like things like sorting out shipments going into Amazon stock levels, inventory levels, I've got somebody who looks after that…we've got someone who looks after FBA orders, someone who basically processes our website orders via Amazon... I have three people who do my social media as well." “I don't do anything that I'm not supposed to. I mean, I might bend the rules slightly, so I might change my title and not follow the exact, you know, Amazon formula, because it doesn't work as well… I do bend the rules as much as I can, in terms of listings, images, EBC, but I don't ever do anything that's actually dodgy.” "One big, big problem is the black hat tactics take for recent companies… there was a period where we got attacked 16 times with false patterns, false images, false counterfeit claims just to stop momentum. And that's very hard for a seller." "I know people who do black hat techniques, I know of them. I've met them in conferences and things like that and they do stuff that's highly dodgy, highly illegal. But then they make a lot of money, so they don't care so much, and they probably don't get called." "It doesn't matter what category you're in, this is going on… there's two different sorts of things, people that are attacking you, specifically… And then there's other people that aren't attacking you directly but they're doing things to cheat you out of sales" "I guess it is a sense of firefighting... you can't really work in a strategic sense if you're just putting out fires all the time." "One of the things that probably holds me back from growing my business the most is the amount of time I have to spend on sorting out Amazon crap." “I have a lot of plans of action that we've had to do typically, within 24 hours. I have to spend a couple hours and figure out why… and then plan, do check, adjust kind of thing. I have to call into Amazon and try to explain to somebody who might be… he doesn't really care. So, you're kind of at their mercy… you kind of have to eat crow at the first one and then just go down the list.” “To contest or to reply to this Blackhat you have to resort to Amazon, or you have to know how to handle yourself, you have to just email harass them two or three times.” “We've been around since 20[XX], we got [Y] brands. One of them we don't sell on Amazon deliberately, so you may see that.” “We are looking to diversify to other channels, because I would say 90% of our profit comes from Amazon exclusively.” “I think the position that we're at in our journey… we're probably not going to do this forever. When I come to sell this company, I'm conscious of two things, what our profit is, and what multiple we could command for that, then I think that we have a business that's probably worth a reasonable amount of money.” 72 Essays on Power in Digital Platforms - Evade the platform’s intermediary role - Create alternative communication channels with customers - Collect and own customer data - “Amazon is trying to remove the ability for sellers to have any contact and real relationship with what it considers its customers… The way I get around that is I make sure that I put personality into the packaging… I use a very personal stroke” - “The market values customers and retention, you know, when you can get valued on a multiple of sales as opposed to EBITDA. I think that shows there's actually a real importance to having that customer data.” Source: Based on interviews 73 Essays on Power in Digital Platforms Play by the rules of the platform’s declared governance First, to a large extent, sellers play by the platform’s declared rules, adapting to them and following the defined activities to develop their businesses and make a profit. This includes adapting to the platform’s rules and requirements and developing and profiting within the platform’s declared rules Adapt to the platform’s rules and requirements Sellers describe that one of the keys to success is constantly being on top of the Marketplace’s rules. This requires them to assign enough resources to remain up to date on Amazon’s policies and rules changes. More importantly, sellers must then ensure compliance with the rules, which requires specialised experience and training. Develop and profit within the platform’s declared rules Participation in the Marketplace has been fundamental in growing all our informants’ businesses. Within the framework and tools offered by Amazon, sellers engage in different business development activities, such as attaining visibility and differentiation by investing in marketing and brand development, optimising supply chain operations, and adequately balancing the outsourcing of activities. The platform facilitates this through its own additional services and by enabling third parties through API integration. Address and exploit the platform governance contradictions We observe that sellers both address and exploit contradictions between the declared and the undeclared governance. In doing so, they introduce new practices into the platform ecosystem and adopt additional roles than those set in the declared governance. We observe two sets of practices: Bend, elude or hack the platform’s rules We find evidence that sellers systematically engage in forbidden activities for profit and obtain competitive advantages in the platform. Within the Marketplace, these practices are known as ‘Grey Hat tactics’ and ‘Black Hat tactics’ (see Table 12 for detailed examples). Grey Hat tactics consist of subtly bending or eluding platform rules to obtain minor advantages, such as infringing product descriptions or picture rules to increase the position in search results. On the other hand, Black Hat tactics involve practices explicitly prohibited by the platform, including illegal ones, for example, paying customers for favourable reviews. These practices are ubiquitous in conversations with our informants, events and online forums, and industry reports. For instance, in two surveys to sellers in 2020 and 2021, between 12% and 16% of respondents declared engaging in some sort of Black Hat tactic to help their Amazon business (Jungle Scout 2020, 2021). Further, some of these specific practices, like the use of fake product reviews and how markets are organised around them, have been studied in previous research (Choi, 2020; He et al., 2022). 74 Essays on Power in Digital Platforms Table 12. Detail of Sellers’ adaptive practices for bending, eluding, or hacking the platform’s rules Type of practice - Grey Hat (subtly bending or eluding platform rules to obtain minor advantages or otherwise gaming the system) - Black Hat (explicit violations of Amazon’s Terms of Service, including illegal practices) Practice - Optimising product images - Optimising product description wording - Using special characters and emojis in the product description - Using review services/farms - Using review groups/rings - Request customers for a specific type of feedback - Selling counterfeit products - Use of Zombie accounts - Listing Hijacking Description - Adding banners, texts, and callouts on images to catch the eye of shoppers, which infringes Amazon’s rules for product images. - Amazon defines strict rules for the order in the wording of product titles and descriptions. Sellers have found that changing this order can be beneficial for driving consumers’ attention. - Amazon restricts the use of special characters. Using emojis can attract customers, and using the slash after the first words can help with Google SEO optimisation - Companies hired by sellers to acquire their products and leave favourable reviews (or unfavourable reviews to competitors) - Groups of sellers that leave each other favourable reviews - Asking explicitly for positive reviews is forbidden by Amazon. Some sellers include discounts or coupons for future purchases in exchange for a positive review - The existence of knockoff goods of various quality has been a constant in the Amazon Marketplace. - Creating fake accounts to search for and buy products, which influences Amazon’s algorithms. Use of these accounts might include attacking other sellers by buying their products, returning them, and leaving negative reviews - If a product sold on Amazon is not registered with a brand and protected by intellectual property rights, other sellers can access the product listing description and change it. A seller can then change the product's image or description by including offensive language to get the listing suspended. Source: Based on interviews, participatory observation, and multiple archival sources Defend from the ecosystem’s undeclared practices Sellers adopt defensive practices in the face of Amazon’s unreliable rules enforcement. One part of this involves learning how to deal with Amazon’s undeclared governance. In particular, our informants highlight the importance of preparing for unexpected account suspensions by Amazon and learning how to revert them. The extent of this problem for sellers has led to the appearance of specialised actors in the ecosystem, providing accountsuspension-specific legal advice and account-suspension insurance. At the same time, sellers learn to defend themselves from other sellers’ attacks, as Amazon fails to meet its responsibilities in protecting them. Several Black Hat tactics, mentioned before, aim to block a competitor’s account or product in the Marketplace, for example, by sabotaging a product’s page or intimidation through false claims of intellectual property rights infringement. Surveys to sellers show that protecting from Black Hat tactics is a constant preoccupation, with answers ranging from 40% to 66% (Jungle Scout 2020, 2021, 2022). Fight platform dependence We observe that even sellers with large and steady sales volumes engage in activities to fight their dependence on Amazon to reach their customers. In addition to Amazon having durable market power in some of the markets in which it participates (Nadler & Cicilline, 2022), dependence is attested (and enhanced) by the fact that unless there is a problem with the transaction, sellers are almost invisible and unreachable by consumers: sellers names 75 Essays on Power in Digital Platforms are barely displayed on the product page, and Amazon strictly forbids sellers to attempt to communicate directly to customers. We find that sellers try to fight this dependence by: Reduce exposure to the platform A common trend among sellers is the attempt to reduce their exposure to Amazon. First, they look for ways to diversify their sales channels, as the Marketplace tends to represent a very high percentage of their total income. In addition, we observed in 2018 the emergence of funds specialised in acquiring and operating third-party Amazon seller accounts, creating an additional feasible exit strategy for sellers. Evade the platform’s intermediary role Even when the legal agreement states that Amazon is not an intermediary between the sellers and its buyers, Amazon tightly controls that interaction. Because of this, sellers attempt to create alternative communication channels with their customers, even when this risks rule infringement, and look to drive buyers to their own sites to collect and own customers’ data. 3.4.3. Two-faced governance In our study of the Amazon Marketplace, we observed two opposed yet coexisting “worlds” (consistent sets of modalities of engagement): One is based on collaboration, clear rules and roles, precise and automated processes, and transparent behaviour metrics, leading to collective value creation. The other is conflicting and oppressing, with uncertain rules and changing roles, an unreliable and uncertain decision-making environment, where opaque processes lead to collective rule transgression. The active creation and maintenance of the contradictions between these two worlds are the fundamental activities that sustain the platform’s two-faced governance as a deliberate strategy to increase platform power. This uncertain space, rife with contradictions, facilitates the platform owner’s lack of accountability and weakens participants’ position as they do not know which face of governance they will encounter next. We have therefore shown that the platform ecosystem’s dynamics are only partially described by relying on the declared governance. The platform owner and its complementors follow this declared governance to some extent, but their behaviours are also guided by practices that contradict it. To build a complete picture, we integrate these different elements in a process model of what we call “two-faced platform governance” (See Figure 5). 76 Essays on Power in Digital Platforms Figure 5. Two-faced platform governance: A process model Platform Firm’s practices Declared Facilitate and expand value creation - Attract and keep complementors by creating opportunities for profit - Attract and keep consumers by controlling and improving its buying experience - Ensure favourable conditions for value creation Ecosystem’s Governance Complementors’ practices Undeclared - Lack of complementor support Rules uncertainty Discretionary policing Create an unpredictable and precarious environment for complementors - Impose use of services to complementors Unfair competition with complementors Profit from asymmetric position - - Unwritten rules - Transformed roles - Declared rules and roles Play by the platform’s declared rules Favors Address and exploit governance’s contradictions - Adapt to the platform’s rules and requirements - Develop and profit within the platform’s declared rules - Bend, elude, or hack the platform’s rules Defend from the ecosystem’s undeclared practices Fight platform dependence - Reduce exposition to the platform - Evade platform’s intermediary role Two-faced platform governance refers to the platform owner’s practice of pursuing simultaneously seemingly contradictory governance practices, creating ambiguous conditions for platform participation. We observe that this practice serves various objectives: it sustains and increases the power asymmetries between the platform owner and the complementors; it is a way for the platform owner to balance between value capture and value creation; and it is a way for both the platform owner and the complementors to engage in practices that are potentially illegal or that would otherwise raise authorities’ attention. In the two-faced platform governance, the platform owner engages in declared and undeclared governance practices. On the one hand, the declared governance aims to facilitate and expand the value-creation processes. It defines the proposed rules and roles that platform participants are expected to follow. On the other hand, the undeclared governance practices aim to increase the asymmetry of power with the complementors and profit from it. The two-faced platform governance seems consistent with a platform owner’s deliberate generating and sustaining of a form of superior bargaining power over complementors. In turn, the discrepancies between the declared and the undeclared governance favour complementors’ practices that address and exploit these contradictions. Overall, both the platform owner and the complementors’ undeclared practices reshape the ecosystem’s governance, transforming the roles and creating unwritten rules for participation. 3.5. Discussion Our study of governance in the Amazon Marketplace holds important theoretical and practical implications for platform research. 77 Essays on Power in Digital Platforms Our study offers four main contributions: First, it contributes to the platform strategy literature by providing new evidence that challenges central claims in this literature, such as the assumptions that it is always in the best interest of platforms to satisfy all their sides when designing their governance, that governance rules should be clearly defined, or that the platform owners should enforce the governance rules they set. Second, it enriches the construct of platform governance by identifying that there are two types of governance practices: declared and undeclared. It also analyses how the two interact. Third, it enriches our understanding of how platform power can be identified and exploited. It provides evidence that platform power can be exerted in ways not previously identified by focusing on the strategic mismanagement of the rules and the creation of uncertainty. Fourth, it offers managerial and policy implications and opens up avenues for future research. 3.5.1. Contribution to the platform strategy literature By integrating the undeclared practices as part of the platform governance and analysing how they interact with the declared governance, we have provided challenging evidence to the central claims of the platform strategy literature described in the literature section. This evidence holds significant implications for theory, which we summarise in Table 13 and develop in detail below. Overall, our findings contribute to refining thinking on what “good platform governance” means, including which criteria are relevant for measuring governance performance and the related trade-offs. Table 13. Challenging evidence to the literature’s central claims on platform ecosystem governance and consequences for theory Literature Claims - It is in the platform owner’s best interest to design the governance in a way that satisfies all the platform sides Challenging Evidence - We observe platform sides’ members operating constrained in an unpredictable environment, and this proves beneficial to the platform owner - Rethink to what extent and how the platform owner keeps each side satisfied and balances conflicting interests, how virtuous the cycles of value creation really are, and to what extent complementor participation is voluntary - Platform rules might not be clearly defined by the platform owner, be often changed by the platform owner, and prove hard to understand and to follow by members of the platform’s sides - Governance rules and roles can be deliberately made to be unclear and generate uncertainty without necessarily creating coordination issues. - Complementors have to assume roles not included in the declared governance - We observe the platform owner not consistently enforcing their own declared rules and not meeting their role’s responsibilities. - Platform owners not consistently enforcing the governance rules they set can benefit themselves and some complementors. - In our study, sustained violation of the governance arrangements has not hindered seller participation - Sustained violation of the governance can be part of sustainable platform owners’ and participants’ strategies. - Inobservance of the declared governance can be used as a competitive strategy - Reassessment of the platform owner’s challenges and strategies for enforcing governance of aligning interests and resolving conflicts - Governance rules and roles should be clearly articulated and precisely defined - Platform owners should enforce the governance rules that they set - Sustained violation of the governance will lead to platform failure Consequences for Theory - Reconsider the roles typically assigned to the platform owner as an enforcer and its incentives to do so. 78 Essays on Power in Digital Platforms Literature claim 1: It is in the platform owner’s best interest to design the governance to satisfy all the platform sides. Our study indicates that Amazon sellers are frequently unsatisfied (they feel constrained, operate in an unpredictable environment, and are subject to substantial power asymmetries), therefore showing that having some sides constrained can prove beneficial to the platform owner. These results call for a revision in our theories of to what extent and how the platform owner keeps each side satisfied and balances conflicting interests and to what extent complementor participation is voluntary (Adner, 2017; Kretschmer et al., 2022). As we observe one of the platform sides struggling while the others profit, our findings also suggest the need to reconsider how virtuous the cycles of value creation really are (Parker et al., 2016). Literature claim 2: Governance rules and roles should be clearly articulated and precisely defined. We offer evidence that platform rules governing some of its sides might not be clearly defined and that by changing frequently, they can be harder to understand and follow by complementors. In addition, we show evidence of complementors having to assume roles not included in the declared governance to participate successfully in the platform. These results imply that uncertainty and lack of clarity in rules and roles can be part of the governance design (Eisenmann et al., 2009; Saadatmand et al., 2019). Literature claim 3. Platform owners should enforce the governance rules that they set. Our study offers evidence of the platform owner purposely not consistently enforcing the rules they declare to abide by, not meeting their role’s responsibilities, further creating an ambiguous and uncertain ecosystem environment. Moreover, we show that this practice can benefit the platform owner and some complementors. These findings challenge the prevailing idea that certainty stemming from rules being properly enforced is always a good thing for the platform and its ecosystem. These results also pose questions about what we know about the roles typically assigned to the platform owner, its incentives to enforce rules, and what its informal authority is built upon (Tiwana et al., 2010). Literature claim 4: Sustained violation of the governance will lead to platform failure. Our study shows that sustained violations of some of the governance arrangements in the Amazon Marketplace have not hindered seller participation. On the contrary, between 2018 and 2021 (the period of our analysis), the gross merchandise volume (GMV) sold by sellers grew 135%, or USD 216 billion, even more than Amazon’s own GMV growth (83%, USD 96 billion) (Amazon, 2019, 2020, 2021, 2022a). We also find evidence that the inobservance of the declared governance has been used to profit both by the platform owner and by part of the complementors. These observations re-signify the platform owner’s challenges in enforcing governance and resolving conflicts (Adner, 2017; Parker et al., 2016). In the platform governance literature, the owner plays this role as both a self-interested and benevolent agent (Hurni et al., 2021). We offer a different view, where the owner might deliberately explore how conflict plays out and impacts the platform’s dynamics and act accordingly to profit from the outcome. 3.5.2. Enriching the construct of platform governance 79 Essays on Power in Digital Platforms Our study enriches the construct of platform governance by identifying and integrating a new set of practices (the undeclared governance practices) that were hitherto ignored. We know of no other studies that identify and analyse the consequences of ambiguous or seemingly contradictory governance practices. By explaining how actors engage in seemingly contradictory governance practices, we suggest that these practices are better interpreted as abuses of the declared governance, yet at the same time, can be an intrinsic part of the platform governance. We argue here that two-faced governance is a deliberate practice that can serve the platform owner and participants to a lesser extent. Consequently, unclear and changing rules and unreliable enforcement can be part of a high-performing governance strategy rather than a flawed governance design or failed governance implementation. In this way, our model offers a way to understand how different governance practices – some of them seemingly contradictory – interact towards achieving some of the platform agents’ goals. Our model can, therefore, be useful to integrate into a platform governance framework extant research describing deceiving and concealed platform practices, such as the use of “dark patterns” to manipulate user choice (Gawer & Srnicek, 2021) or the production of “reputational insecurity” in gig platform workers as a way to increase the platform owner’s power (Wood & Lehdonvirta, 2022). 3.5.3. Contributions to the literature on platform power Our study contributes to the literature on platform power, offering an enriched understanding of how such power can be identified and expressed. Previous research has provided different theories explaining the sources of platform power and how it affects competition. Scholars identified three main foundations of platform market power – gatekeeping, leveraging of resources, and information exploitation - (Jacobides, 2021; Khan, 2018) and explained how the power accumulated by large platforms can then lead to abusive practices, such as imposing supra monopoly prices or discriminatory clauses (Busch et al., 2021; Jacobides, 2021; Khan, 2018), restricting access to the platform, or using data to create barriers to competition (Crémer et al., 2019; Furman, 2019; Stigler report, 2019). Our model of two-faced governance illuminates a different way platform power operates. Typically, platform power abuse involves the owner inclining the playing field set in the declared governance, for example, by changing a rule or a parameter or charging higher prices. Our study focuses instead on how maintaining rule uncertainty or the discretionary enforcing of rules can be used as a source of power, which does not fit into the previous conceptualisation. We highlight the practice of strategically mismanaging the rules rather than abusing the ability to set the rules. This approach expands the focus of inquiry from how power gets exerted as a clear bargaining power imbalance in terms of transactions or exchanges to how the conditions created for that exchange influence its outcome: in our two-faced governance model, the lack of platform owner’s support for complementors, the creation of an unpredictable environment and the discretionary policing are creating an environment that facilitates imposing the use of services to, and unfair competition with, the platform’s complementors. Our study extends the construct of platform power by providing evidence that it can be exerted in ways that hadn't been previously identified. Consistent with the conceptualization from Chapter 2, we also find evidence of activities through which platform ecosystem participants can exert a kind of counter-power. We contribute to this stream of literature by adding to the repertoire of such activities the complementors’ 80 Essays on Power in Digital Platforms exploitation of the platform governance’s contradictions and their attempts to diminish their dependency on the platform owner. 3.5.4. Managerial and policy implications Our study implies that platform owners may get away with not satisfying all of their platform sides’ members, at least in the short term. We do not advocate that platforms should do so, but we observe no evidence of swift “discipline of the market” that punishes the platform if it chooses to do so. Consequently, platform owners might explore which sides can be left, at least temporarily, unsatisfied – and to what degree – without compromising the platform’s business model. Second, our study suggests that for the platform owner to leave certain aspects of the governance deliberately uncertain or unclear can provide it with a certain kind of manoeuvrability. Participants, in turn, might be able to resist and proactively shape part of their own roles, even within ecosystems where large platforms operate with significant power asymmetries. Third, acting on the evidence that platforms seem to operate successfully, at least in the short term, without consistent rule enforcement has practical consequences. For example, platform owners might exploit this finding by reducing policing activities, which would reduce their costs. Finally, if rule infringement does not necessarily lead to platform failure, platform owners might explore and find “buffer zones” of rule violation, that is, spaces of activities where rule violation is tolerated. These can provide flexibility in adjusting governance, capturing value, or changing the platform's design. These findings are likely to be more applicable, at least in the short term, for platforms that enjoy market power. The more durable the platform’s market power, the more the platform will be able to get away with these practices without a “swift discipline of the market”. At the same time, our research also holds important consequences for policymakers. In order to regulate platforms effectively, policymakers need to have a nuanced understanding of how platform power is expressed. Our empirically based, precise description of the various forms of platform practices that lead to the accumulation of platform power and our nuanced discussion of the dynamic and interdependent consequences on complementors can usefully enrich the empirical basis upon which regulatory enforcers may enforce existing law. Our contribution to policy is particularly timely in a context where the behaviour of digital platforms is under increasingly great scrutiny worldwide and where policy for the regulation of digital markets and their associated platform ecosystems has been evolving rapidly, for example, with the approval of the new Digital Markets Act by the European Parliament in 2022. 3.5.5. Limitations and avenues for future research As with any empirical study based on one case study, our work has limitations. They can be addressed in future research. One of the main limitations of our investigation is that our case consists of a unique platform, Amazon, which has unique characteristics. As such, it has durable market power and has been identified as a “gatekeeper” or 81 Essays on Power in Digital Platforms company with “strategic market status” by regulatory agencies in Europe and in the UK (European Commission, 2022). This raises the question of the generalizability of our findings and opens up avenues for new research. Further research could, therefore, examine multiple cases of platform governance over time and develop insights on the conditions under which two-faced governance is viable and opportune for the platform owner and possibly for the complementors. Other lines of investigation might explore how uncertainty and a lack of clear rules impact different types of platforms and might affect each platform side differently. In addition, our findings invite us to explore the conditions under which rules enforcement can be discretionary. Digging deeper into the practice of enforcement, further studies might explore whether the appearance of a certain degree of enforcement might be sufficient to align participants. For example, such studies could examine whether a recurrent change of rules and a partial degree of rule enforcement create an environment where policing efforts might be reduced while still being tolerated. Other studies might also engage in methodological development to better observe undeclared governance practices. Finally, future research could also examine how our findings relate to the organisational theory of decoupling (Bromley & Powell, 2012; Meyer & Rowan, 1977; Scott, 2008). This theory suggests that organisations often decouple by overtly adopting policies that conform to expectations regarding formally stated goals and operational practices but that, in reality, there is often a discrepancy between formally stated goals and actual behaviour. This theory, therefore, appears to offer a set of concepts that seem to illuminate the conducts we observed and pose the question of whether the unique organisational characteristics of platform ecosystems might bring new dimensions to decoupling. 3.6. Conclusion Our investigation of the Amazon Marketplace has revealed dimensions of platform governance hitherto unidentified in the literature and has provided evidence that challenges central claims in the platform strategy literature. We defined two-faced platform governance and found evidence that it occurs in platform ecosystems. It contains both declared and undeclared governance practices, which have significant effects. We characterised two-faced governance and offered a process model describing how both sets of governance practices interact. Our findings illuminate new aspects of how platforms govern ecosystems and also how, to some extent, governance is transformed by participants’ practices. We have shown that for platform owners, it can appear to be sustainable and even narrowly beneficial, at least in the short term, to sustain uncertainty and lack of clarity as to the rules they impose on complementors. We argue that the platform owner can strategically mismanage its own rules and mismanage its enforcement as an integral part of how it governs its ecosystem. Our two-faced governance model for platform ecosystems expands the notion of governance to new dimensions and practices and contributes to a more precise and nuanced understanding of how platform power gets expressed. We have also derived contributions to the literature on platform strategy, platform governance, and platform power. As digital platforms and ecosystems become the dominant organisational form of the digital economy (Kenney & Zysman, 2016; Gawer, 2022), the subject of platform ecosystem governance is bound to become more central. 82 Essays on Power in Digital Platforms Whether in strategy, information systems, and organisation studies, scholars’ focus is shifting away from the clarification of platform business models and the identification of their success factors toward a more complete and dynamic exploration of platform firms’ behaviour, which will lead to a deeper understanding of platform governance. We hope that scholars will find the results and discussion presented in this paper helpful in providing further insights on platforms’ impact on competition and innovation and that our research will be relevant to both management practitioners and policymakers. 83 Essays on Power in Digital Platforms 4. The evolutionary dynamics of platform orchestration6 4.1. Introduction Platform ecosystems have unique ways of coordinating constitutive agents and facilitating transactions among them, leading to novel forms of value creation triggered by network effects (Parker & Van Alstyne, 2017; Rochet & Tirole, 2003). Complementors are fundamental actors in platform ecosystems since they build on the basic functionalities provided by the platform owner (PO) to multiply the value creation process (Gawer & Henderson, 2007). Consequently, one of the critical decisions in platform design is how to adequately balance the incentives, rules, and tools available for complementors so they can create value and thrive in the platform, i.e., how to set the conditions for successful complementors’ business model (Hagiu, 2006; Parker et al., 2016). Overall, complementors’ participation and success depend on how the PO designs the ecosystem’s architecture and orchestrates the ecosystem (Adner & Kapoor, 2010; Autio, 2021). Complementor orchestration has been identified as one of the PO’s key tasks for making or breaking success in platform ecosystems (Adner, 2017; Gawer & Cusumano, 2014b). Extant research has identified some of the key trade-offs orchestrators face, such as between securing control and enabling generativity (Cenamor & Frishammar, 2021; Eaton et al., 2015) or facilitating platform growth and capturing value themselves (Lan et al., 2019). A burgeoning stream of research is exploring how PO orchestrates complementors and the consequences of orchestration activities for the platform and the complementors. These studies have contributed to building a robust understanding of how complementor orchestration influences platform growth and performance (Cennamo & Santalo, 2013; Liang et al., 2022) and explored some of its critical consequences over complementors, such as how it influences their performance (Rietveld et al., 2020), engagement (Saadatmand et al., 2019), and participation practices (Zhang et al., 2022). Despite these considerable efforts, some intriguing aspects of platform orchestration remain unstudied. In particular, there has been so far a lack of attention to the possible connections between orchestration and the appearance of new complementors’ business models in the platform. In other words, if the PO’s orchestration can go beyond influencing complementors’ outcomes and manage to change the complementors themselves over time. In addition, I know of no studies that explore how changes in the complementors’ business models can, in turn, affect back the PO. In this article I ask: How do platform owners orchestrate the appearance of new complementors’ business models? And how does the platform owner benefit from this? To answer this question, I studied the Amazon Marketplace from its launching in 2000 to 2021, using an inductive grounded approach based on semi-structured interviews, participatory observation, and secondary data sources. I focused on Amazon’s orchestration activities and how sellers from the Marketplace adapt and respond to those This chapter was presented as a conference paper in the PhD workshop of the 2023 EU-DPRN conference and accepted to be presented at the 2023 ICIS Conference. 6 84 Essays on Power in Digital Platforms activities. I find evidence that Amazon successfully orchestrated the appearance of new sellers’ business models. Further, I discover a pattern of dynamic adjustments between Amazon and the sellers’ business models, resulting in an evolutionary interplay where both are transformed. This interplay is triggered by Amazon’s orchestration activities, which, by providing functionalities, structuring seller competition, and increasing the competitive intensity sellers face, influence the sellers’ business model transformation. In turn, this transformation is instrumental to improving Amazon’s value proposition and increasing and creating new ways in which it captures value in the platform. This study makes two main contributions to platform research: First, by showing that a PO can orchestrate the change of its complementors’ business models, I contribute to expanding our understanding of the PO’s set of capabilities and responsibilities. Second, I contribute to the stream of literature focused on complementor orchestration by identifying new orchestration activities and illuminating new complexities and trade-offs of orchestration. The remaining of this chapter is structured as follows: in the next section, I review the conceptual foundations underpinning the research problem and discuss previous contributions on complementors and platform orchestration. This is followed by the research methods, describing the empirical setting, data sources, and analysis methods. Next, I present the findings, describing the appearance of new sellers’ business models, Amazon’s orchestration activities, and how they interact. In the final section of this chapter, I offer a discussion on how my findings contribute to extant research on platforms and provide some concluding remarks. 4.2. Literature - Related Research Platform ecosystems have been described as novel organisational forms that facilitate the coordination of autonomous agents towards achieving system-level goals without relying on formal authority (Adner & Kapoor, 2010; Gawer, 2014). Platforms harness great potential for value creation, both by facilitating innovation (such as Android or iOS) and by enabling and coordinating transactions (such as Shopify or Uber) (Cusumano et al., 2019). This potential is founded in two properties characterising platform ecosystems: a modular design and a networked structure where interactions occur. A modular design is a defining aspect of a platform’s technological architecture (Gawer, 2014), characterised by the existence of a group of fundamental and stable “core” functionalities (typically provided by the platform owner) that are used in combination with complementary functionalities (typically provided by third parties, also called complementors) to generate a family of products (Ulrich, 1995). Interfaces are the rules governing interactions among the parts (Baldwin & Clark, 2000); they contribute to reducing a system’s complexity and greatly simplify the scope of information required to develop each functionality (Gawer, 2014). Consequently, modular architectures lead to economies of scale (through the production of greater volumes of core functionalities, fixed costs amortisation, and recombinant innovation) and economies of scope (through reducing the cost of new product development), both of which are strong incentives to use modular design strategies (Baldwin & Woodard, 2009). Critically, the extent of the economies achieved, and the platform's potential to foster innovation will depend on adequately defining the modular design. Next, platforms coordinate transactions by acting as intermediaries that bring together (match) and facilitate exchanges between different groups of networked actors (Caillaud & Jullien, 2003; Rochet & Tirole, 2003). As an intermediary, a platform can drastically reduce transaction costs by configuring an adequate price-setting system, 85 Essays on Power in Digital Platforms providing secure means of exchange, generating a reliable reputational system, and establishing rules for good behaviour (Evans, 2003; Parker et al., 2016). By providing a transactional space and facilitating interactions, a platform also creates value through network effects insofar the value derived from platform usage increases exponentially with each additional user (Katz & Shapiro, 1985). Complementors are fundamental actors in the platform ecosystem, as they contribute by creating, expanding, and delivering value, which results in a complex set of relations with the PO (Gawer & Henderson, 2007). Because in platform ecosystems consumers can usually choose among functionalities and – sometimes – how they are combined (Jacobides et al., 2018), PO rely on complementors and have to strategically decide when to cooperate and when to compete with them (Yoffie & Kwak, 2006). Further, complementors’ actions can substantially affect a platform, for example, as they might dispute the platform’s leadership by building on the PO’s resources (Cusumano & Gawer, 2002; Gawer & Henderson, 2007), and by how complementors define their offering within and across different platforms (Tavalaei & Cennamo, 2021). Nevertheless, in large platform ecosystems, the POcomplementor relationship is typically characterised by a significant power asymmetry, as complementors depend on the PO to access their customers, and the PO controls the technological infrastructure and defines the rules for interaction (Curchod et al., 2019; Cutolo & Kenney, 2021). Complementor orchestration has been identified as one of the PO’s key tasks for making or breaking success in platform ecosystems. As argued by Agarwal et al. (2023), a central problem in platform research has been how, in addition to designing the “ecosystem value blueprint” and creating adequate conditions for participants, PO actively manage complementors. Orchestration can be defined as "the set of deliberate, purposeful actions undertaken by the hub firm as it seeks to create value (expand the pie) and extract value (gain a larger slice of the pie) from the network" (Dhanaraj & Parkhe, 2006, 659). To succeed in establishing the platform ecosystem, these PO’s actions to orchestrate third parties need to follow a set of coherent strategic options concerning firm scope, technology design, type of relations with the complementors, and internal organisation design (Cusumano & Gawer, 2002; Gawer & Cusumano, 2014a). Further, the absence of formal authority in the ecosystem means that orchestration involves managing the tensions arising among participants and persuading them to join and behave in a way aligned with the PO’s offering (Adner, 2017; Autio, 2021). Table 14 summarises the main complementor orchestration issues raised in the literature review and its key findings. Extant research has identified some key trade-offs orchestrators face and how these change according to the ecosystem’s stage. One fundamental tension is between securing control (over participants’ behaviour and outputs) and enabling generativity (understood as the ecosystem’s ability to facilitate unprompted innovative outputs from participants [Zittrain, 2009]) (Cenamor & Frishammar, 2021; Cennamo & Santaló, 2019; G. Parker & Van Alstyne, 2017; Wareham et al., 2014). Orchestrators also need to balance between facilitating platform growth and capturing value themselves (Lan et al., 2019), mediate among multiple sides with divergent interests, and propitiate complementors’ investment in the ecosystem while granting them autonomy (Jacobides et al., 2022). In early or nascent ecosystems, orchestrators’ primary responsibilities have been associated to optimising connectivity and maximising generativity while attracting participation and driving network effects within a “coopetitive” environment (Autio, 2021; Daymond et al., 2022). In contrast, in mature ecosystems, the PO’s 86 Essays on Power in Digital Platforms priorities should be controlling bottlenecks (i.e., complementors aiming to occupy a strategic and irreplaceable position in the ecosystem) and creating the conditions for contained contestation to achieve fruitful outcomes from conflicts arising among participants (Autio, 2021; Daymond et al., 2022). A burgeoning stream of research is exploring how PO orchestrates complementors and the consequences of orchestration activities for the platform and the complementors. Part of this research has focused on the outcomes in terms of platform growth and performance of different complementor orchestration strategies. For example, some of these studies discuss the platform performance trade-offs when pursuing growth-oriented versus differentiation-oriented orchestration strategies (Cennamo & Santalo, 2013) or when employing different strategies for complements’ development and commercialisation (Cenamor & Frishammar, 2021). Other scholars have identified the key orchestration activities required to build ecosystem competitive advantages (Williamson & De Meyer, 2012) and how these change in different platform configurations (Liang et al., 2022; Zeng et al., 2021). Given the challenges in aligning and securing complementors’ contributions to the ecosystem, the design of incentives plays a central role in orchestration. Researchers have explored the best strategies to incentivise high-quality complements while edging low-quality ones (Claussen et al., 2013), the outcomes of alternative incentives design for complementors - including no payment for their contributions - (Boudreau & Jeppesen, 2014), and identified the optimal configurations of platform openness and duration of complementors’ property rights (Parker & Van Alstyne, 2017). Finally, another way for PO to orchestrate complementors is by deciding whether to promote complements and which ones, resulting in alternative strategies that involve complement quality, exclusivity, and timing considerations (Rietveld et al., 2019), as well as how to avoid creating complementor bottlenecks (Agarwal et al., 2023). Another group of studies have focused instead on how complementors are affected by the platform’s orchestration. First, complementor performance and how it changes with different orchestration strategies has recently received more scholarly attention. Some studies have explored, for example, the changes in complementors’ performance and behaviour as they obtain a certification or badge by the PO (Rietveld et al., 2021) or when the PO’s governance strategies shift according to their market position dominance (Rietveld et al., 2020). PO and complementors sometimes share the same market space, and different approaches to orchestration have been shown to influence the PO’s modes of entry, thus affecting complementor performance (Young Kang & Suarez, 2022). Next, one of orchestration’s main goals is achieving high levels of complementor engagement, as this means mobilising their resources for value creation within the platform ecosystem. Complementor engagement has been found to vary according to how the interplay of platform architecture and governance is configured (Saadatmand et al., 2019) and how platform boundary resources are deployed (Engert et al., 2022). Finally, recent research has started to explore how some of the complementors’ participation practices are impacted by orchestration activities, showing, for example, how changes in platform access control can transform complementors’ interactions among themselves (Zhang et al., 2022) and how complementors struggle to adapt during technological transitions, increasing the risk of platform failure (Ozalp et al., 2018). In summary, extant research has highlighted the centrality of platform orchestration, contributed to building a robust understanding of how complementor orchestration influences platform growth and performance, and 87 Essays on Power in Digital Platforms explored some of its key consequences over complementors, such as how it influences their performance, engagement, and participation practices. Despite these considerable efforts, some intriguing aspects of platform orchestration remain unstudied. In particular, there has been so far a lack of attention to the possible connections between orchestration and changes in the complementors’ business models. In other words, if the PO’s orchestration can go beyond producing changes in complementors’ outcomes and manage to change the complementors themselves. Recent studies have begun to recognise the importance of the complementor pool’s composition, heterogeneity, and characteristics (McIntyre et al., 2021; Rietveld et al., 2021; Zhang et al., 2022). These studies mainly discuss the consequences of complementor heterogeneity, such as what type of contributions and offerings different types of complementor bring to the platform (Cennamo & Santaló, 2019; Deilen & Wiesche, 2021; Wang & Miller, 2020), how participation strategies vary with complementor size (Hukal et al., 2022), or how complementors’ organisational form and size impact their ability to capture value (Miric et al., 2019; Nambisan & Baron, 2013). Other articles leave open questions, such as whether the degree of heterogeneity in an ecosystem implies limits to its growth (Wareham et al., 2014) or how platforms can attract the right kind and mix of complementors (McIntyre et al., 2021). In addition, I know of no studies that explore how complementors’ transformation can, in turn, transform back the PO. In particular, there are no references to this problem in the studies discussing either the relations between the PO and the complementors (Deilen & Wiesche, 2021; Heimburg & Wiesche, 2022), how orchestrators manage processes of emergence (Daymond et al., 2022), or how platform ecosystems evolve (Alaimo et al., 2019; Holgersson et al., 2022; Rietveld et al., 2020; Srinivasan & Venkatraman, 2020; West & Wood, 2014). Given the importance of complementors and their business models for platform ecosystems, understanding the possible drivers of complementor transformation and its consequences for the PO constitutes a significant gap in the literature. 88 Essays on Power in Digital Platforms Table 14. Complementor orchestration: main findings and issues raised in the literature reviewed Focus Trade-offs faced by orchestrators Orchestration’s impact on the platform Orchestration’s impact on the complementors Orchestration issues raised Main findings Examples in the literature - Securing control vs. enabling generativity Identification of orchestration mechanisms to manage desirable and undesirable generativity and how they change with platform maturity Wareham et al. (2014), Cennamo & Santaló (2019 - Aligning potentially conflicting interests Identification of the types of platform failures that orchestration can address and those that it risks provoking Adner (2017), Lan et al. (2019), Jacobides et al. (2022) - Changes in orchestrators’ role according to the ecosystem’s stage Distinction between the main orchestrator’s responsibilities in nascent ecosystems (optimising connectivity and maximising generativity) and in mature ecosystems (controlling bottlenecks and creating the conditions for contained contestation) Autio (2021), Daymond et al. (2022) - Impact on platform performance Recognition of performance outcomes of different coexisting innovation, growth, and innovation orchestration strategies Cennamo & Santalo (2013), Cenamor & Frishammar (2021) - Building the platform’s competitive advantages Identification of the key orchestration activities required to build ecosystem competitive advantages and how these change in different platform configurations Williamson & De Meyer (2012), Liang et al. (2022), Zeng et al. (2021) - Effective incentives design Expand the orchestrator’s toolbox on how to incentivise high-quality complements while edging low-quality ones, configure platform openness and duration of complementors’ property rights, or identify the conditions under which not paying to complementors remains advantageous for the platform Claussen et al. (2013), Boudreau & Jeppesen (2014), Parker & Van Alstyne (2017) - Whether to promote complements and which ones Identification of the strategic trade-offs to take into consideration in determining whether and when to promote individual complements, and the tools to balance between complementors’ product adoption and avoid creating complementor bottlenecks Rietveld et al. (2019), Agarwal et al. (2023) - Impact on complementor performance Impact of the platform’s owner governance strategies shifting according to its market position dominance, of its different modes of entry to complementors’ market space, and of different complementor certification programs Rietveld et al. (2021), Rietveld et al. (2020), Young Kang & Suarez (2022) - Impact on complementor engagement Impact of different configurations of platform architecture and governance and how platform boundary resources are deployed Saadatmand et al. (2019), Engert et al. (2022) - Impact on complementors’ participation practices How complementors struggle to adapt during platform’s technological transitions, and how changes in platform access control can transform inter-complementor interactions Ozalp et al. (2018), Zhang et al. (2022) Source: Based on the literature reviewed 89 Essays on Power in Digital Platforms 4.3. Research methods I studied the Amazon Marketplace from its launch in 2000 to 2021. I used an inductive grounded approach based on multiple data sources, including semi-structured interviews, participatory observation, and secondary data sources. I focused on Amazon’s orchestration activities and how sellers adapt and respond to them. I adopted an inductive grounded approach, which is best suited for analysing a phenomenon in an exploratory manner (Charmaz & Bryant, 2008; Eisenhardt, 1989). It is oriented towards theory development rather than proving existing hypotheses (Thornberg & Charmaz, 2014). 4.3.1. Empirical setting: The Amazon Marketplace and its sellers’ business models Amazon is a US-based technological company, founded in 1994 as an online bookstore retailer and later diversified into a broad offering of product categories like media, toys, electronics, furniture, and fresh food, among others. As of 2023, Amazon has retail websites in twenty-eight countries, including the United States and the United Kingdom, and is one of the largest Internet-based retailers in the world. In addition, Amazon provides advertising, fulfilment, and cloud computing services, and produces entertainment media content consumer electronics, and is one of the leading providers of smart-home appliances. The Amazon Marketplace is the platform used by Amazon to run its retail business, where consumers can purchase products sold both by Amazon and by sellers. Sellers can be considered Amazon’s complementors because they contribute to expanding and delivering Amazon’s primary value proposition: offering “unlimited selection” and “low prices” to its customers. While Amazon and the sellers often compete with each other, ultimately the presence of both is crucial for attracting customers. Sellers pay Amazon a combination of fees (subscription, referral, closing, high-volume listing, and refund administration) for maintaining a professional account on the platform and listing and selling their products. In addition, Amazon offers several additional services to sellers, such as storage and fulfilment, advertising, and export, among others (see Annex A for a description of all the functionalities– called “programs” by Amazon – mentioned in this article). The Amazon Marketplace is a well-suited empirical setting to conduct this research, being simultaneously one of the longest-standing, largest and most important platform ecosystems worldwide. First, it was launched in 2000, when Amazon allowed third parties to offer their products on the platform Amazon used for its own retail business. This provides a long observation period to study the orchestration activities in place and how they affected sellers over time. Second, gross sales in the Marketplace grew from USD 2.7 billion in 2000 (97% Amazon, 3% Sellers) to USD 590 billion in 2021 (36% Amazon, 64% Sellers), and counts with over 2 million sellers of varying sizes and longevity in the platform. Most of these businesses’ income depends on selling on the platform, implying that sellers are largely affected by changes in the Marketplace. This research focuses on understanding how Amazon influenced the appearance of new sellers’ business models. The field’s informants and the specialised industry media (such as Jungle Scout and Feedvisor) recognise six different types of business models. The models are described in Table 15, where following Chesbrough's (2006) 90 Essays on Power in Digital Platforms definition of business model7, I summarise their key value creation activities and their value capture strategies. I also present the year in which each business model appeared in the Marketplace and the estimated distribution of sellers for 2021 (Jungle Scout, 2021). There is no available information on the evolution of this distribution. The description of the business models is aimed at building ideal types; in practice, sellers might deviate from the description or use more than one business model at the same time. While all the business models are based on selling products through the Marketplace, substantial differences exist in how they create and capture value. One fundamental difference is that some business models are price competition-based (Arbitrage and Wholesale), while others compete based on product differentiation (Private Label, Handmade, and Aggregator). These different competitive strategies are also expressed in which key activities support each business model: pricing and sourcing for price-based competition and marketing, advertising, and quality assurance for differentiation-based competition. Another difference is the minimum scale required: while most business models require a small initial capital investment and can potentially be handled by a single person (Arbitrage, Private Label, Dropshipping, and Handmade), some require larger investments (Wholesale) and a dedicated and professional large team (Aggregator). Finally, there are significant differences regarding the moment when each business model first appeared in the Marketplace: Arbitrage and Wholesale appeared in the first half of the Marketplace’s existence, while in 2012, sellers using Private Label and Dropshipping models joined the platform, followed later by the Online Arbitrage (a variation of the original Arbitrage model), Handmade and Aggregator business models. “At its heart, a business model performs two important functions: value creation and value capture. First, it defines a series of activities that will yield a new product or service in such a way that there is net value created throughout the various activities. Second, it captures value from a portion of those activities for the firm developing the model” (p. 108) 7 91 Essays on Power in Digital Platforms Table 15. Types of sellers’ business models in the Amazon Marketplace Business Model Arbitrage / Online Arbitrage Description Purchasing bargains and discounted goods in retailers/online stores for reselling on Amazon. Wholesale Buying products in bulk from suppliers or manufacturers for reselling on Amazon. Private Label / Brands Creating a label/brand for a product niche that is sourced from a manufacturer. Dropshipping Handmade Aggregator Selling niche products that are directly fulfilled by a supplier or manufacturer Selling products designed and manufactured by the seller Acquiring other sellers’ businesses (typically Private Label) and operating them at a larger scale. Key value creation activities - Price comparison of well-known products. - Sourcing and listing management. Maintaining product consistency and restocking can be a challenge - Demand forecasting, sourcing, inventory management, and pricing. - Products chosen are often well-known (do not require branding or advertising) and with a steady demand (simplifies supply chain management) - Marketing, sourcing, and quality assurance - Branding and advertising - Market research and advertising - Simplifies the operations by avoiding dealing with inventory management, packaging, and shipping products. However, this implies a lack of control over product quality control and shipping times. Value capture strategy - Price competition based on opportunity purchases - Overall margins are usually low - Price competition based on reduced acquisition costs - Individual margins tend to be low; the key is setting large-volume and efficient operations - Differentiation strategy based on identifying product niches, leading to comparably higher margins. - Based on product niche identification and a minimum cost structure - Margins tend to be lower than private label but higher than wholesale and arbitrage - Product design and development - - Manufacturing / Handcrafting and quality assurance Differentiation strategy based on offering unique or customised products - - Marketing, branding, and advertising While individual margins can be comparably higher, volumes tend to be small and uncertain - Same as private label, focusing on the professionalisation of activities. - - Product and brand development Differentiation strategy based on highmargin private label products and efficiency gains derived from scaling up operations Appearance Share* (2021) 2000 / 2015 19% / 17% 2002 26% 2012 67% 2012 9% 2015 6% 2018 5% Source: Based on specialised industry reports (Jungle Scout, 2021). *Total exceeds 100% because some sellers use more than one business model simultaneously 92 Essays on Power in Digital Platforms 4.3.2. Data collection This research was conducted between 2019 and 2023; I collected data from multiple sources describing the Marketplace’s activities between 2000 and 2021 (see Table 16). Table 16. Detail of the data sources Data Source Cases Hours 37 23 Sellers 20 12 Amazon 2 2 Service providers* 9 5 Interviews Industry experts & media** 6 4 17 109 Events 9 93 Forums & social media 8 16 Documents Pages 635 5627 Financial statements 28 2163 Legal agreements 424 1734 Reports and studies Guides and educational materials 30 4 1123 95 Media articles 141 428 5 3 75 9 Participatory observation Archival Legislation Other documents Use in the analysis - Detailed understanding of the Marketplace’s characteristics and dynamics - Identification of Amazon’s orchestration activities and sellers’ responses - Understanding of sellers’ characteristics and general challenges - Triangulation and validation - Timeline of the Marketplace’s functionalities offering and key Amazon’s policy decisions - Understanding of the Marketplace’s operations - Identification of the Marketplace’s changes and stages - Triangulation and validation * Four of the service provider informants were formerly sellers ** Three of the industry experts & media informants were formerly sellers First (October 2019 to April 2020), I reviewed Amazon’s financial statements and industry reports to study the evolution of Amazon’s business and its performance. I also sought to understand the fundamentals of how the Marketplace operates, its fee structure, and the specific role of the Marketplace in Amazon’s overall business. Next (May 2020 to November 2020), through participatory observation in online events and forums, secondary survey data, and educational materials oriented to sellers, I started to identify sellers’ characteristics, and to create strategies for differentiating them. I shortlisted sellers with the initially desired characteristics to interview in the United Kingdom for convenience. I reached other informants from the United States, China, and Germany through snowballing (Parker et al., 2019). The first round of 23 interviews (December 2020 to November 2021) was conducted online (due to the COVID-19 pandemic), averaging 40 minutes each. I used a common questionnaire but kept an open-ended format. I asked how the sellers’ experiences in the Marketplace and their relationship with Amazon changed over time. In this research stage, I assembled an archive containing a yearly reconstruction of Amazon’s functionalities offering, fees, policies, and other relevant activities for each year. This information was gathered mainly using the Internet Archive, a resource that keeps records of websites over time. Based on this, media articles, and specialised reports, I built a database containing contracts such as the participation agreement, policies such as the 93 Essays on Power in Digital Platforms community rules, the sellers’ code of conduct and performance requirements, functionalities descriptions and availability, and fees charged by Amazon. The next phase of data collection (December 2021-March 2022) consisted of participatory observation through the researchers’ involvement in large, in-person sellers’ conventions and meetups. These events counted with several hundred participants consisting mainly of sellers with multiple backgrounds, product offering, size, and seniority in the marketplace, in addition to specialised media and ecosystem service providers. I participated by attending talks, learning about the services being offered to sellers, and engaging in conversations with participants. These events were crucial to triangulate information, validate preliminary findings, and recruit new informants. I then conducted a second round of interviews (March 2022 to May 2022) consisting of 14 cases and averaging 40 minutes in duration. In these interviews, I increased the variety of sellers I interviewed and talked to industry experts with and extended participation in the Marketplace and to Amazon representatives. I focused the questions on the specific periods and activities of interest identified from my previous data. Once I reached saturation in the themes discussed (Rheinhardt et al., 2018), I stopped recruiting participants. Overall, the interviewees’ composition was highly heterogeneous: I counted with informants with direct experience in all the business models mentioned above except Handmade, located in the UK, the USA, and China, and belonging to organisations with very different sizes, varying from unipersonal businesses to large companies with hundreds of employees. Regarding the informants’ experience in the Marketplace, 17% joined before 2004, 17% between 2005-2009, 14% between 2010-2014, and 51% between 2015-20208. 4.3.3. Data analysis I analysed the data following an inductive grounded approach (Charmaz & Bryant, 2008; Gioia et al., 2013; Rheinhardt et al., 2018; Thornberg & Charmaz, 2014). I first analysed the archival resources to understand the Marketplace’s essential characteristics and immersing in the field. I then proceeded to transcribe and analyse the interviews. I open-coded them (Corbin & Strauss, 2008) to identify Amazon’s orchestration activities and processes of sellers’ change and their possible causes. The core of the analysis activities consisted then of triangulating and consolidating the data from the multiple sources in a timeline. This step revolved fundamentally in linking sellers’ characteristics and activities to relevant changes in the Marketplace. Once I identified and described the process of the appearance of new sellers’ business models and the key orchestration activities used by Amazon, I identified the periods where the interaction between them was more clearly expressed. I gathered additional data about these periods and described them. Finally, I built the model to summarise the interaction mechanisms. 8 A detailed description of the interviewees’ characteristics is kept confidential to prevent potential harms. 94 Essays on Power in Digital Platforms 4.4. Findings I organise my findings in three parts. First, I describe the process through which new sellers’ business models appear in the Marketplace. Second, I discuss how Amazon influenced this process by identifying and describing the specific orchestration activities used by Amazon and by illustrating how Amazon influenced the appearance of sellers’ business models in two periods. Finally, I discuss the consequences of how Amazon’s orchestration activities and the appearance of sellers’ business models played out throughout the period of the study and summarise it in a model. 4.4.1. Why do sellers adopt new business models? Throughout the period of this study, I observed a remarkable transformation of the business models that sellers adopt in the Marketplace. I describe this process next (see Figure 6): I find that the appearance of new business models requires two conditions. In the first place, feasibility, i.e., the availability of the functionalities needed to support the business model’s key activities. For example, Fulfilment by Amazon made the Private Label and Handmade models possible, as they typically are one-person businesses created by entrepreneurs or creators who do not count with fulfilment capabilities. At the same time, the availability of the required functionalities is not enough for the appearance of a business model in the Marketplace. The Aggregator model appeared in 2018, but the advertising and brand development functionalities essential for its activities were released years before, from 2012 on. The second is the existence of changes in the competition intensity faced by sellers in the Marketplace that turn existing profitable business models into struggling ones. Increases in competition intensity can be driven endogenously by the arrival of new sellers, attracted by the existing business models’ profitability, and exogenously by Amazon’s actions, as described in the next section. When facing high competition intensity, I observe three typical responses from sellers: the first option is to exit the Marketplace. The second is improving performance to cope with competition, for example, by improving customer service, increasing supply chain service levels, or developing new products. These improvements typically imply using additional optional functionalities offered by Amazon. The third is adopting a new business model. I observe that adoption occurs when both the business model is feasible and pre-existent models' value capture strategies and profitability prove insufficient. My evidence shows that initially, the development and renewal of the business models are done by actors who know the platform well and are capable of identifying and seizing the opportunity. However, once these new strategies are known, the popularisation of the new business model is driven by new entrants. The interviewees describe how these transitions are motivated to overcome the limitations of pre-existing business models: “When I was reselling branded products, I’d say that had a few years where that was a good business to be involved in but then obviously, a lot of people started doing it… And then you’ve got the kind of private label phase” (Ecosystem service provider and former seller) “There’s been a natural kind of selection process of smaller micro brands that have survived and have put themselves in positions in Amazon that is competitive, sort of say. But that doesn’t escapes the fact that 95 Essays on Power in Digital Platforms in order to run an Amazon brand, you need a number of different things that are difficult to muster up in hard times.” (Seller) “Entrepreneurs are creators of some of these businesses, they are typically reaching the capacity of their time, their knowledge, or their capital.” (Seller) Figure 6. How new sellers’ business models appear on the Marketplace Platform conditions Functionalities availability Existing Business Model Competitive Fit Optional functionalities Minimum for feasibility Enables Improves performance Higher usage Struggle Profitability Minimum for feasibility Develops new model Adoption Higher competition intensity Attracts Exit New sellers Causality New Business Model Sellers’ responses Functionalities availability Attracts Business model’s competitive fit 4.4.2. Amazon’s role: orchestrating change 4.4.2.1. Orchestration activities I find that Amazon had a decisive impact on the transformation of the sellers’ business models through the combination of three sets of orchestration activities: 1) functionality provisioning, 2) structuring seller competition, and 3) increasing competition intensity faced by sellers. I describe these activities next and provide detailed examples and summarise them in Table 17. Functionality provisioning It consists of providing platform functionalities that enable new configurations of sellers’ activities for value creation. As described above, each seller business model is characterised by a group of key value-creation activities. Sellers perform some of these activities in the platform using the functionalities provided by Amazon. I denote as compulsory functionalities those that Amazon requires to be used by anyone selling on the platform, such as the payment, basic listing, and order management systems, and the currency converter (for international operations). Most functionalities available in the Marketplace are nevertheless stated as optional by Amazon, and thus sellers decide whether to use them or not. Optional functionalities range from marketing and data analytics tools to fulfilment, international logistics, or tax calculation services, among others. The interviewees describe that 96 Essays on Power in Digital Platforms optional functionalities can contribute to developing and improving sellers’ activities, as they equip them with professional tools tightly integrated with the compulsory functionalities. “[T]his increased complexity, it's also something which it's beneficial or upgrading your business.” (Seller) “…it was all incremental. We were like "this is fantastic" and I'd actually say Amazon improved us in the early years because we weren't sophisticated.” (Seller) The provision of optional functionalities gives sellers the flexibility to decide how to perform their value-creation activities and to create new ways to configure them. While all the sellers’ business models are based on the compulsory functionalities, these only cover a small part of all the activities. Sellers can thus decide which part of the remaining activities to perform through their own capabilities and processes and which through Amazon’s. By expanding the offering of optional functionalities, Amazon increases the possible configurations of activities and thus augments the potential for the appearance of new seller business models. This possibility to customise the sellers’ activities is recognised by the interviewees, who describe Amazon’s role in augmenting business model generativity: “I think what [Amazon] realised very quickly is if you just give people the building blocks to build their own brand, what you end up with is this hyper competitive marketplace, where people are able to challenge the status quo of existing brands and actually build their own brands.” (Seller) In addition to providing sellers with functionalities that support their key value creation activities, I find that Amazon’s activities also affect sellers’ value capture strategies by structuring how sellers compete and managing the competition intensity they face. Structure seller competition First, as the platform owner, Amazon plays a central role in structuring competition in the Marketplace. I observe that this is done through different mechanisms: firstly, by introducing functionalities that compare and sort sellers’ competitive positions. The Buy Box is an example of this, as it ranks sellers who offer the same product and assigns the highest-ranking seller as the default provider when a customer purchases. By changing the ranking criteria, Amazon changes inter-seller competition. Second, by changing criteria for how products are ordered and displayed. For example, while initially, Amazon served search results based on searched terms and product rankings, from 2017 on, search results began to show a larger proportion of advertised products. This changes sellers’ key parameters for competition from price, product quality, and delivery performance to marketing and advertising optimisation. Finally, Amazon also structures competition through rules constraining sellers’ competitive decisions, such as the Fair Pricing policy (so-called “most favoured nation” – MFN – rules), requiring sellers not to price their products on Amazon higher than they price them anywhere else. Increase competition intensity faced by sellers Next, I observe that some of Amazon’s activities can increase the competition intensity faced by sellers. On the one hand, Amazon can increase the threat of new entrants by lowering entry barriers to new sellers (for example, by offering fulfilment or brand design functionalities) and by changing rules that regulate access to the marketplace (notably expanding the list of countries admitted for a seller’s location). Second, I find that Amazon increases competitive rivalry by increasing performance requirements for sellers’ eligibility for distinctive platform 97 Essays on Power in Digital Platforms features such as the Prime badge or the Featured Merchant status (required to rank for the Buy Box) and by directly entering a sellers’ product category with Amazon’s own branded products. Finally, Amazon can increase competitive intensity by leveraging its bargaining power as a provider for sellers, for example, by increasing fees for services that become essential for competing in the Marketplace. Table 17. Amazon’s orchestration activities: Examples, practices, and outcomes Orchestration activities: Examples (1st order concepts) - - Forcing use of Amazon Pay to process payments, or Amazon Currency Converter for international operations Introducing functionalities that expand the range of possible seller activities, such as with marketing (Amazon Accelerator) or export (FBA Export) - Launching the Buy Box, Match Low Price, or Automatic Pricing tools. - Prioritizing advertised products in search results - Requiring sellers not to price their products on Amazon higher than they price them anywhere else (so-called “most favoured nation” – MFN – rules) Orchestration activities: Activities (2nd order themes) - Defining the compulsion and optionality of the platform’s functionalities - Introducing optional functionalities associated with sellers’ activities - Introducing functionalities that compare and sort sellers’ competitive position - Changing criteria for how products are ordered and displayed - Defining rules that constrain sellers’ competitive decisions - Provision of services such as Fulfilment by Amazon or Global Selling - - Accepting new product segments and sellers from countries previously excluded from the Marketplace, notably China in 2015 Lowering entry barriers to new sellers - Changing rules that regulate access to the marketplace Tightening the requirements to be eligible for distinctive platform features, like the Prime and Featured Merchant badges. - Increasing performance requirements for sellers - Entering a seller’s product category - Increasing fees for optional services that become essential for competing - - Launching of Amazon Basics and other Amazon-owned brands - Increasing the fees for Fulfilment by Amazon or for advertising tools like Sponsored Products and Sponsored Brands Orchestration activities: Outcomes (Aggregate dimensions) Functionality provisioning Structure seller competition …increasing the threat of new entrants Increase competition intensity by… …increasing competitive rivalry …leveraging the platform’s bargaining power Source: Based on interviews, participatory observation, and multiple archival sources In Figure 7, I show how Amazon’s orchestration activities influence the process of appearance of new sellers’ business models described in the previous section. First, Amazon makes different business models feasible by providing functionalities that sellers use to support their business models’ key activities. By providing new additional functionalities, Amazon makes new business models, or variations of existing ones, feasible. Second, I observe that changes in the business models in the marketplace occur when the value capture strategies and profitability of pre-existent models prove insufficient. Amazon affects sellers’ profitability by changing the structure of competition and by increasing competition intensity. 98 Essays on Power in Digital Platforms Figure 7. Amazon’s orchestrating role in the appearance of new business models Platform conditions Existing Business Model New Business Model Functionality provisioning Functionalities availability Enables Competitive Fit Optional functionalities Minimum for feasibility Improves performance Struggle Profitability Attracts New sellers Higher usage Minimum for feasibility Develops new model Adoption Higher competition intensity Structure competition Exit Attracts Increase threat of new entrants Increase rivalry Leverage bargaining power Causality Sellers’ responses Functionalities availability Business model’s competitive fit Orchestration activities 4.4.2.2. Orchestration over time: the appearance of the Private Label and Aggregator business models In Table 18, I summarise my observations on Amazon’s orchestration activities over time, detailing the evolution of functionalities availability and the key changes in seller competition. I highlight three trends from this table: First, while the provision of compulsory functionalities has not changed significantly over time, Amazon continuously expanded the offering of optional functionalities, both in the number of functionalities for specific activities (such as delivery or advertising) and in the range of activities involved (see detail in Table 19). As discussed above, this in turn increases the number of feasible seller business models. Second, I observe that Amazon changed the focus of the new functionalities’ area of application from supply chain management (SCM) to marketing and advertising (MA): Between 2006 (when Amazon launched Fulfilment by Amazon) and 2014, Amazon launched 18 SCM and 9 MA new functionalities (respectively 2 and 1 per year on average). Between 2015 and 2021, the focus shifted, launching 6 SCM and 25 MA new functionalities (0.9 and 3.6 per year on average). This shift in focus indicates what type of sellers’ business models Amazon aims to develop in the Marketplace in different periods. For example, functionalities supporting brand development and advertising activities are not relevant for the Arbitrage or Wholesale business models but are crucial for Private Label, Handmade, and Aggregator. In contrast, pricing functionalities such as Match Low Price and Automated Pricing are fitted for Wholesale and Online Arbitrage business models. At the same time, optional functionalities launched in one moment can be generative of the appearance of a business model later. Notably, this is the case of Fulfilment by Amazon, launched in 2006 in a context where sellers used Arbitrage and Wholesale business models, but that years later became the watershed for the emergence of Private Label and Handmade. Third, I recognise a pattern in how the most impactful Amazon’s orchestration activities are sequenced, starting with the provisioning of key functionalities (such as Fulfilment by Amazon) and reducing entry barriers to new 99 Essays on Power in Digital Platforms sellers (like introducing Global Selling and accepting sellers using China’s IP), followed by actions that structure seller competition (such as introducing the Buy Box or prioritising advertised products in search results), and finally increasing competitive rivalry (like increasing performance requirements) and leveraging Amazon’s bargaining power (increasing fees and advertising costs). I identify two periods where the impact of Amazon’s orchestration activities on the appearance of new sellers’ business models is clearly illustrated: The first one is 2012-2013, when the Arbitrage and Wholesale business models faced challenges coping with the increasing competition intensity (Figure 8). After the release of Fulfilment by Amazon in 2006, the platform experienced the fastest-growing period of sellers’ sales in the Marketplace, which ten-folded between 2006 and 2012. In parallel, Amazon launched the Buy Box in 2008, which became a mechanism that structured competition between sellers who offered the exact same products and thus competed only by price. In addition to higher competition rivalry due to additional sellers joining the platform, three Amazon actions in 2012 contributed to increase the competitive intensity: the increase in the requirements to qualify for competing for the Buy Box and the launching of Global Selling, a program that enabled sellers located in other countries to operate in Amazon’s largest marketplaces. High competition intensity forced existing sellers to streamline their operations or exit. At the same time, some of the existing sellers and new entrants developed new business models based on product differentiation (Private Label and Dropshipping) in response to the high competition intensity created by the Buy Box. These business models originated the so-called “Amazon-native brands”, created in the Marketplace exclusively to differentiate an otherwise generic product. Figure 8. The appearance of the Private Label business model in the Marketplace Platform conditions Arbitrage & Wholesale Private Label Functionality provisioning Functionalities availability - Payments - Product page creation Enables Competitive Fit Improves performance New sellers Higher usage Struggle Profitability Attracts - Brand registry - Sponsored products - Match Low Price Develops new model Adoption Higher competition intensity Introduction of Buy Box (2008) Exit Attracts Introduction of FBA & Prime (2006) Introduction of Global Selling (2012) Increasing requirements to win Buy Box (2012) Causality Sellers’ responses Functionalities availability Business model’s competitive fit Orchestration activities The second period is 2017-2018, when the large number of Amazon-native brands started to create a challenge of product discoverability. Previous to this, Amazon allowed in 2015 sellers located in China to use the Global Selling program to sell worldwide – contributing to the increase in Amazon-native brands - and in 2017, started to favour 100 Essays on Power in Digital Platforms the position of advertised products in search results. High competition intensity and the change in competition structure forced existing Private Label and Dropshipping sellers to renew their differentiation strategy. This posed challenges both in terms of capabilities (such as branding and advertising campaign design and management) and capital requirements (to invest in advertising and brand development). In this context, in 2018 emerged the Aggregator business model, based on acquiring existing sellers’ accounts and operating them on a larger scale through capital injection and the professionalisation of activities. Figure 9. The appearance of the Aggregator business model in the Marketplace Platform conditions Private Label & Dropshipping Aggregator Functionality provisioning Functionalities availability Enables Competitive Fit - Seller fulfilled prime - Lighting Deals - Brand registry - Sponsored products Improves performance Struggle Profitability Attracts New sellers Higher usage - Sponsored products - Campaigns Develops new model Adoption Higher competition intensity Ads prioritized in search (2016) Exit Attracts Chinese Sellers’ IP accepted (2015) Increasing advertising costs (2017) Higher capital requirements to compete (2012) Causality Sellers’ responses Functionalities availability Business model’s competitive fit Orchestration activities 101 Essays on Power in Digital Platforms Table 18. Summary of Amazon’s orchestration activities over time: evolution of functionalities availability and key changes in seller competition Amazon's orchestration activities Type Sellers' activities involved Sales services Functionality provisioning Supply Chain Compulsory Management Marketing & Advertising Sales services Optional 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Supply Chain Management Marketing & Advertising Number of available functionalities 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 1 1 1 1 1 1 2 4 4 4 4 4 4 4 4 4 4 0 0 0 1 1 1 5 8 9 9 10 10 13 15 18 19 19 21 22 22 23 23 0 0 0 0 0 0 0 1 2 3 3 3 8 8 9 12 14 20 24 28 32 34 Key changes in seller competition Structure competition Increase threat of new entrants Increase competition Increase competitive rivalry intensity Leverage bargaining power │Launch of Buy Box │Ads prioritized in search results │Launch of Fulfilment by Amazon │Launch of Global Selling and Amazon Seller Central │Chinese Sellers' IP accepted │Amazon Basics │Increasing requirements to win Buy Box │Increasing FBA fees │Increasing advertising costs Sources: The number of available functionalities is based on a yearly reconstruction of Amazon’s offering based on the Internet Archive, including the use of contracts such as the participation agreement, policies such as the community rules, code of conduct, and performance requirements, and functionalities descriptions. The changes in seller competition are based on interviews, specialised industry reports and media articles, and by a reconstruction of the fees evolution based on the Internet Archive. 102 Essays on Power in Digital Platforms Table 19. Detail of functionalities available for sellers in the Marketplace over time Functionality provisioning Type Sellers' activities involved Detailed area of application Number of available functionalities Payments 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Currency converter 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 Supply Chain Management Order management 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Marketing & Advertising Listing management 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Loyalty programs 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Tax calculation 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 Lending 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 Sales development 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 Delivery 0 0 0 1 1 1 2 3 3 3 3 3 3 3 4 5 5 5 6 6 7 7 Pick, Pack, Handle 0 0 0 0 0 0 1 1 1 1 1 1 2 3 4 4 4 4 4 4 4 4 Returns 0 0 0 0 0 0 1 1 1 1 1 1 1 1 2 2 2 3 3 3 3 3 Storage 0 0 0 0 0 0 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 Order management (optional) 0 0 0 0 0 0 0 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Inventory optimization 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Export 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 Disposition 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 Inbounding 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 Warehouse distribution 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 Warehouse optimization 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 Advertising 0 0 0 0 0 0 0 1 2 3 3 3 4 4 4 5 5 6 8 9 10 11 Brand development 0 0 0 0 0 0 0 0 0 0 0 0 1 1 2 3 5 5 7 7 9 10 Bundling 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 Market reach 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 Pricing 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 2 2 3 3 3 Promotions 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 4 4 5 6 6 Listing management (optional) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 Market research 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 Sales services Compulsory 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Sales services Supply Chain Management Optional Marketing & Advertising 103 Essays on Power in Digital Platforms 4.4.3. Evolutionary interplay between Amazon and the Sellers Having described the specific mechanisms by which Amazon influenced in the appearance of new sellers’ business models, I now turn to how the interactions between these actors played out throughout the period of the study. I recognise a pattern of dynamic adjustments between Amazon’s and the sellers’ business models, resulting in an evolutionary interplay where both are affected. I summarise these dynamics in the model in Figure 10: I observe that changes in the sellers’ business models and in Amazon’s value proposition, value capture, and orchestration activities mutually respond and adapt to each other, in a cyclical process. Figure 10. Interplay between the sellers’ business models evolution and Amazon’s value proposition, orchestration, and value capture strategies Revenue from transactions Increases Amazon Structure competition Increase competition intensity Customer service Profitability Sellers’ Business Models Struggle Forces performance improvement Adoption Expands Product selection Enables Increases Functionality provisioning Increases New revenue streams Evolution of sellers’ business models Sellers’ impact on Amazon Amazon’s value proposition Amazon’s value capture Amazon’s orchestration On the one hand, Amazon’s orchestration activities contribute to pushing the sellers’ business models through its stages of profitability, struggle, and the adoption of new models. While the appearance of a new business model does not replace the pre-existing ones, I observe that Amazon’s activities lead to the appearance of increasingly sophisticated models. In each case, they can sustain more value-creation activities by combining more functionalities and withstand higher competition intensity. The comparison between the first and last business model to appear in the Marketplace is illustrative of this: while Arbitrage is simply purchasing a bargain to sell it locally at a slightly higher price, Aggregators count with dedicated product development and marketing teams, operate in multiple countries, and are typically backed by large financial funds. 104 Essays on Power in Digital Platforms At the same time, the sellers’ business model transformation also impacts Amazon by improving the platform’s value proposition to customers and creating new monetisation opportunities that increase Amazon’s revenues and value capture. First, the increasing importance of sellers for delivering Amazon’s value proposition is expressed in how Amazon has changed its participation in the Marketplace: while initially, Amazon acted as the main provider and sellers only complemented the product selection, as of 2021, sellers provide two every three products sold, the other third being provided by Amazon (Figure 11). Further, my informants acknowledge that sellers cope with the Marketplace’s increasing competition intensity by improving their performance, customer service, product selection, and quality, which is at the core of Amazon’s value proposition of “unlimited selection and fast delivery”. “Everything is focused on the end buyer. And it's done by every company, improving the product listings, etc. It's a way for Amazon to target other people invest in getting the work done.” (Ecosystem service provider) “The standards are going up all the time. And Amazon's attitude is very simple: this is what consumers expect.” (Industry expert) Second, the changes experienced by the sellers contribute to increase Amazon’s revenue and value capture. On the one hand, as thriving sellers deliver Amazon’s value proposition, they attract additional customers, increasing the volume of transactions and consequently Amazon’s revenue (Figure 11). On the other hand, the increasing variety and sophistication of the sellers’ business model is based on a higher usage of optional functionalities, for which sellers pay a fee. I observe that Amazon significantly increased the share of the sellers’ revenue it captures, reaching 36% of it in 2021: from 2008 on, the fees for compulsory functionalities (referral and other basic fees) remained relatively stable, the increase in value capture is explained by the use of optional services by sellers. Figure 11. Changes in the sellers’ importance in the Marketplace and of Amazon’s sources for value capture 70% 64% 60% 50% 38% 40% 36% 9% 30% 19% 20% 2% 12% 6% 10% 6% 12% 0% 15% 5% Amazon's capture of sellers' revenue - Advertising services Amazon's capture of sellers' revenue - Fulfilment and other services Amazon's capture of sellers' revenue - Referal and other basic fees Sellers' share of total sales in the Marketplace 105 Essays on Power in Digital Platforms Source: Based on financial statements and specialised industry reports 4.5. Discussion and conclusions In this article, I have shown how Amazon’s orchestration activities influenced the appearance of new sellers’ business models and how that change affected Amazon. First, I describe the process by which new seller business models appear in the Marketplace. Next, I identify the specific orchestration activities by which Amazon influenced the change in the sellers’ business model – functionality provisioning, structuring seller competition, and increasing competition intensity – and explain how they operate. Finally, I present a process model summarising the interactions between Amazon’s and the sellers’ value creation and capture activities. I observe that Amazon’s orchestration activities lead to the sellers' adoption of increasingly sophisticated business models, which in turn contributes to improving Amazon’s value proposition and capture from the Marketplace. Overall, I identify a pattern of dynamic adjustments between Amazon’s and the sellers’ business models, resulting in an evolutionary interplay where both Amazon and the sellers are transformed. This study makes two main contributions to platform research: first, by showing how a PO can orchestrate its complementors’ transformation, I contribute to expanding our understanding of the PO’s set of capabilities and responsibilities. Second, I contribute to the stream of literature focused on complementor orchestration by explaining the dynamic interplay between the PO’s orchestration activities, the complementors’ transformation, and the consequences for value creation and value capture of this transformation. My study also holds implications for practitioners and for regulatory purposes. 4.5.1. Contributions to platform research My first contribution with this study is to expand current views on what POs are capable of and responsible for by showing that they can in fact orchestrate changes in their complementors’ business models. PO have a unique capability to mobilise and align resources that are not formally under their control (Adner & Kapoor, 2010; Gawer, 2014). So far, extant research has recognised the profound economic impact of platform orchestration in how it facilitates value creation by both enabling interactions and fostering innovation (Cusumano et al., 2019; Parker et al., 2016). In addition to creating value, I argue that the PO’s orchestration activities can have another type of economic impact, producing a transformative effect on its participants’ characteristics. Furthermore, I argue that this transformation is not simply a by-product of the activities performed in the platform but a planned output of the PO’s orchestration. Consequently, this research illuminates PO’s capabilities linked to the long-term planning of the ecosystem’s design. I know of no studies addressing this important effect of platform orchestration in complementors’ characteristics and, consequently, in the platform ecosystem itself. Recognising these PO’s capabilities is also relevant for research on platform power (Cutolo & Kenney, 2021; Jacobides, 2021), as it identifies ways in which PO can create forms of complementor dependence: if a firm relies on a platform to exert or develop at least one of its core competences (Prahalad & Hamel, 1990), it then becomes highly vulnerable to changes on features or rules stipulated by the PO. This form of PO power to shape its complementors’ characteristics can also have competition effects in the form of a reduced variety in organisational forms and organisational innovation by firms, ultimately harming consumers. 106 Essays on Power in Digital Platforms Second, I contribute to the stream of research exploring complementor orchestration in several ways. Firstly, I have shown that Amazon’s activities increased its complementors’ heterogeneity, as I find evidence of the successive emergence of new seller business models. This contributes to previous research on complementor heterogeneity (McIntyre et al., 2021; Rietveld et al., 2021) by showing and explaining how orchestration can, in fact, drive heterogeneity. Next, I illuminate new complexities and trade-offs in platform orchestration by recognising the dynamics through which PO can orchestrate the change in the complementors’ business models. As PO might be using orchestration activities to aim for specific compositions of the complementors’ pool, this creates unequally distributed challenges to the different types of complementors, favouring some of them while penalising others. Third, previous research highlighted that PO typically faces a trade-off between platform growth and value capture (Cennamo & Santalo, 2013; Lan et al., 2019). The set of options through which this trade-off can be handled is increased if we include complementors’ characteristics within the scope of the orchestration activities, as I have shown for the Amazon Marketplace. Finally, my results also make visible other important tradeoffs, such as deciding the degree of the platform ecosystem’s technological complexity. I have shown how increasing the offering of optional functionalities was fundamental for driving an increase in the variety of sellers, an orchestration activity I call “functionality provisioning”. By increasing the number of functionalities and the possible interdependencies, the ecosystem’s technological complexity is likely to increase as well. This can create a trade-off between complementor generativity and platform growth, as extant research has identified a decelerating effect of technological complexity on the ecosystems’ value creation (Cennamo, 2016; Cennamo et al., 2018). 4.5.2. Limitations and avenues for future research This study has limitations, some of them derived from using a single empirical case, which opens opportunities for future research. One of the limitations of this study is whether the results are generalisable for different types of platform ecosystems. In particular, it is foreseeable that in different ecosystem configurations and under different bargaining power relations among the actors, the PO might not be able to structure complementors’ competition and manage the competition intensity they face. Further, in my empirical setting, the interdependencies between the PO and the complementors are built over key complementors activities. It is yet to be explored if the PO will successfully orchestrate the complementors’ business models in ecosystems where these interdependencies are built in non-essential or secondary complementor activities. In addition, while this paper focused on the competitive dynamics within the platform, some of the mechanisms I described are likely to be affected by external factors, such as changes in tastes and maturity of the consumer or inter-platform competition. Future research efforts can address some of these limitations and further explore other problems illuminated by this study. For example, while I have shown how a PO can transform its complementors’ business models, other researchers can explore whether orchestrating other complementor characteristics - like size and organisational form - would require different orchestration activities. Other lines of research can study the competitive outcomes of the PO’s orchestration of complementors’ transformation and the consequences this might have for public agencies and regulatory efforts. 107 Essays on Power in Digital Platforms 5. Discussion and general conclusions This thesis seeks to explain the characteristics and dynamics of power within digital platforms, particularly as they emerge as new organisational forms. I approached this problem by structuring the research around three distinct dimensions. The initial dimension pertains to the structure of social interactions, endeavouring to discern how platforms shape power relations and autonomy by influencing agents’ interactions and modulating participatory practices. The subsequent dimension is centred on the platform's governance, with the aim to contrast the principles guiding the governance and how it is enacted, and how this shapes power dynamics within the platform. Lastly, the third dimension turns its focus to the interplay between firms, with the intent to assess the extent of influence wielded by the platform's owner in directing shifts in the value creation and capture strategies of participant firms. The studies in chapters 2, 3 and 4, in addition to making specific contributions already discussed in each chapter, contribute overall to address this thesis’ objectives, advance our knowledge on how we understand platforms, and provide new ways to characterise platform power, its dynamics, and its consequences. In this chapter, I discuss these contributions (summarised in Table 20), as well as this study’s implications for policy, its limitations, and the avenues it opens for future research. I conclude the thesis with some final remarks about the potential practical implications of this research for platforms. 108 Essays on Power in Digital Platforms Table 20. Contributions of this thesis to research on platforms and platform power Area of contribution Conceptualisation of platforms Characterisation of platform power Platform power dynamics Consequences of platform power Contribution Main consequences Examples of literature related to the contribution - Identifying and explaining the process by which platforms in their organising role transform fields of practice and replace pre-existing practices with new ones Recognising how agentic dynamics in platforms involving individuals and organisations are shaped and how platforms work. Adner (2017); Gawer (2014); Kretschmer et al. (2022); Stark & Pais, (2020) - Explaining the processes by which the platform owner produces, encloses, and coordinates agents’ practices. This illuminates new aspects of how hierarchies and relations of ownership and value capture are formed in platforms - Explaining the relationship between structure and agency in platforms and showing how they are mutually transformed through the differences in the platform constituents’ practices and positions of power. I argue that platforms can be more usefully described as spaces of coordination and production of practices that facilitate controlling outside activities and resources and profiting from them. Langley & Leyshon (2017); Srnicek (2016); Zuboff (2019) View of platform power that is multifaceted (i.e., multiple forms of powers coexist) and multidirectional (i.e., interaction of multiple agents with different motivations and strategies). Integrates social, cultural, and symbolic dimensions of platform power Culpepper & Thelen (2020); Jacobides (2021); Khan (2018) Explains power not as an instrument within an asymmetric structure but as a web of interactions resulting from the situated agents’ practices and strategies. Andreoni & Roberts (2022); Cutolo & Kenney (2021); Srnicek (2016); Zuboff (2019) - Illuminating the role of uncertainty and ambiguity as strategies of platform power Expand explanations of power strategies from abusing the ability to set the rules to strategically mismanaging the rules. Busch et al. (2021); Jacobides (2021); Lianos & Carballa-Smichowski (2022) - Illustrating platform organising dynamics that achieve system-level goals not only directly by influencing on behaviours leading to specific outcomes but also by the means of influencing change in the participants' characteristics and its pool’s composition. - Identifies and explains the coexistence of dynamics of platform power and counterpower accumulation Expand current views on what platforms can achieve as organisational technologies… Daymond et al. (2022); Gawer (2014, 2021) what distinguishes them from other organisational forms… Jacobides et al. (2018, 2022) and how they build hierarchical structures within industries Rikap & Lundvall (2021); Srnicek (2016) A conceptualisation of platform power that is dynamic and grounded on the agents’ practices, positions, and strategies. Busch et al. (2021); Khan (2018) - Explaining how power changes vis a vis the platform evolution by power accumulation and power contestation Enriches previous findings on how power dynamics change as the platform changes Cutolo & Kenney (2021); Zuboff (2019) - Providing evidence that challenges central claims in the literature pertaining to mechanisms assumed to keep some of the existing platform power asymmetries under check - Explain the platform’s organising role in producing and reproducing the delegation of autonomous decisionmaking on individuals Show that the balance of incentives and control can coexist with a sustained mismatch between the governance the platform declares to abide by and its actual practices Constantinides et al. (2018); Kretschmer et al. (2022); Saadatmand et al. (2019); Teece (2018) Identify the conditions under which agency is shaped in platforms by their specific structure and its power dynamics Calo & Rosenblat (2017); Lanier (2018); Scherer & Neesham (2020); Stark & Pais (2020) - Identify and explain forms of platform dependence constraining business autonomy and decision-making Show how dependence is rooted not only in the platform’s owner gatekeeping position but also in its decisions about functionalities provision related to complementors’ key value creation activities Curchod et al. (2019); Cutolo & Kenney (2021); Eaton et al. (2015); Khan (2018) 109 Essays on Power in Digital Platforms 5.1. On the nature of platforms as organising agents By exploring power dynamics, this research illuminates new dimensions of platforms and contributes to enriching how we conceptualise them. This thesis first deepens our understanding of platforms as organising agents by highlighting their importance as spaces of coordination and production of practices. The organisational view on platforms has stated the relevance of bringing agency to the forefront of platform studies (Gawer, 2014). Further, this view characterises platforms based on how they organise the participation of autonomous agents based on balancing incentives and rules within the context of a digital infrastructure (Kretschmer et al., 2022). I contribute to this stream by identifying and explaining the process by which platforms in their organising role, transform fields of practice and replace preexisting practices with new ones (see sections 2.3.1.1 & 2.3.4). Paying attention to the organising of practices contributes to refining our understanding of the conditions by which agency is shaped and how platforms work. A focus on practices also supports studying agentic dynamics in platforms involving both individuals and organisations. Describing platforms as spaces of production of practices also illuminates new aspects of how hierarchies and relations of ownership and value capture are formed in these organisational contexts, contributing to political economy approaches (Langley & Leyshon, 2017; Srnicek, 2016; Zuboff, 2019). In section 2.3.1.1, in addition to describing how a platform transforms existing fields by introducing new ways of practice, I argued that this process also involves the platform owner’s attempt to enclose those practices: by encoding the activities to be performed and controlling the technological infrastructure in which they are performed, the platform owner creates control rights (i.e., ownership, Boyle, 2003) over the practices. In other words, since actions performed in platforms can be considered runtime products - i.e., produced in real-time while performed (Yoo, 2023) – ownership of the product design and the infrastructure confers control over the actions. The process of enclosure is analogous to that of “incursion” described by Zuboff (2019), with the difference that in her approach, the focus of the control is data rather than the agents’ practices. My perspective consequently expands on studies focusing on the role of data control and infrastructure ownership to create new ways to capture value (Srnicek, 2016, 2021; Zuboff, 2019). Instead, I argue that platforms can be more usefully described as spaces of coordination and production of practices that facilitate controlling external activities and resources and profiting from them. 5.2. On the characterisation of platform power This thesis contributes to enriching previous characterisations of platform power in multiple ways. First, this study builds a more complete and integrative understanding of platform power by exploring the relationship between structure and agency in platforms. I addressed this problem both when discussing the conceptual relationship between the digital habitus and the logic of the field (sections 2.3.1.3, 2.3.4, and 2.4.2) and when exploring empirically the Amazon Marketplace sellers’ interactions with the platform’s governance (sections 3.4.2 and 3.4.3) and with Amazon’s orchestration activities (sections 4.4.2 and 4.4.3). In these discussions, I show 110 Essays on Power in Digital Platforms how both structural and agency conditions are mutually transformed through the differences in the platform constituents’ practices and positions of power. Overall, I provide a view of platform power that is multifaceted (i.e., in which multiple forms of powers can coexist) and multidirectional (i.e., as it results from the interaction of multiple agents with different motivations and strategies). On the one hand, this view contributes to integrating social, cultural, and symbolic dimensions of platform power, enriching previous conceptualisations (Culpepper & Thelen, 2020; Kenney et al., 2021; Khan, 2018). On the other hand, it also explains power not as an instrument within an asymmetric structure (Andreoni & Roberts, 2022; Cutolo & Kenney, 2021; Srnicek, 2016; Zuboff, 2019) but as a web of interactions resulting from the situated agents’ practices and strategies. Second, to my knowledge, this research is the first to illuminate the role of uncertainty and ambiguity as strategies of platform power. Typically, platform power has been associated to the owner inclining the playing field, abusing its control over the technological infrastructure (Calo & Rosenblat, 2017; Khan, 2016; Srnicek, 2016). When I identify and describe the “two-faced governance” in the Amazon Marketplace (sections 3.4.1.2, 3.4.3, and 3.5.3), I highlight the crucial role of uncertainty, ambiguity, and unreliability as factors explaining Amazon’s power over sellers. This shifts power strategies from abusing the ability to set the rules to strategically mismanaging the rules. Extant research has explored the circumstances that allow platform owners to increase their power based on the lack of transparency, focusing on how algorithms create “black boxes” that conceal processes behind actions and defer responsibilities (Introna, 2016; Pasquale, 2015). Digital technologies can thus hide rationalities, biases, and decisions and exert a structuring effect over the interactions (Alaimo & Kallinikos, 2017; Beer, 2017). These studies provide useful theories that explain how the platform’s technological infrastructure enables abusive practices. This research contributes to these views by describing how the platform owner, based on this lack of transparency, increases its power by actively creating contradictory conditions for participation. Thirdly, this study expands current visions on the scope and reach of platform power, illuminating capabilities linked to the long-range planning of the ecosystem’s design. By showing how Amazon exerts an influence to change the sellers’ business models (section 4.4.2), I illustrate platform organising dynamics that achieve systemlevel goals not only directly by influencing on behaviours leading to specific outcomes (as was already thoroughly identified in previous literature, e.g., Cennamo & Santalo, 2013; Rietveld et al., 2019, 2021), but also by the means of influencing change in the participants' characteristics and its pool’s composition. Illuminating these capabilities contribute to discussions on what platforms can achieve as organisational technologies (Daymond et al., 2022; Gawer, 2014, 2021), what distinguishes them from other organisational forms (Gawer & Cusumano, 2014a; Jacobides et al., 2018), and how they build hierarchical structures within industries (Rikap & Lundvall, 2021; Srnicek, 2016). 5.3. On platform power dynamics This study also contributes to a more precise understanding of how power dynamics play out in platforms. In the first place, consistent with the multidirectional characterisation of power described above, this research identifies and explains the coexistence of dynamics of platform power and counterpower accumulation. While previous studies have described actions of resistance and contestation in platforms (Calo & Rosenblat, 2017; 111 Essays on Power in Digital Platforms Curchod et al., 2019; Eaton et al., 2015), these expressions are often absent in the efforts that systematise explanations of platform power (Busch et al., 2021; Khan, 2018; Zuboff, 2019). I discuss the dynamics of power and counterpower under different forms (sections 2.3.3 and 2.3.4) and also in the empirical context of the Amazon Marketplace, showing that while Amazon abuses its control of the governance to increase its bargaining power over sellers (sections 3.4.1.2 and 3.4.1.3), sellers in turn adapt and respond to Amazon by eluding and hacking the governance rules (section 3.4.2). Overall, this contributes to a conceptualisation of platform power that is dynamic and grounded on the agents’ practices, positions, and strategies. In addition, this thesis also makes contributions to understanding how power changes vis a vis the platform. My results and arguments are consistent with previous research on how power relations evolve in platforms, particularly the studies describing the change in power asymmetries as users become habituated and increase platform usage (Cutolo & Kenney, 2021; Zuboff, 2019). I discuss these problems when describing the power accumulation dynamics that characterise the different stages of platform evolution (section 2.3.4) and when showing how Amazon influences in the appearance of new sellers’ business models (section 4.4.2). At the same time, this study enriches previous explanations by theorising how platforms change both by power accumulation and power contestation (for example, by changing governance and creating “unwritten rules” for participation) and raising questions about the sustainability of platform owners’ power abuses (for example, as I show how discontent can eventually raise ecosystem members’ incentives to resist the platform owner’s actions). In third place, this study provides evidence that challenges central claims in the literature pertaining to mechanisms assumed to keep some of the power asymmetries existing in the platform under check. While scholars have clearly documented the existence of power asymmetries and power abuses by the platform owner (Khan, 2016; Stigler report, 2019), the consensus in the governance literature is that because of the way platforms create value, there are mechanisms limiting this power and how it is used. This is expressed in four central, interconnected claims, described in section 3.2.2: overall, abuses of the power to set the governance are supposed to be partially offset by the lack of hierarchical authority over participants and the need of them to create value. I find, however, that within the organising dynamics of platforms, the balance of incentives and control can coexist with a sustained mismatch between the governance the platform declares to abide by and its actual practices. This is shown when discussing the governance of the Amazon Marketplace in 3.4.1, with the evidence and its consequences for theory summarised in Table 13. The existence of abuses “hiding in plain sight” (evident for some participants -e.g., sellers- but unknown to others -e.g., consumers) evidence power asymmetries and limits to it work differently than previously assumed. 5.4. On the consequences of platform power This research identifies and explains new consequences of platform power, both on individuals and organisations. Regarding the impact on individuals, this study contributes by explaining how platform power dynamics have an organising role in producing and reproducing the delegation of autonomous decision-making. Previous research has explored the impact of digitalisation on individual autonomy from different perspectives and conceptual frameworks (Calo & Rosenblat, 2017; Curchod et al., 2019; Lanier, 2018; Scherer et al., 2020; Stark & Pais, 2020; 112 Essays on Power in Digital Platforms Zuboff, 2019), discussed in detail in sections 2.1 and 2.2.2. This study builds on and contributes to this research by showing how power struggles lead to an emergent process of autonomy delegation and how the conditions of agency are shaped by the specific structure of the platform and its power dynamics. Second, this research also recognises new aspects of how platform power dynamics can shape the relationships between organisations. Similarly to the consequences over individuals described above, I explain how business autonomy and decision-making processes can be constrained. I illustrate this by showing how Amazon’s orchestration activities influence sellers' business models (section 4.4.2). I describe the consequences for sellers, both in terms of performance (related to the platform owner capturing a larger share of value), strategy (options for value creation), and internal processes (molded to conform to platform standards). These results are in line with previous contributions describing how continued participation in platforms creates forms of organisational dependence and can constrain decision-making (Curchod et al., 2019; Cutolo & Kenney, 2021; Eaton et al., 2015). I contribute to these studies by showing how this dependence is rooted not only in the platform’s owner gatekeeping position and control of the complementors’ access to customers (Cutolo & Kenney, 2021; Khan, 2016) but also by the impact of decisions about functionalities provision related to complementors’ key value creation activities. This effect is in line with the notion of practice ownership resulting from the process of field enclosure, discussed above in sections 2.3.1.1 and 5.1. 5.5. Policy Implications9 This thesis provides conceptual tools and empirical evidence that can help policymakers and regulatory agencies identify and set limits on power abuses, especially in the light of platform domination. Informing policymakers of the strategic management of platforms and contributing to the finer understanding of platform power constitutes, I believe, a vital contribution of this work, which inscribes itself in a larger attempt to enrich the current regulatory debate on regulating the power of platforms and ecosystems (Cusumano et al., 2021; Heimburg & Wiesche, 2023; Jacobides & Lianos, 2021). This is because up until recently, regulators and antitrust enforcers have mostly been (perhaps even disproportionately) influenced by economic industrial organisation perspectives that tended to consider market power rather narrowly, as solely observable in terms of high prices, only within specified markets (as opposed to across markets), and which paid little attention to the strategic consequences of platforms’ differentiated business models. In other words, platform power was, in the earlier discussions on the relevance of introducing new platform regulation, mostly considered through the lens of pricing power and its consequences evaluated exclusively in terms of consumer harm. This framework and its inadequacy when applied to competition issues involving platforms have been highlighted by prominent scholars, some of whom have been commissioned for the highest positions in competition agencies in the USA and Europe (Khan, 2016; Scott Morton & Athey, 2021; Scott Morton & Kades, 2021). While the thinking in regulatory circles has evolved in recent years to tackle some of the specificities of digital technologies and multi-sided markets, 9 This section is based on unpublished drafts prepared with supervisor Annabelle Gawer for Chapter 3. 113 Essays on Power in Digital Platforms there is still a lack of consensus in terms of developing a complete framework of how platform power is actually expressed and how these various expressions may constitute harm. In this context, this thesis provides, in the first place, a conceptual framework with an integrated and nuanced explanation of platform power, its dynamics, and consequences. Because platforms transform society on multiple levels and across industry and market boundaries (on which regulation is typically based), holistic assessments of the social consequences of platforms are essential to inform the design of more effective regulation and regulatory scrutiny. An enriched understanding of the consequences of organising our practices through platforms, how they are embedded in social structures, and expressed in the transformation of fields can be challenging to translate into specific regulation but is nonetheless fundamental to start effectively addressing this issue. At the same time, this research offers tools and evidence that can be more directly considered for regulatory purposes. First, empirically derived constructs, such as the “two-faced platform governance” strategy of platform power, can be easily operationalised and assessed in different platforms and regions. In addition, the evidence provided here on Amazon and its complementors’ practices and the discussion of their consequences indicates that the practices identified can also be used by other platforms, especially those that enjoy market power. The list of practices and mechanisms discovered and organised can be helpful to policymakers worldwide as they grapple with the question of platform power. Overall, the empirically based, precise description of the various forms of platform practices that lead to the accumulation of platform power, as well as the nuanced discussion of the dynamic and interdependent consequences on complementors, can usefully enrich the empirical basis upon which regulatory agencies may enforce existing law. It can also inform policy-making in other regions, such as the USA or developing countries. 5.6. Limitations and future research This research has limitations that must be acknowledged when considering the implications of our findings and contributions. At the same time, these limitations can be addressed in coming studies and open new grounds for future research. Some of the limitations of this study are derived from the characteristics of the case chosen. While the pertinence and benefits of selecting the Amazon Marketplace for this study have been argued in detail in sections 1.1.2, 3.3, and 4.3, this selection can also present drawbacks, as some of this platform’s characteristics can veil relevant aspects of the problem in consideration. In particular, the fact that Amazon is one of the largest companies in the world and a dominant platform with durable market power in different regions, while greatly facilitates the task of studying platform power and identifying its dynamics, it also poses questions regarding the generalizability of the findings. Future studies can pursue similar aims as this research but in different platform contexts and explore the validity of the findings. Further, even if the conclusions of this study are not fundamentally challenged by new evidence, it is still highly relevant to understand how the constructs and dynamics described in this work play out in different settings. 114 Essays on Power in Digital Platforms Other limitations of this study are related instead to its focus. First, in the conceptual development of Chapter 2 and the empirical cases in Chapters 3 and 4, most of the attention is directed to how already powerful actors accumulate more power. While one of the contributions of this study is precisely to highlight both dynamics of power and counterpower accumulation by different agents, it is still possible that overall, this thesis builds a depiction of platforms - and particularly their owners – as immensely powerful agents without significant weaknesses. Nevertheless, however powerful an actor might currently be, it is necessary to assess that power's boundaries and understand how conditions might change. Consequently, a fruitful area for future research is identifying and explaining the specific processes and mechanisms by which, in addition to being contested, platform owners lose power and how these processes relate to the platform’s characteristics. Second, in line with the main trends in platform studies so far, this thesis has focused on the informational aspect of digital technologies. However, some of the current tendencies in how digital innovation is developing, such as the location and construction of data centres or the scale required for the evolving competition on computing power, highlight instead the importance of the technology’s physical dimensions. Consequently, factors such as the scale of physical investments, access to raw materials, and control of territories with strategic importance for connectivity might become increasingly relevant to understanding platform power dynamics in the future. This calls for further research on the interfaces between platforms' informational and physical dimensions and how this affects power. This thesis also uncovers opportunities for methodological development when empirically studying platforms. First, the study on platform governance in Chapter 3 introduces new constructs that are useful to characterise platform dynamics. While my inductive research was focused on identifying the phenomenon for theory building, it also creates opportunities for methodological contributions on how to gather information on platform governance and to create ways for measuring it. Second, Chapter 4 studies the long-term changes in the platform’s characteristics and interactions. This long-term approach, while holding great potential for research contributions, is however not abundant in the literature. It is to be noted the paradox that while data abundance is one of the defining aspects of digital platforms, the availability of longitudinal data remains a significant challenge for research, particularly data that might be used to assess agents’ practices and accountability. Indeed, gathering data on how a platform firm - with the importance of Amazon - formally defines the rules for participation over time involved an arduous and meticulous reconstruction through multiple sources and triangulation. These challenges illuminate, however, the importance for researchers and regulatory agencies to devise new methods to systematically collect information to build longitudinal databases on platform activities. Finally, my empirical study of the Amazon Marketplace also illuminates the importance of developing new frameworks for assessing performance in platform ecosystems. In platform research, overall platform performance is usually evaluated in terms of network size (e.g., number of consumers and providers), value creation (e.g., volume of transactions or rate of product innovation), and market share -for inter-platform competition-. However, while an analysis of the evolution of these metrics in the Marketplace would suggest a highly successful platform ecosystem, I also find evidence of governance practices that enable and further favour misbehaviour (section 3.4.3), tendencies of concentration of value capture, and mechanisms that drive organisational dependence (section 4.4.3). These findings call for contributions that expand the current frameworks on how platform ecosystem success is defined and assessed. 115 Essays on Power in Digital Platforms 5.7. Final remarks: repurposing platforms for social change This thesis commenced by arguing that the profound and abrupt economic transformation brought by digital platforms, along with expectations of higher social cohesion, increased freedom, and power distribution, was soon shadowed by the increasing evidence of platform power concentration and abuse. Whereas a significant part of academic research and public policy is now focused on understanding how to create effective regulation to curve platforms’ power – efforts to which this thesis aims to contribute -some scholars have pioneered a different approach, asking what platforms can mean for development (Bonina et al., 2021). While not disregarding the current and potential threats of platform power, these researchers focus instead on how to rethink platforms so they may deliver positive benefits for development and the common good (Bonina et al., 2021; Katsamakas et al., 2022). In line with this motivation, I suggest that this thesis’ contribution to a more precise understanding of digital platforms’ power and organising role can contribute to rethinking the role of platforms in society. Power abuse in platforms can lead to addiction and over-dependence, consumerism, mental health issues, polarisation and hate speech, and democracy impoverishment. However, platforms can also promote integral human development through sustainable consumption, mental well-being, and civic and political enhancement. Can the processes that lead to organised immaturity be “hacked” by society and reoriented towards organised enlightenment? If immaturity is the delegation of individual reasoning to external authority, enlightenment is sapere aude (i.e., having the courage to use our own understanding, Kant, 2019 [1784]). For digital platforms to become vehicles for human development would involve repurposing them as systems reinforcing autonomy and socially desirable values by promoting digital habitus of emancipation. Purposing platforms for development requires identifying designs that can balance successful business models with other non-economic forms of value creation (Katsamakas et al., 2022). This means understanding how to give a voice and articulate the specific constellation of actors involved in different development goals while keeping the potential for open interaction and growth that characterises platforms (Bonina et al., 2021; Chamakiotis et al., 2021). Further, it requires understanding how platforms can address the differences between developed and developing countries’ specific challenges and offer new, situated ways for creating value, something that academic research has only started to acknowledge (Jia & Kenney, 2021; Zhang, 2020). Overall, it requires exploring how to channel digital platforms’ unique problem-solving capabilities and our current knowledge on practices of platform design and governance towards creating mission-oriented platforms. 116 Essays on Power in Digital Platforms References Adner, R. (2017). Ecosystem as Structure: An Actionable Construct for Strategy. Journal of Management, 43(1), 3958. https://doi.org/10.1177/0149206316678451 Adner, R., & Kapoor, R. (2010). Value creation in innovation ecosystems: How the structure of technological interdependence affects firm performance in new technology generations. Strategic Management Journal, 31(3), 306-333. https://doi.org/10/b2wksf Agarwal, S., Miller, C. D., & Ganco, M. (2023). Growing platforms within platforms: How platforms manage the adoption of complementor products in the presence of network effects? Strategic Management Journal, smj.3487. https://doi.org/10.1002/smj.3487 Agrawal, A., Gans, J., & Goldfarb, A. (2019). The Economics of Artificial Intelligence. National Bureau of Economic Research Conference. https://www.press.uchicago.edu/ucp/books/book/chicago/E/bo35780726.html Alaimo, C., & Kallinikos, J. (2017). Computing the everyday: Social media as data platforms. The Information Society, 33(4), 175-191. https://doi.org/10.1080/01972243.2017.1318327 Alaimo, C., Kallinikos, J., & Valderrama, E. (2019). Platforms as service ecosystems: Lessons from social media. Journal of Information Technology, 026839621988146. https://doi.org/10.1177/0268396219881462 Alvesson, M. (2011). Generating Research Questions Through Problematization. Academy of Management Review, 26. Amazon. (2019). 2018 Amazon Annual Report. Amazon. (2020). 2019 Amazon Annual Report. Amazon. (2021). 2020 Amazon Annual Report. Amazon. (2022a). 2021 Amazon Annual Report. Amazon. (2022b). Leadership Principles. US About Amazon. https://www.aboutamazon.com/about-us/leadershipprinciples Andreoni, A., & Roberts, S. (2022). Governing digital platform power for industrial development: Towards an entrepreneurial-regulatory state. Cambridge Journal of Economics, beac055. https://doi.org/10.1093/cje/beac055 Aral, S., & Walker, D. (2012). Identifying Influential and Susceptible Members of Social Networks. Science, 337(6092), 337-341. https://doi.org/10.1126/science.1215842 Armstrong, M. (2006). Competition in two-sided markets. The RAND Journal of Economics, 37(3), 668-691. Autio, E. (2021). Orchestrating https://doi.org/10/gj6xwf ecosystems: A multi-layered framework. Innovation, 0(0), 1-14. Baldwin, C., & Clark, K. (2000). Design Rules: The power of modularity. MIT Press. Baldwin, C., & Woodard, C. J. (2009). The architecture of platforms: A unified view. Platforms, markets and innovation, 32. Baye, M. R., & Morgan, J. (2001). Information Gatekeepers on the Internet and the Competitiveness of Homogeneous Product Markets. The American Economic Review, 91(3), 454-474. JSTOR. https://doi.org/10/dgvnwh Beer, D. (2017). The social power of algorithms. Information, Communication & Society, 20(1), 1-13. https://doi.org/10.1080/1369118X.2016.1216147 Benlian, A., Hilkert, D., & Hess, T. (2015). How open is this Platform? The Meaning and Measurement of Platform Openness from the Complementers’ Perspective. Journal of Information Technology, 30(3), 209-228. https://doi.org/10.1057/jit.2015.6 Bennett, W. L., Segerberg, A., & Walker, S. (2014). Organization in the crowd: Peer production in large-scale networked protests. Information, Communication & Society, 17(2), 232-260. https://doi.org/10/f2356g Berzon, A., Shifflett, S., & Scheck, J. (2019, agosto 23). Amazon Has Ceded Control of Its Site. The Result: Thousands of Banned, Unsafe or Mislabeled Products. Wall Street Journal. 117 Essays on Power in Digital Platforms https://www.wsj.com/articles/amazon-has-ceded-control-of-its-site-the-result-thousands-of-bannedunsafe-or-mislabeled-products-11566564990 Bhargava, H. K. (2021). The Creator Economy: Managing Ecosystem Supply, Revenue-Sharing, and Platform Design. 33. Bonina, C., Koskinen, K., Eaton, B., & Gawer, A. (2021). Digital platforms for development: Foundations and research agenda. Information Systems Journal, n/a(n/a). https://doi.org/10/ghwkg2 Boudreau, K. (2010). Open Platform Strategies and Innovation: Granting Access vs. Devolving Control. Management Science, 56(10), 1849-1872. https://doi.org/10.1287/mnsc.1100.1215 Boudreau, K. J. (2012). Let a Thousand Flowers Bloom? An Early Look at Large Numbers of Software App Developers and Patterns of Innovation. Organization Science, 23(5), 1409-1427. https://doi.org/10.1287/orsc.1110.0678 Boudreau, K. J. (2017). Platform Boundary Choices & Governance: Opening-Up While Still Coordinating and Orchestrating☆. En Entrepreneurship, Innovation, and Platforms (Vol. 37, pp. 227-297). Emerald Publishing Limited. https://doi.org/10.1108/S0742-332220170000037009 Boudreau, K. J., & Jeppesen, L. B. (2014). Unpaid Crowd Complementors: The Platform Network Effect Mirage. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.2475034 Bourdieu, P. (1969). Intellectual field and creative project. Social Science Information, 8(2), 89-119. https://doi.org/10.1177/053901846900800205 Bourdieu, P. (1977). Outline of a theory of practice. Cambridge University Press. Bourdieu, P. (1979). Symbolic Power. Critique of Anthropology. https://doi.org/10/d8636t Bourdieu, P. (1984). Distinction: A social critique of the judgement of taste (11. print). Harvard Univ. Press. Bourdieu, P. (1987). The Force of Law: Toward a Sociology of the Juridical Field. THE HASTINGS LAW JOURNAL, 38, 42. Bourdieu, P. (1989). Social Space and Symbolic Power. Sociological Theory, 7(1), 14-25. https://doi.org/10/fv5vnj Bourdieu, P. (1990). The Logic of Practice. Stanford University Press. Bourdieu, P. (1991). Language and Symbolic Power. Harvard University Press. Bourdieu, P. (1993). The field of cultural production. Columbia University Press. Bourdieu, P. (2005). Principles of an economic anthropology. En The Handbook of Economic Sociology (Second Edition). Princeton University Press. Bourdieu, P. (2011). The Forms of Capital. En The Sociology of Economic Life (3.a ed.). Routledge. Bourdieu, P. (2014). On the state: Lectures at the Collège de France, 1989 - 1992 (P. Champagne, Ed.). Polity Press. Bourdieu, P., & Wacquant, L. (Eds.). (2007). An invitation to reflexive sociology (Repr). Polity Pr. Boyle, J. (2003). The Second Enclosure Movement and the Construction of the Public Domain. LAW AND CONTEMPORARY PROBLEMS, 66, 42. Bradshaw, S., & Howard, P. N. (2018). Challenging Truth and Trust: A Global Inventory of Organized Social Media Manipulation. 26. Bromley, P., & Powell, W. W. (2012). From Smoke and Mirrors to Walking the Talk: Decoupling in the Contemporary World. Academy of Management Annals, 6(1), 483-530. https://doi.org/10.5465/19416520.2012.684462 Bucher, E., Fieseler, C., Fleck, M., & Lutz, C. (2018). Authenticity and the Sharing Economy. Academy of Management Discoveries, 4(3), 294-313. https://doi.org/10.5465/amd.2016.0161 Bucher, E. L., Schou, P. K., & Waldkirch, M. (2021). Pacifying the algorithm – Anticipatory compliance in the face of algorithmic management in the gig economy. Organization, 28(1), 44-67. https://doi.org/10.1177/1350508420961531 Busch, C., Graef, I., Hofmann, J., & Gawer, A. (2021). Uncovering blindspots in the policy debate on platform power. https://platformobservatory.eu/app/uploads/2021/03/05Platformpower.pdf 118 Essays on Power in Digital Platforms Caillaud, B., & Jullien, B. (2003). Chicken & egg: Competition among intermediation service providers. RAND Journal of Economics (RAND Journal of Economics), 34(2), 309-328. https://doi.org/10/ckddg4 Calo, R., & Rosenblat, A. (2017). The Taking Economy: Uber, Information, and Power. SSRN Electronic Journal. https://doi.org/10/gfvmg3 Castelló, I., & Lopez-Berzosa, D. (2021). Affects in Online Stakeholder Engagement: A Dissensus Perspective. Business Ethics Quarterly, 1-36. https://doi.org/10.1017/beq.2021.35 Castells, M. (2011). Network Theory| A Network Theory of Power. International Journal of Communication, 5(0), 15. Cenamor, J., & Frishammar, J. (2021). Openness in platform ecosystems: Innovation strategies for complementary products. Research Policy, 50(1), 104148. https://doi.org/10.1016/j.respol.2020.104148 Cennamo, C. (2016). Building the Value of Next-Generation Platforms: The Paradox of Diminishing Returns. Journal of Management, 44(8), 3038-3069. https://doi.org/10.1177/0149206316658350 Cennamo, C., Ozalp, H., & Kretschmer, T. (2018). Platform Architecture and Quality Trade-offs of Multihoming Complements. Information Systems Research, 29(2), 461-478. https://doi.org/10.1287/isre.2018.0779 Cennamo, C., & Santalo, J. (2013). Platform competition: Strategic trade-offs in platform markets. Strategic Management Journal, 34(11), 1331-1350. https://doi.org/10.1002/smj.2066 Cennamo, C., & Santaló, J. (2019). Generativity Tension and Value Creation in Platform Ecosystems. Organization Science, 30(3), 617-641. https://doi.org/10.1287/orsc.2018.1270 Chamakiotis, P., Petrakaki, D., & Panteli, N. (2021). Social value creation through digital activism in an online health community. Information Systems Journal, 31(1), 94-119. https://doi.org/10.1111/isj.12302 Charmaz, K. (2000). Grounded theory: Objectivist and constructivist methods. Handbook of qualitative research, 2, 509-535. Charmaz, K., & Bryant, A. (2008). Grounded Theory. En The SAGE Encyclopedia of Qualitative Research Methods. SAGE Publications, Inc. https://doi.org/10.4135/9781412963909 Chen, L., Tong, T., Tang, S., & Han, N. (2021). Digital Platforms’ Governance and Design: A Review and Future Research Directions on the New Organizational Form. https://minervaaccess.unimelb.edu.au/bitstream/handle/11343/281409/JOM%20platform%20governance%20and%20de sign.pdf?sequence=3&isAllowed=y Chesbrough, H. (2006). Open Business Models: How To Thrive In The New Innovation Landscape. Harvard Business Press. Chesbrough, H. W. (2006). The era of open innovation. Managing innovation and change, 127(3), 34-41. Choi, S. A. (2020). Amazon Fake Reviews (arXiv:2009.09102). arXiv. https://doi.org/10.48550/arXiv.2009.09102 Claussen, J., Kretschmer, T., & Mayrhofer, P. (2013). The Effects of Rewarding User Engagement: The Case of Facebook Apps. Information Systems Research, 24(1), 186-200. https://doi.org/10.1287/isre.1120.0467 Clegg, S. R. (1989). Organization Theory and Class Analysis: New Approaches and New Issues. De Gruyter. Constantinides, P., Henfridsson, O., & Parker, G. G. (2018). Introduction—Platforms and Infrastructures in the Digital Age. Information Systems Research, 29(2), 381-400. https://doi.org/10.1287/isre.2018.0794 Corbin, J., & Strauss, A. (2008). Basics of Qualitative Research (3rd ed.): Techniques and Procedures for Developing Grounded Theory. SAGE Publications, Inc. https://doi.org/10.4135/9781452230153; http://web.archive.org/web/20200619005647/http://methods.sagepub.com/book/basics-of-qualitativeresearch Costa, R. L. (2006). The Logic of Practices in Pierre Bourdieu. Current Sociology, 54(6), 873-895. https://doi.org/10/fvkz5m Coveri, A., Cozza, C., & Guarascio, D. (2022). Monopoly Capital in the time of digital platforms: A radical approach to the Amazon case. Cambridge Journal of Economics, 46(6), 1341-1367. https://doi.org/10.1093/cje/beac044 Crémer, J., Montjoye, Y.-A. de, Schweitzer, H., European Commission, & Directorate-General for Competition. (2019). Competition policy for the digital era. http://publications.europa.eu/publication/manifestation_identifier/PUB_KD0419345ENN 119 Essays on Power in Digital Platforms Culpepper, P. D., & Thelen, K. (2020). Are We All Amazon Primed? Consumers and the Politics of Platform Power. Comparative Political Studies, 53(2), 288-318. https://doi.org/10/ggcd52 Curchod, C., Patriotta, G., Cohen, L., & Neysen, N. (2019). Working for an Algorithm: Power Asymmetries and Agency in Online Work Settings. Administrative Science Quarterly, 000183921986702. https://doi.org/10.1177/0001839219867024 Cusumano, M. A., Gawer, A., & Yoffie, D. B. (2019). The business of platforms: Strategy in the age of digital competition, innovation, and power (First edition.). Harper Business, an imprint of HarperCollins Publishers. Cusumano, M. A., Gawer, A., & Yoffie, D. B. (2021). Can self-regulation save digital platforms? Industrial and Corporate Change, 30(5), 1259-1285. https://doi.org/10/gn6zv5 Cusumano, M., & Gawer, A. (2002). The Elements of Platform Leadership. MIT Sloan Management Review, 43(3), 51-58. Cusumano, M., Gawer, A., & Yoffie, D. (2019). The Business of Platforms: Strategy in the Age of Digital Competition, Innovation, and Power. HarperBusiness. https://www.harpercollins.co.uk/9780062896322/the-business-of-platforms-strategy-in-the-age-ofdigital-competition-innovation-and-power/ Cutolo, D., & Kenney, M. (2021). Platform-Dependent Entrepreneurs: Power Asymmetries, Risks, and Strategies in the Platform Economy. Academy of Management Perspectives, amp.2019.0103. https://doi.org/10/gg2zt5 Danaher, J., Murphy, M. H., Khoo, S., & Kennedy, R. (2017). Algorithmic governance: Developing a research agenda through the power of collective intelligence. Big Data & Society. https://www.academia.edu/34710228/Algorithmic_governance_Developing_a_research_agenda_through_t he_power_of_collective_intelligence Daymond, J., Knight, E., Rumyantseva, M., & Maguire, S. (2022). Managing ecosystem emergence and evolution: Strategies for ecosystem architects. Strategic Management Journal, n/a(n/a). https://doi.org/10.1002/smj.3449 Deilen, M., & Wiesche, M. (2021). The Role of Complementors in Platform Ecosystems. En F. Ahlemann, R. Schütte, & S. Stieglitz (Eds.), Innovation Through Information Systems (pp. 473-488). Springer International Publishing. https://doi.org/10/gnbhmz Dewey, J. (1939). Freedom and Culture. Putnam. Dhanaraj, C., & Parkhe, A. (2006). Orchestrating Innovation Networks. The Academy of Management Review, 31(3), 659-669. JSTOR. https://doi.org/10/fwhgzg Eaton, B., Elaluf-Calderwood, S., Sørensen, C., & Yoo, Y. (2015). Distributed Tuning of Boundary Resources: The Case of Apple’s iOS Service System. MIS Quarterly, 39(1), 217-243. https://doi.org/10/gfgncr Eisenhardt, K. M. (1989). Building Theories from Case Study Research. Academy of Management Review, 14(4), 532-550. https://doi.org/10.5465/amr.1989.4308385 Eisenmann, T., Parker, G., & Alstyne, M. W. V. (2006). Strategies for Two- Sided Markets. Harvard Business Review, 12. Eisenmann, T., Parker, G., & Van Alstyne, M. (2009). Opening platforms: How, when and why? Platforms, markets and innovation, 6, 131-162. Engert, M., Evers, J., Hein, A., & Krcmar, H. (2022). The Engagement of Complementors and the Role of Platform Boundary Resources in e-Commerce Platform Ecosystems. Information Systems Frontiers. https://doi.org/10/gn3hks Etter, M., & Albu, O. B. (2021). Activists in the dark: Social media algorithms and collective action in two social movement organizations. Organization, 28(1), 68-91. https://doi.org/10.1177/1350508420961532 Eubanks, V. (2018). Automating Inequality: How High-Tech Tools Profile, Police, and Punish the Poor. European Commission. (2022a). The Digital Services strategy.ec.europa.eu/en/policies/digital-services-act-package Act package. https://digital- 120 Essays on Power in Digital Platforms European Commission. (2022b, marzo 24). Deal on Digital Markets Act. https://www.europarl.europa.eu/news/es/press-room/20220315IPR25504/deal-on-digital-markets-actensuring-fair-competition-and-more-choice-for-users Evans, D. (2003). Some Empirical Aspects of Multi-sided Platform Industries. Review of Network Economics, 2(3). https://doi.org/10.2202/1446-9022.1026 Evans, D. S. (2003). The Antitrust Economics of Multi-Sided Platform Markets. Yale Journal on Regulation, 20, 325. Evans, D. S. (2010). Essays on the Economics of Two-Sided Markets: Economics, Antitrust and Strategy. Antitrust and Strategy (November 24, 2010). http://papers.ssrn.com/sol3/Papers.cfm?abstract_id=1714254 Evans, P., & Gawer, A. (2016). The Rise of the Platform content/uploads/2016/01/PDF-WEB-Platform-Survey_01_12.pdf Enterprise. http://thecge.net/wp- Faraj, S., Pachidi, S., & Sayegh, K. (2018). Working and organizing in the age of the learning algorithm. Information and Organization, 28(1), 62-70. https://doi.org/10.1016/j.infoandorg.2018.02.005 Fayard, A.-L., & Weeks, J. (2014). Affordances for practice. Information and Organization, 24(4), 236-249. https://doi.org/10/f6rt42 Flyverbom, M., Deibert, R., & Matten, D. (2019). The Governance of Digital Technology, Big Data, and the Internet: New Roles and Responsibilities for Business. Business & Society, 58(1), 3-19. https://doi.org/10/gf734f Flyverbom, M., Leonardi, P. M., Stohl, C., & Stohl, M. (2016). The management of visibilities in the digital age. International Journal of Communication, 10, 98-109. Fuchs, C. (2007). Internet and Society: https://doi.org/10.4324/9780203937778 Social Theory in the Information Age. Furman, J. (2019). Unlocking digital competition. Digital Competition Expert Panel. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/7855 47/unlocking_digital_competition_furman_review_web.pdf Routledge. Garud, R., Kumaraswamy, A., Roberts, A., & Xu, L. (2022). Liminal movement by digital platform-based sharing economy ventures: The case of Uber Technologies. Strategic Management Journal, 43(3), 447-475. https://doi.org/10.1002/smj.3148 Gawer, A. (2009). Platforms, Markets https://doi.org/10.4337/9781849803311 and Innovation. Edward Elgar Publishing. Gawer, A. (2014). Bridging differing perspectives on technological platforms: Toward an integrative framework. Research Policy, 43(7), 1239-1249. https://doi.org/10.1016/j.respol.2014.03.006 Gawer, A. (2021). Digital Platforms’ Boundaries: The Interplay of Firm Scope, Platform Sides, and Digital Interfaces. Long Range Planning, 54(5), 1-16. https://doi.org/10.1016/j.lrp.2020.102045 Gawer, A., & Cusumano, M. A. (2014a). Industry Platforms and Ecosystem Innovation. Journal of Product Innovation Management, 31(3), 417-433. https://doi.org/10.1111/jpim.12105 Gawer, A., & Cusumano, M. A. (2014b). Platforms and Innovation. En M. Dodgson, D. M. Gann, & N. Phillips (Eds.), The Oxford Handbook of Innovation Management. Oxford University Press. http://www.oxfordhandbooks.com/view/10.1093/oxfordhb/9780199694945.001.0001/oxfordhb9780199694945-e-014 Gawer, A., & Henderson, R. (2007). Platform owner entry and innovation in complementary markets: Evidence from Intel. Journal of Economics & Management Strategy, 16(1), 1-34. Gawer, A., & Srnicek, N. (2021). Online platforms: Economic and societal effects. https://doi.org/10.2861/844602 Ghazawneh, A., & Henfridsson, O. (2013). Balancing platform control and external contribution in third-party development: The boundary resources model. Information Systems Journal, 23(2), 173-192. https://doi.org/10/f2253r Ghosh, A., & Hummel, P. (2014). A game-theoretic analysis of rank-order mechanisms for user-generated content. Journal of Economic Theory, 154, 349-374. https://doi.org/10.1016/j.jet.2014.09.009 Gillespie, T. (2010). The politics of ‘platforms’. https://doi.org/10.1177/1461444809342738 New Media & Society, 12(3), 347-364. 121 Essays on Power in Digital Platforms Gioia, D. A., Corley, K. G., & Hamilton, A. L. (2013). Seeking Qualitative Rigor in Inductive Research: Notes on the Gioia Methodology. Organizational Research Methods, 16(1), 15-31. https://doi.org/10.1177/1094428112452151 Glaser, B. G. (1987). Theoretical sensitivity: Advances in the methodology of grounded theory. The Sociology Press. Glaser, B. G., & Strauss, A. L. (1967). Discovery of Grounded Theory: Strategies for Qualitative Research. Routledge. Goldhaber, M. H. (1997). The attention economy and the Net. First Monday. https://doi.org/1581002842 Gulati, R., Puranam, P., & Tushman, M. (2012). Meta-organization design: Rethinking design in interorganizational and community contexts. Strategic Management Journal, 33(6), 571-586. https://doi.org/10.1002/smj.1975 Hagiu, A. (2006). Multi-sided Platforms: From microfoundations to design and expansion strategies. http://dx.doi.org/10.2139/ssrn.955584 Hagiu, A. (2013). Multi-Sided Platforms: Foundations and Strategy. Harvard Business School, 16. Hagiu, A., Teh, T.-H., & Wright, J. (2020). Should platforms be allowed to sell on their own marketplaces? 42. Hamel, G., & Zanini, M. (2018, noviembre 1). The End of Bureaucracy. Harvard Business Review, November– December 2018. https://hbr.org/2018/11/the-end-of-bureaucracy He, S., Hollenbeck, B., & Proserpio, D. (2022). The Market for Fake Reviews. Marketing Science. https://doi.org/10.1287/mksc.2022.1353 Heimburg, V., & Wiesche, M. (2022). Relations Between Actors in Digital Platform Ecosystems: A Literature Review. Heimburg, V., & Wiesche, M. (2023). Digital platform regulation: Opportunities for information systems research. Internet Research, 33(7), 72-85. https://doi.org/10.1108/INTR-05-2022-0321 Hilferding, R. (2005). Finance capital: A study of the lastest phase of capitalist development. Taylor & Francis Ltd. Holgersson, M., Baldwin, C. Y., Chesbrough, H., & M. Bogers, M. L. A. (2022). The Forces of Ecosystem Evolution. California Management Review, 64(3), 5-23. https://doi.org/10.1177/00081256221086038 Huber, T. L., Kude, T., & Dibbern, J. (2017). Governance Practices in Platform Ecosystems: Navigating Tensions Between Cocreated Value and Governance Costs. Information Systems Research, 28(3), 563-584. https://doi.org/10/gb2492 Hukal, P., Kanat, I., & Ozalp, H. (2022). Different Strategy Playbooks for Digital Platform Complementors. MIS Quarterly Executive, 21(1). https://aisel.aisnet.org/misqe/vol21/iss1/4 Hunsinger, J., & Schrock, A. (2016). The democratization of hacking and making. New Media & Society, 18(4), 535538. https://doi.org/10/gncrvv Hurni, T., Huber, T. L., Dibbern, J., & Krancher, O. (2021). Complementor dedication in platform ecosystems: Rule adequacy and the moderating role of flexible and benevolent practices. European Journal of Information Systems, 30(3), 237-260. https://doi.org/10.1080/0960085X.2020.1779621 Ignatow, G., & Robinson, L. (2017). Pierre Bourdieu: Theorizing the digital. Information, Communication & Society, 20(7), 950-966. https://doi.org/10/gc7n3s ILSR. (2022). Survey Sellers 2022. Introna, L. D. (2016). Algorithms, Governance, and Governmentality: On Governing Academic Writing. Science, Technology, & Human Values, 41(1), 17-49. https://doi.org/10.1177/0162243915587360 Islam, H. A., Farrell, M., Nair, A., & Zhang, J. (2023). Understanding transaction platform governance and conflicts: A configuration approach. Technological Forecasting and Social Change, 189, 122382. https://doi.org/10.1016/j.techfore.2023.122382 Jacobides, M. G. (2021). What Drives and Defines Digital Platform Power? 57. Jacobides, M. G., Bruncko, M., & Langen, R. (2020). Regulating Big Tech in Europe: Why, so what, and how understanding their business models and ecosystems can make a difference. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.3765324 Jacobides, M. G., Cennamo, C., & Gawer, A. (2018). Towards a theory of ecosystems. Strategic Management Journal, 39(8), 2255-2276. https://doi.org/10.1002/smj.2904 122 Essays on Power in Digital Platforms Jacobides, M. G., Cennamo, C., & Gawer, A. (2022). Externalities and Complementarities in Platforms and Ecosystems: From Structural Solutions to Endogenous Failures. https://events.concurrences.com/IMG/pdf/jcg_platform_ecosystem_failures.pdf Jacobides, M. G., & Lianos, I. (2021). Regulating platforms and ecosystems: An introduction. Industrial and Corporate Change, dtab060. https://doi.org/10/gm7xvb Jia, K., & Kenney, M. (2021). The Chinese platform business group: An alternative to the Silicon Valley model? Journal of Chinese Governance, 0(0), 1-23. https://doi.org/10/gh2kzt Jordan, T. (2009). Hacking and power: Social and technological determinism in the digital age. First Monday. https://doi.org/10.5210/fm.v14i7.2417 Joyce, S., & Stuart, M. (2021). Digitalised management, control and resistance in platform work: A labour process analysis. Work and Labour Relations in Global Platform Capitalism. https://www.elgaronline.com/view/edcoll/9781802205121/9781802205121.00017.xml Julien, C. (2015). Bourdieu, Social Capital and https://doi.org/10.1177/0038038514535862 Online Interaction. Sociology, 49(2), 356-373. Jung, Y., & Lyytinen, K. (2014). Towards an ecological account of media choice: A case study on pluralistic reasoning while choosing email. Information Systems Journal, 24(3), 271-293. https://doi.org/10/f5w7qx Jungle Scout. (2020). State of the Amazon Seller 2020. Jungle Scout. (2021). State of the Amazon Seller 2021. https://www.junglescout.com/amazon-seller-report/ Jungle Scout. (2022a). Consumer Trends Report Q2 2022. https://www.junglescout.com/wpcontent/uploads/2022/06/Jungle-Scout-Consumer-Trends-Report-Q2-2022.pdf Jungle Scout. (2022b). The State of the Amazon Seller 2022. Just, N., & Latzer, M. (2017). Governance by algorithms: Reality construction by algorithmic selection on the Internet. Media, Culture & Society, 39(2), 238-258. https://doi.org/10.1177/0163443716643157 Just, S. N. (2019). An assemblage of avatars: Digital organization as affective intensification in the GamerGate controversy. Organization, 26(5), 716-738. https://doi.org/10.1177/1350508419842710 Kant, I. (2019). An Answer to the Question What is Enlightenment. https://nbn-resolving.org/urn:nbn:de:101:12019121915134762598561 Katsamakas, E., Miliaresis, K., & Pavlov, O. V. (2022). Digital Platforms for the Common Good: Social Innovation for Active Citizenship and ESG. Sustainability, 14(2), Article 2. https://doi.org/10.3390/su14020639 Katz, M. L., & Shapiro, C. (1985). Network Externalities, Competition, and Compatibility. The American Economic Review, 75(3), 424-440. JSTOR. Katz, M. L., & Shapiro, C. (1986). Technology Adoption in the Presence of Network Externalities. Journal of Political Economy, 94(4), 822-841. JSTOR. https://doi.org/10/c8dvpd Kelkar, S. (2018). Engineering a platform: The construction of interfaces, users, organizational roles, and the division of labor. New Media & Society, 20(7), 2629-2646. https://doi.org/10.1177/1461444817728682 Kenney, M., Bearson, D., & Zysman, J. (2021). The platform economy matures: Measuring pervasiveness and exploring power. Socio-Economic Review, 19(4), 1451-1483. https://doi.org/10.1093/ser/mwab014 Kenney, M., & Zysman, J. (2016, abril 1). The Rise of the Platform Economy. Issues in Science and Technology. https://issues.org/rise-platform-economy-big-data-work/ Kenny, M., & Fourie, R. (2015). Contrasting Classic, Straussian, and Constructivist Grounded Theory: Methodological and Philosophical Conflicts. The Qualitative Report, 20(8), 22. Khan, L. M. (2016). Amazon’s Antitrust Paradox. Yale Law Journal, 126(3), 710-805. Khan, L. M. (2018). Sources of Tech Platform Power. Georgetown Law Technology Review, 2, 10. https://ssrn.com/abstract=3558741 Kitchens, B., Johnson, S. L., & Gray, P. (2020). Understanding Echo Chambers and Filter Bubbles: The Impact of Social Media on Diversification and Partisan Shifts in News Consumption. MIS Quarterly, 44(4), 16191649. https://doi.org/10.25300/MISQ/2020/16371 123 Essays on Power in Digital Platforms Kolagar, M., Parida, V., & Sjödin, D. (2022). Ecosystem transformation for digital servitization: A systematic review, integrative framework, and future research agenda. Journal of Business Research, 146, 176-200. https://doi.org/10.1016/j.jbusres.2022.03.067 Kölbel, T., Kunz, D., & Lamberty, R. (2023). Dark Sides of the Platform Economy: Market Power and Its Abuse by Platform Orchestrators. ECIS 2023 Research Papers. https://aisel.aisnet.org/ecis2023_rp/274 Kornberger, M., Pflueger, D., & Mouritsen, J. (2017). Evaluative infrastructures: Accounting for platform organization. Accounting, Organizations and Society, 60, 79-95. https://doi.org/10/gc95xr Kretschmer, T., Leiponen, A., Schilling, M., & Vasudeva, G. (2022). Platform ecosystems as meta-organizations: Implications for platform strategies. Strategic Management Journal, 43(3), 405-424. https://doi.org/10.1002/smj.3250 Krona, M. (2015). Contravigilancia y videoactivismo desde la plaza Tahrir. Sobre las paradojas de la sociedad contravigilante. En Videoactivismo y movimientos sociales (p. 17). Gedisa. Lan, S., Liu, K., & Dong, Y. (2019). Dancing with wolves: How value creation and value capture dynamics affect complementor participation in industry platforms. Industry and Innovation, 26(8), 943-963. https://doi.org/10.1080/13662716.2019.1598339 Landini, F. (2016). The evolution of control in the digital economy. Journal of Evolutionary Economics, 26(2), 407441. Scopus. https://doi.org/10.1007/s00191-016-0450-z Langley, P., & Leyshon, A. (2017). Platform capitalism: The intermediation and capitalization of digital economic circulation. Finance and society., 3(1), Article 1. https://doi.org/10.2218/finsoc.v3i1.1936 Lanier, J. (2018). Ten Arguments For Deleting Your Social Media Accounts Right Now. Random House. Larsson, S., & Heintz, F. (2020). Transparency in artificial intelligence. Internet Policy Review, 9(2). https://doi.org/10/gg35rc Lee, M. K., Kusbit, D., Metsky, E., & Dabbish, L. (2015). Working with Machines: The Impact of Algorithmic and Data-Driven Management on Human Workers. Proceedings of the 33rd Annual ACM Conference on Human Factors in Computing Systems - CHI ’15, 1603-1612. https://doi.org/10.1145/2702123.2702548 Lehdonvirta, V. (2022). Cloud Empires: How Digital Platforms Are Overtaking the State and How We Can Regain Control. The MIT Press. Lessig, L. (2009). El Código 2.0. Madrid : Traficantes de Sueños, http://www.traficantes.net/index.php/content/download/25950/245538/file/codigo%202.0completo_donacion.pdf 2009. Levina, N., & Arriaga, M. (2014). Distinction and Status Production on User-Generated Content Platforms: Using Bourdieu’s Theory of Cultural Production to Understand Social Dynamics in Online Fields. Information Systems Research, 25. https://doi.org/10/f6j6q6 Li, Z., & Qi, H. (2022). Platform power: Monopolisation and financialisation in the era of big tech. Cambridge Journal of Economics, beac054. https://doi.org/10.1093/cje/beac054 Liang, X., Luo, Y., Shao, X., & Shi, X. (2022). Managing complementors in innovation ecosystems: A typology for generic strategies. Industrial Management & Data Systems, ahead-of-print(ahead-of-print). https://doi.org/10.1108/IMDS-12-2021-0809 Lianos, I., & Carballa-Smichowski, B. (2022). A coat of many colours—New concepts and metrics of economic power in competition law and economics. Journal of Competition Law & Economics, nhac002. https://doi.org/10.1093/joclec/nhac002 Lupton, D. (2016). The Quantified Self. John Wiley & Sons. Lynskey, O. (2017). Regulating «Platform Power». SSRN Electronic Journal. https://doi.org/10.2139/ssrn.2921021 Lynskey, O. (2019). Grappling with “Data Power”: Normative Nudges from Data Protection and Privacy. Theoretical Inquiries in Law, 20(1), 189-220. https://doi.org/10/gf25rf Lyon, D. (2018). The Culture of Surveillance: Watching as a Way of Life (1st edition). Polity. Maddox, J., & Malson, J. (2020). Guidelines Without Lines, Communities Without Borders: The Marketplace of Ideas and Digital Manifest Destiny in Social Media Platform Policies. Social Media + Society, 6(2), 2056305120926622. https://doi.org/10/gj89f9 124 Essays on Power in Digital Platforms Manyika, J., Ramaswamy, S., Bughin, J., Woetzel, J., Birshan, M., & Nagpal, Z. (2018, octubre). Superstars. The dynamics of firms, sectors, and cities leading the global economy. https://www.mckinsey.com/~/media/mckinsey/featured%20insights/innovation/superstars%20the%20dy namics%20of%20firms%20sectors%20and%20cities%20leading%20the%20global%20economy/mgi_supe rstars_discussion%20paper_oct%202018-final.ashx Margetts, H., Lehdonvirta, V., González-Bailón, S., Hutchinson, J., Bright, J., Nash, V., & Sutcliffe, D. (2021). The Internet and public policy: Future directions. Policy & Internet, n/a(n/a). https://doi.org/10.1002/poi3.263 Martin, K. E. (2021). Algorithmic Bias and Corporate Responsibility: How companies hide behind the false veil of the technological imperative (SSRN Scholarly Paper ID 3905275). Social Science Research Network. https://doi.org/10.2139/ssrn.3905275 McCoy, J., Rahman, T., & Somer, M. (2018). Polarization and the Global Crisis of Democracy: Common Patterns, Dynamics, and Pernicious Consequences for Democratic Polities. American Behavioral Scientist, 62(1), 1642. https://doi.org/10.1177/0002764218759576 McIntyre, D. P., & Srinivasan, A. (2017). Networks, platforms, and strategy: Emerging views and next steps. Strategic Management Journal, 38(1), 141-160. https://doi.org/10.1002/smj.2596 McIntyre, D., Srinivasan, A., Afuah, A., Gawer, A., & Kretschmer, T. (2021). Multi-sided platforms as new organizational forms. Academy of Management Perspectives, 35(4), amp.2018.0018. https://doi.org/10/gg4brh Meyer, J. W., & Rowan, B. (1977). Institutionalized Organizations: Formal Structure as Myth and Ceremony. American Journal of Sociology, 83(2), 340-363. https://doi.org/10.1086/226550 Miric, M., Boudreau, K. J., & Jeppesen, L. B. (2019). Protecting their digital assets: The use of formal & informal appropriability strategies by App developers. Research Policy, 48(8), 103738. https://doi.org/10/gg339s Nadler, J., & Cicilline, D. (2022). Investigation of Competition in Digital Markets. Subcommittee on Antitrust, Commercial, and Administrative Law of the Committee on the Judiciary of the House of Representatives. Nambisan, S., & Baron, R. A. (2013). Entrepreneurship in Innovation Ecosystems: Entrepreneurs’ Self–Regulatory Processes and Their Implications for New Venture Success. Entrepreneurship Theory and Practice, 37(5), 1071-1097. https://doi.org/10.1111/j.1540-6520.2012.00519.x Neyland, D., & Möllers, N. (2017). Algorithmic IF … THEN rules and the conditions and consequences of power. Information, Communication & Society, 20(1), 45-62. https://doi.org/10.1080/1369118X.2016.1156141 Nguyen, N. (2021, junio 13). Fake Reviews and Inflated Ratings Are Still a Problem for Amazon. Wall Street Journal. https://www.wsj.com/articles/fake-reviews-and-inflated-ratings-are-still-a-problem-for-amazon11623587313 Noble, S. U. (2018). Algorithms of Oppression: How Search Engines Reinforce Racism (Illustrated edition). NYU Press. Nocke, V., Peitz, M., & Stahl, K. (2007). Platform Ownership. Journal of the European Economic Association, 5(6), 1130-1160. https://doi.org/10.1162/JEEA.2007.5.6.1130 O’Connor, C., & Weatherall, J. O. (2019). The Misinformation Age: How False Beliefs Spread. Yale University Press. Orlikowski, W. J., & Scott, S. V. (2015). The Algorithm and the Crowd: Considering the Materiality of Service Innovation. MIS Quarterly, 39(1), 201-216. Ozalp, H., Cennamo, C., & Gawer, A. (2018). Disruption in Platform-Based Ecosystems. Journal of Management Studies, 55(7), 1203-1241. https://doi.org/10.1111/joms.12351 Parker, C., Scott, S., & Geddes, A. (2019). Snowball Sampling. SAGE Research Methods Foundations. http://methods.sagepub.com/foundations/snowball-sampling Parker, G., & Van Alstyne, M. (2005). Two-Sided Network Effects: A Theory of Information Product Design. Management Science, 51(10), 1494-1504. https://doi.org/10/dw2pq9 Parker, G., & Van Alstyne, M. (2017). Innovation, Openness, and Platform Control. Management Science, 64(7), 3015-3032. https://doi.org/10.1287/mnsc.2017.2757 Parker, Van Alstyne, M. W., & Choudary, S. P. (2016). Platform revolution: How networked markets are transforming the economy and how to make them work for you. WW Norton & Company. 125 Essays on Power in Digital Platforms Pasquale, F. (2015). The Black Box Society: The Secret Algorithms That Control Money and Information. Harvard University Press. https://www.jstor.org/stable/j.ctt13x0hch Peitz, M. (2023). How to Apply the Self-Preferencing Prohibition in the DMA. CRC TR 224 Discussion Paper Series, Article crctr224_2023_422. https://ideas.repec.org//p/bon/boncrc/crctr224_2023_422.html Petit, N., & Teece, D. J. (2021). Innovating Big Tech firms and competition policy: Favoring dynamic over static competition. Industrial and Corporate Change, 30(5), 1168-1198. https://doi.org/10.1093/icc/dtab049 Philbeck, T., & Davis, N. (2018). The Fourth Industrial Revolution: Shaping a New Era. Journal of International Affairs, 72(1), 17-22. Portes, A. (1998). Social Capital: Its Origins and Applications in Modern Sociology. Annual Review of Sociology, 24(1), 1-24. https://doi.org/10.1146/annurev.soc.24.1.1 Prahalad, C. K., & Hamel, G. (1990, mayo 1). The Core Competence of the Corporation. Harvard Business Review. https://hbr.org/1990/05/the-core-competence-of-the-corporation Rachlitz, K. (2023). Platform Organising and Platform Organisations. 8(1). Reuver, M. de, Sørensen, C., & Basole, R. C. (2018). The Digital Platform: A Research Agenda: Journal of Information Technology. https://doi.org/10.1057/s41265-016-0033-3 Rheinhardt, A., Kreiner, G. E., Gioia, D. A., & Corley, K. G. (2018). Conducting and Publishing Rigorous Qualitative Research. En C. Cassell, A. Cunliffe, & G. Grandy, The SAGE Handbook of Qualitative Business and Management Research Methods: History and Traditions (pp. 515-531). SAGE Publications Ltd. https://doi.org/10.4135/9781526430212.n30 Richards, N. M. (2012). The Dangers of Surveillance Symposium: Privacy and Technology. Harvard Law Review, 126(7), 1934-1965. Rietveld, J., Ploog, J. N., & Nieborg, D. B. (2020). The coevolution of platform dominance and governance strategies: Effects on complementor performance outcomes. Academy of Management Discoveries. https://doi.org/10/gg3k43 Rietveld, J., & Schilling, M. A. (2020). Platform Competition: A Systematic and Interdisciplinary Review of the Literature (SSRN Scholarly Paper ID 3706452). Social Science Research Network. https://doi.org/10.2139/ssrn.3706452 Rietveld, J., Schilling, M. A., & Bellavitis, C. (2019). Platform Strategy: Managing Ecosystem Value Through Selective Promotion of Complements. Organization Science, 30(6), 1232-1251. https://doi.org/10.1287/orsc.2019.1290 Rietveld, J., Seamans, R., & Meggiorin, K. (2021). Market Orchestrators: The Effects of Certification on Platforms and Their Complementors (SSRN Scholarly Paper ID 3131553). Social Science Research Network. https://doi.org/10.2139/ssrn.3131553 Rikap, C. (2022a). Amazon: A story of accumulation through intellectual rentiership and predation. Competition & Change, 26(3-4), 436-466. https://doi.org/10.1177/1024529420932418 Rikap, C. (2022b). The expansionary strategies of intellectual monopolies: Google and the digitalization of healthcare. Economy and Society, 1-27. https://doi.org/10.1080/03085147.2022.2131271 Rikap, C., & Lundvall, B.-Å. (2021). The Digital Innovation Race: Conceptualizing the Emerging New World Order. Springer International Publishing. https://doi.org/10.1007/978-3-030-89443-6 Ringel, M., Baeza, R., Grassl, F., Manly, J., & Kennedy, D. (2019). The Most Innovative Companies 2019: The Rise of AI, Platforms, and Ecosystems (p. 22). Boston Consulting Group. Rochet, J.-C., & Tirole, J. (2002). Cooperation among Competitors: Some Economics of Payment Card Associations. The RAND Journal of Economics, 33(4), 549-570. JSTOR. https://doi.org/10/cn4h2b Rochet, J.-C., & Tirole, J. (2003). Platform competition in two-sided markets. Journal of the european economic association, 1(4), 990-1029. Rodon Modol, J., & Eaton, B. (2021). Digital infrastructure evolution as generative entrenchment: The formation of a core–periphery structure. Journal of Information Technology, 36(4), 342-364. https://doi.org/10.1177/02683962211013362 126 Essays on Power in Digital Platforms Romele, A., & Rodighiero, D. (2020). Digital Habitus or Personalization Without Personality. HUMANA.MENTE Journal of Philosophical Studies, 13(37), Article 37. Rysman, M. (2009). The Economics of Two-Sided Markets. Journal of Economic Perspectives, 23(3), 125-143. https://doi.org/10.1257/jep.23.3.125 Saadatmand, F., Lindgren, R., & Schultze, U. (2019). Configurations of platform organizations: Implications for complementor engagement. Research Policy, 48(8), 103770. https://doi.org/10.1016/j.respol.2019.03.015 Sandberg, J., & Alvesson, M. (2011). Ways of constructing research questions: Gap-spotting or problematization? Organization, 18(1), 23-44. https://doi.org/10.1177/1350508410372151 Scherer, A. G., & Neesham, C. (2020). New Challenges to Enlightenment: Why Socio-Technological Conditions Lead to Organized Immaturity and What to Do About it. SSRN Electronic Journal. https://doi.org/10/gj8mhq Scherer, A. G., Neesham, C., Schoeneborn, D., & Scholz, M. (2020). Call for Submissions Business Ethics Quarterly Special Issue on: Socio-Technological Conditions of Organized Immaturity in the Twenty-First Century. Business Ethics Quarterly, 30(3), 440-444. https://doi.org/10/gj7skk Schmalensee, R. (2002). Payment Systems and Interchange Fees. The Journal of Industrial Economics, 50(2), 103122. https://doi.org/10/fcq83w Schwab, K. (2017). The Fourth Industrial Revolution. Crown. Scott Morton, F. M., & Athey, S. https://doi.org/10.2139/ssrn.3786434 (2021). Platform Annexation. SSRN Electronic Journal. Scott Morton, F. M., & Kades, M. (2021). INTEROPERABILITY AS A COMPETITION REMEDY FOR DIGITAL NETWORKS. SSRN Electronic Journal. https://doi.org/10/gkj6gt Scott, W. R. (2008). Approaching adulthood: The maturing of institutional theory. Theory and Society, 37(5), 427. https://doi.org/10.1007/s11186-008-9067-z Seymour, R. (2019, agosto 23). The machine always wins: What drives our addiction to social media. The Guardian. http://www.theguardian.com/technology/2019/aug/23/social-media-addiction-gambling Srinivasan, A., & Venkatraman, V. N. (2020). Architectural Convergence and Platform Evolution: Empirical Test of Complementor Moves in Videogames. IEEE Transactions on Engineering Management, 67(2), 266-282. https://doi.org/10.1109/tem.2018.2881560 Srnicek, N. (2016). Platform capitalism. Theory Redux. Srnicek, N. (2021). Value, rent and platform capitalism. Work and Labour Relations in Global Platform Capitalism. https://www.elgaronline.com/view/edcoll/9781802205121/9781802205121.00009.xml Stark, D., & Pais, I. (2020). Algorithmic Management in the Platform Economy. Sociologica, 14(3), Article 3. https://doi.org/10/ghxbsr Staub, N., Haki, K., Aier, S., & Winter, R. (2023). Governance Mechanisms in Digital Platform Ecosystems: Addressing the Generativity-Control Tension. Communications of the Association for Information Systems, 52(1). https://aisel.aisnet.org/cais/vol52/iss1/9 Stigler report. (2019). Stigler Committee on Digital Platforms. 336. Stoker, G. (2018). Governance as theory: Five propositions. International Social Science Journal, 68(227-228), 15-24. https://doi.org/10.1111/issj.12189 Striphas, T. (2010). How to have culture in an algorithmic age. The Late Age of Print, 14. Tavalaei, M. M., & Cennamo, C. (2021). In search of complementarities within and across platform ecosystems: Complementors’ relative standing and performance in mobile apps ecosystems. Long Range Planning, 54(5), 101994. https://doi.org/10.1016/j.lrp.2020.101994 Teece, D. J. (2017). Dynamic Capabilities and (Digital) Platform Lifecycles. En Entrepreneurship, Innovation, and Platforms (Vol. 37, pp. 211-225). Emerald Publishing Limited. https://doi.org/10.1108/S0742332220170000037008 127 Essays on Power in Digital Platforms Teece, D. J. (2018). Profiting from innovation in the digital economy: Enabling technologies, standards, and licensing models in the wireless world. Research Policy, 47(8), 1367-1387. https://doi.org/10.1016/j.respol.2017.01.015 The Guardian. (2022, junio 9). Facebook fails test to detect violent hate speech in global ads – again. The Guardian. https://www.theguardian.com/us-news/2022/jun/09/facebook-hate-speech-test-fail-meta Thelen, K. (2018). Regulating Uber: The Politics of the Platform Economy in Europe and the United States. Perspectives on Politics, 16(4), 938-953. https://doi.org/10.1017/S1537592718001081 Thornberg, R., & Charmaz, K. (2014). Grounded Theory and Theoretical Coding. En The SAGE Handbook of Qualitative Data Analysis. SAGE Publications Ltd. https://doi.org/10.4135/9781446282243 Tiwana, A. (2014). Platform ecosystems: Aligning architecture, governance, and strategy. MK. Tiwana, A., Konsynski, B., & Bush, A. A. (2010). Research Commentary—Platform Evolution: Coevolution of Platform Architecture, Governance, and Environmental Dynamics. Information Systems Research, 21(4), 675-687. https://doi.org/10.1287/isre.1100.0323 Trittin-Ulbrich, H., Scherer, A. G., Munro, I., & Whelan, G. (2021). Exploring the dark and unexpected sides of digitalization: Toward a critical agenda. Organization, 28(1), 8-25. https://doi.org/10/ghkst8 Ulrich, K. (1995). The role of product architecture in the manufacturing firm. Research Policy, 24(3), 419-440. https://doi.org/10/d9xf6z Uzunca, B., Sharapov, D., & Tee, R. (2022). Governance rigidity, industry evolution, and value capture in platform ecosystems. Research Policy, 51(7), 104560. https://doi.org/10.1016/j.respol.2022.104560 Vaast, E., Davidson, E. J., & Mattson, T. (2013). Talking about Technology: The Emergence of a New Actor Category Through New Media. MIS Quarterly, 37(4), 1069-1092. https://doi.org/10.25300/MISQ/2013/37.4.04 Valdez, J. (2022). The politics of Uber: Infrastructural power in the United States and Europe. Regulation & Governance, n/a(n/a). https://doi.org/10.1111/rego.12456 van Doorn, N., & Chen, J. Y. (2021). Odds stacked against workers: Datafied gamification on Chinese and American food delivery platforms. Socio-Economic Review, 19(4), 1345-1367. https://doi.org/10/gk78d6 Vasudevan, R. (2022). Digital platforms: Monopoly capital through a classical-marxian lens. Cambridge Journal of Economics, 46(6), 1269-1288. https://doi.org/10.1093/cje/beac045 Verbeke, A., & Hutzschenreuter, T. (2020). The Dark Side of Digital Globalization. Academy of Management Perspectives. https://doi.org/10/ghhn8g von Hippel, E. (2006). Democratizing https://library.oapen.org/handle/20.500.12657/26093 Innovation. The MIT Press. Walker, M., Fleming, P., & Berti, M. (2021). ‘You can’t pick up a phone and talk to someone’: How algorithms function as biopower in the gig economy. Organization, 28(1), 26-43. https://doi.org/10.1177/1350508420978831 Wang, R. D., & Miller, C. D. (2020). Complementors’ engagement in an ecosystem: A study of publishers’ e-book offerings on Amazon Kindle. Strategic Management Journal, 41(1), 3-26. https://doi.org/10.1002/smj.3076 Wareham, J., Fox, P. B., & Cano Giner, J. L. (2014). Technology Ecosystem Governance. Organization Science, 25(4), 1195-1215. https://doi.org/10/gc8sdq West, J., & Wood, D. (2014). Evolving an Open Ecosystem: The Rise and Fall of the Symbian Platform. En Collaboration and Competition in Business Ecosystems (Vol. 30, pp. 27-67). Emerald Group Publishing Limited. https://doi.org/10.1108/S0742-3322(2013)0000030005 Williamson, P. J., & De Meyer, A. (2012). Ecosystem Advantage: How to Successfully Harness the Power of Partners. California Management Review, 55(1), 24-46. https://doi.org/10.1525/cmr.2012.55.1.24 Wood, A. J., & Lehdonvirta, V. (2022). Platforms Disrupting Reputation: Precarity and Recognition Struggles in the Remote Gig Economy. Sociology, 00380385221126804. https://doi.org/10.1177/00380385221126804 Wu, L., Morstatter, F., Carley, K. M., & Liu, H. (2019). Misinformation in Social Media: Definition, Manipulation, and Detection. ACM SIGKDD Explorations Newsletter, 21(2), 80-90. https://doi.org/10/ggnspw 128 Essays on Power in Digital Platforms Yoffie, D. B., & Kwak, M. (2006). With Friends Like These. Harvard Business Review, 84(9), 88-98. Yoo, Y. (2023, abril 2). Designing a Digital Value Loop [Substack newsletter]. Digital https://youngjinyoo.substack.com/p/designing-a-digital-value-loop?utm_medium=wordpress First. Young Kang, H., & Suarez, F. F. (2022). Platform Owner Entry Into Complementor Spaces Under Different Governance Modes. Journal of Management, 01492063221094759. https://doi.org/10.1177/01492063221094759 Zeng, J., Mahdi Tavalaei, M., & Khan, Z. (2021). Sharing economy platform firms and their resource orchestration approaches. Journal of Business Research, 136, 451-465. https://doi.org/10.1016/j.jbusres.2021.07.054 Zhang, L. (2020). When Platform Capitalism Meets Petty Capitalism in China: Alibaba and an Integrated Approach to Platformization. 21. Zhang, Y., Li, J., & Tong, T. W. (2022). Platform governance matters: How platform gatekeeping affects knowledge sharing among complementors. Strategic Management Journal, 43(3), 599-626. https://doi.org/10.1002/smj.3191 Zhu, F. (2019). Friends or foes? Examining platform owners’ entry into complementors’ spaces. Journal of Economics & Management Strategy, 28(1), 23-28. https://doi.org/10/gg26z5 Ziccardi, G. (2013). Resistance, liberation technology and human rights in the digital age. En Resistance, Liberation Technology and Human Rights in the Digital Age (p. 328). https://doi.org/10.1007/978-94-007-5276-4 Zittrain, J. (2009). Law and technologyThe end of the generative internet. Communications of the ACM, 52(1), 1820. https://doi.org/10/bbxcqp Zuboff, S. (2019). The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power: Barack Obama’s Books of 2019. Profile Books. 129 Essays on Power in Digital Platforms Appendix A. Glossary of Amazon’s programs and policies Name Description A+ Content Amazon Accelerator Amazon Anti-Counterfeiting Policy Amazon Brand Registry Amazon Currency Converter Amazon Custom Amazon Exclusives Program A+ Content, also known as Enhanced Brand Content (EBC), is a feature offered by Amazon to enhance product detail pages on their platform. It allows brand owners and sellers to create more engaging and visually appealing content to showcase their products and provide additional information to customers. An initiative by Amazon that partners with manufacturers and brands to create and launch products directly on its platform, offering increased visibility, datadriven insights, and promotional support. A policy designed to ensure that products sold on Amazon are authentic and genuine. Amazon prohibits the sale of counterfeit items and takes steps to remove counterfeit listings and take action against sellers who violate this policy. A program that allows brand owners to protect their intellectual property and create an accurate and trusted experience for customers on Amazon. It provides tools to help brand owners search for potentially infringing products, report violations, and manage their brand's presence on the platform. A service that allows international customers to see product prices in their local currency and make purchases. The converter automatically calculates exchange rates, enabling customers to shop in their preferred currency, with the final conversion being done at the time of purchase. A service that allows sellers to offer personalized products to customers. Through Amazon Custom, sellers can provide customization options such as engraving, printing, or embroidery, enabling customers to create unique, tailored items based on their preferences. A program where selected brands and innovators showcase their unique products, available only on Amazon or through the brand's own website and physical locations. This offers these brands a platform for increased visibility and marketing support while giving customers access to a curated collection of distinctive products. Amazon Giveaway for FBA A promotional tool that allowed sellers to set up sweepstakes-like giveaways for their products to increase product visibility and generate buzz. By using Amazon Giveaway in conjunction with Fulfillment by Amazon (FBA), sellers could easily set up contests, with Amazon handling prize fulfillment. However, as of 2019, Amazon announced the discontinuation of the Amazon Giveaway service. Amazon Imaging Services A service offered to sellers to help them obtain high-quality, professional images of their products for their Amazon listings. With this service, sellers can enhance their product listings with clear and detailed photos, helping to improve the online shopping experience for customers and potentially increase sales. Amazon Launchpad Program Amazon Outlet Amazon Partnered Carrier options Amazon Payments A program designed to assist startups and innovative brands in launching, marketing, and distributing their products on Amazon. It provides tools, expertise, and a platform for emerging brands to showcase their unique products, helping them reach a wider audience and gain traction in the market. A section on Amazon where customers can find discounted overstock items, markdowns, and clearance products from various categories. It offers a place for sellers to reduce their inventory of slow-moving products and for shoppers to find deals. A shipping program where Amazon partners with specific carriers to offer discounted shipping rates to sellers sending inventory to Amazon's fulfillment centers. By using Amazon's Partnered Carrier options, sellers can benefit from reduced shipping costs and streamlined logistics when participating in the Fulfillment by Amazon (FBA) program. A payment service that allows customers to use the payment methods stored in their Amazon account to pay for goods and services on third-party websites and applications. By integrating Amazon Payments, merchants can provide a familiar 130 Essays on Power in Digital Platforms Name Description checkout experience for Amazon customers, potentially increasing conversion rates and customer trust. Amazon Prime A subscription service offering members a range of benefits, including free twoday shipping on eligible items, unlimited streaming of movies and TV shows with Prime Video, access to Prime Music, Kindle e-book rentals, and other features. The program aims to enhance customer loyalty and encourage more frequent shopping on Amazon. Amazon Services Order Notifier A tool that notifies sellers of new orders in real-time. The Amazon Services Order Notifier provides instant alerts to help sellers manage and fulfill their orders promptly, ensuring timely delivery and improved customer satisfaction. Amazon Vine A program where select reviewers, known as Vine Voices, receive products for free in exchange for writing honest and unbiased reviews. Brands can enroll their products in the Vine program to generate early reviews, which can help potential buyers make informed decisions and improve the product's visibility on Amazon. ASIN creation rules A set of guidelines provided by Amazon for creating Amazon Standard Identification Numbers (ASINs). ASINs are unique blocks of 10 letters and/or numbers used to identify items. The rules ensure that products are correctly and uniquely identified on the platform, preventing duplicate listings and ensuring consistency in the product catalog. This helps in maintaining the integrity of the marketplace and providing a streamlined shopping experience for customers. A-to-Z Guarantee A protection offered to buyers who purchase items from third-party sellers on Amazon. If a buyer's order never arrives or if it does but is significantly different than described, the A-to-Z Guarantee provides a way for the buyer to seek a refund. It aims to enhance trust in the marketplace by ensuring a safe and secure shopping experience. Automated pricing A tool provided by Amazon that allows sellers to automatically adjust their product prices based on predefined criteria. Sellers can set pricing rules, such as matching the lowest price or undercutting competitors by a certain amount. The tool continuously monitors prices and adjusts listings accordingly, helping sellers remain competitive and maximize sales. Brand Analytics A feature available to brand owners registered with Amazon Brand Registry. Brand Analytics provides detailed insights and data about a brand's products on Amazon, including search terms, conversion rates, and competitive data. This information helps brands understand customer behaviour, optimize their listings, and make informed marketing and sales decisions. Brand Referral Bonus Buy Box A program where brand owners can earn a bonus by driving external traffic to their Amazon listings. When shoppers are directed from external sources to Amazon and make a purchase, the brand owner receives a referral bonus, effectively reducing the overall selling fees. The program incentivizes brands to promote their Amazon listings through external marketing channels. A feature on Amazon product detail pages that allows customers to quickly add items to their shopping carts. Multiple sellers can offer the same product, but only one seller at a time "wins" the Buy Box, making their offer the default for the "Add to Cart" button. Winning the Buy Box is crucial for sellers as it significantly impacts sales, and Amazon determines the winner based on factors like price, fulfillment method, and seller performance. Campaign managing & optimization A set of tools and practices offered by Amazon to help sellers and advertisers create, monitor, and refine their advertising campaigns on the platform. This includes Sponsored Products, Sponsored Brands, and other ad types. Through campaign management and optimization, sellers can set budgets, adjust bids, target specific keywords or audiences, and analyse performance data to improve the effectiveness and ROI of their advertisements. Community rules A set of guidelines and standards established by Amazon for user interactions within its community spaces, such as product reviews, discussion boards, and Q&A sections. These rules are designed to maintain a positive and respectful environment, prevent misuse, and ensure that content contributed by users is helpful and relevant to other shoppers. Violations can lead to removal of content or further actions against the user's account. 131 Essays on Power in Digital Platforms Name Description Condition Guidelines A set of standards provided by Amazon to help sellers accurately describe the condition of items they list for sale, especially for used or refurbished products. These guidelines ensure consistency and transparency in product listings, helping buyers to know exactly what to expect when making a purchase. The conditions range from "New" to various levels of used or refurbished states, each with specific criteria about the item's appearance and functionality. Coupons A promotional tool offered by Amazon that allows sellers to provide discounts on their products. Sellers can create digital coupons for specific items, which shoppers can "clip" and apply during checkout. This feature helps sellers boost visibility, drive sales, and incentivize purchases, while shoppers benefit from savings on their desired products. Create Product Page A feature on Amazon's seller central platform that allows sellers to list new products not currently found on Amazon. Sellers provide detailed information about the product, including title, brand, price, and product details, to create a unique product detail page. This ensures that customers have all the information they need to make informed purchasing decisions. Properly crafted product pages can enhance visibility and improve sales conversions. Drop shipping policy A policy set by Amazon that allows sellers to list and sell products that they do not physically have in stock. When a customer orders a drop-shipped product, the seller purchases the item from a third-party supplier, who then ships it directly to the customer. Amazon's policy requires that sellers be responsible for the entire customer experience, ensuring items are delivered on time and match the product description. Additionally, the seller must identify themselves as the seller of record and not the third-party supplier. Event Notification Service A service provided by Amazon that sends real-time notifications to sellers about specific events related to their Amazon storefront. These notifications can include updates about orders, shipments, listings, and other pertinent information. By receiving timely alerts, sellers can manage their operations more effectively and respond quickly to any changes or requirements. Environmental marketing guidelines Guidelines established by Amazon to ensure that sellers make accurate, clear, and substantiated environmental claims about their products. This includes claims related to sustainability, recyclability, energy efficiency, and other ecofriendly attributes. The guidelines aim to prevent greenwashing (misleading claims about the environmental benefits of a product) and ensure that customers receive accurate information about the environmental impact of the products they purchase. Fulfillment by Amazon (FBA) Fulfillment by Amazon (FBA) is a service where sellers store their products in Amazon's fulfillment centers. When orders are placed, Amazon handles storage, packaging, and shipping on the seller's behalf. Additionally, FBA products become eligible for Amazon Prime and other Amazon services. This allows sellers to scale their business and reach a wider customer base while outsourcing the logistics to Amazon. FBA Export FBA Inventory Placement Service FBA Labelling A feature of the Fulfillment by Amazon (FBA) program that allows sellers to offer their products to customers outside their home country. Amazon handles the international shipping, customs duties, and customer service for these orders. This enables sellers to easily expand their market reach to international customers without the complexities of cross-border logistics and customer support. A service within the Fulfillment by Amazon (FBA) program that allows sellers to ship their inventory to a single Amazon fulfillment center. Amazon then distributes the inventory across its network as needed. This can simplify the inbound shipping process for sellers, but there might be associated fees for using the service, especially if Amazon has to distribute the inventory across multiple locations. A part of the Fulfillment by Amazon (FBA) program that requires products to have specific labels to be processed and handled correctly within Amazon's fulfillment centers. Sellers can either label their products themselves according to Amazon's guidelines or, for a fee, have Amazon handle the labeling. Proper labeling ensures accurate tracking, storage, and delivery of products to customers. 132 Essays on Power in Digital Platforms Name Description FBA New Selection FBA Onsite FBA Prep Service FBA Repackaging Service A program within Fulfillment by Amazon (FBA) designed to support and incentivize sellers to launch new products on Amazon using FBA. It offers benefits such as free monthly storage, free removals, and free return processing on eligible new ASINs (Amazon Standard Identification Numbers). This program aims to reduce the initial costs and risks for sellers introducing new products to the marketplace. An initiative that combines Fulfillment by Amazon (FBA) and Seller Fulfilled Prime (SFP). With FBA Onsite, sellers store and pack their products in their own facilities, but the inventory is treated as if it's part of Amazon's own fulfillment network. This means items can be listed with Prime 2-day shipping, and Amazon handles the shipping logistics using its own transportation network. The program aims to expand Prime eligible inventory and reduce shipping times. A service offered by Amazon where they handle the preparation and packaging of a seller's inventory to meet the specific requirements of the Fulfillment by Amazon (FBA) program. This can include tasks like bagging, bubble wrapping, and labelling products. Sellers are charged a fee for this service, but it ensures that products are correctly prepared for storage and shipment from Amazon's fulfillment centres. A service within the Fulfillment by Amazon (FBA) program where Amazon repackages a seller's returned products, if deemed sellable, so they can be placed back into the seller's active inventory. This helps to reduce waste and allows sellers to resell products that are still in good condition. The service ensures that items are appropriately repackaged to meet Amazon's standards before being made available for sale again. FBA Returns Processing Fee A fee charged to Fulfillment by Amazon (FBA) sellers when Amazon processes a return on their behalf. The fee covers the handling, repackaging, and restocking of the returned product in Amazon's fulfillment center. The amount charged depends on the size and category of the item. This fee ensures that returned products are efficiently processed and made available for resale if they are in sellable condition. FBA Small and light A program within the Fulfillment by Amazon (FBA) service specifically designed for sellers who offer small and lightweight products that are typically priced under $7. By using FBA Small and Light, sellers can benefit from reduced fulfillment costs and offer fast, free shipping options to their customers. The program aims to make it more cost-effective for sellers to offer low-cost products through FBA. FBA Subscribe & Save A program that allows sellers using Fulfillment by Amazon (FBA) to offer their products as part of Amazon's Subscribe & Save service. Customers can choose to receive regular, scheduled deliveries of eligible products at a discount. Sellers benefit from increased customer loyalty and recurring sales, while customers enjoy convenience and savings on products they frequently purchase. Featured Merchant Status A designation given to certain sellers on Amazon based on their performance metrics and customer service track record. Having Featured Merchant Status can increase a seller's chances of winning the Buy Box. The criteria for achieving this status includes factors like order defect rate, shipping punctuality, and customer feedback. Being a Featured Merchant enhances a seller's reputation and can lead to increased sales on the platform. Feedback rating A system on Amazon where buyers can rate and leave feedback for sellers based on their experience with a purchase. Feedback is given on a scale of 1 to 5 stars, with 5 being the best. The feedback rating is an average of all the ratings a seller has received over the past 12 months. It provides potential buyers with an indication of the seller's reliability and service quality, influencing purchasing decisions. Fulfilling and Shipping Credits A credit system on Amazon for sellers who handle their own order fulfillment rather than using Fulfillment by Amazon (FBA). When a sale is made, sellers receive a shipping credit based on the product's category and the shipping method chosen by the buyer. This credit is meant to help offset the shipping and handling costs incurred by the seller. The actual cost to ship the item may be more or less than the provided credit, depending on various factors. 133 Essays on Power in Digital Platforms Name Description Global selling An Amazon program that allows sellers to list and sell their products across Amazon's international marketplaces. Through Global Selling, sellers can expand their business into new regions, reaching customers in countries around the world. Amazon provides tools and resources to help with listing translations, currency conversions, international shipping, and understanding local regulations and customs duties. This program enables sellers to tap into a broader customer base and grow their international sales. Headline Search Ads / Brands A type of sponsored advertisement on Amazon, primarily available for brand owners. These ads appear at the top of Amazon search results, providing high visibility for promoted products. Advertisers can create custom headlines, select a set of products to showcase, and choose relevant keywords to trigger their ads. When shoppers click on a Headline Search Ad, they are directed to a product detail page or a branded Amazon Storefront. This ad format helps brands increase product visibility and drive sales. Intellectual Property Policy for Sellers Inventory reports A policy set by Amazon to protect intellectual property rights, including patents, trademarks, copyrights, and trade secrets. Sellers are required to sell only genuine products and avoid infringing on the intellectual property rights of others. Amazon's policy provides a process for rights owners to report and request the removal of listings that they believe infringe on their intellectual property. Sellers who repeatedly violate this policy may face account suspension or other penalties. This policy aims to maintain trust in the marketplace and protect both sellers and rights owners. Inventory reports are documents provided by Amazon to sellers, offering detailed information about their inventory stored in Amazon's fulfillment centers. These reports contain data such as the quantities of each product, their storage locations, historical sales data, and other relevant information. Lighting Deals Lightning Deals are limited-time promotions offered by Amazon that allow sellers to offer their products at a discounted price for a short period of time. These deals are typically available for a few hours and are featured prominently on Amazon's Deals page. Lightning Deals can help sellers increase product visibility, drive sales, and attract more customers to their listings. Sellers need to meet certain criteria to be eligible for Lightning Deals. Listing Photos Listing photos are images of products that sellers upload to their product listings on Amazon. These photos provide visual representations of the items for sale and play a crucial role in attracting the attention of potential buyers. Amazon's guidelines specify the size, format, and quality of listing photos to ensure a consistent and appealing shopping experience for customers. Long term storage Long-term storage in the context of Amazon refers to the storage of a seller's inventory in Amazon's fulfillment centers for an extended period. Amazon charges additional fees for items that remain in storage for an extended duration, typically for items that have been in the fulfillment center for more than 365 days. Manage Your Experiments A feature enabling businesses and developers to run A/B tests on their product listings to understand the impact of different variables, such as product images or descriptions, on customer behavior and sales, with the goal of optimizing their listings for better performance. Manage your order Managing Inventory (FBA) Match Low Price Multi-channel fulfillment (FBA) A tool for third-party sellers on Amazon's platform, enabling them to oversee and process customer orders, update shipping details, handle returns, and respond to customer inquiries related to their sales. A tool for Fulfillment by Amazon (FBA) sellers to oversee and handle their inventory stored in Amazon's fulfillment centers. This includes tracking stock levels, restocking products, forecasting inventory needs, and handling stranded or unfulfillable inventory to ensure efficient operations and timely order fulfillment. A feature for sellers to automatically adjust their product prices to match the lowest price on Amazon for the same item. This helps sellers stay competitive by ensuring their items are priced in line with current market rates on the platform. A service offered to Fulfillment by Amazon (FBA) sellers where Amazon fulfills orders coming from sales channels outside of the Amazon marketplace. This 134 Essays on Power in Digital Platforms Name Description means sellers can store their inventory in Amazon's fulfillment centers, and when orders are received from other platforms or websites, Amazon handles the picking, packing, and shipping. One-Day Delivery A shipping option provided to Amazon customers that ensures the delivery of eligible items within one day of placing the order. This service aims to offer fast shipping times, increasing convenience for customers who need their purchases quickly. Availability may vary based on location and product. Opportunity Explorer A tool that provides insights and analytics for Amazon sellers to identify potential product opportunities, market trends, and customer demand within the Amazon marketplace. By analyzing sales data, reviews, and search frequency, it helps sellers make informed decisions about expanding their product range or entering new market segments. Packaging Guidelines A set of standards and recommendations provided by Amazon to sellers and manufacturers regarding how products should be packaged for shipment and storage. These guidelines ensure that items are protected during transit, are easy to open by the consumer, and minimize waste. Compliance can also impact a seller's standing with Amazon and customer reviews. Participation agreement A formal document outlining the terms and conditions that sellers must agree to when using Amazon's platform and services. It covers a range of topics including listing procedures, payment terms, intellectual property rights, dispute resolution, and other essential aspects of doing business on Amazon. Adhering to this agreement is mandatory for all sellers. Policies for bundling Popular Brand Indicator Guidelines provided by Amazon for sellers who wish to create product bundles, which are multiple complementary items sold together as a single package. These policies specify how bundles should be listed and presented, including rules about relatedness of items, listing accuracy, and packaging, to ensure clear and coherent product presentations for customers. A feature on Amazon's platform that highlights well-known and frequently purchased brands, helping customers quickly identify and trust those brands. The indicator serves to differentiate popular brands from others in search results and on product detail pages, enhancing their visibility and indicating their reputation among shoppers. Premium Shipping A set of expedited shipping options available to eligible Amazon sellers who consistently meet high performance and reliability standards. These options include Two-Day, Next Day, and Same Day shipping. Sellers who offer Premium Shipping must adhere to strict shipping and delivery performance requirements to ensure customer satisfaction. Pricing conditions A set of guidelines and terms set by Amazon that sellers must follow when setting prices for their products on the platform. These conditions aim to ensure fair competition, prevent price manipulation, and maintain customer trust. They include rules against price fixing, below-cost selling, and other pricing strategies that could mislead or harm consumers. Prime Exclusive Discount Refunds policy A promotional offer exclusively for Amazon Prime members where eligible products are offered at a reduced price. Sellers can use this feature to attract more Prime customers by offering special discounts, thereby potentially increasing sales. Products with this discount display a unique badge, differentiating them in search results and on product detail pages. Amazon's guidelines detailing the circumstances under which customers can return products and receive a refund. It specifies the conditions items must meet upon return, the time frame within which returns are accepted, and the process for issuing refunds. The policy aims to balance sellers' rights with the commitment to customer satisfaction and trust. Return and Disposal of FBA Inventory A set of procedures and options available to Fulfillment by Amazon (FBA) sellers for managing inventory that is unsellable or unwanted in Amazon's fulfillment centers. Sellers can choose to have these items returned to them or disposed of by Amazon. Seller Fulfilled Prime A program that allows eligible sellers to display the Prime badge on their listings while handling their own order fulfillment. These sellers must meet Amazon's rigorous delivery and performance standards, ensuring Prime members receive 135 Essays on Power in Digital Platforms Name Description the same benefits, like two-day shipping, they would when purchasing from Amazon's own inventory. Seller Lending Program A financing option provided by Amazon to eligible sellers on its platform. It offers short-term loans, based on sales history and other performance metrics, to help sellers grow their business, increase inventory, or address other operational needs. The loan amount and terms are determined by Amazon, and repayments are typically automated from the seller's account. Seller Offenses / Code of Conduct A set of rules and guidelines that sellers on Amazon must adhere to while operating on the platform. It outlines prohibited behaviours and practices, like listing counterfeit products, manipulating reviews, or violating intellectual property rights. Violations can result in penalties ranging from listing removals to account suspensions, ensuring a trustworthy marketplace for consumers and fair competition among sellers. Seller Performance Measurement A system by which Amazon evaluates the performance of sellers on its platform using various metrics like order defect rate, pre-fulfillment cancel rate, and late shipment rate. These metrics reflect a seller's reliability, customer service, and overall compliance with Amazon's standards. Meeting or exceeding specific benchmarks is crucial for maintaining account health and eligibility for certain Amazon programs. Seller's Best Practices Recommendations and strategies provided by Amazon to help sellers optimize their operations and listings on the platform. These practices cover various aspects, from effective product listing techniques, pricing strategies, and inventory management, to customer service guidelines. Following these best practices can enhance a seller's visibility, sales, and customer satisfaction on the platform. Selling Coach A tool that offers personalized recommendations to Amazon sellers to help them identify growth opportunities and improve their business on the platform. It provides insights, such as potential products to stock, pricing suggestions, and inventory restocking reminders, all based on sales data, market trends, and customer behaviors. Shipping expectations A set of guidelines provided by Amazon that outline the expected time frames and standards sellers should adhere to when shipping items to customers. These expectations ensure that customers receive their purchases within the promised delivery window and in good condition. Adhering to these standards is crucial for maintaining customer trust and seller reputation on the platform. Sold by Amazon A program where Amazon takes over the pricing decisions for a seller's products while guaranteeing a certain amount of payout for each sale, regardless of the sale price. This allows sellers to offer competitive prices without risking profitability. Amazon adjusts the product prices dynamically based on its algorithms. Sponsored Display An advertising solution allowing Amazon sellers and vendors to create display ads that appear both on and off Amazon, targeting specific audiences. These ads can appear on product detail pages, search results, and other sites/apps in Amazon's advertising network, aiming to drive traffic, increase brand visibility, and boost sales. Sponsored Products A pay-per-click advertising solution that helps Amazon sellers promote individual product listings. These ads appear in search results and on product pages, giving products higher visibility. Advertisers bid on keywords, and when a customer's search matches the keyword, the ad may be displayed. Sellers only pay when their ad is clicked by a shopper. Storefront A customizable page on Amazon where brands can showcase their product range, tell their brand story, and create a curated shopping experience for visitors. It provides brands with a unique URL, allowing them to drive external and internal traffic. Features like videos, images, and text can be used to enhance the visual appeal and engage customers. Tax Collection Services A service provided by Amazon to automatically calculate, collect, and remit sales tax on behalf of sellers for orders shipped to customers in jurisdictions that have enacted Marketplace Facilitator laws. This relieves individual sellers from managing sales tax collection in these areas, ensuring compliance with local tax 136 Essays on Power in Digital Platforms Name Description regulations. Unplanned Prep Services A service by Amazon where they handle any necessary preparation of inventory sent to Fulfillment by Amazon (FBA) centers that doesn't meet their inbound requirements. If inventory arrives at a fulfillment center and requires additional prep or labeling, Amazon may perform these services for a fee, which is then charged to the seller's account. Virtual Dash Buttons A digital version of Amazon's physical Dash Button. These virtual buttons allow Prime members to quickly reorder frequently purchased products directly from their Amazon home page or the Amazon app. Each button is associated with a specific product, streamlining the purchasing process and increasing convenience for repeat orders. 137