Platforms have, over the past decade, called into question established socio-economic (b)orders, dynamics, and modes of conduct in different parts of the world, leading to proclamations about the rise of the “platform economy” and “platform capitalism.” Yet while platform businesses are certainly determined to dominate their market, if not the entire world, platformization is often a messy and nondeterministic process. If we follow the definition of platformization as “the penetration of infrastructures, economic processes and governmental frameworks of digital platforms in different economic sectors and spheres of life, as well as the reorganisation of cultural practices and imaginations around these platforms” (Poell, Nieborg & Van Dijck 2019: 1), we should keep in mind such “penetration” is bound to be partial, uneven, and incomplete.

Simply put, things often do not go as planned and, because processes of platformization are fundamentally relational, platforms regularly struggle to overcome their dependencies. This struggle to overcome situated dependencies is what sets one platform apart from another, despite sharing certain qualities. It is also what sets a platform apart from itself, as it splits into distinct (yet still centrally governed) operating entities—each with its own terms and conditions, supported by customized user interfaces—to meet the distinct requirements of particular jurisdictions.

Platformization, as such, is never monolithic. Rather, it is an “actually existing” process (Brenner and Theodore 2002; Peck, Theodore, and Brenner 2009: 52) that is institutionally embedded in national and urban settings. But how does a platform become embedded in specific social, institutional, and technological environments? How does it adapt to the needs and demands of a city, a community, a user group, or other stakeholders, while still safeguarding its business interests? In other words, how does it achieve a satisfactory level of interoperability between its own operations and infrastructures, and the legacy systems of the “economic sectors and spheres of life” it aims to reorganize? The Platform Labor team recently introduced the notion of “actually existing platformization” to help formulate some provisional answers to these questions (Van Doorn, Mos & Bosma 2021).

On June 8 2021, the Platform Labor team¹ organised a workshop to elaborate on this notion’s utility and shortcomings, inviting a small group of platform scholars to critically engage with its theoretical and methodological underpinnings². This blog post presents a selection of the proceedings from this workshop, as well as some additional reflections from our team in response to certain questions raised. First, we report on how the participants assessed the conceptual scope and merits of the “actually existing platformization” (AEP) concept. We then turn to a discussion of how AEP may contribute to a rethinking of the spatial dimensions of platformization, before closing with a set of reflections on AEP’s methodological aspects i.e. the way it informs how we study platforms and platformization.

Assessing “Actually Existing Platformization” as a conceptual framework

Coining the term “actually existing platformization” quickly raises the question of what this “actually existing” adjective is contrasted with or opposed to. As Thomas Poell asked, is there also something like “not actually existing platformization”? In defining the concept of actually existing neoliberalism, Brenner and Theodore (2002) posit the AE framework in contrast to neoliberal ideology—"neoliberalism as an ideological project" (2002: p. 352)—"in which market forces are assumed to operate according to immutable laws no matter where they are ‘unleashed’” (p. 351). As Brenner and Theodore note, there is considerable agreement on the basic elements of neoliberalism, however, there is also "a rather blatant disjuncture between the ideology of neoliberalism and its everyday political operations and societal effects" (p. 352).

Importantly, these disjunctures—"between ideology and practice; doctrine and reality; vision and consequence—are not merely accidental side effects...Rather, they are among its most essential features" (p. 353). Likewise, understanding such disjunctures as constitutive of the platform economy allows for the development of new and more nuanced perspectives on how platforms navigate the material and discursive settings in which they are embedded— perspectives that are usually absent in “disruption”-centric narratives.

Accordingly, if the AEP notion, or framework, would have to be contrasted with anything it would be “ideal type” discourses that tend to privilege disruptive change and overarching logics. Two such ideal type discourses on platforms can be identified. First, there is the textbook “management model” (which John Stehlin during the workshop defined as "the Harvard Business School model"), elaborated by management and organization scholars such as Geoffrey Parker, Annabelle Gawer, Micheal Cusomano, and others. This approach aims to reveal the main principles and strategies of successful platform businesses and its theorists tend to agree on a set of prevalent components such as network effects, asset-light scaling, and the production of “matches” in a multi sided market.

Second, as Juliet Schor noted, there is the ideological discourse that platform companies themselves contribute to for public relations purposes, mostly pertaining to their social and economic benefits. Although this discourse has increasingly been challenged by counter-narratives detailing widespread worker exploitation, community destruction, and carbon footprints, platforms continue to operate by re-adapting both their narratives and business models, thereby having to negotiate the disjuncture between ideology and practice.

So rather than positioning AEP in opposition to these ideal type accounts, let alone thinking of them in terms of “not actually existing platformization,” we are more interested in using this framework as an “analytical strategy” (to use Gernot Grabher’s term) that examines how platforms deal with the fissures between their ideological visions and their practical consequences in different markets. Like Kathleen Thelen, we see the potential contribution of AEP in how it frames platformization as a diverse, adaptable, and experimental process. This involves shifting the attention to aspects of platformization that are too frequently left out of focus: how it is, for instance, a path-dependent process with contingent outcomes and involves trial and (importantly) error.

If it is true that “platforms are what platforms do,” as Benjamin Bratton has written, then it is equally true that they are not always what they say they are and they should thus be judged by their situated actions rather than their selective representations. As John Stehlin noticed, this is not an easy task given how the divergence between a platform’s Crunchbase profile or promotional materials and its actual success as a business often does not become visible until an IPO. There are nevertheless alternative ways to approach the research of platform companies and their operations, which are further discussed in the third section.

The variegated spaces of AEP

As noted above, one of the aspects we want to draw attention to in the AEP framework is how platformization is a path-dependent process that is differentially embedded in national, regional, and urban settings. Platformization necessarily happens in space and time. The second session of the workshop thus turned to these dimensions, questioning in particular what the concrete and abstract spaces of AEP are, how they take shape, and where we should look for them. Going beyond geographical space, the discussion revolved around the question of how platformization becomes embedded in institutional, regulatory, and social spaces as well.

Importantly, embeddedness is not a static situation or singular accomplishment, as several participants noted during the workshop. It is thus more fruitful, as John Stehlin and Kathleen Thelen pointed out, to think about embedding instead of embeddedness: AEP involves an ongoing process of embedding, disembedding, and re-embedding. In this process it is not only platforms that are subject to change; their (social and regulatory) environment is usually reconfigured as well. For this reason, Rachel Phillips argued for the value of a “conjunctural analysis” that sees platforms as windows into broader political questions and dynamics of capital accumulation. A conjunctural analysis of how labor platforms become embedded in and across local markets, for instance, may lead to a more expansive critique of labor’s progressive degradation, segmentation, and devaluation.

Yet, as Noopur Raval noted, we should not limit our critique of platformization to the realm of platform-mediated labor but instead extend the scope of our inquiries to investigate modes of platform living (as she excellently does in her doctoral thesis). This involves attending to the un(der)examined or “hidden spaces” of platformization and asking how platforms insert themselves into the daily lives of platform workers and their communities, whereby they paradoxically tend to disappear from view. Platforms like Uber and Zomato commonly first intervene on the scale of people’s everyday struggles to secure a livelihood, in order to then pursue embeddedness on urban, regional, and national regulatory scales.

Tracing processes of (dis/re)embedding necessarily alerts us to how differences across these scales and between concrete sites play an important role in shaping how certain objectives and outcomes of platformization materialize. Accordingly, as Lin Zhang stressed, Chinese rural areas should be treated as sites of theorizing platformization in their own right and not just as deviating “cases” of a “general” or presumed platformization trajectory that ostensibly happens in cities across the Global North (or even the Global South). Similarly, nation states should not be essentialized but need to be seen as institutional assemblages that evolve over time. For example, comparing contemporary China with the EU or US might suggest that the Chinese state is interventionist, but this misses how China took a much more laissez-faire approach in the early days of platforms like Alibaba—one in which regional differentiation was key.

Increasingly, platform companies also seek to embed themselves in supranational regulatory spaces, such as the European Union, in order to improve their position in specific urban markets, but also to prevent regulations that would close profitable tax loopholes maintained by specific member states. The success of such efforts will, in turn, (co-)determine the extent to which these platforms are able to influence the daily lives and life chances of their users as well as platform-adjacent communities.

Across the Atlantic, Kathleen Thelen’s work on the regulation of Uber in the US has shown how the country’s fragmentation of its national political economy into state-level jurisdictions has provided a space of opportunity for regulatory arbitrage. By stimulating competition between jurisdictions, Uber was able to actively shape new regulatory spaces. Notably, however, the app-based mobilization of drivers and customers in Uber’s campaigns of regulatory activism was vital to its strategy, pointing to the multiscalar nature of projects that seek to achieve regulatory change.

Mark Graham has recently argued that platform power is achieved by “being simultaneously embedded and disembedded from the space-times they mediate.” Based on our conversations in this session, we would further add that the power of platforms is rooted in their ability to strategically differentiate and move between the scales and sites where processes of becoming (dis)embedded take place.

While, as Thomas Poell pointed out, platforms like UpWork or YouTube are less (obviously) embedded in physical spaces and consequently less dependent on local regulators, we argue that the labor sustaining these platforms still forces their parent companies to navigate the necessary legal frameworks and establish the supportive resources required to sustain and expand this labor force. These practices continue to ensure forms of institutional embeddedness that are increasingly recognized and become more visible once you know where to look for them, which brings us to our final theme: how to study AEP?

The methodological opportunities and challenges of AEP

The third and final session of the workshop inquired into the methodological questions raised by the AEP framework, or research approach. The participants reflected on the methods that could productively be deployed for studying AEP, how our epistemological assumptions and social positionality shape our case selection strategies, as well as the methodological challenges of transposing the “actually existing” framework to the study of platforms.

One of the recurring questions that surfaced during the discussion was the extent to which data about (and provided by) the platforms themselves was central, or even necessary, for studying processes of platformization. AEP highlights the necessity of understanding the empirically observable practices of platformization, but these efforts are significantly thwarted by the tendency of platforms to evade scrutiny and conceal their activities “on the ground.” As Koen Frenken pointed out, non-publicly traded platform companies in particular are purposefully opaque—their data and algorithms are inaccessible, they seldom provide interviews to journalists and researchers, and the data sets that they do release are frequently filtered and redacted.

As prior research has shown, this opacity is one of the ways in which platforms gain and maintain their power. It is consequently difficult, if not impossible, to employ the same methods for studying platform companies as have been used for the study of conventional firms. Focusing on contextually specific and materially embedded forms of platformization necessitates charting alternative methodologies that are able to circumvent the many efforts of platforms to enforce inscrutability.

One possible avenue for doing so, argued Anne Helmond, was to study platformization as a process of expansion. Rather than focusing on the inner workings of platforms themselves, tracing the ways in which platforms form connections (e.g. through APIs) and partnerships (e.g. through mergers and acquisitions), and/or documenting how they evolve over time (drawing on the Wayback Machine, for example) are all methods capable of offering us rich new insights. Koen Frenken echoed this sentiment, pointing to the benefits of adopting a “distanced” view that methodologically departs from the institutional environment in which platforms are embedded in order to discern their operations, rather than from the platforms themselves.

While digital archives and online sources do yield valuable insights, they nevertheless tend to offer a limited view into the core business practices of platforms and the key ways in which they exert monopoly power, as Gernot Grabher noted. For instance, it is very difficult to trace Amazon’s anticompetitive practices and its reliance on parity teams to crowd out third-party sellers by using a distanced/outside view or drawing on archival sources. We need other methods to scrutinize the monopoly tendencies of platforms while being aware of each method’s blind spots.

While having its own limitations, ethnography represents another productive avenue for studying actually existing forms of platformization, as Rida Qadri argued. Echoing Raval and Van Doorn, she noted that ethnographic research can offer nuanced insights into the rich social relations and practices that develop around platforms, by approaching gig workers as “access points” for examining platform-adjacent lives and livelihoods. AEP may in this context represent a cue for a methodological reorientation in the study of platforms—away from an effort to capture their (largely impenetrable) “inner workings” and dynamics, towards an examination of their situated impacts and modes of becoming embedded through human relationships. (For a highly evocative piece on alternative methods for studying platforms that informed our discussion, see Fields et al. 2020.)

At the same time, AEP also raises a conceptual question with methodological repercussions: can the “actually existing” approach be completely and coherently transposed to the study of platformization? While both AEP and AEN are concerned with articulating the shape-shifting and heterogeneous quality of platformization and neoliberalization, respectively, Jamie Peck pointed to the substantive difference in the processes being studied. While “actually existing neoliberalization” is concerned with understanding the pathways of market-led restructuring of global capitalism, and the different forms that this restructuring takes in different places, “actually existing platformization” seeks to understand the context-specific strategies that platform businesses adopt and the variegated ways in which they embed and expand themselves in different geo-institutional frameworks.

In this sense, as Peck pointed out, AEP seems to ask different questions and thus provokes a different set of answers. Methodologically, this has implications because studying the ways in which platforms seek to embed themselves into diverse contexts requires a different set of methods and analytical tools compared to the ones deployed in the study of market-driven socio-spatial transformations at the core of neoliberalism. As suggested above, these methods and tools will have to be geared toward a sustained interrogation of complex and strategically opaque sociotechnical systems in which algorithmic modes of control and datafication are key to business success. Moreover, as John Stehlin added, strategic opacity extends to how platform companies attract capital, from private equity rounds to the rise of SPACS as an alternative to an IPO. What methods are best equipped to illuminate such developments, which shape the conditions of possibility for platformization on a global scale?

Concluding remarks

“Actually existing platformization,” as articulated in our recent contribution, highlights the importance of zooming in on actors and practices that conventionally have not been readily associated with platformization, thereby seeking to draw attention to platformization as a socio-spatially variegated and institutionally embedded process. While the original article focused primarily on how platforms strategically deploy partnership as a boundary resource to embed themselves in local settings, the workshop discussions drew attention to a broader range of “actually existing” manifestations of platformization. These discussions highlighted the many incongruities between ideal-type understandings of platforms and their real-world materializations, while also questioning some of our team’s assumptions about AEP’s tenets, scope, and potential.

Ultimately, the key themes that emerged from the workshop discussions were (1) the complementary modes of (dis)embeddedness that platforms pursue, (2) the institutional and geographical path-dependency of platformization, and (3) its situated, multi-scalar impacts. By examining these themes in detail, the AEP framework can play an important role in charting new, underexplored territories in platform research. We believe that the key potential of AEP lies in its ability to inform research that seeks to overcome monolithic, reductionist and technologically deterministic conceptualizations of platform power, instead redirecting scholarly attention to the often contradictory, complex, and multifaceted trajectories that give rise to platformization.

At the same time, the workshop indicated the need to further develop AEP analytically, and to build on the initial insights presented in the article. Beyond signaling the need to take the local embeddedness and contextual specificity of platforms seriously, how does AEP help us to better conceptualize platforms and platformization? What are the concrete platform practices that AEP can spotlight in a way other approaches cannot? Conceptually refining AEP should entail specifying what we understand by platformization in a more general sense, before speaking about the particular forms that it takes in different contexts. This requires that we do not treat it as an overly expansive, ubiquitous process. As Gernot Grabher asked: what are the boundaries of or limits to platformization?

Future work should therefore also guide us toward sharpening AEP methodologically and defining the objects of AEP research. A growing body of research has already shed light on various situated impacts of platforms—from the ways in which platform companies adapt to different regulatory frameworks to the local diversity of platform working conditions—and it is from these contributions that we will continue to take cues as we move forward. There is, for instance, much to be learned from recent research on platformization in the Global South, particularly its attention to locally specific institutional dynamics and social practices. This frequently detailed ethnographic research can also usefully complement political economic analyses that chart the development of broader platform ecosystems and their evolving regulatory landscapes.

Near the end of our workshop, someone humorously summarized the discussion as follows: “Platformization, it’s complicated.” This statement at once reflected the Zoom-inflected mood at this point and pointed to the necessity of an ambitious research agenda. We want to warmly thank all our workshop participants for generously and critically engaging with our paper and the proposed AEP concept/framework. We look forward to continuing the dialogue in writing and hopefully also in the form of a future event.

--