Consolidating platforms

With the rapid expansion of platforms in different parts of the world has come a giddying array of commentary and analysis. The arrival of the journal Platforms & Society promises to bring a degree of organization and integration to a fragmented debate with which nobody can keep up. Aside from wryly observing that the journal promises to provide a platform for platform studies, the emergence of this forum comes at a particular moment in the life of platforms. Let me call it a moment of consolidation, even if this process is circumstantial rather than systemic.

The pandemic has strengthened the hold of platforms on labor and life. This boost to platformization has played out unevenly, and not just in the sense that some platforms benefited while others suffered. For fifteen years or more, platform growth has occurred intensively, in the sense that it has reshaped economies and societies, and extensively, in so far as it has been operative across borders and geographical scales. In the pandemic's wake, the relations between these intensive and extensive processes are being reorganized.

Given the heterogeneity of platforms, such generalization is dangerous. Consolidation does not mean that platform economies will cease to be dynamic. New players will appear. Existing platforms will shift in and out of territories. New users will come online. Regulations will shift. Novel forms of labor organization will emerge. AI will have its role. Disruption will remain a byword, at least in management literature and the business press, if not in critical analysis.

Yet some questions and tendencies seem settled, or at least unlikely to change quickly. Singapore is Grab. Amsterdam is Uber. Deliveroo is out of Berlin. Microtaskers go to Amazon Mechanical Turk. Professionals to Upwork. Alibaba has had its wings clipped. WeChat's superapp status looks unassailable.

At stake in these seemingly consolidated scenarios is not merely the monopolizing push of large platforms, which for some commentators registers an anti-market principle that runs throughout the history of capitalism (Peck and Phillips, 2020). Nor is it simply a matter of certain platforms providing an infrastructural ground upon which others can be built, or stacked, as the jargon would have it. Rather this relative stability seems remarkable in the context of the breakneck growth and expansion of the platform economy after the 2007/8 crisis.

Nick Srnicek's Platform Capitalism (2016) observes that platforms favor growth over profit, at least initially. Revenues are channeled back into geographical expansion. Venture capital and minimal ownership of infrastructure and assets enable a drive to market dominance. Network effects combined with publicity and cross-subsidization of supply and demand are important elements of the formula. Without loss, it seems, platforms cannot take off.

What, then, does it mean when firms like Uber and Grab turn a profit, with Uber achieving this milestone in 2023 and Grab doing so in the opening quarter of 2024? Sure, these are only two platforms and their rapid expansion into new sites is not a strategy pursued by all platform firms, even if they are similarly tethered to certain geographical locations. Yet, these profits mark a tipping point that cries out for analysis, not only by those interested in the valuation of platforms but also by those concerned with their relevance for struggles of labor and capital.

Let me turn to the question of valuation. To say that platforms are implicated in financialization is almost to state the obvious. The reliance on platform growth on financial operations is clear, even if logistical modes of monitoring and coordination are evident in their workings. Despite claims for digital innovation, driven from within, platforms could not have expanded without massive injections of venture capital. Along with such financing has come an emphasis on shareholder value, for listed companies, and lean operations.

An inverse tendency is also at work here in so far as platforms alter financial operations, a change evident in fintech. In many instances, digital finance has eroded the role of banks and other financial institutions in favor of tech firms or even telecommunications providers, as Janet Roitman (2023) argues is the case in East Africa. There is an analytical and political onus to ask how these transformations re-engineer the production of value, especially in light of the vast data holdings accumulated by many platforms.

The move from loss to profit is only one aspect of platform consolidation and one that is surely reversible. Nonetheless, it attests an inflection in how the intensive impact of platforms on markets and societies interacts with their extensive operations across borders and scales. For Uber, present in 71 countries and 10,500 cities, the turning of a profit coincides with increased barriers to the company's further expansion across the world map. For Grab, it corresponds with both a near saturation of Southeast Asian markets (the company now operates across all ASEAN countries with the exception of Brunei, Laos, and Myanmar) and the enlargement of its service offerings to the point where it faces growth challenges.

In accounting terms, loss and profit are measured against return on investment. In the case of Uber, data holdings, which are among its most significant assets, are tucked away on the balance sheet under the heading of intangibles. Grab does the same. Valuation is calculated in conventional terms as stock price multiplied by the number of shares outstanding. The value of data is buried, arguably willfully discounted.

Admittedly data value is hard to grasp. Such value remains elusive both for market actors, who struggle to set prices, and for critical thinkers, who must account theoretically for how data is made into a value form.

Writing of food delivery platforms, Niels Van Doorn and Adam Badger (2020) posit a dual value production by which the monetary value produced by service provision is supplemented by the speculative value of the data produced by the associated platform activity. Platform labor generates a monetary rent whose value is set by the fee charged for the service minus the piece-rate labor costs accrued by the worker under conditions of indirect employment, algorithmic control, and more often than not racialized precarity. Significantly, this is the scene of labor exploitation in traditional Marxist analysis, although surplus value is extracted as rent rather than profit.

The analysis thus far abides by the thesis of the becoming rent of profit advanced by theorists of cognitive capitalism such as Carlo Vercellone (2013). But what happens when platform companies begin to post a profit, an advent that in most discussions of platform capitalism is deferred indefinitely to the future? Clearly, this does not imply a return to the industrial setting of “free” wage labor that animates the standard Marxist analysis. Nor does the claim for monetary rent extraction need to be abandoned. Yet the terms of analysis need to shift.

The question of data value is paramount in this regard. Even if we accept that platforms extract data, there is still a need to understand how they produce value from that data in the wake of its extraction. Clearly, one path of value production for platform data is its application to the algorithmic control of labor. However, not all platforms organize the buying and selling of labor or the commodification of the labor relation.

Platform data is productive of value in a speculative mode. It attracts capital investment in so far as it promises system optimization and future opportunities for application. In this sense, data value is performative. Its value lies in an ability to scale or to reach a threshold of accumulation in which its potential to produce more value can expand, not least through the use of machine learning techniques. In this respect, the data value is highly elastic and even indeterminate.

Katharina Pistor (2020) observes that the value of data lies not in its exchange value but in the power conferred on those who control it. In this sense, the business of data is not about markets but about hierarchy. It needs to attract a political analysis rather than one based on technicalities of pricing or transaction costs. Data provides a means of governing others on a scale that rivals the law of nations.

The consolidation of platforms needs to be rethought in light of Benjamin H. Bratton's (2015) claim that they offer a third institutional form, neither state nor market. In this respect, their ability to concentrate power and produce value across territories is relevant. Without their initial growth and often wild global expansion, platforms would not have a ground on which to consolidate. Nonetheless, in such a heterogeneous domain, consolidation always encounters contingency and setbacks. What seems clear is that platformization is by now an ingrained process; one whose resistance will need to be organized politically, encompassing labor action and struggles over data governance but by no means restricted to these spheres of intervention.