Reading passage
A growing share of commerce now occurs on digital platforms, intermediaries that do not primarily produce goods themselves but connect distinct groups whose interaction generates value. Operating systems link software developers with users, online marketplaces link sellers with buyers, and ride-hailing applications link drivers with passengers. The economic logic of such businesses differs fundamentally from the linear supply chains of industrial capitalism, and understanding that difference has become essential for scholars, regulators, and entrepreneurs alike.
The defining feature of platform markets is the network effect: the value of the service to each participant rises with the number of participants on the other side, or sometimes on the same side, of the market. A payment card is useful only if shops accept it, and shops accept it only if customers carry it. Direct network effects arise among similar users, as with telephones or messaging services, while indirect effects connect complementary groups, as when more console buyers attract more game developers, whose titles in turn attract more buyers. These feedback loops produce winner-take-most dynamics, in which an early lead can harden into durable dominance.
Because network effects reward scale, platforms pursue growth with strategies that puzzle observers trained on conventional firms. Many subsidise one side of the market entirely: search engines and social networks charge their users nothing and earn revenue from advertisers, while video platforms pay selected creators to attract audiences whose attention is then sold. Pricing below cost on one side is not necessarily predatory in the traditional sense but can be a rational investment in the participation that makes the other side valuable. Determining which side to subsidise, and by how much, constitutes a core design decision that economists call the platform's pricing structure, a matter distinct from the overall price level.
The same logic explains the aggressive cultivation of ecosystems. By opening application programming interfaces, a platform permits outsiders to build complementary products that deepen user attachment at little cost to the platform itself. Yet the relationship is ambivalent: complementors depend on the platform's rules, which can change without consultation, and successful third-party innovations are sometimes absorbed into the core service, a practice that critics label appropriation and defenders call integration. The boundary between platform and complement thus shifts continuously, negotiated through technical standards and contractual terms rather than through prices alone.
Multi-sided markets also complicate competition law written for single-sided industries. Conventional tests ask whether a firm charges excessive prices, but a platform that charges users nothing cannot easily be accused of overcharging, even while it extracts vast quantities of personal data. Harms may instead appear on the unattended side of the market, in degraded quality, reduced privacy, or foreclosed innovation. Antitrust authorities have consequently begun to weigh data accumulation, switching costs, and the acquisition of potential rivals before those rivals mature, though courts remain cautious about condemning conduct whose competitive consequences are genuinely ambiguous.
Labour platforms illustrate further tensions. By classifying couriers and drivers as independent contractors rather than employees, gig-economy firms avoid the obligations attached to employment while retaining detailed algorithmic control over assignments, prices, and discipline. Supporters emphasise flexibility and low barriers to entry; critics counter that algorithms replicate managerial authority without managerial responsibility, and that apparent autonomy conceals economic dependence. Courts in several jurisdictions have ordered reclassification, prompting platforms to experiment with hybrid arrangements whose long-run viability remains uncertain.
Two further strategies set limits on even dominant intermediaries. Envelopment occurs when one platform enters another's market by exploiting an overlapping user base, swallowing the smaller service into its own bundle. Disintermediation arises when transacting parties, having found one another through the platform, bypass its fee by dealing directly. Meanwhile, advocates of decentralised protocols propose distributing platform rents among participants rather than concentrating them in corporate owners, though no such system has yet matched the convenience of the centralised incumbents at scale.
What seems clear is that platforms are not merely firms with websites but a distinct institutional form, part marketplace, part regulator, and part infrastructure. Their rules govern access to economic opportunity for millions of small businesses and workers, granting private entities powers once associated with public authorities. How societies reconcile the efficiency of orchestrated markets with the accountability of the orchestrators will shape the digital economy for decades to come.