All papers
Group-Forming Networks: Relational Value and Explosive Revenues
Digital platforms have turned interaction into a primary productive input, yet much economic reasoning still treats value as the aggregate outcome of individual preferences and marginal contributions. We propose a conceptual framework in which platform economies are modelled as group-forming networks: higher-order interaction structures in which collective contexts—groups, communities, shared norms, and identities—are economically causal, not residual.
A key clarification is that “group” has two analytically distinct meanings in platform economies. The first is the relational collective: an entity constituted by interaction among its members, sustained by shared norms and mutual recognition, and capable of collective action and governance. The second is the statistical segment: a subset of users identified by the platform from behavioural data for targeting purposes, with no interaction required among members. Conflating these notions obscures the mechanisms of value creation and extraction in platform markets.
Building on this distinction, we identify two structurally different forms of superadditivity. Revenue superadditivity is empirically documented in advertising-funded platforms and is maximised by the platform’s ability to identify and monetise an explosively growing number of targetable segments. Relational superadditivity is theoretically founded in the collective action and social capital literature and is maximised by dense interaction within relational collectives, cross-group bridging, and overlapping membership. These two forms respond to architectural choices in opposite directions. Engagement-optimising interaction architectures tend to reinforce intra-group homogenisation and reduce cross-group bridging, thereby depleting relational superadditivity, while simultaneously producing more homogeneous audiences that fine-grained targeting can monetise effectively.
This divergence explains how platforms can exhibit explosive financial growth while eroding the relational resources that sustain trust, collective action, and democratic discourse. It implies (i) the need for a meso-economic layer that treats relational collectives as productive units rather than aggregates of individuals, and (ii) complementary indicators that track relational capacity and generativity, alongside standard monetisation metrics. Finally, it suggests that tools calibrated for diminishing-returns markets can misclassify superlinear regimes, motivating scale-aware diagnostics and data-access frameworks whose mandate extends beyond competition toward broader welfare measurement.