Network Structure and Collective Intelligence in the Diffusion of Innovation

Abstract

When multiple innovations compete for adoption, historical chance leading to early advantage can generate lock-in effects that allow suboptimal innovations to succeed at the expense of superior alternatives. Research on the diffusion of innovafacetion has identified many possible sources of early advantage, but these mechanisms can benefit both optimal and suboptimal innovations. This paper moves beyond chance-as-explanation to identify structural principles that systematically impact the likelihood that the optimal strategy will spread. A formal model of innovation diffusion shows that the network structure of organizational relationships can systematically impact the likelihood that widely adopted innovations will be payoff optimal. Building on prior diffusion research, this paper focuses on the role of central actors i.e. well-connected people or firms. While contagion models of diffusion highlight the benefits of central actors for spreading innovations further and faster, the present analysis reveals a dark side to this influence: the mere presence of central actors in a network increases rates of adoption but also increases the likelihood of suboptimal outcomes. This effect, however, does not represent a speed-optimality tradeoff, as dense networks are both fast and optimal. This finding is consistent with related research showing that network centralization undermines collective intelligence.