The identification of influencers in socio-economic systems and their application to marketing has attracted enormous attention from scholars and companies. Conventional influencers identification typically relies either on simplified models of information diffusion on social networks, or on network metrics of individuals’ centrality. In this project, we take a different perspective, and use statistical procedures to identify those individuals, called discoverers, who are persistently among the first consumers to make purchases in shops that later become popular. Importantly, our techniques do not rely on simplified models of information dynamics, and the identified discoverers may allow for popularity predictions for recent shops with little past information. We will also compare the discoverers with the influencers as identified by conventional network centrality metrics. The metrics that will be developed in this project could be useful for target marketing and trend prediction with little past information.
Author: Manuel Mariani