Answers are grounded in this paper and cite their sources.
Can we effectively predict top-performing products, services, or businesses through a small set of adopters? Most previous studies focused on the effects and predictive power of the adoptions by central individuals in social networks, but the resulting insights are typically not applicable without relevant network data. However, informative adopters may still be detectable even when social ties are not observed. Here we examine the predictive power of selected adopter groups detected without network data in large-scale datasets from diverse domains: an e-commerce retailer, Yelp, Ethereum, and user-generated content online communities. Despite the differences among the studied domains, we consistently detect a small set of ``discoverer" adopters who offer stronger and more reliable out-of-sample predictive signals than previously-studied adopter groups. Further, the discoverers' early adoptions can improve the performance of statistical-learning models for success prediction. We explore the implications of the found regularities through a two-stage laboratory experiment, whose results indicate that informing the discoverers first on new products generates a larger success difference between high-quality and low-quality products. Based on a multimethod approach, our findings indicate that a small group of discoverers can offer informative early signals of success, and their early awareness of new products might influence the relative success of competing products.
No takes yet. Share an insight, caveat, or question.
Mariani et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: