Collective dynamics behind success

TL;DR

This article synthesizes cross-domain research on the collective dynamics of success, arguing that success emerges from interactions across individuals, teams, and organizations. It organizes the literature around four interrelated effect types—feedback, temporal, network, and identity—and examines how these dynamics play out in creation and reception processes, including mechanisms like the Matthew effect, fitness-driven growth, aging, diffusion, and network contagion. Key findings highlight when success begets more success or when failures can catalyze future gains, the nuanced roles of aging and hot streaks in creativity and careers, and how network structure and identity shape disparities in recognition and opportunity. The work stresses the need for interventions and responsible algorithmic design to curb inequalities, while promoting future research that integrates culture, inequality, experimental causality, and societal values into models of collective success.

Abstract

Understanding the collective dynamics behind the success of ideas, products, behaviors, and social actors is critical for decision-making across diverse contexts, including hiring, funding, career choices, and the design of interventions for social change. Methodological advances and the increasing availability of big data now allow for a broader and deeper understanding of the key facets of success. Recent studies unveil regularities beneath the collective dynamics of success, pinpoint underlying mechanisms, and even enable predictions of success across diverse domains, including science, technology, business, and the arts. However, this research also uncovers troubling biases that challenge meritocratic views of success. This review synthesizes the growing, cross-disciplinary literature on the collective dynamics behind success and calls for further research on cultural influences, the origins of inequalities, the role of algorithms in perpetuating them, and experimental methods to further probe causal mechanisms behind success. Ultimately, these efforts may help to better align success with desired societal values.

Figures (6)

• Figure 1: Success in diverse domains and at different scales. The recent literature on success has examined subjects across disparate domains and at different scales. From bottom to top, the subjects involved can be classified as individuals, teams, or large-scale organizations; these subjects create, adopt, or endorse ideas, products, and behaviors. Recent studies on success often link subjects within scales (solid lines) and across scales (dashed lines). This review presents several instances in which these connections predict or influence individuals', teams', and organizations' success. For example, the network of interactions among potential adopters affects the successful proliferation of new ideas, products, and behaviors; the network of interactions between a team’s members affects the team’s success; and artists' and scientists' early connections with prestigious institutions predict their success. Each gray box provides examples of individuals, teams, and organizations of interest at different scales for a given domain (from left to right: arts, business, online systems, science), as well as examples of the datasets used in the analysis (blue boxes). Icon credits: The icons representing organizations and teams have been realized by user Freepik from Flaticon.com; The icons representing individuals and products have been realized by users Kiranshastry and bsd, respectively, from Flaticon.com.
• Figure 2: Types of effects influencing the dynamics behind success for creation and reception processes. This review explores regularities, mechanisms, and predictive signals in both creation processes (through which individuals, teams, and organizations acquire resources and create new ideas, products, or behaviors, gray boxes) and reception processes (through which a completed product or well-defined behavior is potentially adopted or endorsed by interconnected individuals, teams, and organizations, blue boxes). We identify four interrelated types of effects that influence the collective dynamics behind success for both creation and reception processes: feedback, temporal, network, and identity effects. The boxes highlight key themes found in the literature for each effect type.
• Figure 3: Fitness and success-breeds-success. The concept of fitness was borrowed from ecology to capture the intrinsic appeal of an idea, product, or behavior. Empirical data suggest that fitness typically follows bounded distributions (e.g., exponential kong2008experiencemedo2011temporal, see panel a for an illustration). In a hypothetical world in which recognition is purely determined by fitness, success would also follow a bounded distribution (panel b, red line). However, mechanistic models of success dynamics indicate that even when fitness follows a bounded distribution, success-breeds-success mechanisms can generate fat-tailed success distributions (panel b, blue line) where a few products accrue most endorsements, and many others remain unnoticed. There is ample empirical evidence of fat-tailed success distributions, from wealth and income to scientific citations, from online popularity to book sales, and more (see main text). Here we illustrate: (c) the distribution of and the total number of retweets received by a Twitter user in response to the experimental discovery of the Higgs boson [dataset made publicly available by de2013anatomy]; (d) the number of reviews received by a business in Yelp [dataset made publicly available by yelp]; (e) the number of citations received by a paper published in American Physical Society (APS) journals [dataset made available by the APS aps, version analyzed in mariani2016identification].
• Figure 4: Early signals of the failure-to-success transition. For successful grant applicants in the NIH data (blue line), the average inter-event time between two consecutive failures, $T_n$, is a decreasing function of the number of early failures, $n$, as predicted by Wright's law. This means that it takes less time for the applicant to formulate a new attempt, i.e., the applicant is in the progression regime. Unsuccessful applicants, however, do not follow Wright's law (orange line). The significant difference between the two lines can be leveraged to make accurate success vs. failure predictions in science and entrepreneurship. Data from yin2019quantifying.
• Figure 5: Which network structure favors the large-scale adoption of a new product or behavior? The right-side, low-diameter network features long ties that connect otherwise distant nodes (e.g., the red tie). The presence of more long ties is associated with a lower diameter (i.e., a smaller average distance between the nodes, which implies the small-world property barabasi2016network) and lower clustering (i.e., lower tendency of an individual's social contacts to be connected). By contrast, long ties are absent in the left-side, high-diameter network. Our intuition and the properties of epidemic spreading processes watts1998collective suggest that an idea, product, or behavior would spread faster and farther in the right-side network. While this intuition is correct for simple social contagions, it is not supported for complex social contagions, i.e., adoption processes in which multiple exposures to a new behavior or product are required before an individual adopts. Complex contagions, such as the adoption of health-related behaviors, spread more effectively in the left-side network, as demonstrated both theoretically centola2007complexguilbeault2021topological and experimentally centola2010spread. Data from [Centola, D. The spread of behavior in an online social network experiment. Science 329, 1194–1197 (2010). DOI: 10.1126/science.1185231].