Predicting success in the worldwide start-up network

Authors: ['Moreno Bonaventura', 'Valerio Ciotti', 'Pietro Panzarasa', 'Silvia Liverani', 'Lucas Lacasa', 'Vito Latora']

Year: 2020

Methodology

Sample: Large-scale (worldwide network of start-ups)
Design: longitudinal
Data: Online professional relationship data (Crunchbase/LinkedIn style data implied), Employee flow data, Venture capital performance benchmarks

Network Centrality (Position in the ecosystem) [anecdotal] — Doubling the state-of-the-art performance of VC funds in prediction accuracy

Flow of employees (Know-how transfer) [anecdotal] — Not explicitly quantified in abstract, used as the link weight/basis for network construction

Early-stage ecosystem position [anecdotal] — Predictive of long-term positive economic performance

The position of a start-up within the global network of professional relationships (employee flows) has significant predictive power for its long-term economic success.
Using network centrality measures can potentially double the predictive performance of traditional venture capital screening processes.
The transfer of know-how through employee movement across companies creates a 'start-up network' that serves as an effective leading indicator of success.

Liability of newness: The inherent uncertainty of disruptive products and uncharted markets makes any prediction difficult.
Sunk costs and rapidly changing technological regimes can disrupt historical network patterns.
The study relies on large-scale online data which may have inherent reporting biases or lags.

Extracted by lib/ingest/literature_review.py via gemini-flash