This study provides the first large-scale analysis of face-based impression factors in venture capital. Using machine learning to extract Trustworthiness, Dominance, and Attractiveness factors from founder photos across 57,460 U.S. startups, we find that these factors significantly predict initial VC funding decisions, with magnitudes comparable to founder education and prior experience. Their relative importance varies by founder gender, team composition, industry, and VC experience. The factors persist as predictors of long-term exit outcomes, suggesting that facial cues function as informative signals of venture success, even among experienced investors.
Liudmila Alekseeva, KU Leuven
Silvia Dalla Fontana, ESCP Business School
Caroline Genc, Michigan State University Eli Broad College of Business
Lin Peng, Zicklin, City University of New York, Baruch College School of Business