Gender bias in high stakes pitching: an NLP approach

Investors use heuristics and biases which may disproportionately impact entrepreneurial teams with women when hearing pitches and evaluating early-stage ventures. However, merely entertaining a pitch is not enough—the tone of the conversation also matters. To investigate, we explore non-rationality in the funding process by applying Natural Language Processing (NLP) to Shark Tank: a high-impact television program fosr entrepreneurial pitching. Using sentiment analysis, we show that male judges react more positively to pitches from entrepreneurial teams with women. Importantly, these positive reactions are not indicative of increased deal flow. The opposite is true for female judges who, while no more likely to react positively to teams with female entrepreneurs, are significantly more likely to ink a deal when they do. By showing how non-rational thinking and biases impact both the pitch narrative and the likelihood of securing funding, our findings have important implications regarding heuristic formation and gender bias in early-stage investment.