Pivoting and Performance of “Scientific” Entrepreneurs: Evidence from a Field Experiment

Research Seminars
Academic Areas Strategy
Alfonso Gambardella, Professor in the Department of Management & Technology, Bocconi University & ICRIOS, Milan, and CEPR, LondonProfessor in the Department of Management & Technology, Bocconi University & ICRIOS, Milan, and CEPR, London
April 18, 2019 | 12:00 PM - 1:30 PM | Thursday
AC 2 MLT, Level - 2, Hyderabad, India
Open to Public
This paper argues that if entrepreneurs formulate their problems clearly, develop theories about the implications of their actions, and test these theories, they make better decisions. In our framework, entrepreneurs who adopt this scientific approach enjoy two levels of precision. First-order precision enables them to estimate a smaller variance of the value of the business idea than entrepreneurs who do not adopt this approach. This generates more conservative predictions of the value of business ideas because negative outcomes are bounded by the option to close the firm, while entrepreneurs with lower variance put lower probability weight on positive outcomes distant from the expected value. Second-order precision enables entrepreneurs to select more successful ideas. In particular, it indicates better ideas where the firm can pivot to. The scientific approach then has an ambiguous effect on the decision to close the firm: while first-order precision makes entrepreneurs more conservative, second-order precision encourages them to stay in business and pivot to a more valuable idea. The scientific approach also makes it more likely that if entrepreneurs pivot, they pivot fewer times: second-order precision reduces the benefit of additional pivots because entrepreneurs already reach relatively high outcomes with the first few pivots. We test these predictions using a field experiment that employs 16 data points for 250 start-ups over about one year. We teach the treated group to formulate the problem scientifically, and to develop and test theories about their actions, while the control group follows a standard training. Our empirical analysis corroborates the predictions of our theory. It also shows that scientist-entrepreneurs enjoy higher performance.