Treatment and control groups

Finance

Statistics

Operating system

Machine learning

Heuristics

Odds

Business

Logistic regression

Economics

Marketing

Management

Control (management)

Mathematics

Computer science

Entrepreneurship

Management Science

A classical approach to collecting and elaborating information to make entrepreneurial decisions combines search heuristics, such as trial and error, effectuation, and confirmatory search. This paper develops a framework for exploring the implications of a more scientific approach to entrepreneurial decision making. The panel sample of our randomized control trial includes 116 Italian startups and 16 data points over a period of about one year. Both the treatment and control groups receive 10 sessions of general training on how to obtain feedback from the market and gauge the feasibility of their idea. We teach the treated startups to develop frameworks for predicting the performance of their idea and conduct rigorous tests of their hypotheses, very much as scientists do in their research. We let the firms in the control group instead follow their intuitions about how to assess their idea, which has typically produced fairly standard search heuristics. We find that entrepreneurs who behave like scientists perform better, are more likely to pivot to a different idea, and are not more likely to drop out than the control group in the early stages of the startup. These results are consistent with the main prediction of our theory: a scientific approach improves precision—it reduces the odds of pursuing projects with false positive returns and increases the odds of pursuing projects with false negative returns. This paper was accepted by Marie Thursby, entrepreneurship and innovation.

A Scientific Approach to Entrepreneurial Decision Making: Evidence from a Randomized Control Trial