A fish rots from the head down: How founders lead startup fraud

A new study by Janine Crivelli and Joakim Wincent from the University of St.Gallen, Manuel Hess from Grenoble Ecole de Management, and Dean A. Shepherd from the University of Notre Dame shows that the answer lies not only in the actions of a fraudulent founder. Startup fraud can become embedded through the interaction of concentrated founder influence, limited organizational safeguards, and employees’ strong loyalty to the startup and its mission. 

Published in the Journal of Management, the study examines how founders shape employee responses to fraud and how employees react as their concerns develop over time. The researchers analyzed the Theranos case (one of Silicon Valley's most notorious corporate fraud scandals) over a period of 15 years, drawing on court testimony, internal emails, private messages, interviews, and media reports. 

The findings show that founders can influence not only what employees know, but also how they interpret questionable practices. In startups, founders often control important information, decisions, and access to external stakeholders. At the same time, they are the main source of the startup’s vision, identity, and sense of purpose.

Founders have considerable structural and symbolic influence

This combination gives founders considerable structural and symbolic influence. By controlling information and presenting questionable actions as necessary for the startup’s mission or survival, founders can make it difficult for employees to recognize when ambitious experimentation has crossed the line into fraud. 

This is especially important in startups because they operate under high uncertainty and often lack established routines, clear benchmarks, formal controls, and independent reporting channels. Practices that would appear clearly unacceptable in an established corporation may initially be interpreted as temporary compromises, technical setbacks, or part of the uncertainty of building something new. 

Employees are therefore not simply passive followers. Nor do they usually make one clear decision either to support or reject the fraud. Instead, the study shows that they repeatedly reassess what is happening around them,” said Crivelli.

Employees may remain committed because they believe in the founder, the technology, or the social purpose of the startup. Their work may feel meaningful, and they may have invested considerable time, emotion, and personal identity in the venture. These bonds can lead employees to question their own judgment when problems emerge. 

The study describes employees’ responses as a fluctuating moral evaluation process. Employees may initially rationalize irregularities, minimize potential harm, or accept temporary workarounds, while colleagues reinforce the belief that questionable practices are necessary for the startup’s mission. As technical failures, secrecy, customer harm, and misleading claims accumulate, however, doubts become harder to suppress. Pressure, intimidation, and the repeated dismissal of concerns may initially silence employees, but over time they can erode trust in the founder and push employees from condoning the fraud toward condemning it. 

Employees who reach this point may distance themselves from the startup, preserve evidence, resign, resist leadership, or report the misconduct. Whistleblowing is therefore often not an immediate reaction. It may be the result of a prolonged process in which loyalty gradually gives way to moral condemnation.

Direct implications for founders, boards, and investors

The findings have direct implications for founders, boards, and investors. Flat hierarchies should not be confused with open communication. A startup may have few formal management levels while still concentrating on information and decision-making in its founder. 

Boards and investors should therefore not rely exclusively on founder-provided information or high-level performance figures. They need independent access to employees, technical evidence, and external verification. Startups also require credible channels through which employees can raise concerns without depending on the founder or those who are personally loyal to the founder. 

The study ultimately shows that startup fraud is not merely the action of one unethical individual. It can become a social and organizational process in which vision, loyalty, uncertainty, and weak safeguards gradually draw employees into practices they may later come to reject. 

The study A Fish Rots from the Head Down: How Founders Lead Startup Fraud was published  by the Southern Management Association und Sage Journals.