Learning from Failure, or Failing to Learn
Philippe Baumard ([email protected]) and William H. Starbuck ([email protected]), “Learning from Failures: Why It May Not Happen,” Long Range Planning, vol. 38, no. 3, June 2005, available from the authors or click here to purchase.
In their agenda-setting book A Behavioral Theory of the Firm (second ed., Blackwell, 1992), Richard M. Cyert and James G. March asserted that organizations learn more readily by confronting problems than by experiencing success. Since that book first appeared in 1963, the importance of learning from failure (for individuals as well as corporations) has entered the canon of management wisdom.
But do companies actually live by this generally accepted wisdom? To find out, Philippe Baumard, a professor of strategy at the University of Aix-Marseille III and on the faculty of the Center for Catastrophic Risk Management at the University of California at Berkeley’s Haas School of Business, and William Starbuck, the ITT Professor of Creative Management at New York University’s Stern School of Business, analyzed 14 strategic failures that had occurred at a large, unnamed European company over 20 years as it grew from a domestic to a global telecommunications firm. They ruled out failures that could be attributed to managerial inexperience and selected seven “small failures” and seven “large failures” to analyze. Twenty of the company’s executives contributed their thoughts on the failures.
Among the small failures examined was the introduction of what was labeled “Net TV.” Someone with television experience joined the company and proposed Internet television. The CEO liked the idea, and another manager was a strong supporter, but other managers regarded the proposal with skepticism. After four months, because of inadequate technology and low revenues, the venture had racked up losses amounting to four times the initial investment. The two executives who had championed the idea left the company.
In another venture — a large failure — the company launched a satellite broadcasting service that would have to compete against two rivals controlling 80 percent of the market. Despite the long odds, the project seemed at first like a good idea because the company would receive a government subsidy for its efforts and would be expanding its portfolio of communications technologies. But management made a host of mistakes. The company broadcast analog signals rather than far-superior digital signals, and three of its four satellites were technologically obsolete by the time the service was operational. Moreover, the company was ill equipped to deal with price cutting from the established satellite TV providers. Before long, the company concluded that it needed to focus instead on terrestrial technologies.
The authors found that the small failures, such as Net TV, generally attracted the most attention within the company. Yet when these failures challenged the basic direction of the firm, they were dismissed as simply being tactical mistakes. As for large failures like satellite TV, the company’s managers tended to brush them aside, producing a litany of excuses that mostly blamed external forces — “exceptional or historical conditions,” “society was undergoing large, dramatic change,” and so on. No relationship was seen between recent large failures and previous large failures, however similar. Instead, each large failure was regarded as unique.
From their research into the telecommunications firm’s inability to learn from failure, the authors produced a series of lessons that may be useful for other organizations: First, the outcomes of ventures should be aligned with the compensation and promotion of managers, creating a balance between rewarding success and penalizing failure so that risk taking is not discouraged. Second, managers should treat with skepticism internal attempts to blame external forces. Third, the level of cynicism and self-interest among executives should not be underestimated. They are only human. Fourth, unlearning successes — moving beyond past successes — may be a requirement for being able to learn from failure. The researchers note that managers who have taken part in failures too often tend to express great confidence that they will be able to reproduce past successes.