Authors Kay-Yut Chen and Marina Krakovsky look at how research in human behavior saved HP millions of dollars, and offer lessons for other companies
Hewlett-Packard Co. (HPQ) employees, it turns out, can do a good job predicting sales patterns at the computer maker. At least they did in the late 1990s, when Kay-Yut Chen, a researcher at HP Labs, was exploring how worker decision-making could help the company save money and avoid risk. Chen, along with science writer Marina Krakovsky, details that research in his new book, Secrets of the Moneylab: How Behavioral Economics Can Improve Your Business. In experiments from 1996 through 1999, Chen proved that by placing bets in a stock-market-like environment, employees could produce a more accurate sales forecast than the ones issued by HP. The challenge, though, was getting employees to participate. Even doling out $50 in betting cash wasn't enough to get some staffers to play along as frequently as researchers would have liked. "Overall, the results were encouraging, beating official forecasts in accuracy six out of eight times," Chen and Krakovsky write in Secrets of the Moneylab, published this month by Portfolio Penguin. "But these modest improvements were not worth the costs, which included participant training and their trading time." The fact is, humans don't always behave rationally, and sometimes not even in their own best interest, Chen finds. That theme has been outlined in previous books including Predictably Irrational and Nudge. Yet, Secrets of the Moneylab departs from these two books by showing how businesses can use and even profit from unpredictable human behavior. Harnessing Collective Intelligence
The findings also contribute to a growing body of knowledge surrounding so-called collective intelligence, or the trove of knowledge represented by a company's workforce. Companies as varied as AT&T (T), Electronic Arts (ERTS), and Pitney Bowes (PBI) aim to get more adept at soliciting ideas from their employees, organizing that information, and then predicting which suggestions are likely to prevail. While Chen worked with groups of about 10 to 30 workers, companies such as AT&T cast a wide net, enlisting thousands of employees. They nevertheless can learn from Chen's exploration of human motivation and his explanation of how HP overcame employee reluctance to make predictions. HP designed a game called BRAIN that required fewer participants and less executive time but nevertheless gave the company valuable insight into players' attitudes toward risk. HP and some of its customers have since used BRAIN for revenue forecasts and other predictions. Economic experiments conducted by Chen are one of the book's biggest strengths. Yet, the authors go into extensive detail on those experiments in the first chapter, slowing the early narrative of this 246-page tome. Reputation Lessons
The pace picks up in later chapters, which focus on such themes as fairness and reciprocity. In Chapter Five, Chen and Krakovsky explore reputation. The authors note that a great reputation doesn't ensure future performance and show how Bernie Madoff was able to trade on his reputation, cultivated over many years. The lesson is that the better a reputation, the greater the opportunity for exploitation. Another chapter examines the link between trust and prosperity, referencing a study of 344 automaker-supplier relationships in the U.S., Japan, and Korea. "The researchers found that automakers with the lowest perceived trustworthiness have transaction costs five times higher than those with the highest perceived trustworthiness, spending far more of their face-to-face time with suppliers on haggling and ironing out contract details." Secrets of the Moneylab is chockablock with similar insights likely to be useful to any manager who wants a better understanding of why employees, customers, and suppliers behave the way they do.