There has been a lot of work in the area of organisational psychology around challenging goals, and this has underpinned calls for adopting “stretch goals” to increase overall organisational performance. Indeed, the leadership of chief executive Jack Welch and his implementation of stretch goals is often credited with the success of the US conglomerate General Electric in the 1990s. The idea is that goals around metrics such as profit, market share and the share price would be set so high that achieving – and even attempting them – would inspire innovation that would not only boost short-term performance, but transform the company. In describing stretch goals to his executives, Welch’s favourite example was reportedly Japan’s high-speed Shinkansen bullet trains. The goals for the bullet train project required the development of completely new technologies that would never have been attempted if the stretch goals had not been in place. Despite their success at GE, and at many other organisations, stretch goals have remained controversial. In some cases, they have created both success and failure. Toyota, for example, gave its engineers a year to create a vehicle that increased fuel efficiency by 100 per cent. The team tried out 80 hybrid technologies before narrowing the list down and creating the highly successful car, the Prius. But at the same time, stretch targets at Toyota around market share, introduced with the aim of overtaking General Motors as the world’s No. 1 car maker, led to some less than optimal production techniques, which saw thousands of cars recalled. To examine this issue, case studies of organisational goal setting are useful, but also have important limitations because they cannot give a picture of what would have happened if the organisation had not implemented stretch goals. Field experiments with organisations willing to systematically vary performance goals would be a great approach for studying this issue, but most organisations are not eager to set and stick with different goal levels across similar business units. The alternative is a laboratory experiment, which is the path pursued by our research team from UNSW Business School and the Massachusetts Institute of Technology. So, are there right and wrong ways to set stretch goals? Are they better in some situations than in others and how best can they be implemented? These were questions we considered in our collaborative research project. In two experiments using a management simulation of a start-up airline, we sought to understand how stretch goals affect risk-taking, motivation and, ultimately, influence performance. We found goals can provide focus and motivation for individuals, and teams to put in the amount of effort they need to do well. The trouble is that when you scale this up to the organisation level, there has not really been much research on what the effects are – either positive or negative – of setting stretch goals. In our study, goals were varied from stretch to moderate in order to understand how different goal levels impacted performance, commitment and risk appetite. In the first experiment, participants were divided into teams, and in the second they worked individually. The key insight is that stretch goals do benefit a small fraction of the organisations that adopt them, but they are certainly not a “rule for riches” for every organisation. Overall, stretch goals led to lower risk-adjusted performance. We also see a vicious negative spiral with some of the managers in our experiments. If they started to believe that the goals were simply unattainable, their confidence would erode, their commitment would plummet and their anxiety levels would increase. This contrasted with experimental participants who had been assigned moderate goals, and whose commitment to the assigned goals remained high. Our results also give us some indication about situations where stretch goals might be more likely to succeed, and are also suggestive about the different types of organisations that will gravitate to stretch goals and see positive results. For example, venture capital or private equity firms with a portfolio of investment companies are well suited to adopt stretch goals, because they need for only one in 10 investments to pay off. Shooting for the moon makes sense in such situations. But a medium-sized family business where the majority of the family wealth is tied up in the business does not generally want to adopt stretch goals for performance, because the risk of failure is greater. You have to have multiple balls in the air and if one of them works, then the benefits can be massive. The accepted view is that one out of every seven start-up ventures is successful, but when you begin you do not know which one that will be, so you have to be resilient to the possibility of failure. And while failing six times and succeeding the seventh can be a good result for some organisations, for others this would be a disaster. While some teams and individuals responded well to stretch goals, the research project also highlighted their potential negative effects. There is also an ethical dimension to setting stretch goals that we did not examine in our study. Managers under extreme pressure to achieve unrealistic goals can sometimes make unethical decisions. The fake bank account scandal at Wells Fargo bank in 2016, for example, has been attributed to the potentially dysfunctional effects of stretch goals. Managers were incentivised with aggressive performance targets, and to reach them they opened multiple accounts for customers, many of whom had no idea the other accounts in their name even existed. So, when considering whether stretch goals work in your organisation, try to define stretch goals that maintain high commitment to goals, high motivation to achieve the goals, reduce the likelihood of internal goal erosion and encourage the level of risk-taking appropriate in the organisational context. The goals should be ambitious, but not so difficult to achieve that they set off the vicious, downward spiral feedback effects we found in our study. Once people start to fall far short of stretch goals, it can be very disheartening. While they might work for some organisations, stretch goals do not work for all and are by no means a magic silver bullet. Shayne Gary is an associate professor and AGSM fellow at UNSW Business School