At one point in my career, I took a job as a hedge fund analyst. It was interesting because the metrics of success were so radically different from normal corpo

At one point in my career, I took a job as a hedge fund analyst. It was interesting because the metrics of success were so radically different from normal corporations. In a hedge fund, if you succeed, say, 55% of the time, you're a rock star - like Ted Williams batting 400. Psychologically, this was hard to take - and I realized it was because in most corporate environments, success is expected. Inside most companies, there is an assumption of 100% success. You can see it everywhere - but particularly in goal setting. The idea is simple: you ask people to set goals, often about areas they don't know much about, then you measure them on their ability to hit those goals. So one way to solve the problem, in my experience, is to make it clear that goals are not expected to be 100% achieved - otherwise they wouldn't be goals. If you make 100% of your goals, then you likely sandbagged the goals. But to do this, you need to remove the connection between goal and financial outcome. Not easy to do. The way I've been doing it is for well known/understood areas (aka "mature"), goals should be aggressive and you should be measured on them. However, if the project or product is very nascent, you should shift goals to focus on iterations - namely that you do your best estimate, but recognize that you don't know much, and therefore you are more than likely to be wrong. This has worked (thought there are always exceptions) - but it is challenging for managing inside a corporation, mainly because some senior exec will sh*t all over a failed project not understanding the difference between mature and developing projects.