Darts was more than a pastime; for him, it was a style of management. Stand as close to the situation as possible, use your years of experience to gauge your shot, and throw into the general area. Sometimes you hit, sometimes you miss, but you always have two more tries and you usually get close enough that the wall isn't covered in holes.
Even after years of working with Smitty, I always was frustrated that the difference between triple 20 and 2 was a metal bar thinner than your car key. Too often I was on the wrong side of that bar. If you are not a dart player, triple 20 is actually the bull's-eye. It's worth more than the red dot in the center in most games. Inexperienced players shoot for the middle, but you know you're around a dart player when they go after the triple 20. In IT management, it's like when application developers track the number of lines of code as a performance measure. Increasing the number of lines is not shooting for the real goal.
Performance management is more like artillery fire than darts and it's a better way to lead. With artillery, there are coordinates and weather factors you enter in, carefully aim, fire, and adjust the cannon to get to your mark with the next shot. It's a repeatable cycle proven to get results and remove some of the human error you get in things like darts.
In performance management the cycle is: Measure performance, set goals, make improvements to hit those goals and measure performance to see if you met them. Lather, rinse, repeat. The cycle serves as the basis for how improvement is made and should be in the DNA of a well managed organization.
In IT you can pick any process and apply the cycle -- like data center utilization. By measuring the times you are running at peak capacity, you can adjust run schedules to maximize efficiency. Data center managers can use their experience to gauge this, but to really hit the mark, you need the measures to show which jobs cause the peaks, and to understand exactly how changes will impact performance. Is one better than the other? Depends on how good your data center manager is, and how long until he or she retires.
Most times, the absence of this cycle has inherent risks with real consequences for IT shops. However, knowing performance management methodology is more important than just running your shop. Knowing if and how your customers use it is just as vital to your success. Let's go back and look at the darts.
Dart managers like Smitty look to improve work based on their own instincts. Nowadays it seems like instincts are touting technology as the key to improvement. After all, there isn't a business process left that doesn't hit the network at some point, and everyone believes the next system, enhancement, Web site or Facebook/iPhone application is going to hit the bull's-eye.
Just like darts, with new technology, you can miss the center by less than a quarter inch and go from bull's-eye to just a single point. Rarely do we miss completely, but we're not getting what we were aiming for. Next time you see a dartboard, throw a dart and see if you hit where you were aiming, or turn on ESPN2 in the early morning hours and watch the pros. With IT, it's tens of millions of dollars per throw. Wouldn't you want a better success rate than even the pros get?
Our rush to use technology as an improvement effort throws the performance management cycle out of whack. By skipping measurement and goal setting and going straight to improvement, we fail to improve the right things, the right way. Performance management requires us to use the entire cycle: knowing how well the work is getting done, using the numbers to set goals and working to make improvements (whether technology-based or not) that move the needle on those numbers. It's a lot more reliable than tossing darts.
