A colleague pointed to the video and article by Fred Kofman at Linkedin, Doing Your Job May Be Hazardous To Your Career. It's a great discussion of systems thinking and ties into much of what the folks in the Theory of Constraints camp talk about. In the first minute, he says several times some version of, "To optimize the overall system, you must sub-optimize the subsystems."
It's obvious, maybe, but why people can't we do it? It's built into the way we are paid. Management by Objectives, Pay for Performance, you name it. Almost all of these are limited to the sub-system in which they operate, not the overall.
Unfortunately, global incentives - at least the first pass version - don't work very well either. They solve some of the local incentive problems, but then they generate some of their own. They tend to be too complex and they tend to inspire free-riders. It seems that Global incentives are harder make work than Local incentives. Every incentive system deludes you into thinking that your job is to achieve the metric. That's where they fail. Your job is not the metric, it's to help the company succeed. Kofman suggests that while all companies struggle with this problem, those that "suck less" at it will tend to beat their competition.
Kofman uses the fun analogy of a sports team. The team goal is to score more than the opponent. But if the defense is paid to prevent scores, they might be happier losing 0-1, instead of winning 5-4. And if the offense is paid to score, they might be happier losing 4-5, instead of winning 1-0. As Kofman says, "Multiply by a million, and you have a company - every company."
Here is the introduction to the talk. Go have a listen if you are intrigued.
Your job is not what you do, but the goal you pursue. Your real goal, however, is not your individual KPI, but the team’s goal. Whether you play defense or offense, your real goal is to help your team to win the game.