1. Motivation & Performance evaluation
But the basic idea about incentives is the solution to the agency problem, as it's usually described, you're trying to reward people for doing the right thing. You want them to finish this project, they get paid so much when they finish a project.
The idea of a tournament as a way to structure pay is another kind of economics based argument that is about relative compensation, and this mimics the way sports tournaments work. So for example if you're a golfer or somebody playing an individual sport, you know that if you win the tournament, you get a big prize. (缺少合作太多竞争，适用于晋升设计；还让大家偏向于加入弱者竞争环境)
Rewarding A while hoping for B 类似OKR目标导向
大家的本性是在不知道做什么的时候模仿别人，所以如果是新人，可以放在表现比较好的位置，他会模仿周围的人。
Another example of this is something known as the Pygmalion effect. The idea that you can shape people's behavior by what you expect of them. So if you set higher goals for them and higher expectations, rather, for them, they are inclined to achieve more.
Now, one way to do this is with internship right?. Now you might think that the typical reason for internships is the candidate can scope you out and decide or you can scope the candidate out, rather, and decide whether this is the kind of person you want. But a lot of companies have figured out that the more important use of that is the candidate can scope you out. So that if you offer them a job and they say, yes, they know what they're getting into. It's going to have a retention device to make sure you're not going to lose people as soon as you hire them.
There's a huge literature on this, but let me summarize it briefly. The things that matter the most are things associated with past behavior, tends to be the best predictor of future behavior. If you could actually find out how they performed in a similar job before, that would be great, right? That would be maybe the best thing you could get.
In this case, there's a big time delay, especially if the performance exercise is the end of the year. It's pretty difficult for a supervisor to remember everything you did accurately, let alone to give you feedback about that.
supervisor和employee都不喜欢评价，因为supervisor如果给到员工的反馈不及员工预期（很多人都会overconfident），那么员工会很气觉得评价者不对，于是产生了矛盾。
And this a bias that we have when we look at human behavior to assume that the behavior is the result of dispositions. If somebody does something bad, it’s because they’re a bad person, rather than the circumstances.
And the second aspect of bias, which is just as important, is self-fulfilling prophecies. And that means if we've got an employee coming in to our workplace who we believe is a very good worker, people say great things about him, we tend to manage them differently. We give them more opportunities.
And we also know there are a lot biases related to relationships. We give better scores to people who are more like us. We give better scores to people who are physically more attractive, etc.
增加conversation频率，把一年的评价拆成好几个小评价，提高评价频率，让evaluation更合理一点。
如果反馈是negative，需要很小心地去反馈，否则会给员工很强的心理负担。
Tasks vs. Jobs
Job = a series of Tasks
员工希望工作可以和别人有交集，以及可以control自己的工作，而不是完全被动，另外不想单一做一件事。
Making good and timely management decisions
不要等100%有结论再行动，传递宗旨给下属，让他们自己做决定，容忍第一次决策错误，但manager不能不敢于做决定：
永远不要过于自信，不断地改进才可以实现更好的品质——Toyota case
Designing and Changing Architecture
If we want performance, we want to reward people for performance. If we want performance in our organizational design, we want promotion. Not just pay to go with the performance that we're expecting.
That's a way of saying as we move to pay and promotion as a design feature of the organization that we run. We want people regardless of background, gender, racial identity, anything you can think of that doesn't relate directly to performance. Once we move in that direction, lots of evidence says it's a great way to go. Companies all over the world are moving in that direction as well. We open up a kind of Pandora's box of the de-motivation that can come with inequity.
如果没有晋升激励，老员工可能hold住业绩，反正业绩好没有奖励，下一年的要求可能还会更高。
Change is one of the great enemies of organizational design：
