But for all creative jobs we would pay one incredible employee at the top of her personal market, instead of using that same money to hire a dozen or more adequate performers. This would result in a lean workforce. We’d be relying on one tremendous person to do the work of many. But we’d pay tremendously.

This is the way we have hired the majority of employees at Netflix ever since. The approach has been remarkably successful. We have exponentially increased our speed of innovation and our output.

At most companies, the boss is there to approve or block the decisions of employees. This is a surefire way to limit innovation and slow down growth. At Netflix, we emphasize that it’s fine to disagree with your manager and to implement an idea she dislikes. We don’t want people putting aside a great idea because the manager doesn’t see how great it is. That’s why we say at Netflix: DON’T SEEK TO PLEASE YOUR BOSS. SEEK TO DO WHAT IS

INNOVATION CYCLE STEP 2: FOR A BIG IDEA, TEST IT OUT

Most successful companies run all sorts of tests in order to find out how and why customers behave the way they do—and the results of those tests usually influence the corporate strategy. The big difference at Netflix is that the tests take place even when those in charge are dead set against the initiative.

Even at Netflix we have pockets of the company where safety and error prevention are our primary goals and there we fence off an area to build a little symphony orchestra that plays pitch-perfect rules-and-process.

Take, for example, employee safety and sexual harassment. When it comes to protecting our employees from injury or harassment, we invest in error prevention (training) and hotlines; we have strong processes to make sure all claims are properly investigated; and we use process-improvement principles to drive the incident rates down to zero.

In select instances like these, where error prevention is clearly more important than innovation, we have loads of checks, processes, and procedures to ensure we don’t screw anything up. In these moments, we want Netflix to be like a hospital where there are five people verifying the surgeon is operating on the correct knee. When a mistake would lead to a disaster, rules and process isn’t just nice to have, it’s a necessity.

With this in mind, you can consider your objective carefully before deciding when to opt for freedom and responsibility and when rules with process would be a better choice. Here are a set of questions you can ask in order to select the right approach:

• Are you working in an industry where your employees’ or customers’ health or safety depends on everything going just right? If so, choose rules and process.

• If you make a mistake, will it end in disaster? Choose rules and process.

• Are you running a manufacturing environment where you need to produce a consistently identical product? Choose rules and process.