One of the most important jobs of every leader is to define how you  measure success. You've probably heard the phrase you get what you measure. More important, you get what you reward.

If you reward your team for success, they will try to do more things like whatever it was that got them the reward. So, be careful about what you measure, and be careful what you reward. Get it wrong, and your team will probably fail. 

This all seems obvious, but in truth, it's pretty complicated. The reason is that the way most leaders define success misses one of the most important points. In an interview at Stanford University's "View From the Top" speaker series, Google CEO Sundar Pichai explained it like this: 

You have to encourage innovation. ... You know, one of the counterintuitive things is companies become more conservative as they grow. You have a lot more cash, you have a lot more resources but, you know, companies tend to become more conservative in their decision-making.

As a result, Pichai says his job is "encouraging the company to take risks and innovate and be OK with failure--and reward effort, not outcomes." 

That last part is really important. In fact, those four words are one of the most important lessons for every leader. As Pichai points out, that can be difficult since "people tend to reward outcomes."

Here's why it matters:

If you measure people only on outcomes, they will do whatever it takes to avoid negative outcomes. They'll play it safe and fall back on whatever they already know that works. They'll keep giving you exactly what you've always gotten, nothing more or less. 

The thing is, avoiding a negative outcome is not a great measure of success. Avoiding negative outcomes doesn't lead to greatness.

On the other hand, if you want people to work really hard, reward them for that. If you want people to take risks, try new things, and come up with new ways of doing things, incentivize them for their effort. 

Those things don't always lead to what we think of traditionally as success. Sometimes they are messy. Sometimes they fail. Sometimes the effort will cost you a lot of money. The point is, that's not necessarily a bad thing if what you're trying to do is build something new. 

If you're trying to do something worthwhile, you're going to fail a bunch of times anyway. You're going to break things and you're going to figure out how to put them back together. You might as well encourage people to learn from it all and reward them for their effort. 

Obviously, this doesn't mean you don't expect a lot from your team. It doesn't mean outcomes don't matter. It just means that the outcome you're looking for might be less obvious than a 5 percent increase in your stock price year over year.

It really comes down to the type of culture--and, ultimately, the type of company you're trying to build. If what you want are incremental growth and a small but consistent increase in quarterly earnings, that's fine, but don't expect anything revolutionary to happen. 

If, however, you want to build a team that is willing to do hard things, reward them for their effort. Reward them when they try and fail, and when they learn and succeed. Ultimately, that's the best possible outcome anyway.