one of my jobs is to encourage people to be bold. It’s incredibly hard. Experiments are, by their very nature, prone to failure. A few big successes compensate for dozens and dozens of things that didn’t work. Bold bets — Amazon Web Services, Kindle, Amazon Prime, our third-party seller business — all of those things are examples of bold bets that did work, and they pay for a lot of experiments.
I’ve made billions of dollars of failures at Amazon.com. Literally billions of dollars of failures. You might remember Pets.com or Kosmo.com. It was like getting a root canal with no anesthesia. None of those things are fun. But they also don’t matter.
More here.
I had lunch recently with a guy that runs a large insurance broker. He said that every year his sales force budgets a big drop in revenue and every year they grow.
Sales production is an incredibly risky and uncertain business and I’ve observed that really good sales people are often extremely risk averse. They know how uncertain it can be and that’s what drives them to kill themselves on every lead. Service is born from fear.
But at the top of the organization there is diversification and meta knowledge of the process. It makes sense to have a lot of people swing for the fences because home runs pay everyone’s bills. But you can wind up with a small group of superstars and an army of losers, which is unfair. Risk adjusting your response to results means going easy on the failures and not lauding the successes as much. After all, luck shouldn’t be worth much.
Attribution bias makes a measured response incredibly hard to pull off. It’s very easy to buy into the story that the winners and losers earned the result on their own.
It takes a great culture to keep a steady hand in a risky business.
Not Unreasonable 
I love this quote from the article:
– – On the other hand, if you’re the person who gets up in the morning and says, in the shower, “What can we invent for customers? What can we do differently? How can we improve that experience?” —
We do not do this in insurance. The standard homeowners insurance product for instance hasn’t change much in decades and it is still a product only a lawyer would love! No one arrives at their desk thinking, how can we make this better? How can we make the policy language easier to understand? How can we make the pricing more transparent? How can we add coverages so that we are assuming the risks customers need?