The Secrets of "Are Right, a Lot" - the Intriguing Amazon Leadership Principle About Not Always Being Right
Are Right, a Lot is more than just being correct frequently. It's about being open and interested in learning as well.
Amazon's "Are Right, a Lot" leadership principle sounds like the most obvious of the principles. It's almost like a company saying, "Our plan is to make money."
Are Right, A Lot - Leaders are right a lot. They have strong judgment and good instincts. They seek diverse perspectives and work to disconfirm their beliefs.
Of course we all expect leaders to be right a lot. Yet a lot of people miss what the principle is fundamentally asserting.
Years ago on one of my teams, a principal engineer had proposed a new project to improve recommendations for customers. I looked at the schedule, which included a significant amount of time up front to instrument the feature with metrics, and validate the hypothesis before rolling out the new feature.
"At this point, we trust that you're doing the right thing." I said to Courtney, our principal engineer. "You could save a ton of time and roll out your improvement now. I'll back you up if you'd like to move forward."
Courtney shook her head vigorously, "I have a hypothesis instead of a conclusion, because we don't have enough data to prove my proposal will work. So we need more data. Plus, this won't be the last improvement we'll want to roll out. We need to have mechanisms to verify we're right, not just use our judgement."
Being right a lot as a leader is not simply about having a high percentage chance of having the correct judgement. The principle isn't accurate without the second half of the clarifying text: "They seek diverse perspectives and work to disconfirm their beliefs."
Bias towards our perspective
Imagine you're that principal engineer, and you're looking to roll out an improvement in recommendations for customers. You know your new machine learning algorithm has a higher accuracy than your previous one.
Many people's instincts would be to find fellow principal engineers, and have them debate if the new machine learning algorithm does indeed have a higher accuracy. They may even consult a machine learning scientist to discuss if the calculations used to calculate the relevance is the right one.
However, the perspectives of other software engineers or scientists is only a part of the story. While I can absolutely learn from other experts in my field, I'm even more blind to what expertise in other fields can bring to a discussion. Everyone is biased towards their own field, because you speak the same language, and have the same priorities.