Jeff Bezos: Here’s how I make Amazon’s highest-stakes decisions

I like to putter in the morning. I get up early. I go to bed early. I like to read the newspaper. I like to have coffee. I like to have breakfast with my kids before they go to school. So my puttering time is very important to me.

That’s why I set my first meeting for ten o’clock. I like to do my high-IQ meetings before lunch. Anything that’s going to be really mentally challenging is a ten o’clock meeting because by 5 p.m., I’m, like, I can’t think more about this issue today. Let’s try this again tomorrow at 10 a.m.

Then on to eight hours of sleep. I prioritize sleep unless I’m traveling in different time zones. Sometimes getting eight hours is impossible, but I am very focused on it, and I need eight hours. I think better. I have more energy. My mood is better.

And think about it: As a senior executive, what do you really get paid to do? You get paid to make a small number of high-quality decisions. Your job is not to make thousands of decisions every day.

So let’s say I slept six hours a day, or let’s go really crazy and say I slept four hours a day. I’d get four so-called productive hours back. So if before I had, say, twelve hours of productive time during any waking day, now all of a sudden I have twelve plus four—I have sixteen productive hours. So I have 33 percent more time to make decisions. If I was going to make, say, one hundred decisions, I can now make thirty-three more. Is that really worth it if the quality of those decisions might be lower because you’re tired or grouchy or any number of things?

Now, it’s different if the company is a start-up. When Amazon was a hundred people, it was a different story, but Amazon’s not a start-up company, and all of our senior executives operate the same way I do. They work in the future. They live in the future. None of the people who report to me should really be focused on the current quarter.

When I have a good quarterly conference call with Wall Street, people will stop me and say, “Congratulations on your quarter,” and I say, “Thank you,” but what I’m really thinking is that quarter was baked three years ago.

Right now I’m working on a quarter that’s going to reveal itself in 2023 sometime, and that’s what you need to be doing. You need to be thinking two or three years in advance, and if you are, then why do I need to make a hundred decisions today? If I make, like, three good decisions a day, that’s enough, and they should just be as high quality as I can make them. Warren Buffet says he’s good if he makes three good decisions a year, and I really believe that.

Making Decisions

There are ways to increase the speed of decision making, and it’s super important. If I were to be so bold as to advise other senior leaders, I would say one of the things to watch out for—I see it at Amazon—is junior executives modeling senior executives in their decision making. And that’s normal. But you’re always looking up at the senior people and modeling. And a lot of it is even subconscious. The problem with that modeling is that it may not take into account the fact that there are different types of decisions.

There are two types of decisions. There are decisions that are irreversible and highly consequential; we call them one-way doors, or Type 2 decisions. They need to be made slowly and carefully. I often find myself at Amazon acting as the chief slowdown officer: “Whoa, I want to see that decision analyzed seventeen more ways because it’s highly consequential and irreversible.” The problem is that most decisions aren’t like that. Most decisions are two-way doors.

You can make the decision, and you step through. It turns out to have been the wrong decision; you can back up. And what happens in large organizations—not in start-up companies but in large organizations—is that all decisions end up using the heavyweight process that is really intended only for irreversible, highly consequential decisions. And that’s a disaster. When there’s a decision that needs to be made, you need to ask, “Is it a one-way door or a two-way door?” If it’s a two-way door, make the decision with a small team or even one high-judgment individual. Make the decision. If it’s wrong, it’s wrong. You’ll change it. But if it’s a one-way door, analyze it five different ways. Be careful, because that is where slow is smooth and smooth is fast.

You do not want to make one-way-door decisions quickly. You want to get consensus or at least drive a lot of thought and debate.

What can also really speed up decision making, in addition to asking whether a decision involves a one-way or two-way door, is teaching the principle of disagreeing and committing. So you’ve got passionate missionaries, which you need to have. Everybody cares, and if you’re not careful, the decision process can basically become a war of attrition. Whoever has the most stamina will win; eventually the other party, with the opposite opinion, will just capitulate: “Okay, I’m exhausted. We’ll do it your way.”

That is the worst decision-making process in the world. It leaves everybody demoralized, and you also get a kind of random result. A much better approach is for the more senior person to escalate to even more senior leaders. Controversial decisions need to be escalated quickly. You can’t let two junior people argue for a year and exhaust themselves. You have to teach those junior people.

When your team is really at loggerheads, escalate—and escalate fast. And then you, as the more senior person, hear the various points of view, and you say, “Look, none of us knows what the right decision is here, but I want you to gamble with me. I want you to disagree and commit. We’re going to do it this way. But I really want you to disagree and commit.”

And here’s the important part: Sometimes this disagreement happens between the more senior person and a subordinate. The subordinate really wants to do it one way, and the senior person really thinks it should be done a different way. And it’s often the case that the more senior person should disagree and commit. I disagree and commit all the time. I’ll debate something for an hour or a day or a week. And I’ll say, “You know what? I really disagree with this, but you have more ground truth than I do. We’re going to do it your way. And I promise I will never tell you I told you so.”

It’s actually very calming really because it’s acknowledging the reality that the senior person has a lot of judgment. That judgment is super valuable, and that’s why sometimes you should overrule subordinates even when they have better ground truth. But that’s your judgment. And sometimes you’re, like, “I know this person, or I’ve worked with them for years. They have great judgment. They really disagree with me, and they have way better ground truth. I’m going to disagree and commit.”

Reprinted by permission of Harvard Business Review Press. Excerpted from Invent and Wander: The Collected Writings of Jeff Bezos. Invent and Wander is co-published by PublicAffairs, an imprint of Perseus Books, and Harvard Business Review Press. Copyright 2021 Jeffrey P. Bezos. All rights reserved.