You’re reading What They Said, a project by Chris Gallo.

Jeff Bezos


Jeff Bezos is the Amazon.com Founder & CEO. This is a short chat brought to you by the Internet Association. The association’s President & CEO, Michael Beckerman, discusses business, the early days of Amazon, space, and the future.

It’s a conversation that gets more interesting as it goes on because it’s full of wisdom. Read through some bits and pieces below, and go watch the full 37-minute discussion here.

Did you anticipate what would happen over the last 22 years at Amazon? 1:53

The answer is God no. So Amazon started as a small company. It was me and a few other people. I was driving all the packages myself to the post office in my 1987 Chevy Blazer.

And when I raised money for Amazon, I had to raise a million dollars. Which I raised from 22 different investors, $50,000 dollars each. They got 20 percent of the company for the million dollars.

40 people told me no. So I had to take 60 meetings to get 20 yes’s. The first question was always what is the Internet? And I had to walk through that. This was 1994, early 1995.

So did I anticipate, fast forward to today and the current version? No.

It has been one foot in front of the other. And I think that is true for most businesses. Where you kind of proceed adaptively, it’s step-by-step. You have a success and then you kind of double-down on that success.

Then you kind of figure out what you can do, what customers want. Everything we’ve done and all the success we’ve had, at it’s root, is primarily due to the fact we’ve put customers first.


How should consumers be thinking about Amazon? 4:14

That is something that people get confused about sometimes. What is Amazon?

We do such a diverse array of things, from producing original content at Amazon Studios, to Amazon web services where we sell startups and enterprises computing infrastructure, to the things that most people know about which is our consumer offering where we deliver things in little brown boxes.

This things seem some disparate. How is it that we’re doing all of them?

The common thread, and really what it is, is an approach. We have a very distinctive approach that we’ve been honing, thinking about, and refining for 22 years.

It’s really just a few principles that we use that we go about these activities.

The very top of the list is one I’ve already mentioned, but you’ll probably hear it 10 times tonight because it’s so central. And it is customer obsession.

It’s customer obsession instead of, for example, competitor obsession or business model obsession or product model obsession or technology obsession.

There are many ways to center a business. And by the way, many of them can work. I know, and have friends who lead competitor obsessed companies. And those companies can be successful.

It’s not a bad strategy. But you have to be really good at close following. You identify winners when you watch your competitors very carefully. If they latch onto something, you do it as quickly as possible.

It’s a very good strategy in some ways because you don’t have to go down blind alleys. If you’re not pioneering, you’re only close following. It has some advantages, but it has some disadvantages too.

And I like the customer obsession model. I think it’s the right one. It’s better than product obsession even. Product obsession is not bad, but I think customer obsession gets you there in a healthier way.

So that’s one of the principles, the approaches we take in every single that we do.


Invent and Pioneer 6:26

The second one is that we are willing, and even eager to invent and pioneer. I think that goes along, it marries really well with customer obsession.

Customers are always dissatisfied, even when they don’t know it, even when they think they’re happen they actually do want a better way and they just don’t know yet what that should be.

That’s why I always warn people, customer obsession is not always just listening to customers. Customer obsession is also inventing on their behalf because it’s not their job to invent for themselves.

And so you need to be an inventor, and a pioneer.

The third one that is really central to the way we think about all of our business problems and a bunch of things at Amazon is we’re long-term oriented.

I ask everybody not to think in two to three year time frames, but to think in five to seven year time frames. When somebody says to me, congratulates Amazon on a good quarter, which is a very common thing to say when looking at your results, I say thank you.

But what I’m thinking to myself is that quarter, those quarterly results were pretty much fully-baked about three years ago. So today, I’m working on a quarter that is going to happen in 2020. Not next quarter.

Next quarter, for all practical purposes, is done already. And it’s probably been done for a couple of years.

If you start to think that way, it changes how you spend your time, how you plan, where you put your energy, your ability to look around corners gets better.

So many things improve if you can take a long-term view. By the way, it’s not natural for humans. It’s a discipline that you have to build.

The kind of get rich slowly schemes are not big sellers on infomercials. So that’s something you kind of have to steal yourself for, discipline for and teach over time.

So what is Amazon?

I would say really it’s a collection of principles and it’s an approach that we deploy. And it’s fun. I dance into work.


What worries you? 9:50

For me, the thing I worry about the most is that we would lose our way on one of those things. That we would lose our obsessive focus on customers or we would somehow become short-term oriented or start to somehow become overly cautious. Failure adverse and therefore kind of unable to invent and pioneer.

You cannot invent and pioneer, if you cannot accept failure. To invent, you need to experiment.

And if you know in advance that it’s going to work, it is not an experiment. So that’s a very important thing.

They are inseparable twins, failure and invention.

You have to be willing to do that. And it’s embarrassing to fail. It’s always embarrassing to fail, but you have to say, no that’s not how this works.

If I say to you, you have a 10 percent chance with a particular decision, a 10 percent chance with a 100x return. You should take that bet every time, but you’re still going to be wrong nine out of 10 times. It’s going to feel bad nine out of 10 times.

With technology, the outcomes or the results can be very long-tailed. The payoff can be very asymmetric, which is why you should do so much experimentation.

Everybody knows if you swing for the fences, you hit more home runs, but you also strike out more. But with baseball that analogy doesn’t go far enough, because with baseball, no matter how hard you connect with the ball, you can only get four runs. The success is capped at four runs.

But in business, every once in awhile, you step up to the plate and you hit the ball so hard you get a thousand runs. So when you have that kind of asymmetric payoff, one at-bat can get you a thousand runs, it encourages you to experiment more.

It’s the right business decision to experiment more. It’s also right for your customers. Your customers like the successful experiments.


Customer adoption versus disruption 11:57

By the way, this is a giant misconception in a lot of young entrepreneurs, inexperienced entrepreneurs.

One of the things that is very fashionable is how disruptive their business plan is going to be. But invention is not disruptive.

Only customer adoption is disruptive.

At Amazon, we’ve invented a lot of things that customers did not care about at all. And believe me, they were not disruptive to anyone.

It’s only when customers like the new way, that anything becomes disruptive.

So if somebody comes to you with a business plan that they claim is disruptive, you should ask them to explain it to you in simpler language.

The simpler language is why our customers going to adopt this? Why are they going to like it? Why is it better than the traditional way?


What do people not know about Amazon? 13:10

First of all, we’ve been over 20 years on balance been treated very fairly by the media. Not every story written about us is the way we would have written it, if we were writing it ourselves. But on balance, very fairly treated. So wrong might be too strong.

There is a piece of our business that is probably not fully appreciated, the scale of it and what’s happening, the energy and the dynamism around it, which is the third party seller part of our business.

Today, roughly half of the company is really third party sales. It’s small businesses. There are 100,00 businesses on Amazon where the businesses make $100,000 or more a year.

It’s turned into a gigantic service for them. We even let them stow their products in our fulfillment centers, so their products become Prime eligible, which increases their sales. And a lot of people make their living that way.

We’ve extended that into other areas too. We have a platform called Kindle direct publishing, so authors can write books and publish them directly. That’s become very successful.

People make livings. I have so many emails from authors saying because of this program, I’ve been able to quit my job. They already have 25 rejection letters from traditional publishers, because their book didn’t fit in a category or didn’t seem right. But readers disagreed.

I think of this more like gardening. You get all the nutrients right in the soil, and then you see what comes up. And that’s what happens with these third party businesses.

We don’t plan which books people are going to write, we just provide a service that makes it really easy to contact and reach readers. And then they create new things.

Same thing with Amazon Web Services. We create a set of tools, and software developers and companies that employ software developers, they figure out surprising and amazing things to build with these tools.

Empowering these third party tools to do things, starting with our third party sellers, authors and AWS, probably the biggest thing that most consumers probably don’t appreciate about Amazon.

Good investors know it very well.


Overnight success 16:08

Another overnight success. I’ve noticed all overnight successes take about 10 years.


What will Amazon look like in 10 years? 16:31

That’s a very good question. On the outside, I think the observable Amazon could change quite a bit.

As you said, 10 years ago, you didn’t think AWS would be such a significant contributor to the business. There could be more things like that in a 10 year time frame that would don’t know about.

The one thing I hope will not change is that approach that I outlined at the beginning. The customer obsession. The willingness to invent. The patience, letting things develop, accepting failure as a path to getting success.


Types of failure 17:07

By the way, one thing I should point out about failure, this is a fine point internally, we take it and we don’t need to know about it much.

But there is a different kind of failure, which is not what you want. That’s where you have operating history, and you do know what you’re doing, and you just screw it up. That’s not a good failure. That’s just bad operational excellence.

We’ve opened a 130 fulfillment centers. We’re on generation eight of our fulfillment center technology. If we open a new fulfillment center, and just woof it, we have to do some internal examination. That’s not an experiment. That’s just bad execution.

So there’s different kind of failure, and you need to make sure you’re making the right kind of failure.

The right kind of failure should be invention. It’s an experiment. You don’t know if it’s going to work. And you know upfront, you don’t know if it’s going to work.

That shouldn’t be opening a new fulfillment center for us.


Big ideas and things that will stay the same 18:16

Hopefully the approach will stay the same. And the other thing I would advise any entrepreneur or large business or large organization or government organization, you need to identify your big ideas.

There should only be two or three of them. A senior leader, the main job of a senior leader is to identify two or three big ideas and then to enforce great execution on those big ideas.

The good news is the big ideas are usually incredibly easy to identify. You shouldn’t need to think about them very much. You already know what they are.

Let me give you an example, for Amazon, the consumer business, the three big ideas are:

  1. Low prices
  2. Fast delivery
  3. Vast selection

The way you know they’re the big ideas is because they’re so obvious. The big ideas should be obvious.

And by the way, it’s very hard to maintain a firm grasp of the obvious at all times. Little things can distract you from the obvious.

But you have to backup, and say these are the three big ideas.

How do we always deliver things a little faster?

How do we reduce our cost structure, so we have prices a little lower?

And the good thing about these big ideas is that they will be stable in time. So I know for a fact, that 10 years from now, customers are still going to like low prices.

No matter what happens with technology and everything else, people are going to like faster delivery.

It is impossible for me to imagine a scenario where 10 years from now, a customer comes to me and says, “Jeff, I love Amazon. I just wished you delivered a little more slowly.”

This is so inconceivable that you can have great conviction as a leader to continue to put energy in driving speed of delivery.

I know in AWS, customers like low prices. They like availability or they don’t want the services to be down. They like data security.

It’s not very hard to figure out what the big ideas are, and then you can keep putting energy into those things. You can spin up those flywheels and they’ll be still paying you dividends 10 years from now.

And what I’m saying, I’m putting it in kind of a business context, for those of you that are in government, these principles would apply identically to a government organization.

You should figure out what the big ideas are, spin up flywheels, get those things rolling. Make sure you pick things that will still be true 10 years from now or 20 years from now.


Amazon’s approach to Artificial Intelligence 21:14

Artificial Intelligence, machine learning, this is, it is a renaissance. It’s a golden age.

We’re now solving problems with machine learning that were kind of in the rum of science fiction for the last several decades. And natural language understanding, machine vision problems, it really is an amazing renaissance.

And for decades, AI researchers kind of struggled and made very low progress. Over the last 10 years, it just looks like one of the S-curves. You’re just making a lot of progress over the last 10 years.

Machine learning and A.I. is a horizontal enabling layer. It will empower and improve every business, every governmental organization, every philanthropic organization. There is no institution in the world that cannot be improved with machine learning.

So at Amazon, some of the things we’re doing are super obvious. Superficially obvious. They’re interesting, kind of cool, and you should be paying attention to them.

I’m thinking about things like Alexa and Echo, our voice assistant. I’m thinking about our autonomous Prime Air delivery drones. Those things use a tremendous amount of machine learning, machine vision systems, natural language understanding, and a bunch of other techniques.

Those are kind of the showy ones. I would say a lot of the value we’re getting from machine learning is kind of happening beneath the surface.

It is things like improved search results, improved product recommendations for customers, improved forecasting for inventory management. And literally hundreds of other things beneath the surface.

I would say the most exciting thing that we’re working on in machine learning is that we’re determined, through Amazon Web Services where we have all of these corporations and software developers, to make these advanced techniques accessible to every organization even if they don’t have kind of the current class of expertise that’s required.

Right now, deploying these techniques for your particular institution’s problems is difficult. It takes a lot of expertise.

So you have to go compete for the very best PhD’s in machine learning. It’s difficult for a lot of organizations to win those competitions.

We’re in a great position because of the success of Amazon Web Services to be able to put energy into making those techniques easy and accessible. We’re determined to do that. I think we can build a great business doing that for ourselves, and it would be incredibly enabling for organizations that want to use these sophisticated technologies to improve their own organizations.


How do you interact with elected officials? 21:39

I think probably the normal ways. We have a fantastic public policy organization here. We have a bunch of people here in Washington, overtime we’ve grown that group.

We also do things at the state level. We’re doing certain things at municipal levels too. We work hard at it. It’s really important to work hard at it. You can’t just pretend that people know everything and you don’t have to educate them.

The main thing we try to do, and I think we’re very good at it, is stay very issued focused. So we try to stay on the things that are specific to Amazon and our customers, and our employees.

As a company, our employees are free to have opinions about every issue they choose to have an opinion about. They can have opinions about anything they want.

I feel like Amazon needs to be a little more laser focused than that. On things that are important to our customers and to our employees, as a group.

Our team does that, they keep us very disciplined in that way.


Sustainability and environment commitments 26:25

For us, we have done a couple things that I think everyone inside the organization is proud of. It’s been hard work, but I think it’s starting to pay off.

We have a program called frustration free packaging. We’ve been working on it for roughly 10 years. It’s a simple idea, but incredibly hard to execute.

We go to manufacturers and say you’re producing this thing in a four-color printed package with a cellophane window and lots of twisty wire ties. Or worst case, the clamshell packaging where they seal it in indestructible plastic.

I actually looked it up, there is a Wikipedia entry for packaging rage. There is a term for this, and that Wikipedia entry has the number of people that go to the Emergency room per year trying to open that indestructible, bubble packaging.

But why do manufacturers put things in that packaging?

That packaging is not cheap. It’s not cheap. It’s expensive, has a bunch of flaws. But it shows off the product well.

So when it’s hanging on a hook on store shelf, you can actually see the product as you walk by because it’s in clear plastic.

And of course, at Amazon, that doesn’t matter at all. We’re going to show you great pictures of the product and videos of the product, and customer reviews of the product. The product can be sealed in a tiny little cardboard box that is easy to open.

So we started this program, and we teach people how to make Internet packaging that is very efficient, easy to recycle, easy to open.

And something like just in the past year, we have saved 55,000 tons of waste as a result of this program.

We have now started putting solar cells on the roofs of all our fulfillment centers. We’ve done about 15 this year, these are big fulfillment centers, they’re about a million square feet each. So we’re covering them with solar cells and that will cover about 80 percent of the fulfillment center’s energy budget on an annual basis.

We’re the largest corporate purchaser of renewable energy. I don’t know the figure, but it’s something like 3.6 million megawatt hours per year. We’ve built solar farms and solar wind farms. We keep throwing this out.

We have a goal long term of being completely renewable, so it will take a long time.

The good news is these are not only great for the environment. They’re good business decisions too. The technologies have improved to build a solar farm or a solar wind farm now. These are competitive technologies with natural gas plans, so you can actually do it and feel like you’re making a sensible business decision.

So I think a lot of companies will do this in the future. I think they should do it. We’re having fun with all of that.


Space and Blue Origin 30:27

First of all, I love space. I have been a space love since I was a five-year old boy.

I feel like I won the lottery with Amazon. I know I won the lottery. And now I’m investing those lottery winnings in Blue Origin, which is the space company.

We’re flying a tourism vehicle, and hopefully start taking people up in 2018. That’s coming right up. We’ve been working on it for more than 10 years.

This overnight success is taking longer than 10 years.

We’re also building an orbital vehicle called New Glenn. New Shepard is named for Alan Shepard, the first American in space. New Glenn is named after the American hero, John Glenn, who was the first American to orbit the Earth.

These are reusable boosters, fully reusable. That’s the key to lower the cost.

Our vision at Blue Origin is we want people millions of people living and working in space.

My personal hope is I live long enough to see kind of the dynamism in space. I want to see the whole economy, the entrepreneurs in space that I got to witness in the last 20 years on the Internet.

The problem with being an entrepreneur in the space arena is the price of admission is so darn high. To do interesting things in space, the kind of beginning or entry level is a few hundred million dollars.

You’re not going to get two kids in their dorm room doing something interesting in space. Whereas that’s literally what happened to Facebook.

On the Internet, the price of admission is so low, you can get these amazing experiments. Where one kid in a dorm room does something and it turns into Facebook.

Or Amazon’s case, I started this with an incredibly small amount of capital and was able to grow because we didn’t need a lot to begin with. The heavy lifting was in place for Amazon.

I didn’t need to build a transportation network. It existed already. It was the UPS, Royal Mail, US Postal Service, and Deutsche Post.

I didn’t have to build a telecommunications backbone to connect with my customers. It was there. It was called the Internet.

And in fact, the Internet was resting on top of the long distance phone service of that time. You guys remember the dial-up modem. Some of you do. Some of you should go to a museum to see a dial-up modem.

I recently showed a group of youngsters a payphone. I came across one,and I was like, my god guys, come over here. This may be the last one. You have to see it!

The point is Amazon got to rest on top of, we didn’t have to build a payment system. It already existed. It was called the credit card.

Computers were already on every desk thanks to Microsoft, IBM, and Apple. But what were they being used for? To play games, not to buy things on Amazon.

So all that heavy lifting was in place.

I would have such a good feeling if I could be an 80-year-old guy, laying there thinking about my life, and say there is now a bunch of entrepreneurs in space because I took my Amazon lottery winnings and built the infrastructure that does take billions of dollars in CapX to lower the cost to access to space. That’s the way you get millions of people living and working in space.

By the way, we need that. Those of you that think about the future, at all, you can do a simple calculation. We can argue about limited resources on Earth, so on and so on.

But here’s the calculation that you can’t argue with. That is you take current baseline energy usage on Earth, compound it by just a few percent a year for just a few hundred years, and you have cover the entire Earth in solar cells.

We’re going to have to decide do we want a society of pioneering, invention, expansion, growth? Or do we want a society of stasis?

Because the Earth is finite.

And personally, I don’t believe that stasis is compatible with freedom. For me, that’s a big problem.

Second of all, it’s going to be dull. Stasis is really dull. You don’t want to live in the stasis world.

Of course, we’re going to get more efficient. For the past hundred years, we’ve been getting more productive, more efficient. That trends is going to continue. But even so, we’re going to way to use more energy. More energy per capita.

And also, I don’t want to stabilized population. I want there to be a trillion humans in the solar system. With a trillion humans, we’d have a thousand Einsteins and a thousand Mozarts.

It would be an incredible system. Don’t you want that dynamism? It would be so much more interesting.

This is for your great, great, great grandchildren. What kind of world do you want them to live in?

I want them to live in that expansive world that is learning more about the universe and expanding out into the solar system. So we have to do it.

It’s fun to work on that.

I won so many lotteries in my life. My mom, Jackie, and my dad, Mike. My mom had me when she was 17 in Albuquerque, New Mexico. She was still in high school. It was not cool. And she did an amazing job.

My dad, Mike, is a cuban immigrant that came here when he was 16 without his parents. Did an amazing job.

So those are the things that are your little gifts in life.

My greatest admiration is held for those people who had terrible parents and still somehow made it through. That’s hard.

I had the opposite, so I got lucky.


That was a fun conversation.