Amazon CEO on leadership, innovation and customer obsession

- Welcome to another special episode of our "Capital Ideas Podcast," part of the limited series, "The Power of Advice."

At Capital Group, we believe individuals are best served through financial advisors, advice that can be critical to protect and grow their assets.

Today, we have the pleasure of speaking with Andy Jassy, president and CEO of Amazon and the founder of Amazon Web Services.

Andy has led Amazon from strength to strength during his tenure, and we talk about the advice and decisions that shaped his path to becoming CEO and how he fosters a culture of innovation with speed, intentionality and long-term focus.

Andy, welcome to "The Power of Advice."

- Thanks for having me. I appreciate it.

- I'm glad you're here.

We have a lot of A's to talk about: Amazon AI advice.

But I want to start with the A for Andy.

And I think everyone, or most people listening, are Amazon consumers.

I am, I know a lot of people are.

- Thank you.

- Which is a good thing.

It's super user-friendly.

What's the one thing that you've bought on Amazon recently that people would say, "God, I can't believe the CEO of Amazon bought that?"

- I don't know if I've bought anything that surprising.

Maybe Ruffles potato chips every Sunday.

During the football season, I have three friends who grew up in New York who moved to Seattle around the same time I did, and we have this group called Giants Anonymous, and we're New York Giants fans, and we've watched- - You're no longer anonymous.

- I know, well, this has been a tough year for us, but we have some pieces, just wait.

It's coming, Mike. - All right, I'm waiting.

- We've been watching the games together since 1998.

And so, I have to stock the, they're 10:00 a.m. games on the West Coast, so I have to stock it with bagels and then an array of potato chips.

People forget we have the third largest grocery business in the U.S.

It's a $150 billion business in groceries.

I often wake up early Sunday morning and order it, and it's there by game time.

- That's a good tradition.

Early on for you, you were an athlete.

Soccer, tennis, other sports, I'm guessing.

What made you say, "You know what?

I'm going to stop."

What caused you to give up on that?

- Reality.

(both laughing) - Talent?

- Yeah, I just wasn't good enough.

I mean, growing up, I really wanted to be a professional athlete.

- That was your goal, then?

- That was my, I have a very sports-heavy family.

My dad is a sports fanatic.

From the time I was three years old, he was taking me every Sunday to the Giants games and to the New York Rangers games.

I played as well and I wanted to be a professional athlete, but as I became a teenager, and then I played a little bit of soccer in college, I just wasn't good enough to be anywhere near being a professional.

And so, that's why I stopped.

But I feel like sports taught me so many amazing lessons that have been so useful in my life.

Tennis really taught me the value of hard work, what hard work looks like and what you get from it, and what happens when you don't.

And then, I also think that it teaches you how to move on quickly.

You can have your best win of the year at 9 a.m. on a Saturday, have lunch, and go out and play a match at 2 p.m.,

and it's all gone.

You lose to somebody you shouldn't and you play terribly.

It's totally different from the morning.

And so, you have to learn when things go well, you have to move on, because you have another match to play.

And when things, when you don't win, which is most of the time, you also have to find ways to get back up and to move on and turn the page and keep moving and take the good from the experiences.

And so, I just feel like sports has been incredibly impactful in my life.

I wasn't good enough to be a pro, but I'm glad I had the chance to do it.

- So, you've tried a bunch of things, and before you got to Amazon, sportscaster, maybe paralegal, investment banking.

Is there anything you didn't try that you looked back and said, "I wish that was one of the things on the list that I did try?"

- Well, I did, I tried so many things, and it wasn't really by design.

I just didn't know what I wanted to do.

And I tried, I wanted to be a sportscaster.

Then, I did sports production.

I coached, I was the assistant coach of my high school soccer team.

I worked in a golf retail store in New York City.

I tried sales.

I was a paralegal one summer.

I did investment banking.

I started a couple businesses.

I think when, if you don't know what you want to do, it's useful to try.

And I think at that age, it's almost as useful to know what you don't like doing as it is to know what you do like doing.

- All right, let's move on to another A.

Let's start talking about Amazon.

When you started working with Jeff, I know there were times where you had to go in there and you're presenting ideas.

And at one point, you had a really long deck.

And I watched you on another podcast, and it was a really interesting story where there were a few things with the deck that weren't great.

So, whether it was the accuracy or whether it was the length, it wasn't the best deck you'd ever presented, but you had a good story behind it.

Can you share that with us?

- Yeah, at the time, I was co-leading our marketing team.

This was before we moved away from PowerPoint to doing narratives.

And we had about a, I don't know, maybe a 180-page PowerPoint presentation.

- Super long, super long deck.

- Crazy, and it wasn't that unusual at that time.

- No, I mean, people- - And it was marketing for the whole company.

And so, the meeting was a six-hour meeting, and it was our operating plan.

And the plan was I was supposed to present the first 80 of these, and then my partner was supposed to do the next 80.

I was about six or seven slides in, and I started to present a slide on customer retention, and Jeff interrupted and said, "All the numbers on this slide are wrong."

And I said, "Why do you say that?"

And he started to go through a couple of the numbers.

And you know when someone is, like when you have something wrong and someone starts to dissect it, and you realize, "Uh-oh."

- There was that moment, like, "Oh, yeah, that's wrong."

- As he was talking, I realized, "He's right."

And so, he got done.

I said, "You're right.

The numbers are wrong."

And he said, "Well, why should I trust anything I hear the rest of this presentation?"

I said, "Well, I hope you will, because we have like 173 more slides to this deck."

- If you didn't like what you saw on page three, I got 173. - There's a lot more coming!

And we got through the presentation, and I would not say it went well.

It wasn't as bad as that the whole time, but there were all sorts of things we didn't get quite right.

But in many ways, it was a real positive turning point for me at the company and in my own professional career, because probably my biggest fear every time I got to different levels in the organization was presenting to new people, was that they would think I was dopey and that was going to be the end of my career.

- Imposter syndrome, right?

- Yeah, so my worst fear always was that I would be up in front of people and they would realize I didn't know what I was doing or didn't know what I was talking about and what would happen.

And here was, really, my worst fear, and it was okay.

I got through it.

I think we represented ourselves well.

I think we got a number of the points we needed to get across.

And then, I got back to work the next day, and people seemed to still treat me like I belonged on the team.

And that was a really good, it's an obvious lesson, but I think it is a fear that a lot of people have.

And most good companies with most good leadership teams realize that who you are is broader than just one experience.

- I think that's right.

Let's dial back to AWS, because I think Amazon Web Services is pretty remarkable.

Talk us through how you saw a problem that you thought AWS could solve.

And today, obviously, it's where it is, which is an industry leader.

But how did that come about?

- Well, interesting, just following on the story I was just telling.

A year later, Jeff asked me to, Jeff Bezos, our founder and CEO at the time, asked me to be his, what he called at the time, his shadow, which really was most closely akin to a chief of staff role.

And in this role, I spent all my time with Jeff.

I spent time in all his meetings, including his one-on-ones, which didn't really make them one-on-ones anymore.

And then we would talk every week about the things that we felt like we needed to drive differently and divide and conquer, how to do that.

And because I'd been at the company for five years and knew a lot of the people on the team, I was able to help Jeff with that.

One of the things that was really frustrating us around that time was, even though by most companies' standards, we moved pretty quickly at Amazon, we just felt like it was too slow.

I would say the time I've been at Amazon, which is 28 1/2 years, we have always felt like we moved too slow, even though we move a lot faster, I would say, objectively, than most companies.

It's never fast enough.

And so, this was really bothering Jeff.

We didn't understand why we couldn't go quicker.

And I would talk to a lot of my peers, and they would say, "Look, I think it's great that you and Jeff think these projects should take three months end to end, but I spent three months just on the compute solution, or just on the storage solution, or just on the database piece."

And all my peers are doing the same thing.

We're all recreating and reinventing the wheel, and we spend about 70% to 80% of our time on all the infrastructure pieces, and only about 20% to 30% of our time on what differentiates what we're trying to build.

And again, in retrospect, it seems fairly obvious, but at the time, we really didn't understand that.

And that made us think, if we could flip that equation on its head, we could move so much more quickly.

So, if we could take what really is the undifferentiated heavy lifting of infrastructure- - The foundation, the infrastructure.

- Yeah, you have no technology if you don't have the right compute and storage and database and analytics and messaging and content distribution.

All those things really matter, but they rarely are the pieces that differentiate your customer experience.

So, if we could take that undifferentiated heavy lifting and do it- - Scale that.

- in a scalable, operationally performant, cost-effective way so that all our own builders could focus on what was unique about the customer experience they were trying to build, we figured that would help us move much more quickly.

And as we started defining what that might look like, you pretty quickly realized that other companies have the same problem, the same challenge.

And we thought, if we had this issue and we were a strong technical company, a lot of other companies would probably find benefit and solutions there.

And that's how we started going about building AWS.

We launched it in early 2006, and it grew quicker than we thought, and it's become a pretty big business.

- But you were owning that in your mid-30s.

I mean, that's a pretty transformational project for a company the size of Amazon to own.

Even though Amazon was smaller back then, it's still a material company, and you were owning this massive project in your mid-30s.

- Yeah.

- That's an important leadership role.

So, you're a shadow for Jeff, but you're also driving what now is a real driver for Amazon organization.

- I left that shadow job to go lead that team, build that team, and build a set of services.

And in the beginning, for the first few years of AWS, people inside the company and outside the company, but at first, nobody knew it outside the company.

People in the company thought it was kind of nutty.

- What are we doing here?

- Yeah. We're a retailer, what is this?

This has nothing to do with retailers.

It's actually developers who are customers.

What do we know about developers?

Early on in AWS when we were building the first set of services, almost every month, one of the best people on my team would come in, a different person, by the way, a half dozen would come in and say, "Why are we doing this?

Why don't we just build APIs for e-commerce?

That's what we know.

Who's going to buy compute from us?

Who's going to buy database from us?"

I would have these conversations and I would say, "Well, look, we can always go back to doing the commerce APIs, but if we want to go after the broader infrastructure opportunity, we have to be first to market or we probably don't have a chance."

Because people didn't know us as an enterprise technology provider.

I remember, I would go home at night, and my wife would ask me how my day was.

And I would say, "I feel like, every day, like I'm almost doing therapy for the team, for myself." And she would say, "Well, are you sure this is going to work?"

And I would say, "I have no idea if it's going to work.

I really don't know."

But I think for- - Truthful answer, but- - I think we, there were some signs internally that this was needed, but whenever you're doing something new, you don't really know.

But I think when the team and I started to get focused on what we could control, we were much better off.

We couldn't control if people like the services, if people were going to use the services, if people would stop building on-premises data centers and use these cloud services.

But we could control who we hired, what products we built, the technology we used, the pricing, all the go-to-market pieces, the customer experience, the security, operational performance, the delivery of it.

And so, once we got focused on controlling what we could control, we just had a lot more success and a lot more fun.

And so, being in my 30s at the time, I just thought it was another new business I had a chance to be a part of starting.

I don't think anybody, including our team, thought it was going to be as big as it turned out to be.

I don't think anyone was feeling like, "Oh, my God, we're working on this job."

I mean, it is now a $132 billion annual run rate for AWS, but back then, we were just consuming a lot of capital and operating income and hoping we could build a business that really mattered for customers.

- One of our core values at Capital Group is client focus.

One of your 16 principles at Amazon is customer obsession.

Let's talk about that a little bit.

You have a lot of different customers, you have these different businesses.

How do you keep this massive organization focused on the customer?

Because the customer's not stagnant.

It's a moving target.

Their preferences change.

How do you keep the team focused on that?

- Well, I think there are several ways you try.

The first thing is who you hire is the starting point.

When we hire people, we try to disproportionately hire for people who seem to really have passion about providing amazing customer experiences, and that is the compass for what drives what they want to do and what they want to accomplish.

So, I think we inherently have a number of customer-focused people, and you have to find ways that reinforce it at every turn.

When we hire people, when we promote people, when we give people's reviews, it's always based on how they perform against the leadership principles.

So, if somebody is not being customer-focused enough, they're unlikely to get hired, they're unlikely to get promoted, and they're going to see it in their performance.

- So, it's built into the system.

- It's built into all of it.

The second thing is we have all sorts of mechanisms to reinforce it.

So, if you were to stop in the weekly business review of AWS, as an example, it's an unconventional business review.

It's not just going through a deck of all the numbers.

The first 30 minutes of that 90-minute meeting tend to be around key customer issues.

Has nothing to do with any of the metrics of the business.

It's around, what are the key issues that we're seeing customers face?

And the leaders in that meeting, if there's something that looks wrong, the leaders effectively tell the team to stop what they're doing, drop it, and go fix it.

And sometimes, even just leave that meeting and go fix it.

And that is very clear.

When you model the behavior that you want, and you're teaching people the expectations, if you're constantly reinforcing that what matters most is the customer experience, you start to get everybody in the organization to behave that way.

And then, the way we build products is also instructive of this, which is we won't write code until we work on these working backwards documents, which are a press release and a frequently asked questions doc.

And the press release is intended to write out, if we built this product, why will people find it remarkable?

Because a lot of times, these products take twists and turns, and you get to the press release, the real press release, and you ask yourselves, "Why do we think anyone's going to care about this product?"

- So, you're starting with the why before you get to the how.

- Exactly, we start with, "What is this that we're building and why will people care about it?"

And we argue with one another and debate.

Does this matter for customers that much?

Is it moving the, is it really what customers want?

These are all parts of trying to reinforce that what matters most is what you build for customers.

- And like you say, leading by example, starting those meetings with the customer first, you're going to get to that customer outcome.

I think that's a great way to do it.

Let's transition to AI.

Everyone wants to discuss it for obvious reasons.

You've said before it is likely the most transformative technology in our lifetimes.

And I don't disagree with you, but why do you think that?

- Because the technology is so powerful and enables you to do things that you haven't been able to do before.

And so, if you look at, today, if you look at what's happened with AI, it's been incredibly good.

Now, I'll say this generation of AI.

AI has happened, been happening for 25 years.

But with transformers and generative AI, it's allowed you to take giant swaths of information that's been all over the place and time-consuming, and very quickly be able to pull that data together to be able to find needles in the haystack for what people are asking for, to summarize it, and then to fluently communicate back.

And that is, that by itself has been very remarkable.

You see that a lot in some of the chatbots, and ChatGPT has obviously been a runaway success, and things like Alexa+, which is our latest version of Alexa, or Rufus, which is our shopping assistant.

These assistants where you can ask it a question instead of you having to search all over the place and click on links, and it just takes the information you want and finds what you want, summarizes what, compelling.

It also does it visually, by the way.

If you look at our "Thursday Night Football" telecast, just the ability to look at all the data of all the formations of this team in prior games and be able to highlight in real time, with a circle, which players are likely to blitz the quarterback based on where they're lined up, that is really interesting.

- I've that, that's pretty cool.

- Or to look at the pocket of a quarterback with the offensive line and being able to predict, assess and then predict which pockets are healthy and which are about to collapse.

Those are things that you just couldn't do quickly before without AI.

Increasingly, now you're starting to see AI be able to take actions for you.

So, whether, look at how much coding today is being done when you tell AI, a developer tells AI, "Here's what I want to actually accomplish."

Today, we have an agentic coding service called Kiro, where you can tell Kiro, "Here's what I'm building, here's what I want to build, here's the spec.

Go do the coding for me."

As you're getting into this more advanced reasoning, you're able to do things like these frontier agents that we just launched, where you can point an AI agent at all your documentation, all the code you've built, the repository, and then show them the list of bugs and initiatives you want to get done, and they can just go in and do the work.

That's a different level of reasoning that you're starting to get to now with AI.

And you're going to see over time just the intelligence and the capabilities of the reasoning and this sorting through how to get to trillions of combinations, which are the right ones.

It's going to ultimately help us solve some of the worst health problems of our lifetime.

It's going to let us do materials sciences and create.

The power of the data and the models as you can train it on more and more data, and as the reasoning gets better, it just changes what's possible.

And there will be all sorts of things we learn along the way, and it won't be perfect, and some will be better than others earlier.

There'll be new jobs that people start to do.

But it's going to be very different, in my opinion.

- Yeah, I totally agree with you.

You talk about five years from now, drones and the delivery you're working on there.

I saw some stat, you can correct me, but that hundreds of millions of deliveries you expect by 2030 via drone.

- Yeah. - Actually,

I'm going to give you a mug showing how grateful I am that you're here soon.

That mug came through that door via drone.

We were practicing this morning.

Came right through that door, and it landed on this table.

What's your over/under?

You're a sports person.

What's your over/under on 250 million deliveries via drone through Amazon five years from now?

- Over.

But we'll see.

I mean, the number we gave is we expect to have a half billion by the end of the decade.

I think people felt like, in the beginning, that the drones technology was a little bit of a science project.

And I would say it took a while to really figure out the right architecture for it to deliver at the cost and the speed and the right environmental thoughtfulness and noise level.

But we have a design now that really works, and we're using it in a number of cities in the U.S.

- Let's get into "The Power of Advice."

It's what we call this special segment, and it's where we get advice from folks like you who are generous enough to come on here and share that with us.

What do you tell recent graduates?

Recent graduates, there's a lot of data that's showing that it's more difficult for recent graduates to get jobs.

What advice do you have for them?

- The things I would probably share is, I think, number one, I don't think you have to know what you want to do at 22.

I have a 22-year-old son.

I love his college friends, but a lot of them, I think, feel pressure to know what they're going to do for their life.

And I think it's great to have an idea.

But as I was saying earlier, it's- - You tried a lot of things.

- very useful to try a lot of different things to figure out which things you don't like, which things you do like.

So, I would really encourage people to try things, not to be worried if you don't know exactly what you're going to do, to pick some paths and pursue them.

You never know which things you're going to like.

In my lifetime, I have not predicted the things that I have loved.

I thought I was going to do sports.

I would not have predicted that I would work on a set of technology infrastructure services.

I think the second thing is, if you can get over the idea that every time you're being exposed to somebody new, that it's a pass/fail referendum on your competence, you're going to be better off.

It just puts undue pressure on people, and it's just not the way the real world works.

- Yeah, agreed.

- I think, relatedly, you just have to have resilience.

And there are going to be a lot of times where things don't work out the way that you'd hoped.

You are going to face adversity and you are going to fail and things aren't going to work out, and you just have to realize it's okay.

It happens to everybody.

You wake up the next day and you start over.

But you have to be resilient.

I think another thing is, I was talking with a young person in college the last week or so, very smart person, asking them about their, what they wanted to do in the summer, and maybe full time.

And we were having this conversation, and they were rejecting certain things under the guise of, "Well, it's not really my dream job."

And I would just say, at 20- - It's a bit early for your dream job.

- You probably, you could get your dream job, but that would be complete luck.

I think, early on, you really have to be willing to pay your dues.

You have to be willing to start at the bottom.

You have to do whatever people ask you to do, within reason.

But you have to get good at the details.

You have to learn to deliver.

You have to learn to be reliable.

And all that learning, apart from being the right way, I think the right approach to have, all those skills are skills that you learn from having to do all the little, hard, detailed jobs.

And they build a skill set, they build knowledge, they build pattern recognition.

But if you aren't willing to start at the bottom and pay your dues, it's unlikely that you're going to ever be successful.

Yeah, I think the last thing I would say, and maybe the most important thing of all, is you just have to be a learning machine.

I would say the biggest difference in the people I started working with 28 1/2 years ago at Amazon and what they're doing today was their ability to learn.

And it seems obvious that you should learn, but it is- - But not everyone does that.

Some people get stuck in their job and they don't get, they're not curious enough and they're not trying to live outside of the four walls and learn things that people might know outside of Amazon.

- You're 100% right, Mike.

If you're constantly learning, if you're hungry to learn, you take feedback differently, you accept failure differently, you listen to people in meetings or around differently, you get counsel, you read, you're just ravenous to figure out how to learn.

I see this with a lot of senior people, too, which is they get to a certain point, and somehow they've decided they know what there is to know.

To me, the second that you think you know what there is to know is the second you're unraveling.

It may take you a while, a couple years to unravel, but that is the beginning of the end.

And I look back every three to six months at what I'm doing, and I always think, "Wow, I've learned so much.

It's so different than I thought."

And so, I just think, for young people, having that approach that you want to constantly be learning will help you get better and will help you get feedback that makes you better.

- Great. I'm going to finish us up with a handful of questions.

We started with Andy, we're going to finish with Andy.

Favorite tennis player?

- Current or past?

- Both.

- I was always a big John McEnroe fan.

Growing up in New York, playing tennis, almost everybody from New York worshiped John McEnroe.

- Got the haircut, too.

- I wish, I wish I had the hair still.

(Mike laughing) - Close to it, a lot more than I got.

- And probably current, probably Alcaraz.

I just love his enthusiasm and athleticism.

He seems to love the sport.

- He's got a long runway as well.

Band you'd want to be a member of?

- Foo Fighters, it's my favorite band.

It's hard to not love Dave Grohl and that whole band.

They are, in my opinion, very underrated.

I think that may be the best rock band of our lifetime.

- Chicken wings, how many can you eat in a sitting?

I know you're into quality, not just quantity.

What's your record?

- My record is 57 wings.

We have an eating club that we started at Amazon.

When we first got to Seattle, we didn't know anybody, so we used to go for Buffalo wings every Tuesday night.

And there were about a dozen of us at work that did this.

And then we eventually turned that into a eating competition once a year.

So, it's a two-person competition, and I had 57 wings and I really had difficulty standing when it was done.

(Mike laughing) - All right, fun!

A leader outside of Amazon that you admire?

- I'd say my dad.

Yeah, my dad, he is one of the very smartest people I know, incredibly insightful thinker.

He has high integrity.

He's humble and he's very level-headed.

And I have looked up to him for a long time and try to emulate whatever I can.

- That's a great mentor to have.

New York Giants 2026, how many wins?

- Seven.

I hope it's more, but I'll say seven would be good progress.

- Seven would definitely be progress.

I'm going to end with the mug, the aforementioned mug.

And we always end with the mug on the podcast we do with our own investment professionals.

And it says, "Obsessively grateful."

And we ask folks, "What are you grateful for?"

And I have a view that all of us, this is not a Capital Group, an Amazon comment, it is a societal comment that we could have more gratitude.

So, what are you grateful for?

And I'll give you this and we'll send it to you.

We won't make you travel with it.

- Thank you, I appreciate it.

- But what are you grateful for?

- Gosh, I'm grateful for so many things.

Let's see if I can keep the list manageable.

I mean, starting with my family, I grew up with amazing parents and great siblings.

I have an unbelievable wife who's my best friend, and I would not be doing what I'm doing without her.

Amazing two kids.

One of the luckiest parts for me over time has been the friends I've made.

I don't know how I got so lucky, but I have just an amazing group of friends who are almost like siblings to me, and I get to see them a lot.

And the people I work with.

I used to think I could work with anybody.

And then, you work at a bunch of places, and you realize there are differences.

And I work with such a group of smart, ambitious, hungry, customer-focused, inventive, missionary people, that it's why I'm still at Amazon 28 1/2 years later.

I feel grateful that I have my health.

You always take it for granted until you don't have it, and then you've got nothing.

I'm grateful that I have been able to experience all sorts of adventures throughout my life, and they continue to pile up.

But as much as I work and as passionate as I am about work and as much time as I spend there, I've been lucky enough outside of work to have a lot of different adventures and to have some balance.

- I appreciate that.

Andy, thank you again for spending the time with us, sharing some wisdom and advice.

- Great to be with you, thanks for having me.

- Appreciate it. - Appreciate it, Mike, yeah.

(upbeat music)