Jeff Bezos – March 1998, earliest long speech

Good evening and welcome to the annual A.B. Dick lecture on entrepreneurship at Lake Forest College. Lake Forest College, 32 miles north of Chicago, was established in 1857 as a private co-educational liberal arts college. Lake Forest College engages our eleven hundred and fifty students in the breadth of the liberal arts and the depth of the traditional disciplines. Lake Forest College draws its students from 43 states and 44 foreign countries. Faculty at Lake Forest College know their students by name and encourage students to read critically, reason analytically, and, above all, to think for themselves.

The Alfred Blake DIC endowment recognizes the long-standing connection between the Dick family, Lake Forest College, and our community. It is a particular pleasure to note that members of the Dick family are in attendance this evening. Our distinguished speaker this evening validates three key words that are hallmarks of contemporary culture, and they are risk, strategy, and technology. Jeffrey P. Bezos likes to tell his story this way: he has always been interested in anything that can be revolutionized by computers. Upon learning about the dramatic increase in the use of the internet, he focused on creating a new way to retail: the worldwide web.

Unique ability to deliver incredible amounts of information quickly. From this came a well-calculated business strategy that culminated in the founding of Seattle-based amazon.com just four years ago. It is also important to share the observation offered by a correspondent writing for The Wall Street Journal. Mr. Bezos, they wrote, has quietly built a fast-growing business where the world's mightiest merchants mainly have racked up failures. Amazon.com has pinpointed one of the few products that people really want to buy online: ebooks.

Jeff Bezos is a Phi Beta Kappa graduate, class of 1986, of Princeton University. He earned a Bachelor of Science degree, summa cum laude, with majors in electrical engineering and computer science. Now, just before we begin, I wish to warmly acknowledge the presence this evening of Jeff Bezos's aunt and uncle, Nonie and David Brock, members of Lake Forest College class of 1966. The fact is that they had something to do with attracting him to our campus this evening. [Applause]

And it is with great pleasure that Lake Forest College welcomes, as this year's A.B. Dick lecturer on entrepreneurship, the founder and CEO of Amazon.com, Jeffrey P. Bezos, who will discuss building Earth's biggest bookstore. Welcome! [Applause]

Well, thank you! It's a great pleasure to be here. Thank you for the very nice introduction. The same Wall Street Journal correspondent that you mentioned said so many nice things in that article that I actually forgave him for also noting in the very same article that I have thinning hair and a receding hairline. So there you have it!

I actually can't resist at any gathering like this, even though it's anecdotal and there's a huge bias in the crowd of people who come. But how many people here have been to Amazon.com?
Raise your hands. Does anybody buy anything at Amazon.com? Yeah, okay, that's not bad! Thank you, we really appreciate that. And after the lecture, I invite those people to come to me and get a handshake of thanks. Also, I'd love to hear how we can improve the customer experience for you.

I can tell you a little bit about Amazon.com, a little bit of the story of sort of how it came to be. The company was conceived in the spring of 1994. I came across a startling fact: in Spring '94, web usage was growing at 2,300 percent a year. You have to keep in mind that human beings aren't good at understanding exponential growth. It's just not something we see in our everyday life. But things don't grow this fast outside of Petri dishes; it just doesn't happen.

When I saw this, I said, "Okay, what's a business plan that might make sense in the context of that growth?" I made a list of 20 different products that you might be able to sell online. I was looking for the first best product, and I chose ebooks for lots of different reasons, but one primary reason, and that is that there are more items in the book space than there are items in any other category by far. There are over 3 million different books worldwide in all languages. The number two product category in that regard is music—there are about 300,000 active music CDs.

When you have this huge catalog of products, you can build something online that you just can't build any other way. The largest physical bookstores, the largest superstores, and these are huge stores, often converted from holding alleys and movie theaters, can only carry about 175,000 titles. There are only a few that large, and in our online catalog, we're able to list over two and a half million different titles and give people access to those titles.

Being able to do something online that you can't do in any other way is important. It's all about the fundamental tenet of building any business, which is creating a value proposition for the customer. Online, especially three years ago, but even today and for the next several years, the value proposition that you have to build for customers is incredibly large. That's because the web is a pain to use today. We've all experienced the modem hang-ups, the browsers crashing, and there are all sorts of inconveniences. Websites are slow; modem speeds are slow. So if you're going to get people to use a website in today's environment, you have to offer them overwhelming compensation for this primitive infant technology.

I would claim that that compensation has to be so strong that it's basically the same as saying you can only do things online today that simply can't be done any other way. That's why this huge number of products looks like a winning combination online. There is no other way to have a two and a half million title bookstore; you can't do it in a physical store, but you also can't do it in a print catalog. If you were to print the Amazon.com catalog, it would be the size of more than 40 New York City phone books.

So it's also very important when planning a business like this, I think, to sort of look at what is the brand promise that you're going to make to customers. The brand promise that you make has to actually coincide very closely with the things that you can deliver. The kind of important, but sometimes overlooked, component, I think, in our case led to the name of Amazon.com. It's a mistake versus the biggest river, Earth's biggest bookstore. We wanted to use this large selection to be able to build an authoritative store. Authoritative means a few things: authoritative selection, authoritative prices, and authoritative information about books so that you can make a purchase decision. And the internet allowed us to do all those things.

We could have more selection. We could also have lower prices because we don't have to have retail real estate in high-traffic areas. That's very expensive. Inventory in those spaces is very expensive. We get to inventory our books in very low-cost warehouse space. It's also possible to have more information online about products than you could ever have in a physical shopping environment, so you can really build an authoritative store.

And just for those of you who don't read National Geographic nearly often enough, the Amazon River is ten times the size of the Mississippi River in terms of volume of water that flows through it. And there are these great pictures every decade or so in National Geographic, where they superimpose a cross-section of the Amazon in a cross-section of the Mississippi, and the Mississippi looks like a little creek.

So we opened the store in July of 1995, almost three years ago now, after doing about a year of software development, and the store has grown very rapidly. At this point, our sales are about the equivalent of 50 of those big chain superstores. The interesting thing is that we opened the equivalent of twelve of those superstores in just the last three months, so things are growing fast. We ship books regularly to over 160 different countries. In our first 30 days of being open, we shipped books to more than 45 countries in all 50 states.

There are some interesting stories about the international aspect. For example, we ship four orders a day at least to Bulgaria every day. To show you the sort of the exhibit: even though Bulgaria only has the equivalent of 41 lines of bandwidth coming into the whole country, that's significant. Amazon.com has about 60 times as much bandwidth as Bulgaria coming into our one company.

We also had people; it's very difficult to get English-language books outside the US. About 22% of our sales are outside the US. People go to great lengths because they want access to information; they want access to things that help them and educate themselves and so on.

We had an order from Romania, and the person didn't have a credit card, which is the most common way that we get paid. This was the most unusual way that we had ever gotten paid. The person sent us cash, which we don't encourage. They sent it from all the way from Romania. It was US currency, but it was very interesting. They had taken two crisp $100 bills, folded them up into a tiny little package, opened up the door of a floppy disk, and put the money inside a floppy disk. Then they mailed us the floppy disk with a note on the cover of the floppy disk that said, “The customs inspectors steal the money, but they don't read English. It's inside the floppy disk.”

Sure enough, our erstwhile accounts receivable people pried open the floppy disk, and there was $200. The person got their books, and it's a happy ending.

I believe that the primary success factor for Amazon.com has been an obsession over the customer and a focus on customer service. This is clearly important in any business but especially important online. The reason for this is that word-of-mouth is incredibly powerful online. If you make a customer unhappy on the internet, they won't tell five friends; they will tell 5,000 friends on Usenet newsgroups, listservs, and so on. Likewise, of course, if you can make a customer happy by meeting or exceeding their expectations, they can also tell 5,000 people and therefore be evangelists for you. They can use the internet as a megaphone.

So it's very important to do the things that you actually are going to claim to do, and what this means we see is the balance of power has shifted away from the merchants and toward the consumer, which is a great thing, I think, for everyone involved. It's even great for the merchants as long as the merchant realizes it.

One of the statistics that we track most carefully at Amazon.com, because we believe that this is so important, is the fraction of orders we receive every day that are from repeat customers. That number has consistently gone up and it's now over 58% of our orders that we receive every day, even in the context of very rapid new customer growth. One of the great things online is that the customers help you figure out what you're doing wrong and how to do things better. Email is this great medium for receiving feedback because somehow, email turns off that little piece of everyone's brain that causes them to mostly be polite.

So, what happens is, you know, if you go into a restaurant and have a bad meal, very rarely are you trying to have a good time. You ever go into the kitchen, grab the chef by the scruff of the neck, and say, “You know, you really shouldn't be cooking!” But online, with email, that will happen in a heartbeat. We've benefited greatly from that, and I believe from the three years that we've actually been doing business, I suspect we've received more feedback, both positive and negative, people helping us improve our service, than probably most business establishments have received in the last 20 years.

Just very briefly, since a lot of you have actually visited our website and noticed how great we're doing, let me just briefly hit on some of the key features of Amazon.com. We do have two and a half million titles, and that is something... At all these things, we try to package the other things that you can't do without the internet because that's where you'll create real value: two and a half million titles. We try to focus on convenience, so we have something called one-click shopping. This is a new feature we just launched a few months ago that’s been very successful and allows repeat customers to literally order with one click of the mouse.

When we did focus groups and tested this new feature before launching it, the biggest problem was that people didn't really think they had finished the order. So we had to actually change the surrounding of this thing so that not only would it say "Thank you for your order," but in parentheses, it now says, “Yes, it really did place an order.” This works because we keep the shipping address on file, we keep the credit card number on file—something that you can do online that's very difficult to do in any other way because we have the low cost of not having retail real estate and also the efficiencies; the personnel efficiencies are very high of doing business this way.

We generate, for example, more than $300,000 a year in revenue per operating employee today, even though we're still inefficient in honing our techniques compared to physical bookstores that generate about $95,000 a year in operating revenue per employee. Now, because of that, we can afford to charge lower prices, and basically what ends up happening is we offer bestselling level discounts, but instead of on the 65 New York Times bestsellers, on 400,000 bestsellers.

We change our website every day, and the reason we do that is to encourage people to come back. We're happy always when people come to our website, even if they don't buy anything because occasionally they will. One of the interesting things about the book space, and every business environment is different, and every industry is different, is that people don't just shop for books for convenience. They don't shop for them just because they need them. It's actually an incredibly fun thing to do. Many people will spend two hours on a Sunday afternoon in a bookstore.

There are other products that aren't like this at all. For example, pharmaceuticals. You know, nobody browses the preparation age counter. You go in, you go out, you're done. You want that to be a painless experience. So that wasn't meant to be funny, but I mean well.

But we do work very hard, and we'll never make Amazon.com fun and engaging in the same ways great physical bookstores are. You'll never be able to hear the bindings creak and smell the books and have tasty lattes and soft sofas at Amazon.com, but we can do completely different things. We'll blow people away and make the experience an engaging and fun one. I'll talk about some of the stuff that we can do in the future a little later.

One of the things that we've done that's been much commented on, and I think has been very important, is we let regular old customers review books. It turns out our customers are very smart, and the reviews are extremely helpful to other customers in terms of making purchasing decisions. I'm also convinced that there's a little piece in the human brain that can immediately tell by reading somebody else's writing whether the person is smart or crazy within like the first five words.

So you can actually weed out the reviews that you don't want to pay attention to really well. One of the things that surprises people, sometimes upsets publishers and authors, is that we let people write negative reviews. They say, "Why do you do this? You only make money if you sell the books. Why don't you read out the negative reviews?" The reason is because we're taking a different approach of trying to sell all books. We want to make every book available, the good, the bad, and the ugly.

When you're doing that, you actually have an obligation if you're gonna make the shopping environment one that's actually conducive to shopping, to sort of let truth loose. That’s what we try to do with the customer reviews. So we do actually remove reviews that are off topic or even if they're funny. We also let, for example, authors review their own work so they can say, “I’m the author,” and click here. If you're the author, you can review this book, and we scan all of these.

We have a team of people who do this and matches them up, and we have to have the email address, and we can verify things with the publisher. But occasionally things slip through. We did have God review the Bible. There were a couple of other amusing things. We let, let's see, who was, Karen Burrow was very, very upset that Jane Austen had gotten two miniseries.

What is the name of the writer, the contemporary of Jane Austen, who was jealous of her all the time? It slipped my mind right now. Yeah, that’s Bronte, thanks to my wife, speaking of saving me yet once again. Anyway, she also reviewed one of her own works and said, “I just can't believe that Jane Austen has gotten two miniseries and a full-length motion picture in one year.”

So even those funny ones we take away. We should probably post them on a special page; that would be a fun thing to do. Finally, what people don’t see sort of behind the scenes at Amazon.com: you can see a lot of things on the website, but you can’t see the stuff that the software development, all the effort that goes into the back-end tools. Amazon.com is very much a technology company. In fact, I think of us in many ways as sort of a small AI company.

Artificial intelligence is something that’s gotten a bad rap. There are some good reasons for that and some bad reasons. One of the bad reasons is that once the AI guys work really hard and accomplish something, it's not AI anymore; it’s just a parlor trick. What’s interesting is there are some very simple but sophisticated techniques that are working well to help people discover things online, like collaborative filtering. I’ll talk a little more about that in a minute, but this is all subterranean.

So we’re starting to do things on our site now that aren’t very visible, like customizing the homepage for particular customers using their past activity and their stated preferences to guide us. You don’t even notice this unless you sit right next to someone and see that their version of Amazon.com is slightly different from yours. We’ve just started doing this. This is something we’ve wanted to do forever. And the goal is to make the perfect store for everybody; we don’t have to have the average store for the mythical average customer.

We’ve also worked hard on all the customer service tools, the things that process electronic mail. If you’re doing business in a mail-order company, for example, there are dozens of off-the-shelf systems you can buy to handle inbound phone calls into a call center and so on. We get thousands of email messages from our customers every day, and there aren’t any great off-the-shelf inbound email center software packages yet. There will be.

So we've had to write our own. In fact, about 85% of all of our software development to date has actually gone into the back-end systems that are completely invisible to customers. The logistics systems that handle two and a half million different items and six different availability categories. Early on, we looked at could we buy a software package that would do the logistics for us? The problem is that if you look at the available packages that are built for mail-order companies, they typically have just two availability categories: in stock and backordered.

We actually have six because we have books that are shipped within 24 hours, two to three days, one to two weeks, four to six weeks, not yet published, shipped when available, and when we're out of print, shipped within one to three months if we can find it at all. So some of this stuff just has to be done the hard way. We've also done one more thing that I think is interesting enough to mention, which is our Amazon.com Associates program.

There are actually several other places doing this now, but at the time I believe it was new. The idea was, let's make it possible; let's use the inherent features of the web to make it possible for any website, no matter how big or how small, to have its own bookstore in association with Amazon.com. Today, we have over 30,000 of these associated Amazon account associates out there on the web, ranging in size from on the big side: AOL, Excite, Yahoo, Netscape, AltaVista, Geocities, to the small.

Somewhere in the middle, Dr. Ruth Westheimer sells sex-education books from her website through Amazon.com to the very tiny, including one of my favorites, which actually is a website that bills itself as the web's—the world's—the web's oldest and best place to buy meteorites. You can indeed find everything on the web if you look hard enough, and now they can sell meteorite books along with this.

It's interesting because it's a win-win-win. It's a win for the associate because we pay them a share of revenues; we pay them 15% of revenues for all the books that they sell. It's a win for our customer because they now have, in many cases, a specialized website, like the meteorite one or one on yellow Labradors and so on, that puts these books in an editorial context. You know, we can't say with editorial authority that this is the Bible of Labrador Retrievers, but there's a website out there that can say that. And finally, it's a win for us because we get new customers this way, and these sites help us with branding.

So, what I'd like to do now is sort of go back in time. We have over a thousand employees now, which startles me to no end because three and a half years ago we had seven. So things... I want to sort of walk you through some of the just really kind of little anecdotes about Amazon.com from the very beginning in 1994. When I came across this fact that the web was growing at 2,300 percent a year, I was working on Wall Street, at a very specialized kind of investment bank called a quantitative hedge fund, a company called D. E. Shaw.

When I decided I was gonna go do this crazy thing, I went and told my boss. He said, “No. Hmm. Let's go for a walk.” So we were on a two-hour walk in Central Park, and he actually thought it was a good idea. Some very smart guys named David Shaw is one of the few people I know who has a fully developed left brain and a fluid right brain, just a super smart guy, and he said, “This is a great idea, but it would be a better idea for somebody who didn’t already have a good job.” That was actually a compelling argument to me.

So I went away. He convinced me to think about it for two more days, and it was very hard to make that decision until I found the right framework in which to make the decision. I think because I know there are a lot of people who are interested in entrepreneurial activities out there, I think this could be useful. The right framework for me, anyway, to make that kind of decision was a regret minimization framework.

So I projected myself into the future, and I’m 80 years old, and I’m looking back on my life. What do I want to have done at that point? One of them was to minimize all the regrets that I have. I knew that when I was 80, there was no chance that I would regret having walked away from my 1994 Wall Street bonus or knew that kind of thing. I wouldn’t even remember that. But I did think there was a chance that I might regret not having participated somehow in this thing that I thought was going to be very exciting called the internet.

I also knew that if I tried and failed, I wouldn’t regret that. So once I thought about it that way, it became incredibly easy to make that decision. It's also easy for me because my wife, who’s sitting right here, is so supportive and also beautiful. Thankfully, she’s not geographically tied to a particular area, so we were able to pick up and move to the best place to do Amazon.com, which turned out to be Seattle. But there's actually a story there, so when you get bored of these anecdotes, let me know, and I will stop.

When somebody's growing 2,300 percent a year, you have to move fast. A sense of urgency becomes your most valuable asset. So we did do that. I wanted to turn out, the problem of where to locate Amazon.com was highly constrained by three factors. It had to be in an area with a large pool of technical talent, it had to be in a small state because only the residents of the state where you have what they call Nexus have to charge those customers sales tax.

So I wouldn't want to be in California and so on, bigger states. Finally, it had to be nearby a major book wholesaler. It turns out the largest book warehouse in the world is in Roseburg, Oregon. So I was doing this analysis and I had tons of statistics about cities all over the country. Even I can tell you to this day that there are about 3,000 flights a week out of the Seattle Airport, which makes it sort of a medium-sized hub.

There were minor considerations, but trying to make this decision, the problem was I hadn't quite made an arrow down to four cities: Portland, Oregon, Seattle, Boulder, and Lake Tahoe on the Nevada side of the border. The movers were here, and they had packed up our stuff and wanted to know where to take it. So I told them just to head west and call us tomorrow, and we'll tell you. And that's what they did.

My wife drove to Seattle, and I tapped out the first draft of the business plan in the car on the way there. I also called our attorney, I called a friend in Seattle and said, “Can you recommend an attorney?” and retained an attorney by cell phone, had that person set up bank accounts, and so on, so those were all ready by the time we got there. I had him incorporate the company, and of course he asked, “What should the name of the company be?” I knew he was going to ask that question, so without even missing a beat, I said Cadabra, Cadabra, Inc., like “abracadabra.”

Well, I instantly knew that was an incredibly dumb name because he came back to me right away with “Cadaver?” “No, no! Not cadaver! Cadabra!” and he's like “cadaver?” “I’ve no,” so anyway, the company was incorporated as Cadabra, Inc. right there, and three months later, was changed to Amazon.com, a name that actually perused an entire section of the dictionary.

It turns out that things are always listed online almost always in alphabetical order. Kind of like the yellow pages where you get the triple-A auto repair phenomenon, you start to see that in fact we do have a competitor or a small competitor online called, like, triple-A one books or something like that. It was hard to pick a name because even then, in 1994, most of the good domain names were gone already.

So Amazon.com, people considered to be early, but actually it was not that early in this race. Mosaic Communications, which would later become Netscape, had already been started several months before. Wired Magazine had already started a hotwired. Halsey Miner had already started Scene. These things were happening very quickly.

Let’s see; try to actually pick the more interesting anecdotes here. We found a house that could serve as the first business, and it had a garage. It seemed the right thing to do since there’s a history of startups in garages. This wasn’t fully legitimate because it was a converted, enclosed garage, but I considered it somewhat legitimate because it wasn't insulated.

So to solve the cold problem, the owners of the house we were just renting had installed a big pot-bellied stove right in the center of this garage. This converted garage, later, when we got the fifth person moving into this small, what had been a two-car garage, we had to move the pot-bellied stove out, and the thing must have weighed 800 pounds.

They were not impressive offices, and so when we had meetings with third parties, the most convenient place to have the meetings was a little café just down the street, only about a mile away. The only—and it’s not really a problem—but the interesting thing about this café is that it was inside a Barnes and Noble store. So in fact, we did do all of our early third-party meetings at Barnes & Noble.

The only other interesting thing from that first office is that again, by the time we got about five people, there wasn't enough electricity in that garage because it just had one circuit breaker serving the whole garage. We had the place full of computers and other equipment, and so we had taken long extension cords and run them from every room in the house into the garage.

So we were siphoning the entire house's power into the garage to the point where, you know, my wife couldn't turn on a hairdryer; I couldn't vacuum the living room. Anything like that happened, and all the computers would turn off. So we were really getting to a place where we needed to move to a real office to finish up our software development, and we did that.

This was a small office, actually about a thousand square feet of office space, and we had also our first we built our first warehouse space, about 400 square feet in the basement of this facility. It was actually the Color Tile building in sort of an industrial section of Seattle. I remember the day before we launched, we had everything set up and we had shelves from the warehouse, and there were no notebooks because we were using, at that time, an almost just-in-time inventory technique.

We would order from wholesalers after customers ordered from us. Wholesalers would deliver to us in big boxes, and we’d break them down to ship them out in small boxes. Well, one of our programmers, I remember, was looking at this little tiny 400 square foot warehouse that we had set up the day before we opened. He said, "I can't figure out if this is incredibly optimistic or hopelessly inadequate." I looked at him, and I said, "Yeah, you're right."

In fact, we believed that it was going to be incredibly optimistic. We had very low expectations for starting off and thought it would take a long, long time for consumer habit to adopt buying online at all. In fact, we felt so strongly about this that we were trying to convince these wholesalers to make smaller shipments to us.

They had what seemed to be a very small minimum shipment size, which was 10 books; you had to order at least 10 books. But we were convinced it would take months and months to get to ten books a day that we would be selling. We wanted it to be even that fast. We said, "What, can we pay like a $20 fee or whatever to get one book?" They said, “No, no! We didn’t want to just go buy them at Barnes & Noble because we wanted to actually test and exercise our systems and so on."

So we found a loophole; it was a great loophole! It turns out that with both of the wholesalers we were using at that time, you just had to order ten books if you ordered ten books, but nine of them were books they didn't have in stock. They would ship you the one book.

So we figured out that there was this loophole. Both of them sensibly carried this particular obscure book on lichens, but they didn’t have it. So we tested all of our systems by ordering one book and nine copies of this lichen book. Well, we did launch shortly after that comment was made, and an amazing thing happened. Three days after launching, we got an email message from one of the founders of Yahoo!

At the time, Yahoo! was still a very small company. They probably had four or five people at that time, and the guy said, "Wow! I saw your website; I think it's incredibly cool, and I'd like to put it on the Yahoo! What's Cool list," which, at that time, was the most trafficked web page on the internet. Still, you know, Yahoo! of course, the most trafficked site. I'm not even sure if they have the 'What's Cool' page anymore; things have changed a lot in the near past.

But we sort of sat around, and he said, "But, you know, you're going to get a lot of traffic, and do you want to wait? If this isn’t the right time to send all these people to you, let me know, and we can do it in a month or whenever you want." Very informal! Today, to get Yahoo! to do something like that, you'd have to pay them tens of millions of dollars, but then, it really was just an email exchange.

So we sat around eating Chinese food. There were maybe seven or nine of us in the company at this point in the office, discussing whether or not we were ready to have sort of this huge influx of traffic. Shel Kaphan, who was our vice president at the time and still is now our chief technology officer, said at the time that it might be a bit like taking a sip from a firehose.

I thought that was right, and we sort of sat around, we talked about it for five minutes, and said, "Yeah, let's do it." So we did! We had programmed a bell to ring every time we got an order because it was very exciting to get an order, and basically every computer terminal in which there were only about nine would ring when we got an order, and a little thing would pop up interrupting whatever you were doing showing the contents of the order and what had happened, and a big cheer would go out.

Well, as soon as this Yahoo! thing hit, all of a sudden the bell became extremely annoying. So we quickly unprogrammed that, and instead, somebody wrote a little script that we could run at any time we wanted to see what the sales had been in the last hour or the last five minutes. The thing got used a lot in the first few days, and so anyway, we were still, you know, compared to what kinds of levels of business exceed Amazon a lot on the internet in general today—very, very small—but they were huge compared to what our expectations had been.

In fact, everybody in the company, every single programmer, all nine of us, were shipping and packing books out of this 400 square foot thing. We were trying to figure out how we were going to get enough bandwidth to actually hire more people to do this while doing it, literally working until midnight every night shipping out, you know, 100, 200, 300 packages a day.

We had not planned perfectly; we didn’t even have packing tables in this little 400 square foot warehouse. So this was all done on our hands and knees on a cement floor, and I remember we got used to it so quickly that two weeks later, when we actually got packing tables moved into that space, we thought, “Wow, these packing tables are really good. This is a great idea!”

So, let’s see; at that time, and still for the next couple, we’ve been in five different locations now, and in just, you know, about three years. So that’s either poor planning or unusual growth, whichever way you want to think about it. In our first couple of warehouses, we used to... now we use a logistics technique we reprogrammed to our computer systems called random style, but before we’d had this really cool, less efficient, but very cool technique where we, as we got the books in, we put them directly into bins that were by customer order.

You could wander through the warehouse and literally see any grouping of books, and that would be a customer’s order. You could see what people were buying in combination; it was very interesting, and there were a couple of interesting vignettes from that.

One, I walked through one day, and I saw this order that was “a hundred and one ways to make love to your husband,” “a hundred and one ways to make love to your wife,” and the third book was a “Hawaii vacation planning guide.” I looked at this order and said, “I think we know what's going on here!”

Somewhere around this point, we had gone through a couple of warehouses. We were still a very, very small company in the scheme of companies but starting to actually be noticed online. We had a revenue run rate of about five million dollars a year, and we were featured on the front page of The Wall Street Journal in the lead story, which was good.

It was good because it generated a lot of exposure for the company, and you have to keep in mind that our first year we didn’t spend a dollar on advertising. All anybody found out about us was either through word of mouth or through PR, so PR was important to us. In fact, since then, we spent a significant amount of money on advertising, but word of mouth is still the largest source of our customers. Again, it’s incredibly powerful online.

But having this article was good because it generated lots of awareness, and we got lots of new customers as a result of this article. Those people, you know, if we treated them right, did become evangelists for us. It was bad also because when you get a big article like that, it’s not just your customers who read it, but so do your competitors.

One of the things that was the wake-up call, whereas the sort of the 2,300 percent a year growth in web usage and our wake-up call, that Wall Street Journal article was Barnes and Noble’s wake-up call. So which was basically, that was May of, I guess, ’96. It took them about a year from that point to launch a website, which they launched back in May of this past year.

They launched that website which is a very reasonable website. It’s not as good as Amazon.com in any important respect, but it is a good first effort. It pays, as Andy Grove has told countless audiences, to be paranoid. As I was discussing with one of, we had dinner earlier tonight with a few of the people in the audience, and it pays to be humble and paranoid.

It would be hard to be actually paranoid enough if you’re not humble. We were very worried about what Amazon.com... I mean, sort of what Barnes & Noble would be able to do with their purchasing power and their brand name. We decided instantly that we weren’t going to let the purchasing power get in the way, so we were going to fund any purchasing power difference.

We would have the same prices no matter if our margins were lower, and then we would adopt a strategy of getting big fast so that we could eventually level the playing field in terms of purchasing power. That became a consistent focus of ours that we've articulated now to Wall Street and everybody else. It's actually at this point quite a well-understood strategy.

I would say, but the interesting thing, of course, is that Barnes & Noble launched their website about five days or so before our initial public offering, and I’m sure it was a complete coincidence of timing. But it didn’t upset anything. Even though at the time, it was very interesting because there were several pundits, particularly one in particular, who was making the lecture circuit, talking about now that Barnes & Noble was here, was going to be Amazon toast instead of Amazon.com.

And this was actually, we even found this very funny; it’s an amusing thing to say, but we were worried, and we didn’t know exactly what to expect. We knew that we were going to remain customer-obsessed. Our mantra became... It was interesting, too; one of the things I was most worried about is that Amazon.com, for two years, had been a company with virtually no competition. At this point, we already had over 300 employees.

So we were sort of...I didn't know whether all the employees were ready for competition and whether, you know, in very previous jobs and so on, people had always been in competitive environments, and it was a competitive group of people. But we let our mantra be that we were going to obsess over our customers and not our competitors.

So we would watch our competitors, learn from them, see what things they’re doing good for customers, would copy those things as much as we can, but we were never going to obsess over them. That became our watchword, and it’s actually worked very well.

Since Barnes & Noble launched their site, we've gone from having a revenue run rate of about 60 million dollars a year to a revenue run rate of having 260 million dollars a year. This has happened in just 10 months. We have also gone from having 340 thousand customers to having over one and a half million customers in that same period of time.

I am convinced that, I mean, there are two things a company like Amazon.com can only compete against a much larger company if you really do offer a better service. We know by anecdotal evidence, focus groups, and quantitative research that there are three things that are important to our customers: selection, ease of use, and convenience and price.

So every day we work on making sure that each of those three things are better at Amazon.com than they are anywhere else, and we have a saying that people ask me sometimes, "Are your customers loyal to you?" I say, “Absolutely! Right up until the second that somebody else offers them a better service.” So we work on that very hard, and then I think the thing that makes it work for us is the word of mouth, and it's that 58% of repeat customers every day.

Those people are our evangelists and help us sign up new customers. Let’s see. I would say the most interesting thing, though, is I firmly believe this is day one. This is a time when we know very little about e-commerce and very little about merchandising. There’s a huge amount to be done. If you look at the kinds of things we’ll be working on in the near future, we’re gonna work on geographical expansion, so we want to make it possible for people to buy a Japanese language book anywhere in the world, for people to buy a German language book anywhere in the world, just as we now do that for American books.

We’re also working on product expansions, so we want to leverage our customer base, our competencies, and our brand name into products like music and videos. If you search Usenet news groups, you’ll find at least once a week somebody will ask the question, “Who’s the Amazon.com of music?” We’re desperate to say, “We are!” and so far we can’t say that! So those are two of the things we work on, but I think one of the most interesting things is this notion I spoke about earlier, which is redecorating the store for each and every individual customer.

This is the future of e-commerce; we’ve stopped calling it e-commerce and started calling it emergentizing because commerce is the simply find it, buy it, ship it sort of be a...action, and merchandising is much more about customer behavior online. In the physical world, there are all sorts of things that are understood about customer behavior.

So something called the hard right, for example. You walk into a store, 90% of customers turn right; that’s why the cash registers are used on the left, merchandise on the right. There are a whole bunch of those kinds of things. Nobody knows what those things are online, and we view Amazon.com as an experimental laboratory to try and figure out some of those things at the same time.

We can use advanced technology—and we are, as I said, starting to do this now—to understand our products on a product-by-product, individual basis but to understand our customers on a customer-by-customer, individualized basis so that we can... The goal is to enhance the discovery process. The goal is to make it, you know, discovery is an incredibly powerful motivator for people.

All of us like to discover things, and when you walk into a bookstore, and you serendipitously stumble across something that blows you away and say, "Wow, I really want to read this book," that’s very powerful. All of us have read half a dozen books in our lives that have made a major impact on us in some way, and I would posit that there are 500 books in the Amazon.com catalog right now that could have that same sort of impact on you if you could only find them.

I believe that we can use advanced technology to dramatically improve the odds that you can find those books because we’ll not just let readers find books, but we’ll let books find readers. What really is the case is that we know 2% about all this stuff that we will know ten years from now. This is absolutely the Kittyhawk era of e-commerce and emergentizing, and I thank you for listening to me, and I am happy to take questions if there are any. [Applause]

When asking questions, will you please rise, and please state your name as well? Thank you! Speak into the boom.

“Yes, I’m Dave. One question about other businesses that you think are relevant for commerce on the Internet beyond just records and music in some way in the future.”

There are certainly a lot of other businesses that, besides those three, that are even in the near future. They're going to be very successful online—computer hardware and computer software sales are going to be a big deal. A lot of companies are using the internet very effectively not to sell necessarily, but to save significant amounts of money. Federal Express is a great example with their package tracking website. They have many fewer calls to their 1-800 number attracting packages.

So there are a lot of opportunities for many, many companies to offer customer self-service. The interesting thing is that if you look at all the studies, customers prefer self-service when it's done properly to being helped by somebody else because they have control over the situation.

I think that I may have this slightly wrong, but I know the hotel industry does satisfaction surveys every year, and I believe Embassy Suites always gets the highest customer service ratings, even though they don’t have any. They just let everybody know, "You have your own coffeepot and your own things," and people love that.

“Michael McCaskey, when I visit Amazon.com and I start to look at a review of a particular book, you bring up suggestions for three other books. I wonder in what direction you'll be able to take that as the artificial intelligence component becomes more sophisticated. I know some of the things I'd like you to be able to do, but I'm curious about where you think it's going to be in four years or ten years, or choose a time frame?”

And by the way, please come find me afterward and tell me what you think we should do. There are huge possibilities for this one of the techniques that's being used today—we're using on our website—it’s called collaborative filtering, and the way collaborative filtering works is basically it looks for your affinity group.

So if we have something called instant recommendations, it looks at the books that you've bought in the past from Amazon.com, and then it finds other customers who've bought the same books. Then it looks for the things that that group of customers have bought that you haven't, and it recommends those things to you.

So it's basically looking for people who have the same tastes or interests that you do and then using them as a guide to what we should recommend to you. There are other techniques that can be used. People think that neural networks are going to be valuable in this space. There are other, simpler, but sometimes the simple things work best: just simple statistical techniques that can be used—there are genetic algorithms that can be used.

A lot of this, what’s going to be done is a lot of experimentation. We’re gonna find out what works best.

“You mentioned you list 2.5 million books on your website, but really only about 1.5 million are in print currently. Could you mention how Amazon.com goes about finding those remaining million books for those who should have to order them?”

If you look at the catalog, there are two and a half million books, about 200,000 bestsellers. We actually inventory ourselves in our own warehouses between 200,000 and 400,000. We go to a network of about a dozen different wholesalers to get those books. Between 400,000 and 1.5 million, we go to over 20,000 different publishers directly to get those.

Then, between 1.5 million and 2.5 million, are out-of-print books, and we use a network of search services. There are companies that have specialized for years. This actually isn’t even a new industry; it’s just something that the internet is allowing people to access more easily, that we tap into. Those people search for the books.

There are two basic ways that this has traditionally been done: one way you charge the customer a search fee, and the other way you don’t, but you mark up the book. We do the second thing, and we actually promise to keep looking forever, so we keep it on the search list. We re-initiate it every once in a while. It’s theoretically possible that you could put in an order, and, you know, seven years later we’d email you and say, “You know, we did find your book.”

There are some interesting stories about out-of-print books found, but I won’t go into those.

“Good evening, my name is Howard Sanderman, and I was wondering how you infuse your employees with your entrepreneurial spirit and vision, in essence, especially in a growing company that’s gone from seven to 1,000 employees?”

So, in essence, how do you cultivate internal entrepreneurship or intrapreneurship? I definitely do not claim to have sort of a secret recipe for doing that. Some of it is self-reinforcing. So when a company is doing well and it’s growing rapidly, that naturally tends to excite everybody in the company. I also think that corporate cultures are incredibly important, and they’re one thing that can never be copied.

Competitors can never copy a culture, and the unfortunate thing, I think, for businesses and entrepreneurs is that you can set out from the very beginning with a particular culture that you want to form in mind, as we did. Our watchphrase for that was "intense and friendly.” We always talked about it in the sense that you have to be intense, and in fact, if you ever had to give up friendly in order to have intense, we would do that. If we needed to be intense and combative, we’d do that before we’d be not intense.

You can start out with the way you want things to be, and then really your early employees become the beginning that you carry around. They carry around the flame of what your culture begins. So I think mostly it’s a blend of sort of 30% what you set it out to be, 30% who your employees, your early employees happen to be, and 40% random chance.

The bad thing about the random chance part is that once it's set, you’ve got it. There’s really no way to change a corporate culture.

“You talked about your being customer-obsessed as one of your key strategies. I was wondering what other strategies you're going to need to break into the music industry?”

Our biggest disadvantage is that we don't have a first-mover advantage in that space, and Amazon.com itself has demonstrated the first-mover advantages on the internet are incredibly powerful. Companies like CD Now and In2Ke have the first-mover advantage.

What we have is a customer base of more than a million and a half people who we have polled, know that many of them will buy music from us. That customer base is much larger than any other internet retailer.

The strategy really is to leverage that customer base, leverage the brand name, which we worked very hard to associate with high-quality service, low prices, and ease of use, an authoritative selection. So those are the kinds of things that you’ll see us do to try and be successful in that space, and there are no guarantees whatsoever that we’ll be able to do it.

“Brad Bradford, do you have any anecdotal things on people who've been unhappy and wanted to send stuff back, or has everybody been happy?”

I would love to give you an anecdote like that, but there just aren’t any. No, there are many. In fact, oftentimes when I speak, I carry around a few of those messages because the more hostile ones are actually funny. I mean, there’s always a good reason for why the person is so upset, but it’s a good, you know, sort of cheap way to get a laugh from the audience.

But I've gotten email messages from customers in all capital letters and upset for various reasons. The one that springs to mind is a customer who was extremely unhappy that he had spent a large amount of time—hours and hours—filling his shopping basket, but he wasn’t quite ready to purchase the books.

So his online electronic shopping basket at Amazon.com—then after 30 days had passed, we, if there’s no activity on account for 30 days, we clear out the shopping basket, which he thought was an incredibly stupid policy. In fact, we have changed that. I think it’s at the disk space actually isn’t that costly; that’s probably was a stupid policy.

He also thought that we should at least have warned him—also probably a good point—but he wasn’t polite in his message, I can assure you. We did apologize, and we actually went back through our logs, our HTTP logs, which are these huge massive files and found his shopping basket and sort of like raw data form and sent it to him, which made him very happy.

But, you know, there are countless stories like that.

“Where do you see yourself in five years?”

Can continue to be a pioneer in electronic commerce, electronic merchandising. There are lots and lots of stories of companies that were pioneers and then disappeared from the face of the earth. Visicalc is a great one that, you know, entrepreneurs should have nightmares over if their business was taken away by Lotus. And then, of course, the same thing happened to Lotus when Microsoft took their business away.

So we are gonna work like crazy to expand, not to defend the territory we have now, because I think that strategy wouldn’t work. What happens if you just do that is that this is a scale business. This is a business that only works in large size because of their large fixed costs and the very low variable costs.

What you need to do is have a business of sufficient size so that you have purchasing power and the brand name that’s self-sustaining. Really, I think if we work on that combined with the discovery technology that I was talking about earlier, if we can do those two things in the next five years, that will be a huge success, and Amazon.com, we’ll all be able to be very proud of ourselves.

“Margaret Nugent, given what you were saying about the people who are carrying the flame at the beginning and how critical they are, could you talk a little bit about who you selected those seven people who were in that garage and whether they've evolved with the company or left?”

All the original employees, only one has left, and his wife finished her doctorate, and they had to move across the country. Basically, when we were looking for people, we were looking for intense, hard-working smart people who want to be a part of what we’re doing, so passion is a part of it. Just plain old smarts, you know, sort of high IQ brainy people—we really like that—and people who can hire other great people.

When I interview somebody, I actually spend about a third of the interview asking questions designed to ascertain whether or not they can hire great people. So it’s a sort of a third of the meta-interview. Different businesses have different criteria, but in this business, we’ve reached an interesting inflection point where I would say 70% of the risk now to Amazon.com is execution risk. So it’s inside the company; it’s, you know, our ability to stumble. We’ve basically gotten past the point where 70 to 80% of the risk was external, and where we needed a huge amount of luck to get to where we are now.

Now, all we need is a clear, consistent vision and the ability to execute on it very, very well at high speed, and that second piece comes from having large numbers of talented employees with lots of executive bandwidth to help guide them.

Individuals do not win. We’re very lucky we have the unarguably the most talented venture capitalists in the United States on our boards— a guy named John Doar. John is very well—he talks about the fact that teams win; individuals don't. We have built a great management team and a great group of people underneath that.

If you look at companies like Microsoft, that’s the main way that they’ve won. If you look at all their assets, they’ve got more than ten billion dollars in cash in the bank. They’ve got an operating system which is a virtual stranglehold on the industry. I would say all those things are great, but the thing that really makes them such an awesome competitor and the reason that they have the position they have today is because they don't just have Bill Gates; they have 40 people at the top of that company who are smart, dedicated, and hardworking.

If you look down below that sort of incredibly deep executive team, they’ve got more people that are waiting in the wings to take on that kind of responsibility. So we're trying to build that kind of team at Amazon.com. We've got a great start, and we do have a thousand entrepreneurs at Amazon.com.

“There are a number of discussions going on about the future of the book now that there’s the computer. I'm wondering what your thoughts are on that, and also, relatedly, I guess, with the growth of the superstores, the Borders and Barnes and Nobles, they've really had an impact on, I think, on publishing in some ways. With no deals with the publishers, what kind of an impact are you having on publishing and in books?”

Great questions! Well, impact... I’m terrible; first of all, terrible multi-part questions, so if I forget one of the parts while answering the other, just you remind me.

The impact on publishers is especially Amazon.com's impact is especially positive on small and independent presses. The reason is that this extensive catalog—the ability that the internet gives us infinite shelf space—so that we can actually merchandise to a worldwide audience titles that otherwise simply could not and cannot get distribution in physical bookstores, so that has been a positive impact.

The second thing has been reducing returns. Turns out most people don’t know this, but 38% of all books that get shipped out from servers into the retail distribution channel are eventually returned to the publishers as unsold merchandise; that’s an incredible statistic. Amazon.com returns to publishers less than 4% because of our centralized distribution model.

Your first question: What’s the future of the book? I firmly believe that at some point, the vast majority of books will be in electronic form. I also believe that’s a long way in the future. The thing that holds many more than ten years in the future and the reason that that gets held back today is that paper is just the world’s best display device. It turns out that today, with the state of the art in display devices, dead trees just make the best display devices.

They’re higher resolution; they’re portable; they have high contrast; and so someday, computer displays will catch up with that, and then I think electronic books will be extremely successful.

[Music]

“Hi, my name is Will Price. I'm just wondering what you're doing to drive consumer behavior or access to technology in terms of growth avenues. That's something you're actively trying to work on, or as you know, the inherent growth in the industry is sufficient to cater to?”

Right now, well, we don’t... we aren’t trying to grow the internet. What we target—the customers that we target with our advertising and our other promotional activities are already online, and they're heavy book buyers. So at this stage of the game, we’re not trying to convince people who aren’t online to come online. We’re also not trying to convince people who don’t read to read; we will save that for a future more glorious day.

“Hart Rogers, you spoke about regret minimization. Do you have other advice for people contemplating moving into entrepreneurship from the normal track?”

Yeah, I do; I think that I almost started a company straight out of school. I would say there are many examples of people who've done that successfully—some very high-profile examples of people who’ve done that successfully.

I actually advise people not to do that. I almost did it; who knows how it would have worked out, but I think I’ve benefited tremendously from working at a couple of best practices companies that were very entrepreneurial and fairly small at the time that I did it, and I did that for about ten years.

I believe that the experience of doing that and working with people like David Shaw at D. E. Shaw were just unbelievably helpful to me in terms of being prepared to do this. I think it just dramatically increases your odds, and when you're doing things like this, the problem with being an entrepreneur and undertaking a business initiative is that companies don’t fail quickly.

If they failed fast, that’d be okay because then you could, you know, you could try 20 times. But in fact, they take years and years to fail. If you look at the failure statistics of startup companies, very rarely do they fail in the first year. It’s usually seven, eight, nine years.

So it means that you're dedicating a significant percentage of your life to this thing, and you really want to make sure you've optimized, you know, maximized the odds that you're going to be able to succeed.

“I’m Kitty Lansing. I'm curious about what percentage of the books you sell are children’s books because I think of children’s books as being so dependent on illustrations. How do you see your way getting around that?”

Yeah, well, the direct answer to your question is we don't disclose any breakdowns on sales. We’re competitors, I can’t actually tell you what percentage of our sales are children’s books. But what I can tell you is that the strategy we’ve taken with children’s books is it turns out with children’s books, people just want to be—parents are mostly shopping for children, none running friends of parents and so on—adults shopping for children.

They basically want to be told what to buy. Other things, people know because you’re buying for yourself; you know what you want. But when you’re buying something for a four-year-old, guidance is very much appreciated. That’s the way that we merchandise children’s books online.

[Music]

“John McKinnon, I’m curious to know how Amazon's gonna respond to some of the large strategic relationships that Barnes and Noble is building with say America Online or CNN?”

Well, really, if you look at the large relationships that have been built online, all of the significant ones are actually built with Amazon.com. So, you know, this is one of the advantages of being a first mover.

We have multi-year exclusive agreements with five of the top six websites including AOL. Barnes & Noble has an exclusive agreement with AOL's proprietary network. We have an exclusive agreement with AOL.com where we have a permanent homepage banner— for example, is the single most visited webpage on the internet.

So, you know, literally, if you take the PC meter data, which is to where the Nielsen service is on the internet, and look at the top six websites; we have exclusive, multi-year arrangements with all of them—vault with all five of the top six. The one that’s not there is Microsoft; they don’t have a relationship with anybody.

We're gonna bring the formal part of our program to a conclusion, although I'm sure Mr. Bezos will be glad to chat with people informally for a few moments. I want to thank you for coming a long way this afternoon to be with us, and it's made this year's A.B. Dick lecture at Lake Forest College a truly memorable experience, and we'll be looking for you. [Applause]