kieran Levis.com

Kieran Levis.com

 
 
 
 

Amazon

Jeff Bezos became the most celebrated of all Internet entrepreneurs, partly because his business was one that most outsiders thought they could understand. In the late 1990s his face was on the cover of countless magazines and he had a ready supply of memorable quotes for the media, suddenly obsessed with the Net and the instant accumulation of wealth. He described himself as a nerd, but he had grasped the commercial potential of the Internet long before most people, and acted quickly and systematically to seize the opportunity it presented.

He was an outstanding student with a flair for quantitative analysis but at Princeton University he realised that although he was amongst the top 25 students in physics, “there were three people in the class who were much, much better at it than me”. He switched to engineering and computer science and graduated, summa com laude, in 1986. Bezos was offered jobs by blue chip companies like Intel and Andersen Consulting, but decided to join a small Wall Street start-up, Fitel, where he developed and marketed an information service for cross-border trading by brokers, investors and banks. He then moved to Bankers Trust, where he became its youngest ever vice-president at the age of 26. He led a programming team that developed a communications network that enabled clients to monitor the performance of their investments. Traditionalists at Bankers Trust disliked PCs, and in 1990 Bezos decided he would move to a company where he could conduct “second-phase automation – which would allow him to fundamentally change the underlying business process and do things in a completely new way.”

D.E.Shaw & Co was a quantitative hedge fund, founded only two years earlier and described by Fortune as “the most intriguing and mysterious force on Wall Street today. . . a place where the avant-garde meets arbitrage, and intellectualism and profit-seeking mix harmoniously”. It was an ideal environment for Bezos and he was put in charge of exploring new markets.

Opportunity

Early in 1994 he conducted a study of ways to make money from the Internet, and was staggered to discover that Web usage was growing at 2,300% per annum. Clearly this was going to create enormous business opportunities, though their nature and scale were difficult to grasp. “Human beings aren’t good at understanding exponential growth . . . it’s invisible today and ubiquitous tomorrow”. He drew up a list of 20 types of product that he thought could be sold efficiently online and decided that books and music were the most attractive. They were standardised products that customers understood, so they would know exactly what they were buying. Of the two, he favoured books, as there were more titles and “no eight-hundred-pound gorillas” in the fragmented publishing and distribution industry.

It was clear to Bezos that an online business could offer customers a wider selection than the biggest stores and a more convenient way of buying: the very largest stores stocked 175,000 titles, yet there were 1.5 million books in print in the US. An online store could offer very much greater choice, and without the real estate and staffing costs of physical stores, it should be possible to achieve both lower prices and higher margins. “So that became the idea . . . easily find and buy a million different books”. The inefficiency of the industry was exemplified by the fact that 35% of the 460 million books shipped in 1994 were returned to publishers. Bezos came to realise that this was “not a rational business. The publisher takes all the return risk and the retailer makes the demand predictions”.

D.E.Shaw decided that it did not want to get into this business, so Bezos decided to do it himself. Characteristically, he constructed what he called a “regret minimisation framework” to decide whether to jump. “I knew that when I was 80 there was no chance that I would regret having walked away from my 1984 Wall Street bonus in the middle of the year. . . . But I did think there was a chance that I might regret significantly not participating in this thing called the Internet that I believed passionately in. I also knew that if I tried and failed, I wouldn’t regret that. So, once I thought about it in that way, it became incredibly easy to make that decision.”

Start-Up

Bezos decided to locate the business somewhere where there was a large pool of technical talent and near a large book wholesaler. According to Amazon legend, he wrote the business plan on his laptop, whilst his wife, Mackenzie, drove them West. In fact, they flew from New York to Texas, and drove from there, while the business plan was not written for another year. He did however choose a house with a large garage, in the hallowed tradition of start-ups.

He chose the name ‘Amazon’, so that it would come up early on alphabetical lists and the brand could stand for anything. From the outset he had the idea of moving beyond books at some stage. He insisted that the company should always be called ‘Amazon.com’, which later helped to differentiate it.

His first recruits were computer specialists. Sheldon Kaphan, who was to become Amazon’s CTO, was an experienced software engineer. He was initially reluctant to commit himself to another start-up, but Bezos eventually convinced him that Amazon had a chance of being a big winner. He was later to say that finding Kaphan was one of his luckiest breaks. Paul Burton-Davies was younger, had experience of the Web, but like Kaphan was chosen primarily for his intellectual ability. Neither of them had experience of writing business software for users. This became a pattern – Bezos insisted on getting the brightest possible people, regardless of their specific experience. Scarcely any early recruits came from the book business. Together Kaphan and Burton-Davies constructed the first web site over the next year, working on two Sun workstations in the converted garage. The fourth member of the team was Mackenzie Bezos, who handled the accounts and administration until 1996.

While they beavered away, Bezos took himself to a four-day course on book selling and was inspired by the emphasis on customer service, above and beyond the call of duty. He decided that he must make customer service “the cornerstone of Amazon.com”, the best possible shopping experience that could be provided online. This idea subsequently crystallised into the goal of making Amazon “the most customer-centric company on the planet”.

Amazon was not the first online bookstore. Computer Literacy Bookshops had been selling technical books by e-mail since 1991, when only the technically adept were using the Net, and overt commerce was widely abhorred. In 1992 Books.com started up, using bulletin boards for customer communications, and in 1994 launched a rather crude web site with 400,000 titles at discounted prices.

The Amazonians felt that they could do better, but were happy to adopt many of Books.com’s features such as reviews, excerpts and attempts to cultivate community. They also hedged their bets, building an e-mail based system for ordering books as well as one for the Web. In 1994-95 however, nobody knew whether many consumers could be persuaded to make purchases online. Amazon anticipated that many would prefer to give credit card details over the telephone.

Bezos played an active part in defining the look of the web site, the operating system interface and the database. His own software experience made him a demanding but realistic client. Barton-Davies said that one of the reasons for Amazon’s success was “Jeff’s insistence on everything being done right”. The standard software packages used by mail-order companies would not be good enough for them.

The database was constructed mainly from Bowker’s ‘Books In Print’, a catalogue of 1.5 million titles. Converting it from CD-ROM was a lengthy, labour-intensive process. Amazon’s operating system made it possible to store details on the thousand most popular books in computer memory, which made most responses very fast. When the site was launched, the database contained more than a million titles. Each customer entering the site was assigned an ID so that browsing and buying habits could be analysed.

Amazon knew that it had to be seen to have a secure system for dealing with credit cards, but had no direct experience or knowledge of how card transactions were processed. Security was something they had to work out from first principles, a long learning process. A crucial innovation, much imitated subsequently, was the idea of the ‘shopping basket’. In 1995 the very simple, five-step ordering process was a big breakthrough. Scarcely any other early sites were as well planned and easy to use.

Bezos insisted on keeping transactions as simple as possible, both to make it easy for customers and because scarcely any of them had high-speed connections. Fancy graphics for users with slow dial-up access made many early web sites painfully unusable. Amazon pages, consisting mainly of text, were designed to be downloaded in seconds. Amazon’s site stood out by its stripped-down approach, part of its rigorous focus on the customer experience. For many customers its main advantage was speed and convenience. Bezos believed that “in the late twentieth century, the scarcest resource is time. If you can save people money and time, they’ll like that”.

Launch

After a lengthy and surprise-free beta test, the site was made public in July 1995. The logo was a large ‘A’ with a river running through it and the tagline ‘Earth’s Biggest Bookstore’. The page mainly consisted of text and navigational bars. ‘Spotlight’ was a daily feature on a book with a 40% discount. The top twenty best sellers carried a 30% discount, and about 300,000 books a discount of ten per cent.

They were expecting a trickle of orders in the first few weeks. However, three days after launching they received an e-mail from Yahoo!, asking if they could include Amazon on their ‘What’s Cool’ list, then the busiest page on the Web. Kaphan wondered if accepting might be like taking a sip of water through a fire-hose – in its first week, Amazon took hundreds of orders, but was only able to send out a handful of books. During the first month, they shipped books to every state in the Union and to 45 countries.

These were very much bigger numbers than Bezos had planned for. “I think one thing we missed was that the Internet was exclusively made up of early adopters at that time. So all the people online, even though it was a relatively small number compared to today, were those who liked to try new things.” Bezos had believed that it might take years for large numbers of people to feel comfortable about buying books online and that sales would build slowly. In fact Amazon’s challenge became that of managing very much greater demand than anyone had dreamed of. It was one that it handled brilliantly well, by a mixture of inspired improvisation, systematic analysis and competitive drive.

In the launch period, scarcely any provision had been made for packing, so Bezos and everybody else had to work until midnight every night getting packages ready. They were mostly working on their hands and knees until they had the bright idea of getting tables. Returns were another thing they had not really planned for. Bezos wanted to be generous in order to develop customer loyalty, and gave customers 30 days to change their minds, and the first returns created accounting chaos – yet another set of operational skills and processes that had to be learned from scratch.

Partly for reasons of frugality, and even more to symbolise it, Bezos made himself a desk from a door. He wanted to demonstrate that the company was only spending money on things that mattered to customers. He had to be talked out of putting stickers on all the furniture to show how little money had been spent on it. His desk however proved a sound investment. It featured in nearly every media interview and in a Vanity Fair photo shoot. Soon everyone at Amazon had door desks.

Funding

Initially Bezos paid his colleagues out of his own pocket, putting $60,000 dollars into the company in 1994. His parents invested another $250,000. Money was very tight – the loss for the year would be $303,000, and Bezos anticipated several more like that. He desperately needed more funding and decided to raise it from ‘angel’ investors in Seattle, rather than venture capital.

Bezos subsequently said that raising this first round of just short of a million dollars, mainly in lots of $30,000, was the most difficult task he ever had. He had to pitch to 60 different people to find 20 investors. He told them, “I know nothing about the book industry, nothing . . . But I know that I can get the books here, and I can get them to the customer and forget about bricks and mortar.”

One investment adviser did extensive research on the industry and decided that Barnes & Noble would crush Amazon when they moved online. Most investors raised the objection that people liked going to bookshops. Bezos initial response was that books were one of the few categories where computers were already an aid to selling. But his main argument was that an online bookstore offered customers choice and convenience that conventional stores could not. A survey by the American Booksellers Association showed that book buyers were almost twice as likely as average to use the Net and that half of book buyers under the age of 50 were disposed to buy books online.

Amazon’s business model was also very much more efficient. Conventional American bookstores turned over their inventory less that four times in a year – often much less. Amazon would achieve 150 turns on much of its stock. Furthermore it would receive payment from credit card companies 53 days before it had to pay its suppliers. Conventional stores had to pay for their books on average 79 days before they got their money, mainly because each book sat on their shelves for 161 days before it was sold.

This was a major advantage to Amazon in terms of working capital. It is more debatable whether Amazon really achieved enormous economies of scale, as Bezos argued it would. Distribution and customer service were hardly fixed costs. And as Amazon grew, it ceased to be a virtual organisation, needing to build large warehouses of its own.

Bezos’ concern was that customers would require a long period to be ‘educated’. His business plan in 1995 showed two scenarios. The moderate one projected sales of $11.5 million by 1997, with a profit of $50,000. His fast-growth scenario was for sales of $17.7 million and a profit of $143,000. Sales would in fact almost reach the larger of these two estimates in the first quarter alone of 1997.

By November 1995, Bezos managed to convince 21 individual investors to stump up a total of $981,000. A few months later he was approached by General Atlantic Partners, a large venture capital company on the East Coast. With sales revenues now heading for $5 million, GAP valued the company at around $10 million.

Bezos and his colleagues, however, were realising that Amazon had the chance to dominate not just a large online book sales market, but many others. It could quickly become a multi-billion dollar business, and it should not be too difficult to raise the finance to achieve this. Eric Dillon, one of the early investors and now a financial advisor, saw this as “a crossover point. . . We changed our whole focus to one that was driven by momentum. We had the momentum and we had to keep the momentum. We needed to bring in money. We needed to be the first ones to use national advertising.”

In May 1996, Amazon was effectively granted this wish, free. The Wall Street Journal ran a front-page story under the headline “How Wall Street Whiz found a Niche Selling Books on the Internet”. The article could have been written by Amazon’s PR department. “Its site on the World Wide Web has become an underground sensation for thousands of book lovers around the world, who spend hours perusing its vast electronic library, reading other customers’ amusing online reviews and ordering piles of books.” Orders doubled overnight and continued to accelerate.

The article also triggered a swarm of calls from the venture capital industry. John Doerr of Kleiner Perkins, who had previously been ignoring Eric Dillon’s calls, decided to come to Seattle. His was the firm that Bezos most wanted, for the prestige it would bring, and its contribution to momentum: “Kleiner and John are the gravitational center of a huge piece of the Internet world. Being with them is like being on prime real estate”. Doerr not only invested $8 million for 13% of the company, valuing it at $60 million, he agreed to join the board.

Get Big Fast

The goal now was to grab market share as quickly as possible, regardless of short-term profitability. The strategy was expressed in three words emblazoned on company t-shirts: ‘Get Big Fast’. Bezos wanted to make Amazon’s brand as big and as valuable as Disney’s, and make it difficult for later entrants to online retailing to dislodge it. When Barnes and Noble announced a deal with AOL, George Colony of Forrester Research, pronounced Amazon “dot.toast” – as accurate a judgment as Forrester’s many inflated forecasts.

Amazon used Kleiner Perkins’ money to launch an advertising campaign in the New York Times, Wall Street Journal and USA Today, and did deals with The New Yorker, Atlantic Monthly and Wired. It also continued to get lots of free, almost entirely uncritical publicity. Everyone wanted to write about the Web, and Bezos made Amazon into a story everybody could understand and admire. In December 1996, Fortune declared, “Amazon is truly virtual. Though it has become a multi-million-dollar business that employs 110, there’s still no storefront and little inventory.” In fact, the month before, Amazon had quietly opened its own 93,000 square feet warehouse in Seattle to store copies of popular titles. Having got pretty big very fast, it no longer made any sense to wait to order every book in response to customer orders.

Amazon appointed Deutsche Morgan Grenfell as its financial advisors, and in March 1997, filed a prospectus. This positioned Amazon as a glamorous technology company, rather than a low-margin retailer. It talked of “virtually unlimited online shelf space” and “lack of investment in expensive retail real estate and reduced personnel requirements”. It also disclosed losses to date of $6 million and said that it expected losses to continue for the foreseeable future and at significantly higher levels. It was paying on average $16 to buy each book it sold for $20, and spending $8 on advertising. It acknowledged that barriers to entry were low and that it expected competition to intensify. It could not say when the company would become profitable.

In normal times this would have been put off most investors, but these were not normal times. There was no shortage of experts like Mary Meeker of Morgan Stanley to convince investors that market share was much more important than profits, that companies like Amazon would enjoy big “first mover advantages” and that they would quickly come to dominate their markets. At that point they would be able to extract good profits. Many people were convinced by the ‘winner takes all’ and ‘land rush’ theories and that was what was driving stock market sentiment.

In the first quarter of 1997 Amazon’s sales reached $16 million, more than for the whole of 1996. Losses were almost $3 million, but enthusiasm for the IPO enormous. The IPO in May raised $54 million, valuing the company at $548 million. By September the stock had more than doubled, and Amazon secured credit of $75 million for further expansion. The publicity from the IPO rebounded directly on sales – revenues in 1997 hit $148 million, almost ten times their level the previous year.

Bezos was not just thinking big about book sales. “Our strategy is to become an electronic commerce destination. When someone thinks about buying something online, even if it is something we do not carry, we want them to come to us. We would like to make it easier for people online to find and discover the things they might want to buy online, even if we are not the ones selling them.”

In July Amazon made its long-planned move into music, with 130,000 CD titles available, divided into 14 genres and 280 sub-genres. As usual, great care had gone into the software and the support systems. Customers could sample any of 225,000 songs online. The New York Times rated the search system for songs the best on the Web. Within three months, Amazon had overtaken CD-NOW, the established leader in online music sales, incidentally disproving the universality of the first mover advantage and land rush theories. Almost certainly, this success owed much to Amazon’s customer base and high public profile, but its ability to deliver on the promotional promise was critical.

In 1997, Amazon also started a programme of acquisitions that was later to reach a frenzy. It started with IMDB, an online database of information about movies, in preparation for a move into videos. PlanetAll was a ‘contact management service’ that collected personal information from consumers. Junglee Corp was an early search engine that enabled consumers to make price comparisons. All of these were complementary to Amazon’s main business. It also formed alliances with most of the main players on the Internet at the time – Yahoo, Excite, AOL, Netscape and @Home.

Inside Amazon

The new funding enabled Amazon to take on a lot more people. However, Bezos remained extremely picky – he only wanted the very best, people who had “been successful in everything they had done”, who would commit themselves to the intense corporate culture he was trying to create. He liked to tell potential recruits, ‘‘You can work long, hard, or smart, but at Amazon.com you can’t choose two out of three’’.

John Doerr encouraged him in this, though it occasionally frustrated Kaphan and Burton-Davis who were often desperate to get people on board. Dillon recommended four friends whom Bezos rejected as not quite good enough. Everyone was interviewed several times, with Bezos always having the last word in the early days. “Jeff would find things wrong or right with people that no one else spotted”, according to Glenn Fleishman, the catalogue manager. When he joined in 1996, he felt “it was like going to the world’s best college. I was surrounded by smart people”.

Bezos thought a lot about culture. He told recruits he wanted Amazon’s to be “intense and friendly . . In fact, if you ever had to give up ‘friendly’ in order to have ‘intense’, we would do that. So, if we needed to be ‘intense’ and ‘combative’, we would do that before we’d be ‘not intense’.” He was modelling himself on his neighbour, Bill Gates, but hoped to avoid the ferocious internal competition that was rife at Microsoft. Two key executives, came from the Beast of Redmond. David Risher became VP of product development and Bezos’ right-hand man; Joel Spiegel became VP of engineering.

We have a fascinating account of what it was like to work at Amazon from James Marcus, who joined in 1996 and spent five years there as an editor, mainly writing book reviews. His rather snooty perspective is very different from that of the business journalists, breathless with admiration for Jeff Bezos. Despite his distaste for the ‘Culture of Metrics’ and his difficulties maintaining the “ever more precarious balance between art and commerce”, he paints a largely sympathetic portrait of Bezos, certainly in the early days:

“I had already succumbed to his brand of anti-charismatic charisma, which would have mortified a Great Man of a century ago, but which seemed just right for our nerd-driven meritocracy. Jeff it should be stressed was a likeable and normal person. . . the habit of humourous self-effacement kept his Napoleonic side under wraps.”

Bezos did not just want a metric for customer enjoyment, he felt that they should eventually be measuring customer ecstasy. He believed that ‘content’ would be an essential tool for Amazon and Marcus was one of dozens of professional writers he hired to create it. Marcus liked the “delightful uncertainty about what the company was, when it could be taken for a bookstore or bulletin board or electronic agora, a revolutionary enterprise or (as Jeff continued to say) a destination.” He graphically describes the feverish pace, the constant improvisation, the endless interviews of prospective staff, the boundless ambition. He reports with gloom the arrival of droves of MBAs at the end of 1997, bringing with them a plague of jargon: “Pulling on revenue levers meant making more money. If we leveraged our verbiage correctly, the division would soon reach an inflection point (we would make more money). The main thing in any case was to monetise those eyeballs.”

By 1999, Marcus felt that “monetising those eyeballs had become a lynchpin of Amazon’s corporate philosophy”, and as time went on, he was observing the company from the outside, where he clearly intended to be as soon as he could exercise his stock options.

Customer-Centric

Bezos frequently declared his ambition to make Amazon “the most customer-centric company in history”. Certainly he made it more responsive to customers than any other e-commerce business, none of whom ever actually met any of their customers. He believed that the Web required a completely different approach. In the physical world a merchant might spend 30% of his time creating a good customer experience and 70% shouting about it. On the Web, the proportions had to be reversed.

He set out to engineer the customer experience, systematically planning every step, every element, from the time to download Web pages and the taking of the order through to delivery. He wanted the brand above all else to stand for a great customer experience. Precisely because the early Internet was primitive and slow, the value proposition needed to be ‘overwhelming’, to persuade potential customers to change their habits. Hence the metric for customer enjoyment and the musings about ecstasy.

He acknowledged that “We will never make Amazon.com fun and engaging in the same way as the 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 that will blow people away and make the experience an engaging and fun one.”

In the early years, editors like Marcus tried giving customers the kind of advice they might find in an independent bookstore. Marcus’ boss, Rick Ayre, said he wanted users to “get a sense of the quirky, independent, literate voice, and that behind it all you’re interacting with people, and that it’s people who care about these things, not people who are trying to sell you these things.” The emphasis had changed by 1999, but there is little evidence that many customers were put off. Given that the vast majority were using Amazon for the choice and convenience, the sophisticated editorial offerings were a nice bonus for some, rather than an essential feature. What was important was having enough information about the book, and Amazon was happy to get this from any reputable source, notably the publisher, and reviews written by customers. Whilst forthright critical reviews came to be discouraged, as they generated hundreds of complaining emails. Bezos himself stirred up a controversy in 1999 when he refreshingly summed up a book on Internet business planning as “Stupid book . . . don’t waste your time”.

Increasingly, Amazon started to make suggestions to individual customers, based on previous purchases, part of its shift towards personalising their experience. These new approaches made use of relational software and collaborative filtering, which told customers of books chosen by other customers who had bought the same title. In 1998, Bezos declared that most e-commerce was just about “find-it, buy-it, ship-it”. What Amazon was doing was ‘e-merchandising’. “We can use advanced technology to not only understand our products . . . but to understand our customers”

Amazon invested heavily in market research, particularly focus groups, to get a detailed picture of customer likes and dislikes, and before introducing new features like ‘1-click shopping’. This was another significant step towards making the process of placing an order as simple and easy as possible, indeed too simple for some, who did not believe that they had actually completed it, until Amazon made the confirmation more explicit.)

The term ‘e-merchandising’ never quite caught on, but the idea of systematically organising the approach to the customer experience did clearly distinguish Amazon from most of its competitors, direct and indirect, in the early years of e-commerce. Like Dell and Starbucks, it made its marketing operations almost scientific in their precision. In the early days, it took four days to ship books that distributors had in stock, but they aimed to get that down to 24 hours. To ensure that every book arrived in perfect condition they ‘over-packed’.

The brightest people in Amazon worked in customer service. They were trained to know every step in the company’s operation, and to rectify the impersonal aspect of buying books online. If there was a problem, Bezos was determined that Amazon would not only be responsive but demonstrate an exceptional commitment to helping customers. Patricia Seybold recounted receiving a package with a hand-written note: “ ‘We know you ordered the softcover version of this book, but it’s out of stock so we sent you the hardcover version for the same price’. That little hand-written note cemented my love affair with Amazon.”

Delusions of grandeur

The remarkable growth of the Internet in the late nineties and the even more extraordinary rise in share prices went to the heads of most entrepreneurs, and Bezos was no exception. Given that he was not the most modest of men to begin with and that he was being hailed as a business genius by almost the entire media industry, this is hardly surprising. In November 1999, Business Week wrote breathlessly:

“When we try to comprehend something as vast, amorphous and downright scary as the Internet, it’s no wonder we grope for familiar historical precedents – the railroads, the interstate highway system, the telephone network. But none of these really captures the Internet’s earthshaking impact on the business world. For that, we must take the advice of Internet commerce pioneer, Jeffrey P. Bezos.”

It was in this year that he had started to implement his dream of making Amazon ‘the Wal-Mart’ of the Internet, and dominating not just sales of books and music, but just about anything. He believed that eventually 15% of all retail sales would move online, and he intended Amazon to have a large chunk of that $500 billion.

In March 1999, Amazon attacked the rising star, eBay, with its own auction site and much better customer service. The previous year, eBay executives had gone to Seattle to propose collaboration on cross-promotions. To their astonishment, Bezos informed them that with 8 million customers, Amazon would soon crush their little operation.. Amazon’s auction site was positioned as a service that helped consumers to find products, and was aimed rather more at buyers than sellers. “We want to build a place where people can come to find anything they might want to buy online. You realise very quickly that you can’t sell everything people might want directly. So instead you need to do that in partnership with thousands and indeed millions of third party sellers in different ways. To try to do that alone, in strictly a traditional retailing model, isn’t practical.”

In June, following eBay’s acquisition of Butterfield’s, the US auction house, Amazon bought a stake in Sotheby’s. They set up a joint auction site for up-market items, which was launched in November, but t was not a success. Traditional Sotheby’s customers had less need of an online service than the smaller scale dealers and buyers who had been flocking to eBay. Their collectors valued the atmosphere of live auctions in swish locations.

With its stock riding high, Amazon embarked on a shopping spree of acquisitions and investments. In February 1999 it bought 46% of Drugstore.com and in March, 50% of Pets.com. Both had business models clearly inspired by Amazon’s, but with far less competitive advantage.. In April, Amazon acquired outright e-Niche, which ran two marketplace sites, and Accept.com, which developed software that enabled person-to-person Web transactions. In May it acquired 35% of Home Grocer.com for $42.5 million in cash; in July, 49% of Gear.com; and in September, 20% of a wedding gift registry.

Its biggest move towards becoming the ‘electronic commerce destination’ was the launch of zShops in November 1999. It had been working all year on building a network of thousands of affiliates. ZShops was to be an online shopping mall, where customers could find everything bar firearms, live animals, pornography and tobacco. “Sixteen months ago, we were a place where people came to find books” said Bezos. “Tomorrow, we will be a place to find anything, with a capital ‘A’.”

This particular venture, unlike most of the others and the core business of books and music, really did enjoy minimal marginal costs, since Amazon itself, like eBay, did not have to handle or deliver anything, but simply collect commissions on every transaction. The underlying logic of the new strategy of getting many times bigger than anyone had previously dreamed of was increasing returns to scale. Yet in 1999, Amazon opened five new warehouses, in addition to the two it had in Seattle and Delaware. These employed thousands of people, not all of whom were counted as employees. It was a logical step in providing better service to customers, but it showed that Amazon was not as virtual as it had first appeared and nor was its business model anything like as radical as eBay’s.

This paradox was highlighted by the financial results. In the third quarter of 1999, sales rose 130% from the same quarter of 1998, to $356 million. Losses however rose more than fourfold to $197 million – customer acquisition costs had risen to $35. For the first time, several prominent analysts started to question Amazonian economics. Henry Blodget, whose bullish pronouncements normally made Mary Meaker sound conservative, downgraded Amazon’s rating to ‘accumulate’, and several others marked it down. The chinks in the armour were clear to many now.

Nonetheless, the year ended gloriously, with Bezos’ face on the cover of Time magazine, as their ‘Man of the Year’, and the share price at a new peak of $113.

Crisis

The crisis that hit Amazon in 2000 was primarily a general collapse of confidence in Internet business models and valuations. Ominously, in February, Pets.com’s IPO flopped and Amazon announced losses for the previous quarter of $323 million, seven times their level a year previously. Bezos described this as a “high water mark”. From now on he would be leading a “drive towards profitability” and specifically to reduce the percentage loss on sales to single figures by the end of the year. For a few weeks the market seemed reassured by this, but these were the last days of the bubble.

In March Nasdaq, the stock exchange where most Internet stocks were traded, reached its peak and then started its rapid fall. Amazon’s share price plummeted to $52 in April, but there was much worse to come. In June, Ravi Suria of Lehman Brothers wrote that Amazon “shows the financial characteristics that have driven innumerable retailers to disaster throughout history”. He predicted that it could run out of working capital by the end of the year. Amazon called the report “hogwash”, but clearly Suria was not entirely wide of the mark. However, Bezos had been serious about improving profitability. In the third quarter, losses were reduced from 22% of sales to 11%. The books, music and video division earned a profit of $25 million on $400 million sales. In the new mood of deep investor pessimism, this brought only temporary relief. The share price ended the year at $15.

The rate of attrition was frightening. Pets.com, which was totally dependent on external funding to cover its huge losses was the first publicly traded Internet company to close down, taking with it the $60m Amazon had invested. HomeGrocer.com was off-loaded at a loss.

However, Amazon was slowly turning the corner. Although losses for the year as a whole were $1.4 billion, they were reduced in each quarter and there was still $1.1 billion in the bank. In January 2001, the company closed down two of its distribution centres and laid off 1,300 people. The era of land grabs and Get Big Fast was over, but the core business was now quite healthy.

The company achieved the highest ever score for a service business in the University of Michigan’s Customer Satisfaction Index in 2000 and maintained the same score in 2001. Sales growth stagnated for a while in the US, but grew by 72% internationally. Losses were reduced to $567 million, but the share price continued to drop to $6 by the end of the year. That was as much as anything else a reflection of general business confidence in that traumatic year. In fact, online sales did much better than conventional retailing in the period after September 11.

Postscript

Amazon’s former preeminence as the e-commerce business has waned slightly with the many other retailers who now have an online presence, but it remains dominant in its core Media business of books, music and DVDs, where sales reached $9.2 billion in 2007. However intense competition in this sector will make it difficult ever to achieve high levels of profitability.

Since 2001, sales have grown steadily, particularly internationally, in non-media products and through third parties, reaching $14.8 billion in 2007. The company achieved a tiny operating profit in 2002, and in 2003 its first ever net profit, of $35 million, less that 1% of sales. By 2007, profits had climbed to $588 million, 4% of sales. Margins would have remained considerably lower had it not been for sales through third parties, where Amazon’s variable costs are negligible. By 2004, there were more than 100,000 third party sellers – Amazon Marketplace, for part-timers, and Amazon Merchants for other retailers – and these accounted for nearly a third of total revenues.

For a while it appeared to be eclipsed by the growth of eBay, which became the largest and most profitable e-commerce business . However, Amazon’s networked model, with its emphasis on buyers rather than sellers, began to look more durable than eBay’s rather grasping one, which was alienating many of its traders. Certainly Amazon’s brand shines more brightly, commands strong customer loyalty and wins high levels of repeat business.

In 2004, Amazon started to reinvent itself in a way that surprised even Bezos. It realized that its enormous computing infrastructure, built to cope with Christmas peaks, could be used by other firms. Starting with its retail partners, by 2008 it had signed up 370,000 customers, from tiny websites to the New York Times, who rent computing capacity instead of building their own data centres.

Judged by most standards, Amazon’s achievements are astounding. It is only in relation to the over-ambitious goals it set itself in the late nineties, and the ridiculous hype of that era, that its success is a qualified one. To go from inadequately funded start-up to sales of $150 million in two years is impressive by any standards. To reach revenues of $15 billion in another ten is extraordinary.

On the success factors for market creation, Amazon scores almost top marks. Its strategic vision was radical and crystal clear, and it adapted quickly and boldly to new circumstances. Its competitive advantage was built solidly on a set of highly distinctive capabilities, which it has taken care to renew and enhance. It was exceptionally entrepreneurial and ambitious, but intensely disciplined from the outset. It understood from the outset the importance of a compelling value proposition and was consistent in articulating and justifying it.

Jeff Bezos, unlike the vast majority of Internet entrepreneurs, has proved an exceptionally smart strategist. He is also one of the very few business founders to have led his company all the way from start-up to multi-billion dollar success.