How did Amazon grow (focus on products/services): - During the dot com boom (late1990s to the early2000s) ? - After the dot com bust (early2000s to mid2000s) ? - Recently ?

Where are the similarities across the eras ?

Why are they able to grow successfully within each era ?

Describe the management practices that you believe will be important for Amazon's successful continuation of their growth.

What do you think is the role of the devices, e.g. the Kindle, in their plan for growth?

Amazon Article:

Jeff Bezos' store in the sky is hard to beat for books, CDs, and a zillion other products. It's also great for quick technology fixes. Say you need a fat HP server for hosting the too-moronicto-fail Facebook app you plan to launch next week. Only $1,300 and change! Hit 1-Click. Select expedited shipping. What's for lunch? But there's a cheaper, faster, better way to satisfy your hardware jones. Tucked over on the left side of the page, the nerd gnomes in Beacon Hill, Seattle, have embedded an option that blows computer shopping into, well, the clouds. Click on "Amazon Web Services." Key in your Amazon ID and password and behold: a data center's worth of computing power carved into megabytesized chunks and wired straight to your desktop. Clones of that HP tower cost 10 cents per hour 10 cents! and they're set to start spitting out widgets as soon as you upload the code. Virtual quad cores are a princely 80 cents an hour. Need storage? All you can eat for 15 cents per gigabyte per month. And there's even a tool for monitoring your virtual stack with an iPhone. No precious cash tied up in soon-to-be-obsolete silicon, no 3 am runs to the colo cage. Outsource your infrastructure to Amazon! There's a name for this: cloud computing. Or less poetically, utility computing, or even the unfortunate acronym HaaS, meaning hardware as a service. Whatever it's called, Jeff Bezos is loving it. On a sunny winter day, with Mount Rainier and the Olympic range on spectacular display, the bald billionaire sits in a conference room in the converted VA hospital that serves as Amazon headquarters. He's ticking off all the reasons why Amazon Web Services and its industrial-strength suite of online processing power, storage, and other geeky treats are the best thing to come out of the company since 1-Click. AWS capitalizes on Amazon's combination of computational skills and operational savvy. It piggybacks on a multibillion-dollar IT infrastructure. And it pulls in a whole new category of customers looking for rock-solid scalable computing on demand blue-chip startups like Zillow and PowerSet, kids in garages building the next Google, even adventurous corporate IT jocks looking to offload some of the drudgery. But hold on, Jeff. Back in the material world of FedEx boxes and holiday wish lists, some 80 million Amazon customers this year are expected to drop $20 billion nonvirtual dollars into the Beacon Hill till. Computer hardware is right up there on the hit parade with Harry Potter novels and boxed DVDs of The Simpsons. For years, Wall Street and Silicon Valley alike have rolled their eyes at the legendary Bezos attention deficit disorder. What's that secret pet project? Spaceships! Earth to Jeff: You're a retailer. Why swap pricey stuff in boxes for cheap clouds of bits? Cue one of Bezos' famed mirth bursts: "That's a really good question!" Honk! But Bezos being Bezos ex-hedge-fund quant, cool Seattle Spock to Silicon Valley's high-flying Captain Kirks there's also a really good answer, and it goes something like this: Don't get hung up on ecommerce. Don't fixate on Amazon Prime, media downloads, or Kindle eReaders. And never, ever listen to Wall Street. Because while the pinstripes gnashed their teeth about costs more than $2 billion spent on IT! Amazon was bolting together the world's broadest, deepest, smartest, most versatile global technology platform, ready for anything from running data centers to delivering cat litter. Now Bezos' globe-spanning empire is primed to lead the migration of computing power from the desktop to the Net. It's a shift potentially as dramatic as the desktop revolution that another Seattle-area company often similarly underestimated rode to power two decades ago. Jeff Bezos' aerial department store leafing out into a digital cloud forest. Amazon, indeed. If you're an Amazon customer living between the Golden Gate Bridge and the western edge of the Rockies, there's a good chance your latest shipment originated in a big box rising from the sagebrush in Fernley, Nevada, a half hour drive up I-80 from Reno. If a spreadsheet turned into a building, this would be it: 14 windowless football fields of wide-open concrete. Inside, it's a Donkey Kong world of overhead conveyor belts, automated chutes, and robotic tilt trays. The floor is surrounded by two-story metal storage racks that could double as a data center's hard disk arrays. Tom Taylor, an operations maestro visiting from Seattle, reaches toward one of the shelves and pulls out a hefty tome entitled Introductory Physics With Algebra as a Second Language: Mastering Problem-Solving. It's lightly used, with a barcode label pasted neatly on the cover. "The seller could be anybody," Taylor half shouts through the din and mandatory earplugs. "You need a scanner to know whose it is." Amazon is notoriously tight with operational data. But the company does provide a few figures, and one shines a light on how far Amazon has come from the old Wal-Mart-in-the-sky concept. Almost a third of the store's total number of sales last year were made by people selling their stuff through the Amazon machine. The company calls them seller-customers, and there are 1.3 million of them. They're just one wave of outsiders to have washed over Bezos' threshold. The next is already breaking: Web developers itching to make their mark on the Net. The transformation from storefront to platform happened by evolution rather than design. It began way back in the dialup dark ages, when Bezos and his crew were looking for a cheap and easy way to expand the store's reach. One solution was to let Web developers link directly to Amazon products on their own pages and give them a 15 percent bounty on sales they generated. Today that program, Amazon Associates, is the Net's definitive distributed sales network, turbocharged by the world's biggest archive of product information, customer reviews, and recommendations. One spooky Associates service dubbed Omakase Japanese for "Leave it up to us!" offers up Amazon product ads keyed to your site's content, your audience's response to earlier ads, and (gulp) the specific viewer's past Amazon purchases. Google AdSense with a kickback. Turning other people's sites into Amazon storefronts set a stage for the next leap: letting outsiders sell their own stuff through Amazon's platform. Third-party selling began in 1999 as a jittery response to eBay, which looked smart for avoiding handling things physical altogether. Amazon's auctions were a bust, but they kicked off what is now the thriving Amazon Marketplace ("Sell Your Stuff"). The company earns a cut of up to 15 percent from sellers who opt for plain-vanilla product listings ("12 used & new from $58.99"). High rollers can sign up for top-of-the-line Fulfillment By Amazon, which parks their goods on the shelves (45 cents per cubic foot per month), then slips them into Amazon's own shipping stream. If you're Target or the NBA (or, until recently, Borders), you can sit back and rent a world-class ecommerce operation, from shopping cart to warehouse to delivery. The rationale for all this, in Jeff-speak, is "totally not abstract." Meaning it's a no-brainer. "If you're trying to have not just every book but every out-of-print book, every indie-band CD, and every vinyl record, there's no way you're going to do it on your own. You have to open up." That turns out to be harder than it sounds. When Amazon plunged into auctions, its simple dotcom-era server-and-database combo was already staggering under the combined weight of growing complexity recommendation engines, transaction tools, a steady rain of new features and mushrooming traffic. Beset by problems with speed, reliability, and scalability, engineers first tried migrating the whole hairball to a mainframe. It didn't help. Then they gambled on a Net-centric idea just beginning to bubble through the computer science world: loosely joined software modules, each performing a specific function while exchanging information with the others. Bingo. Log in to Amazon's gateway today and more than 100 separate services leap into action, crunching data, comparing alternatives, and constructing a totally customized page (all in about 200 milliseconds). The new architecture brought a huge improvement along with an unexpected opportunity. "We were already building these really incredible, robust services for ourselves," says Andy Jassy, who wrote the original Amazon Web Services business plan four years ago and has been running the operation ever since. "It wasn't a huge leap to realize they could be valuable to other people." Thus Amazon's next new business: opening not just the store in the sky, but the very silicon and software that keeps it aloft. Just across Highway 101 from the Googleplex in Mountain View, California, a banner fluttering from a nondescript building heralds yet another high tech dream: Ooyala. Inside, a dozen software engineers wearing headphones pore over screenloads of code while grazing on M&Ms and organic treats served in plastic tubs. The three founders, all Google emigres, are chasing a suitably big idea interactive "hypervideo" with seamlessly embedded links. What they're not doing is blowing capital on shiny hardware. They built their television killer using little more than the desktops in front of them, plus major slices of computing power courtesy of Mr. Bezos. "You can't execute fast when you're running to the data center every night to fix machines," says Ooyala engineering chief Sean Knapp. "Infrastructure is the big guys' most powerful asset. This levels the field." Ooyala that's "cradle" in India's Telugu language epitomizes Amazon's vision of utility computing. Last December, Knapp and his crew beat more than 900 other AWS-powered startups in a demo derby, taking home $50,000 in cash and $50,000 in usage credits. (Bonus prize: a decommissioned Amazon server plus a gold-painted, Bezos-autographed sledgehammer to bash it with.) Ooyala joins a who's who of high-profile AWS clients, including Justin.tv, which serves up a half-million live videostreams daily thanks to Amazon hardware. Real estate site Zillow recalculated the values of 67 million US homes in several massive runs, and the gray old New York Timescrunched 150 years' worth of newspapers into searchable PDFs. An estimated 10,000 new developers are signing up monthly, with the total closing in on 400,000. As New York venture capitalist Fred Wilson Twittered during a brief outage of Amazon's S3 storage service in February, "amazing how many of the services i use are reliant on S3." Among the temporary casualties that morning was Twitter itself, which stashes user profile images on Amazon's servers. Utility computing is Web 2.0's version of rocket fuel. "You don't generate your own electricity," Bezos says. "Why generate your own computing?" The forces driving online apps Internet bandwidth and reliability also mean that, in terms of data per dollar, servers in your closet or colocation facility can't compete with industrial-scale bits piped in from hundreds, even thousands, of miles away. Amazon's developer-customers reap several advantages. Capital efficiency is one. (Trendy Mountain View- and Delhi-based startup SlideShare cheekily dubbed one of its presentations "Using S3 to Avoid VC.") Another is instant scalability. AWS users can add or delete server and storage capacity without getting up from their Aerons. AWS trades heavily on the fact that Amazon itself is the biggest customer. "We tend to build the things people are asking for internally," Jassy says. And the idea that AWS is mostly about wringing extra bucks (especially off-season) out of Amazon's data centers? "We've far exceeded the excess capacity of our internal system," Jassy says. "That ship sailed 18 months ago." For a company at which operational data is a state secret, that's a telling detail: AWS is now big enough to be piling up its own silicon. A museum of Amazon memorabilia decorates the Beacon Hill building's lobby. One cabinet enshrines a faded copy of the May 31, 1999, issue of Barron's. "Amazon.bomb," the cover screams. Barnes & Noble, then the King Kong of booksellers, had already followed Bezos online. Wal-Mart was gearing up for a Net offensive, too. And Amazon's numbers, Barron's smirked, were "gimmicky." Nearly a decade later, Beacon Hill is still on top. "We can live with being underestimated," Bezos says. That's easy to do, for the moment. Wall Street's best guesses for AWS's 2007 revenue don't even reach $100 million the proverbial pimple amid nearly $15 billion in overall sales. Projections are doubly hard to gauge because a big part of the appeal to Web-focused startups is what Bezos calls "paying by the drink." Upload an app that no one uses, and Amazon doesn't charge a penny. Stable startups like widget maker Renkoo talk about spending around $1,000 a week on AWS; storage-heavy customers like photo site SmugMug can push that into five figures. But there is one indisputably positive indicator, for utility computing in general and maybe for Amazon in particular: the pack of other companies building data centers capable of delivering similar services. Salesforce.com already offers Force.com, a platform for online business apps. Storage giant EMC recently paid $76 million for the online backup service Mozy. IBM has a project dubbed Blue Cloud, teed up to launch this spring, that reportedly targets financial services companies. Then there's Google, which has amassed a purported motherlode of computing power and bandwidth. And across Lake Washington there's Microsoft, orchestrating a global data center buildout. Bill Gates' anointed successor as Redmond's software king, Ray Ozzie, recently credited Amazon with "some fairly provocative, interesting services." But he also predicted "big volleys" from the rest of the pack over the next 18 months. B ezos doesn't disagree. "I'd be surprised if no one else does this," he says, pausing for effect. "It's a really good idea!" And there may be an ace up his sleeve. Any economist will tell you that a commodity business storing and processing data, for instance is a mug's game, with prices that plunge inevitably toward the cost of production (in the case of bits, pretty close to zero). That's music to Bezos' ears. "Commodity businesses don't scare us," he says. "We're experts at them. We've never had 35 or 40 percent margins like most tech companies." That's not Amazon's only edge. Chasing those fat margins, both Google and Microsoft are building their own Web apps the same market indie developers can enter on the cheap using AWS. Sergey and Larry will gladly toss you the API for Google Earth but if you're building a better mapping app, Amazon is your place. And even if Google or Microsoft eventually does embrace utility computing, would anyone building, say, a better search engine want to park their code in Mountain View? As VC Fred Wilson suggests, "Amazon is out-Googling Google." One sure sign of where things are heading is the flock of geeky enablers circling Amazon's cloud. In November, Red Hat released a version of Enterprise Linux tuned for Amazon's services, opening a trove of 3,000 Red Hat-certified apps to AWS developers. The Ruby on Rails crowd is all over Amazon's instant infrastructure, as are communities pushing things like Erlang (an open source language for parallel processing) and Hadoop (an open source equivalent of Google's core software). Many startups, like Elastra and Coghead, are building first-wave AWS tools to take developers to the next level even faster. "Web services are part of a tectonic shift," says Adam Selipsky, who heads the AWS product team. "Fortress-style software development is history." Bezos likes to say that Amazon is a software company at heart. That has a corollary: The more open it is, the more powerful it is. Consumers continue to get the world's best retail experience. Long-tail retailers get the world's best back office, plus shelf space in the mother of all malls. Developers get a cheap, instant, essentially limitless computing cloud. And there's the real disruption. Instead of building cute apps and ladling them out to the masses the Google and Microsoft model Amazon is delivering silicon power to the people. Bezos is fueling a flotilla of nimble, aggressive entrepreneurs, including frontier types chafing inside gilded prisons like the Googleplex. For them, AWS is a launch pad, not just for the next million Facebook apps, but also for personal live TV channels, virtual desktops, pay-by-the-mile auto insurance, and no doubt plenty of things no one has thought of yet. When Bezos wrote the company's original business plan famously careening toward Seattle in a Chevy Blazer, his wife MacKenzie at the wheel he knew little about books and cared not much more: They were merely a vehicle for a larger vision. "We're willing to plant seeds and wait five to seven years for them to turn into trees," Bezos says. "We focus our strategies on delivering fundamental services better than anyone else. Technology changes. Competitors change. But 10 years from now, nobody's going to say to us, We love what you do, but it's too fast.' Or It's too reliable,' or too cheap.'" Someone is watching the results particularly the 30-plus percent annual sales growth across Amazon's sprawling bazaar. Eighteen months ago, half the analysts following the company rated the stock "sell" or "underperform" Wall Street's equivalent of an F. Since then, the cost of an AMZN share has doubled, while the rest of high tech is relatively flat at best. Which is one reason why Bezos, who owns nearly 100 million shares worth more than $7 billion gives every appearance of being the happiest man in high tech. "People in this business spend a lot of time looking at ideas and asking, why do that? But sometimes the more powerful question is, why not?"