Tech Book of the Month

This book is a great deep dive on the history of Amazon and how it became the global powerhouse that it is today.

Tech Themes

The Everything Store: Jeff Bezos and the Age of Amazon

By Brad Stone

The Birth of AWS. We’ve looked at the software transition from on premise, license maintenance software to SaaS hosted in the cloud, but let’s dive deep into how the cloud came to be. The first ideas of AWS go back to 2002 when Bezos met with O’Reilly Media, a book publisher who in order to compete with Amazon, had created a way to scrape the latest book rankings off Amazon’s website. O’Reilly suggested creating a set of tools to let developers access Amazon’s rankings, and in 2003 Amazon launched Amazon Web Services (AWS) to create commerce API’s for third parties. Around this time, Amazon had centralized its IT computing resources in a separate building with hardware professionals operating and maintaining the infrastructure for the entire company. While parts of the infrastructure had improved, Amazon was struggling internally to provision and scale its computing resources. In 2004, Chris Pinkham, head of the infrastructure division, relocated to South Africa to open up Amazon’s first office in Cape Town. His first order of business was to figure out the best way to provision resources internally to allow developers to work on all types of applications on Amazon’s servers. Chris elected to use Xen, a computer that sits on top of infrastructure and acts as a controller to allow multiple projects access the same hardware. This led to the development of Elastic Compute Cloud (EC2). During this time, another group within Amazon was working on solving the problem of storing the millions of gigabytes of data Amazon had created. This team was led by Alan Atlas, who could not escape Bezos’ laser focus: “It would always start out fun and happy, with Jeff’s laugh rebounding against the walls. Then something would happen and the meeting would go south and you would fear for your life. I literally thought I’d get fired after everyone one of those meetings.” In March 2006, Amazon launched the Simple Storage Service (S3), and then a few months later launched EC2. Solving internal problems can lead to incredibly successful companies; Slack, for example, originally started as a game development company but couldn’t get the product off the ground and eventually pivoted into the messaging giant that it is today: “Tiny Speck, the company behind Glitch, will continue. We have developed some unique messaging technology with applications outside of the gaming world and a smaller core team will be working to develop new products.”
A9. In the early 2000s, Google arrived on the scene and began to sit in between Amazon and potential sales. Around this time, Amazon’s core business was struggling and a New York Times article even called for Bezos to resign. Google was siphoning off Amazon’s engineers and Bezos knew he had to take big strategic bets in order to ward off Google’s advances. To do that, he hired Udi Manber, a former Yahoo executive with a PhD in computer science who had written the authoritative textbook on Algorithms. In 2003, Udi set up shop in Palo Alto in a new Amazon subsidiary called A9 (shorthand for Algorithms). The new subsidiary’s sole goal was to create a web search engine that could rival Google’s. While A9.com never completely took off, the new development center did improve Amazon’s website search and created Clickriver, the beginning of Amazon’s advertising business, which minted $10B in revenue last year. Udi eventually became VP of Engineering for all of Google’s search products and then its Youtube Division. A9 still exists to tackle Amazon’s biggest supply chain math problems.
Innovation, Lab126 and the Kindle. In 2004, Bezos called Steve Kessel into his office and moved him from his current role as head of Amazon’s successful online books business, to run Amazon Digital, a small and not yet successful part of Amazon. This would become a repeating pattern in Kessel’s career who now finds himself head of all of Amazon’s physical locations, including its Whole Foods subsidiary. Bezos gave Kessel an incredibly abstract goal, “Your job is to kill your own business. I want you to proceed as if your goal is to put everyone selling physical books out of a job.” Bezos wanted Kessel to create a digital reading device. Kessel spent the next few months meeting with executives at Apple and Palm (make of then famous Palm Pilots) to understand the current challenges in creating such a device. Kessel eventually settled into an empty room at A9 and launched Lab126 (1 stands for a, 26 for z – an ode to Bezos’s goal to sell every book A-Z), a new subsidiary of Amazon. After a long development process and several supply chain issues, the Company launched the Kindle in 2007.
Business Themes
Something to prove: Jeff Bezos’s Childhood. What do Jeff Bezos, Steve Jobs, Elon Musk and Larry Ellison (founder of Oracle) all have in common? They all had somewhat troubled upbringings. Jobs and Ellison were famously put up for adoption at young ages. Musk’s parents divorced and Elon endured several years of an embattled relationship with his father. Jeff Bezos was born Jeffrey Preston Jorgenson, on January 12, 1964. Ted Jorgenson, Bezos’s biological father, married his mother, Jackie Gise after Gise became pregnant at age sixteen. The couple had a troubled relationship and Ted was immature and an inattentive father. The couple divorced in 1965. Jacklyn eventually met Miguel Bezos, a Cuban immigrant college student, while she was working the late shift at the Bank of New Mexico’s accounting department. Miguel and Jacklyn were married in 1968 and Jeffrey Jorgenson became Jeffrey Bezos. Several books have theorized the maniacal drive of these entrepreneurs relates back to ultimately prove self-worth after being rejected by loved ones at a young age.
Anti-Competitive Amazon & the Story of Quidsi. Amazon has an internal group dubbed Competitive Intelligence, that’s sole job is to research the products and services of competitors and present results to Jeff Bezos so he can strategically address any places where they may be losing to the competition. In the late 2000s, Competitive Intelligence began tracking a company known as Quidsi, famous for its site Diapers.com, which provided discount baby products that could be purchased on a recurring subscription basis. Quidsi had grown quickly because it had customized its distribution system for baby products. In 2009, competitive intelligence reached out to Quidsi founder, Marc Lore (founder of Jet.com and currently the head of Walmart e-commerce) saying it was looking to invest in the category. After rebuffing the offer, Quidsi soon noticed that Amazon was pricing its baby products 30% cheaper in every category; the company even tried dropping prices lower only to see Amazon pages reset to even lower prices. After a few months, Quidsi knew they couldn’t remain in a price battle for long and launched a sale of the company. Walmart agreed in principle to acquire the business for $900M but upon further diligence reduced its bid, which prompted Lore to call Amazon. Lore and his executive team went to meet with Amazon, and during the meeting, Amazon launched Amazon Mom, which gave 30% discounts on all baby products and allowed participants to purchase products on a recurring basis. At one point, Amazon’s prices dipped so low it was on track to lose $100M in three months in the diapers category alone. Amazon submitted a $540M bid for Quidsi and subsequently entered into an exclusivity period with the Company. As the end to exclusivity grew nearer, Walmart submitted a new bid at $600M, but the Amazon team threatened full on price war if Quidsi went with Walmart, so on November 8, 2010, Quidsi was acquired by Amazon for $540M. One month after the acquisition, Amazon stopped the Amazon Mom program and raised all of its prices back to normal levels. The Federal Trade Commission reviewed the deal for four months (longer than usual), but ultimately allowed the acquisition because it did not create a monopoly in the sale of baby products. Quidsi was ultimately shut down by Amazon in 2017, because it was unable to operate it profitably.
The demanding Jeff Bezos and six page memos. At Amazon, nobody uses powerpoint presentations. Instead, employees write out six page narratives in prose. Bezos believes this helps create clear and concise thinking that gets lost in flashy powerpoint slides. Whenever someone wants to launch new initiative or project, they have to submit a six page memo framed as if a customer might be hearing it for the first time. Each meeting begins with the group reading the document and the discussion begins from there. At times, especially around the release of AWS, these documents grew increasingly complex in length and size given the products being described did not already exist. Bezos often responds intensely to these memos, with bad responses including: “Are you just lazy or incompetent?” and “If I hear that idea again, I’m gonna have to kill myself” and “This document was clearly written by the B team. Can someone get me the A team document? I don’t want to waste my time with the B team document.” Its no wonder Amazon is such a terrible place to work.

Dig Deeper

Only the Paranoid Survive: How to Exploit the Crisis Points That Challenge Every Company

By Andrew S. Grove

This book details how to manage a company through complex industry change. It is incredibly prescient and a great management book.

Tech Themes

The decoupling of hardware and software. In the early days of personal computers (1980s) the hardware and software were both provided by the same company. This is complete vertical alignment, similar to what we’ve discussed before with Apple. The major providers of the day were IBM, Digital Equipment Corporation (DEC - Acquired by Compaq which was acquired by HP), Sperry Univac and Wang. When you bought a PC, the sales and distribution, application software, operating system, and chips were all handled by the same Company. This created extreme vendor lock-in because each PC had different and complicated ways of operating. Customers typically stayed with the same vendor for years to avoid the headache of learning the new system. Over time, driven by the increases in memory efficiency, and the rise of Intel (where Andy Grove was employee #3), the PC industry began to shift to a horizontal model. In this model, retail stores (Micro Center, Best Buy, etc.) provided sales and distribution, dedicated software companies provided applications (Apple at the time, Microsoft, Mosaic, etc.), Intel provided the chips, and Microsoft provided the operating system (MS-DOS, then Windows). This decoupling produced a more customized computer for significantly lower cost and became the dominant model for purchasing going forward. Dell computers were the first to really capitalize on this trend.
Microprocessors and memory chips. Intel started in 1968 and was the first to market with a microchip that could be used to store computer memory. Demand was strong because it was the first of its kind, and Intel significantly ramped up production to satisfy that demand. By the early eighties, it was a computer powerhouse and the name Intel was synonymous with computer memory. In the mid-eighties, Japanese memory producers began to appear on the seen and could produce higher quality chips at cheaper cost. At first, Intel saw these producers as a healthy backup plan when demand exceeded Intel’s supply, but over time it became clear they were losing market share. Intel saw this commoditization and decided to pivot out of the memory business and into the newer, less-competitive microprocessor business. The microprocessor (or CPU) handles the execution of tasks within the computer, while memories simply store the byproduct of that execution. As memory became easier to produce, the cost dropped dramatically and business became more competitive with producers consistently undercutting each other to win business. On the other hand, microprocessors became increasingly important as the internet grew, applications became more complex and computer speed became a top selling point.
Mainframes to PCs. IBM had become the biggest technology company in the world on the backs of mainframes: massive, powerful, inflexible, and expensive mega-computers. As the computing industry began to shift to PCs and move away from a vertical alignment to a horizontal one, IBM was caught flat-footed. In 1981, IBM chose intel to provide the microprocessor for their PC, which led to Intel becoming the most widely accepted supplier of microprocessors. The industry followed volume - manufacturers focused on producing on top of Intel architecture, developers focused on developing on the best operating system (Microsoft Windows) and over time Intel and Microsoft encroached on IBM’s turf. Grove’s reasoning for this is simple: “IBM was composed of a group of people who had won time and time again, decade after decade, in the battle among vertical computer players. So when the industry changed, they attempted to use the same type of thinking regarding product development and competitiveness that had worked so well in the past.” Just because the company has been successful before, it doesn’t mean it will be successful again when change occurs.

The six forces acting on a business at any time. When one becomes outsized, it can represent a strategic inflection point to the business.

Business Themes

Strategic Inflection Points and 10x forces. A strategic inflection point is a fundamental shift in a business, due to industry dynamics. Examples of well known shifts include: mainframes to PCs, vertical computer production to horizontal production, on-premise hardware to the cloud, shrink-wrapped software to SaaS, and physical retail to e-commerce. These strategic inflection points are caused by 10x forces, which represent the underlying shift in the technology or demand that has caused the inflection point. Deriving from the Porter five forces model, these forces can affect your current competitors, complementors, customers, suppliers, potential competitors and substitutes. For Intel, the 10x force came from their Japanese competitors which could produce better quality memories at a substantially lower cost. Recognizing these inflection points can be difficult, and takes place over time in stages. Grove describes it best: “First, there is a troubling sense that something is different. Things don’t work the way they used to. Customers’ attitudes toward you are different. The trade shows seem weird. Then there is a growing dissonance between what your company thinks it is doing and what is actually happening inside the bowels of the organization. Such misalignment between corporate statements and operational actions hints at more than the normal chaos that you have learned to live with. Eventually, a new framework, a new set of understandings, a new set of actions emerges…working your way through a strategic inflection point is like venturing into what i call the valley of death.”
The bottoms up, top-down way to “Let chaos reign.” The way to respond to a strategic inflection point comes through experimentation. As Grove says, “Loosen up the level of control that your organization normally is accustomed to. Let people try different techniques, review different products. Only stepping out of the old ruts will bring new insights.” This idea was also recently discussed by Jeff Bezos in his annual shareholder letter - he likened this idea to wandering: “Sometimes (often actually) in business, you do know where you’re going, and when you do, you can be efficient. Put in place a plan and execute. In contrast, wandering in business is not efficient … but it’s also not random. It’s guided – by hunch, gut, intuition, curiosity, and powered by a deep conviction that the prize for customers is big enough that it’s worth being a little messy and tangential to find our way there. Wandering is an essential counter-balance to efficiency. You need to employ both. The outsized discoveries – the “non-linear” ones – are highly likely to require wandering.” When faced with mounting evidence that things are changing, begin the process of strategic wandering. This needs to be coupled with bottom-up actions from middle managers who are exposed to the underlying industry/technology change on a day to day basis. Strategic wandering reinforced with the buy-in and action of middle management can produce major advances as was the case with Amazon Web Services.
Traversing the valley of death. The first task in traversing through a strategic inflection point is to create a clear, explainable, mental image of what the business looks like on the other side. This becomes your new focus and the Company’s mantra. For Intel, in 1986, it was, “Intel, the microcomputer company.” This phrase did two things: it broke the previous synonymy of Intel with ‘memory’ and signaled internally a new focus on microprocessors. Next, the Company should redeploy its best resources to its biggest problems, including the CEO. Grove described this process as, “going back to school.” He met with managers and engineers and grilled them with questions to fully understand the state and potential of the inflection point. Once the new direction is decided, the company should focus all of its efforts in one direction without hedging. While it may feel comfortable to hedge, it signals an unclear direction and can be incredibly expensive.