Amazon is truly a company worth studying, and its legend is not only about Jeff Bezos’ personal achievements but also about its continuous self-disruption and innovative attempts that break boundaries. Ultimately, what’s most important about Amazon is the management philosophy it has built—a narrative and reusable philosophy. In this article, let’s take a closer look at the logic behind Amazon’s management philosophy.
Your AI-powered meeting assistant — Huddles
01-Amazon’s Core Fourteen Management Principles
Amazon’s early management philosophy only existed in Jeff Bezos’ mind and was not formalized. As the company continued to grow and evolve, it gradually developed the following fourteen core principles:
- Customer Obsession: Leaders start with the customer and work backward when making decisions. While the company pays attention to competitors, its focus is always on customers.
- Ownership: Leaders act on behalf of the entire company. They think long term and don’t sacrifice long-term value for short-term results. They act on behalf of the company and never say “that’s not my job.”
- Invent and Simplify: Leaders expect and require innovation and invention from their teams and always find ways to simplify. They are externally aware, look for new ideas from everywhere, and are not limited by “not invented here.”
- Are Right, A Lot: Leaders are right a lot. They have strong judgment and good instincts.
- Learn and Be Curious: Leaders are never done learning and always seek to improve themselves. They are curious about new possibilities and act to explore them.
- Hire and Develop the Best: Leaders raise the performance bar with every hire and promotion. They recognize exceptional talent and willingly move them throughout the organization.
- Insist on the Highest Standards: Leaders drive their teams to deliver quality products, services, and processes. They work to continually improve standards.
- Think Big: Thinking small is a self-fulfilling prophecy. Leaders create and communicate a bold direction that inspires results. They think differently and look around corners for ways to serve customers.
- Bias for Action: Speed matters in business. Many decisions and actions are reversible and do not need extensive study. We value calculated risk-taking.
- Frugality: Accomplish more with less. Constraints breed resourcefulness, self-sufficiency, and invention.
- Earn Trust of Others: Leaders are sincerely open-minded, genuinely listen, and are willing to examine their strongest convictions with humility.
- Dive Deep: Leaders operate at all levels, stay connected to the details, and audit frequently. No task is beneath them.
- Have Backbone; Disagree and Commit: Leaders are obligated to respectfully challenge decisions when they disagree, even when doing so is uncomfortable or exhausting.
- Deliver Results: Leaders focus on the key inputs for their business and deliver them with the right quality and in a timely fashion. Despite setbacks, they rise to the occasion and never settle.
Amazon’s Management Guideline Execution Mechanism: Amazon’s planning for the next year begins in the summer, which is a high-intensity and painful process. S-teams (Senior Leadership Team) initially set high-level goals, based on goals set by the CEO or CFO. S-team’s goals have more details, such as breaking down revenue targets by different regions.
Teams set OP1 and OP2: Once the top-level goals are established, each team starts setting their own OP1, which includes a review of the past year’s performance, goals for the next year, detailed revenue and profit statements, and resource requirements. Each team aligns their OP1 with the company’s finance and HR departments, which can often lead to multiple iterations. Before the fourth-quarter holiday promotion season, goals are modified to reflect the latest business developments, resulting in OP2. Finally, S-teams select the most important goals from each team’s OP objectives to establish S-team Goals, prioritizing the team’s objectives. Amazon’s executives are well-informed about these details (Dive Deep).
Amazon’s Long-Term Incentive Approach: Long-term thinking is one of the foundations of Amazon’s survival, and the company’s incentive system is based on the same principle. The highest base salary for employees is $160,000 (regardless of level), and the rest is distributed in the form of stock. This approach aligns the company’s long-term future with employee earnings and naturally screens out employees lacking a long-term perspective.
Amazon’s Hiring: Bar Raiser System
Traditional hiring limitations: Randomness and bias. Common problems in traditional hiring processes include lack of preparation, with even the smartest recruiters asking random, unfocused questions. Another common issue is bias, as people tend to hire candidates with similar backgrounds, often unconsciously. According to data from Sequoia Capital, Silicon Valley startups spend 990 hours to recruit 12 engineers, which is over 80 hours per hire, representing a significant time cost.
The Birth of Amazon’s Bar Raiser System: When it comes to Amazon’s hiring, the most famous aspect might be the random “brain-teaser” questions like, “How many passengers does LAX airport in Los Angeles transport annually?” These questions test a candidate’s ability to respond immediately but don’t necessarily help assess their fit with the company’s culture and values. As Amazon grew, there was a growing need for a scalable hiring system to ensure the company’s operation. This is where the Bar Raiser system came into play. As the name suggests, a Bar Raiser is a recruiting system and a collective term for a type of recruiter who plays a pivotal role in Amazon’s overall hiring process. Bar Raisers do not receive additional compensation, but they have a special logo in the internal system (an honorary title).
Specifically, Amazon’s hiring process can be divided into several steps:
- Job Description: The core difference here is that Amazon’s JDs are not written by the HR department but by the recruiting department head.
- Resume Review: Screening suitable candidates through a structured process. If a target candidate is screened out, the JD may need to be reviewed for accuracy.
- Phone Screen: The recruiting department head conducts a 45-minute interview with pre-set behavioral questions. The primary goal is to identify candidates who align with Amazon’s core values. The last 15 minutes allow candidates to ask questions. Generally, the pass rate at this stage is around 25%.
- In-House Interview Loop: A 5-7 hour on-site interview where candidates meet with about 7 interviewers. The hiring manager, HR manager, and Bar Raiser must all approve the candidate. It’s essential to avoid having interviewers who are more than one level below the target position, and it’s crucial to avoid a situation where a candidate interviews with their future supervisor.
- Written Feedback: This includes a decision on whether to hire or not and must be completed promptly after the interview, often within 15 minutes, to ensure feedback completeness.
- Debrief Hiring Meeting: This meeting is for deciding whether to hire or not. The hiring manager makes the final call, but the decision must be approved by the Bar Raiser. The latter rarely exercises veto power. In an internal Amazon study of 4,000 hiring processes, there were only three instances where a Bar Raiser disagreed with hiring.
- Reference Check: Department heads call referees to further verify the candidate’s background. However, this step is not used as often nowadays.
- Offer Through Onboarding: Once an offer is approved, the hiring manager for the applicant’s department personally makes the offer call. It’s important to note that hiring is a two-way process, and excellent candidates often have multiple choices.
Amazon’s Organizational Guidelines – Single-Threaded Leader (Specialist Leadership System)
Scale is always the enemy of efficiency. Since 1994, Amazon has demonstrated continuous and astonishing growth, leading to the rapid expansion of its organization. As interdependencies among teams increased, employees spent more and more time on synchronization and collaboration, resulting in decreasing efficiency. The problems Amazon faced seemed to be the same as those encountered by most growing companies. The author, Colin, led a team at Amazon to develop a project called Amazon Associates, which aimed to allow other websites to feature Amazon products and earn a commission through the CPS program. During the development process, they faced challenges related to code synchronization (each change needed to be synchronized with other departments to avoid system crashes) and data coordination (development data required multiple layers of approval). This led to lower development efficiency.
The solution was not more collaboration. Faced with these issues, the natural instinct was to increase the frequency of synchronization and collaboration. However, Amazon was not satisfied with this approach. They believed that the collaboration between teams didn’t need improvement; it needed elimination. Jeff Bezos introduced the concept of APIs to describe the relationships between different teams. He believed that the ideal relationship between teams should involve each team modularizing its functions and collaborating loosely with other teams, rather than tightly interconnecting. This way, each team would have autonomy and innovation efficiency.
The first solution was New Project Initiatives (NPI). In simple terms, Amazon created an internal NPI process for prioritizing internal projects across the company. Each team had to prepare an NPI project proposal, including a project overview, collaborating teams, P&L, and more. The most challenging part was predicting user demand in advance, so the forecasts provided by project teams were often broad, such as funding ranging from $4 million to $20 million. Overall, the NPI project design process was cumbersome, and the evaluations were challenging, making it unpopular within Amazon.
The second attempt was the Two-Pizza Team. Bezos believed that instead of trying to enhance collaboration, they should eliminate it at its roots. He suggested breaking down software engineering teams into smaller, independent teams that didn’t require collaboration with others. Eventually, Amazon developed the theory of the Two-Pizza Team, which had the following key elements:
- Size: Must be smaller than 10 people.
- Autonomy: Has decision-making authority within its scope.
- Evaluation: Fitness function serves as the core method for evaluating team progress.
- Real-time monitoring: Progress is monitored in real-time.
- Owner: The first responsible person within its scope.
- Competent leadership: Demonstrates leadership skills in various aspects, including recruiting.
- Self-sustaining: Capable of taking responsibility for profit and loss.
This approach required prior approval from the S-team.
Challenges faced by the Two-Pizza Team approach. Initially, the Two-Pizza Team approach was implemented in software engineering and product design teams. However, it didn’t permeate all Amazon departments, such as retail, HR, and legal teams, where interdependencies weren’t as strong. Defining a suitable fitness function to evaluate the effectiveness of independent teams proved to be a challenge. Moreover, independent teams required higher leadership capabilities than regular teams, and it became evident that, in some cases, larger teams were needed, highlighting the importance of leadership rather than just the organizational structure.
Single-Threaded Leader (Specialist Leadership). Through a series of practices, Amazon found that adopting a specialist leadership approach was the most effective way to address critical cross-departmental collaboration issues. For instance, the FBA (Fulfillment By Amazon) project was initially designed to provide a service to merchants, allowing them to use Amazon’s logistics infrastructure. Both the retail and logistics teams saw the potential in this idea, but for over a year, no one took ownership of the project, resulting in slow progress. In 2005, Amazon appointed a specialist VP for FBA, and by September 2006, the project was successfully launched, becoming a crucial part of Amazon’s flywheel. In fact, Amazon has used the specialist leadership approach in subsequent digital business initiatives and product development.
Internal Communication at Amazon: Ditching PowerPoint for Embracing Word
If you were to walk into any Amazon meeting room right now, you would be surprised by one thing: the astonishing 20 minutes of silence at the start of the meeting. During these 20 minutes, participants are engrossed in reading relevant materials.
The prehistoric Amazon S-team meetings. Typically, core discussions revolved around several S-goals, and relevant teams would present their findings using PowerPoint (PPT) slides. Often, these presentations did not meet expectations, as the PPT format made it difficult to focus on the real issues. Additionally, as the business became more complex, linear PPT presentations struggled to represent the true facts. PPTs sometimes unintentionally steered meetings toward unimportant details, wasting time and lowering efficiency. Finally, PPT presentations placed a high demand on presentation skills, which was not aligned with the core purpose of the presentations.
These issues prompted Bezos to question whether the presentation format itself was the problem.
A paper led to change. In 2004, during a flight, Bezos and Colin read a paper by a Yale University professor that analyzed the drawbacks of PPT presentations and suggested a shift back to text-based communication, particularly using Word documents. In June 2004, Colin sent an email to Amazon’s entire S-team, announcing that future meetings would use Word documents instead of PPTs. Initially, many members did not understand the change, but gradually, Amazon reached a consensus on the format of presentation materials: a Word document of no more than six pages, without any formatting tricks like shrinking font size. Less important information could be placed in appendices.
Key points of a good Word presentation. One of the benefits of the six-page Word format is that it forces presenters to think deeply, summarize key points, and avoid unnecessary details, which is more challenging with PPTs. Generally, Amazon’s six-page document includes sections like problem statement, solution approach, major achievements, key failures, next-phase focus, resource requirements, P&L, FAQs, and appendices. The length of the material is related to the meeting’s duration. A six-page document suits a one-hour meeting, with the suggestion that one-third of the time is spent reading, and two-thirds are allocated for discussion. In a half-hour meeting, preparing a three-page document is recommended.
Providing effective feedback based on the material is key to meeting efficiency. Jeff Bezos always manages to offer distinctive and effective feedback because he assumes that every word in the document is wrong until proven otherwise.
Amazon’s Product Development: Starting with the Goal, Working Backwards (Working Backwards)
Amazon’s digital strategy. In 2004, Amazon decided to embark on a digital strategy. Bill, the author, transitioned from the retail business to become the VP of digital business. The question was how Amazon could start a completely new business. Initially, Bill and his team, as usual, gathered some MBA students to conduct market space and opportunity research. They presented detailed research findings to Jeff Bezos. After several serious discussions, Bezos asked a crucial question: “Where are your samples?” In reality, Bill’s team didn’t have any samples, and Bezos was primarily concerned about their product and its differentiation from competitors. After several meetings with unsatisfactory results, Jeff Bezos suggested starting with a product concept paper, which became the unique PR/FAQ approach that Amazon uses to initiate new business and product development.
PR/FAQ approach: The PR section typically includes some images and text, kept to less than a page, while the FAQ section can be as long as five pages (total length less than six pages, as per the rule). The initiator is usually the creative mind or project leader, and during a one-hour discussion meeting that follows, everyone shares their thoughts (with the most senior person speaking last to avoid influencing others). The PR section typically includes headings, subheadings, context summary, problem statement, solution description, supporting evidence (e.g., customer quotes), and more. The FAQ section typically addresses market space and customer demand questions, economic questions (P&L), resource coordination requirements (e.g., reliance on carriers), operational feasibility questions (challenging aspects to overcome), and more.
Amazon’s Process Control: Focusing on Input Metrics, Not Output
Amazon’s WBR (Weekly Business Review) system. Amazon always closely tracks its business. A key regular practice is the weekly review, and the first question was what to focus on. Amazon’s answer was to concentrate on input metrics (e.g., supply) rather than output metrics (e.g., stock price).
How are input metrics defined? Amazon focuses on leading indicators rather than lagging indicators. Amazon’s retail business has always focused on three core elements: selection (a wide variety of choices), convenience (faster and friendlier user experience), and price (affordability). Bezos’s flywheel effect has consistently emphasized these three core elements (as shown in the diagram below). The logic sequence of the flywheel effect is: good user experience leads to more buyers, which attracts more sellers, leading to a richer selection, resulting in a better experience. This is the first-level flywheel. The second-level flywheel involves larger revenue scale, optimized cost structure, lower prices, improved user experience, and larger revenue scale again. The flywheel effect focuses on typical input metrics.
Another example of Input Metrics – changing metric definitions over time: Amazon focused on providing users with a wider selection of products. Initially, one of the metrics was defined as increasing the number of product pages from the supply side. However, Amazon later realized that teams were creating many unnecessary product pages to meet this KPI. The metric was then optimized to measure the consumption traffic of relevant pages (pages that nobody viewed were considered useless). It was further refined to measure the percentage of consumption traffic for pages of products that were in stock (pages for out-of-stock products were excluded). Finally, it was optimized to measure the percentage of consumption traffic for pages of products that were in stock and available for delivery within 2 days. This demonstrates that Amazon’s Input Metrics are continuously evolving and adapting.
DMAIC process: Amazon follows the Define-Measure-Analyze-Improve-Control process for monitoring its business. As mentioned earlier, the Define step involves clearly defining objectives. In the Measure step, it’s crucial to collect real-time data on key metrics. The Analyze phase involves input from various teams to analyze the data and understand where the issues lie. Once the problems are identified, the next step is to Improve the situation. Finally, Control is about ensuring ongoing efficiency and maintaining process control.
Key points of WBR (Weekly Business Review): At the beginning of each WBR, Amazon presents a series of core data that are compared in detail with previous data to reflect the operational details and changing trends of the business. To fully understand these data, Amazon executives must have an in-depth knowledge of the business details (Amazon’s Dive Deep culture). Additionally, the core structure of the materials should remain consistent to allow participants to quickly grasp and understand changing trends. The central focus of WBR is on changing trends and analyzing the reasons behind them. Routine metrics are not the primary focus (nor should they be over-interpreted, as explaining random fluctuations can consume unnecessary time).
02-Amazon’s Theoretical Practice Cases
KINDLE: Amazon’s Digital Strategy Vanguard
Influenced by Steve Jobs, Amazon was determined to embrace digitization. In 2003, Colin, Jeff Bezos, and an Amazon SVP were invited to visit Apple’s headquarters and were received by Jobs himself. Jobs showed them the first application that Apple was about to launch on Windows, claiming it to be the best Windows application ever created. It allowed users to connect their iPods to any computer, enabling them to consume digital music anytime, anywhere. Jobs predicted that physical products like CDs and tapes would disappear in the future, and Amazon would be the last respectable place to buy CDs, still enjoying high profit margins as CDs became scarcer. Bezos must have had mixed feelings upon hearing this, as becoming the exclusive seller of CDs in the digital age didn’t seem like a very appealing business proposition. A few months later, Bezos appointed a dedicated leader for Amazon’s digital business, Steve Kessel, who would report directly to Bezos instead of reporting through the Amazon retail division. This highlighted the strategic importance of the digital business. Additionally, the author of this book, Bill, also became a core member of Steve’s team.
After conducting research, Amazon ultimately chose the path of e-books. Following six months of market research, Amazon identified three core digital directions: books, music, and video. Amazon’s first challenge was to make a strategic choice: be a fast follower in the industry or a leader? In Jeff Bezos’s view, imitating the iPod was unlikely to make Amazon a star, and digitalizing movies and TV shows was technically challenging at the time (due to immature infrastructure like broadband). Consequently, Amazon selected e-books as its core direction. According to Bezos, being merely a content aggregator (as shown in the diagram below) didn’t require significant barriers in the digital arena. To establish competitive barriers, Amazon needed to expand into either content or end devices. Thus, Amazon initially chose hardware, embarking on the development of a unique e-book reader, which was the genesis of the Kindle.
Choosing internal development to incubate ideas and capabilities. Bezos’s decision sparked significant internal skepticism because Amazon had never ventured into hardware development before. It was seen as a massive and uncertain investment. Based on the “Bias for action” principle, once the top executive made a decision, subordinates were expected to fully execute it. During the PR/FAQ phase, the digital team determined the goal of seamlessly integrating software ecosystems (like the app store) with hardware. Furthermore, Amazon chose to develop hardware internally rather than outsourcing, primarily because internal development could incubate creativity and development capabilities, which was challenging to achieve through outsourcing. To this end, Amazon recruited Greg Zehr as the head of the hardware development team. Zehr had previous experience at companies like Apple and played a crucial role in establishing Amazon’s hardware lab in Silicon Valley.
Full commitment. Amazon devoted its full resources to the development of the Kindle, hiring top talents for the project. Bezos himself got deeply involved and was even jokingly referred to as the “Chief Product Officer” of the Kindle. In 2005, during an OP1 S-team meeting, it became apparent that the investment in the Kindle project had exceeded initial expectations in terms of both time and money. Someone asked Bezos, “How much money are you willing to invest in the Kindle?” Bezos responded with a question to the CFO, saying, “How much money do we have in the bank?” This meant going all-in.
Two Killer Features of Kindle. One was instant content updates, inspired by the Blackberry. Bezos was a fan of Blackberry’s design, which allowed users to check email anytime, anywhere (email was the equivalent of QQ and WeChat for Westerners for a long time). Bezos believed that requiring users to connect to a computer every time they wanted to update content, as was the case with the iPod, led to a poor user experience. He conducted research that showed the average iPod user updated their device only once a year. Consequently, Bezos insisted that the Kindle must have wireless connectivity (3G was gaining popularity at that time) to allow users to update content anytime, anywhere. This feature, known as “Whispernet,” was implemented, and due to the small size of e-books, it incurred manageable additional costs for Amazon. The other key feature was the use of E Ink displays. Despite E Ink’s limitations, such as displaying only black and white and having a slower refresh rate, it met Amazon’s definition of an excellent e-reader: making the book “disappear” in the hands of the reader (providing a reading experience similar to physical books) and offering long battery life. Kindle’s choice of using an E Ink display ultimately became one of its core advantages, along with a vast selection of content (initially 90,000 e-books) and low pricing (selling books for $9.99 each, almost at zero margin).
Kindle’s Instant Success. Kindle was launched in November 2007 with a retail price of $399, $100 more than Sony’s competing product. However, thanks to its excellent user experience, Kindle quickly gained popularity. In 2008, Oprah Winfrey publicly declared the Kindle as her favorite product on her show, further boosting its sales.
Prime Membership: Amazon’s Long-term Ambition
Starting with an Urgent Email. In mid-October 2004, several senior Amazon leaders received an email from Bezos. The email emphasized that Amazon should not settle for the status quo and should continue improving the user experience, especially in the delivery aspect. Bezos urged the launch of a free shipping membership program by the end of the year. Like a cry in the calm night sky, Bezos’s voice became a “royal decree” that had to be executed within the company, even though it seemed like an impossible task (as the fourth quarter is the busiest season for promotions).
The Birth of Amazon Super Saver Shipping. Bezos’s email was not without foundation. In fact, the idea of offering free shipping had been brewing within Amazon for a long time and had already undergone pilot testing at scale. In early 2002, Amazon introduced a service offering free shipping for orders over $99 (later reduced to $25). This service primarily targeted customers who were not time-sensitive about delivery, as it took 3-5 days to deliver the items. As expected, offering free shipping led customers to purchase more items, increasing the average order value. However, the Super Saver Shipping program had not been widely adopted on the platform. Many customers were unwilling to wait for extended delivery times, and price-sensitive customers were not willing to buy more items just to qualify for free shipping. Bezos believed that Amazon should offer customers the choice of “Fast and expensive” and “Free and slow,” with the latter being the preferred option for users.
Prime as the Solution to Growth Pressure. In the third quarter of 2004, Amazon’s revenue grew by 29% year-on-year, and free cash flow increased by 76%. While it seemed impressive on the surface, compared to the same period in 2003, the growth rates in various aspects had significantly slowed down (as shown in the graph below). Amazon was equally concerned about growth, and the company tried various methods to reinvigorate it, such as nationwide marketing. However, these efforts did not yield the expected results (they couldn’t justify the advertising costs, and Bezos disliked advertising). The company chose to return to the Input Metrics, focusing on the core input factors of Price, Selection, and Convenience. During discussions, the company concluded that Amazon excelled in providing diverse choices and lowering prices but still had room for improvement in providing convenience to users. The most crucial aspect of convenience was reducing users’ shipping costs. Research also showed that offering free shipping provided more significant sales incentives than equivalent price discounts.
From a Slow Process to Bezos’s Decisiveness. In reality, Amazon had been discussing the need for a loyal membership program internally since 2000. The necessary conditions for such a program included: 1) affordability for users, 2) promotion of user-Amazon relationships leading to more transactions, and 3) higher capital efficiency, fostering Amazon’s flywheel effect. Regarding the specific implementation plan for the membership, there were many discussions within the company. One software engineer, Charlie, proposed the idea: “Why not have consumers pay an annual fee and provide them with free shipping?” Charlie eventually became responsible for the technology, user experience, and finances of Prime membership. The biggest uncertainty was whether the membership fee would be sufficient to cover the additional shipping costs. If not, the service would not be sustainable and would be meaningless to users. “Institutional No” is Amazon’s term for the tendency of large companies to opt for conservative approaches to avoid risks. Initially, the Prime membership project progressed slowly within Amazon, and a significant reason was the difficulty and operational risk associated with it (Bezos had proposed offering free shipping for high-value items, but it was technically challenging).
Ultimately, Bezos’s commitment to Prime didn’t waver despite its low system efficiency. All of this led to the email sent in mid-October 2004. The Prime membership program was finally launched in February 2005 (Amazon even postponed its earnings conference call for this). Initially, some of Amazon’s heavy users (those who purchased frequently and incurred high shipping costs) were the first to join Prime, even though Prime membership increased the platform’s costs. However, through long-term persistence, Prime membership became a crucial part of Amazon’s flywheel. In 2018, Amazon announced that Prime membership had reached 100 million, and in 2021, it announced that it had surpassed 200 million members.
Prime Video: Amazon’s Persistence Pays Off
Early Failed Attempt: Amazon Unbox. Amazon’s foray into video business dates back to the mid-2000s. When Amazon announced its digital strategy, video was one of the three main directions, corresponding to Amazon’s physical DVD business. In the early 2000s, Netflix’s DVD rental business was thriving, while Amazon believed that video downloads were the future of the industry. Amazon developed Amazon Unbox. However, video downloads faced several limitations at that time, such as taking 1-2 hours to download a movie (even with broadband), and the difficulty of connecting computers and TVs, meaning movies downloaded on a computer couldn’t be watched on a TV. Additionally, Amazon wanted to offer some unique features, like the ability to share content libraries across multiple devices (allowing a single movie to be downloaded multiple times). However, Amazon Unbox ultimately failed due to slow download speeds and numerous bugs in Microsoft’s DRM encryption software that prevented many users from downloading.
Bill Gave Himself a “D,” but Amazon Gave Him Another Chance. In his self-evaluation that year, the author of this book, Bill, gave himself a “D” and believed that his performance in 2006 was terrible, primarily because Amazon Unbox failed for various reasons. If it were at another company, Bill would probably have been fired. However, perhaps because video was not a significant focus of Amazon’s digital strategy (as explained in the previous Kindle section) and, more importantly, due to Amazon’s long-standing belief in the necessity of accepting failure to innovate, Bezos saw it differently. Bezos thought, “I’ve invested millions of dollars in you, and you need to make those investments pay off next time.”
From OTT to Living Room Market with Initial Success. Amazon’s video business initially faced many challenges, including the difficulty of obtaining content due to content windows (after movies left theaters, they went to different distribution channels like DVDs and premium cable, making it hard for Amazon to secure content rights). During a 2006 business review meeting, an employee named Josh, who had expertise in content, business, technology, and more, suggested a partnership with OTT service provider TiVo. He recommended that Amazon enter the big-screen market through the OTT platform (rather than the small-screen PC market). This way, Amazon’s content could be pre-downloaded onto TiVo’s set-top boxes, providing a better experience for users. In 2007, Amazon released Unbox on the TiVo platform, offering a feature that allowed users to download while watching, and achieved initial success.
Amazon’s Video Business Struggles Amid Pressure from Competitors. In 2007, Netflix rose to prominence in the streaming video space by combining subscription and streaming models. However, like other players in the industry, Amazon initially didn’t pay full attention to this competitor until later when Netflix established an unbeatable advantage (highlighting that even Amazon, like most large companies, had typical problems). In 2008, Amazon relaunched Amazon Video On Demand, an integrated video app also introduced on television platforms. However, Amazon’s competitive relationships prevented its video application from being introduced on Microsoft, Sony, and other competitor’s hardware (such as gaming consoles). Netflix, on the other hand, faced no such challenges (at the time, 95% of Netflix users either signed up through the official website or via the three major gaming console channels: Xbox, PlayStation, and Wii). Hulu, with its ownership by NBC and Fox, had vast content rights and outperformed Amazon’s video service.
Bezos’s Decision: Making Video Business a Bonus for Prime Members. Amazon’s video business had a challenging journey, and internal teams explored various paths to highlight their uniqueness, such as positioning specific content categories. In a meeting, Bezos suggested, “Let’s consider video service as an extension of Prime membership, just as Netflix’s streaming service started as a bonus for DVD rentals.” Bezos’s proposal solved Amazon’s video business cold start problem – the paradox of having enough content and enough users. Bezos believed that having video as part of Prime would make Prime more competitive, and neither e-commerce nor streaming competitors could offer exactly the same comprehensive service to users. In 2011, Prime Video was officially launched as part of the Prime membership (before this, free shipping was the core value of Prime membership).
Based on a similar strategic thinking as Kindle, in the digital content field, being just a content aggregator platform lacked a competitive advantage. Amazon had to either enter content
production or hardware manufacturing, and Amazon chose to do both. First, Amazon developed its own hardware, starting with the Kindle Fire Tablet and later the Fire TV, among others. Additionally, Amazon formally established its own studio to develop original content. In Bill’s view, Hollywood could provide a large supply of creative talent (supply-side prosperity), making it the greatest guarantee for Amazon to enter the original content market. The biggest challenge was how to identify or select the best scripts or ideas. Furthermore, Amazon experimented with innovations like releasing pilot episodes ahead of time on the platform and making decisions about whether to recommend projects based on viewing data.
From a trend perspective, Amazon’s streaming business still holds great promise. On one hand, it has invested heavily in creating the most expensive series in history, “The Lord of the Rings” (reportedly with a budget of $500 million for a single season). On the other hand, Amazon recently solidified its copyright holdings with the $8.45 billion acquisition of MGM.
AWS: From Humble Beginnings to a Thriving Ecosystem
The Origin: Developer Tools for Amazon’s CPS Business. In early 2000, Amazon still generated 75% of its revenue from book, CD, and DVD sales. The author of this book, Colin, was primarily responsible for the Amazon Associate program (Amazon’s CPS program, where other websites referred traffic to Amazon in exchange for commissions). As the business grew, website developers began to demand more aesthetically pleasing and well-designed Amazon plugins. Colin led the development of foundational SDK tools that allowed website developers to design their own plugins’ appearances. This tool unexpectedly gained popularity among developers, who came up with many creative ideas that even Amazon hadn’t considered. After learning about these developments and seeing impressive data, Bezos realized that it was a significant opportunity, and the company should double down on this area. In July 2002, Amazon launched the initial version of AWS, primarily offering API services to website developers working with Amazon’s CPS program. This API service was surprisingly welcomed by Amazon’s internal engineers (it was considered more user-friendly than some internal development tools). Eventually, this API service allowed Amazon to provide early website development services to other brick-and-mortar retailers like Target. So, early AWS had little to do with what we now commonly associate with AWS.
AWS’s Transformation into a General Toolset. In the summer of 2003, Colin transitioned to becoming Bezos’s personal assistant, and the former assistant, Andy Jassy, became the head of AWS (he also became Bezos’s successor). Because AWS was so well-received by developers, it gradually expanded beyond its initial focus on Amazon’s retail system APIs to become a general-purpose tool for developers. In reality, Amazon also recognized that some “undifferentiated heavy lifting” was becoming a common burden for startups. For example, the CEO of Flickr once said that half of his time was spent dealing with technical infrastructure issues. On the other hand, Amazon had accumulated the capability to store and process massive amounts of data from years of work in the retail industry, which provided a foundation for Amazon to become a cloud service provider. Amazon also realized that these advantages were not unique to them, so the “Bias for action” principle was applied. Amazon had to seize the first-mover advantage. In fact, by the time other competitors recognized the opportunity in cloud computing, Amazon AWS had already established a significant lead.
Among the many new features, Amazon S3 (Simple Storage Service) and Amazon EC2 (Elastic Compute Cloud) became core offerings. Jeff Bezos met with the AWS team at least every two weeks to drive the development of new business. In the pricing process of AWS, Amazon mainly used a “Cost Following” method, where pricing was mainly based on costs (with storage and bandwidth costs as key factors). These decisions were part of Amazon’s PR/FAQ development process.
Of course, AWS has grown into a massive business with annual revenues of $500 billion, and there’s much more to the story than covered here. Perhaps because the author of this book was only involved in the early stages of the business, there’s more to discover about the rise of AWS, which awaits further research.
In Conclusion: Taoism teaches that “opposites are the way of the Tao,” meaning that things will inevitably move to their opposites. The implication is that a company’s key to long-lasting success is to embrace its own opposite. Amazon’s innovative approach exemplifies this. When its primary revenue source was physical books, DVDs, and CDs, it ventured into digital business, showcasing its ability to adapt and evolve. Amazon is undoubtedly a company worth studying repeatedly, whether from an industry perspective or a management one.