Venture Capital & Investing

Bezos’s 2015 Shareholder Letter: AWS Hit $10 Billion Faster Than Amazon Did

Share this article
LINEFacebook

Bezos shareholder letter (Chinese translation by Vinchent; official English original) + Robin’s introduction

Storage information

Bookmarks and reading positions are stored in this browser. Manage and back them up in My Reading.

Collected in Robin Shen’s Field Notes · Content type and attribution are shown above

English version published 2026-11-08 with AI assistance. Attribution, source notes, and verification limits are retained in the text. 中文原文

Reading guide

Part 2 of the 2015 letter: AWS is reaching $10 billion in annual sales at age ten—faster than Amazon.com did. The second half is about the discipline of an “invention machine”: make bold bets, separate one-way from two-way door decisions, and own the failures.

Key points

  • AWS: 70+ services, 722 new features (+40%)
  • Aurora: 5x the performance at 1/10 the cost
  • Employees: 30,000 to 230,000
  • Part 1 is on this site: “One Culture, Two Businesses”
Venture Capital & Investing thematic illustration
Venture Capital & Investing · Thematic illustration · AI generated

Full text — English translation

(Continued from Part 1, “Bezos’s 2015 Shareholder Letter: One Culture, Two Businesses”: /articles/post-700874454616724/) In his 2015 letter, Bezos drew the line between reversible and irreversible decisions. If a decision is reversible, you can try it and try again, so your team should move fast and experiment. The two-door framing has been widely discussed ever since. ---- /Amazon Web Services/ Just over 10 years ago, AWS started in the U.S. with its first major service, a simple storage service. Today, AWS offers more than 70 services for compute, storage, databases, analytics, mobile, Internet of Things, and enterprise applications. We also offer 33 Availability Zones across 12 geographic regions worldwide, with another five regions and 11 Availability Zones in Canada, China, India, the U.S., and the U.K. to be available in the coming year. AWS started with developers and startups, and now is used by more than a million customers from organizations of every size across nearly every industry – companies like Pinterest, Airbnb, GE, Enel, Capital One, Intuit, Johnson & Johnson, Philips, Hess, Adobe, McDonald’s, and Time Inc. AWS is bigger than Amazon.com was at 10 years old, growing at a faster rate, and – most noteworthy in my view – the pace of innovation continues to accelerate – we announced 722 significant new features and services in 2015, a 40% increase over 2014. Many characterized AWS as a bold – and unusual – bet when we started. “What does this have to do with selling books?” We could have stuck to the knitting. I’m glad we didn’t. Or did we? Maybe the knitting has as much to do with our approach as the arena. AWS is customer obsessed, inventive and experimental, long-term oriented, and cares deeply about operational excellence. Given 10 years and many iterations, that approach has allowed AWS to rapidly expand into the world’s most comprehensive, widely adopted cloud service. As with our retail business, AWS is made up of many small teams with single-threaded owners, enabling rapid innovation. The team rolls out new functionality almost daily across 70 services, and that new functionality just “shows up” for customers – there’s no upgrading. Many companies describe themselves as customer-focused, but few walk the walk. Most big technology companies are competitor focused. They see what others are doing, and then work to fast follow. In contrast, 90 to 95% of what we build in AWS is driven by what customers tell us they want. A good example is our new database engine, Amazon Aurora. Customers have been frustrated by the proprietary nature, high cost, and licensing terms of traditional, commercial-grade database providers. And while many companies have started moving toward more open engines like MySQL and Postgres, they often struggle to get the performance they need. Customers asked us if we could eliminate that inconvenient trade-off, and that’s why we built Aurora. It has commercial-grade durability and availability, is fully compatible with MySQL, has up to 5 times better performance than the typical MySQL implementation, but is 1/10th the price of the traditional, commercial-grade database engines. This has struck a resonant chord with customers, and Aurora is the fastest-growing service in the history of AWS. Nearly this same story could be told about Redshift, our managed data warehouse service, which is the second fastest growing service in AWS history – both small and large companies are moving their data warehouses to Redshift. Our approach to pricing is also driven by our customer-centric culture – we’ve dropped prices 51 times, in many cases before there was any competitive pressure to do so. In addition to price reductions, we’ve also continued to launch new lower cost services like Aurora, Redshift, QuickSight (our new Business Intelligence service), EC2 Container Service (our new compute container service), and Lambda (our pioneering server-less computing capability), while extending our services to offer a range of highly cost-effective options for running just about every type of application or IT use case imaginable. We even roll out and continuously improve services like Trusted Advisor, which alerts customers when they can save money – resulting in hundreds of millions of dollars in savings for our customers. I’m pretty sure we’re the only IT vendor telling customers how to stop spending money with us. Whether you are a startup founded yesterday or a business that has been around for 140 years, the cloud is providing all of us with unbelievable opportunities to reinvent our businesses, add new customer experiences, redeploy capital to fuel growth, increase security, and do all of this so much faster than before. MLB Advanced Media is an example of an AWS customer that is constantly reinventing the customer experience. MLB’s Statcast tracking technology is a new feature for baseball fans that measures the position of each player, the baserunners, and the ball as they move during every play on the field, giving viewers on any screen access to empirical data that answers age-old questions like “what could have happened if…” while also bringing new questions to life. Turning baseball into rocket science, Statcast uses a missile radar system to measure every pitched ball’s movements more than 2,000 times per second, streams and collects data in real-time through Amazon Kinesis (our service for processing real-time streaming data), stores the data on Amazon S3, and then performs analytics in Amazon EC2. The suite of services will generate nearly 7 TB of raw statistical data per game and up to 17 PB per season, shedding quantitative light on age-old, but never verified, baseball pearls of wisdom like “never slide into first.” About seven years ago, Netflix announced that they were going to move all their applications to the cloud. Netflix chose AWS because it provided them with the greatest scale and the broadest set of services and features. Netflix recently completed their cloud migration, and stories like theirs are becoming increasingly common as companies like Infor, Intuit, and Time Inc., have made plans to move all of their applications to AWS. AWS is already good enough today to attract more than 1 million customers, and the service is only going to get better from here. As the team continues their rapid pace of innovation, we’ll offer more and more capabilities to let builders build unfettered, it will get easier and easier to collect, store and analyze data, we’ll continue to add more geographic locations, and we’ll continue to see growth in mobile and “connected” device applications. Over time, it’s likely that most companies will choose not to run their own data centers, opting for the cloud instead. /Invention Machine/ We want to be a large company that’s also an invention machine. We want to combine the extraordinary customer-serving capabilities that are enabled by size with the speed of movement, nimbleness, and riskacceptance mentality normally associated with entrepreneurial start-ups. Can we do it? I’m optimistic. We have a good start on it, and I think our culture puts us in a position to achieve the goal. But I don’t think it’ll be easy. There are some subtle traps that even high-performing large organizations can fall into as a matter of course, and we’ll have to learn as an institution how to guard against them. One common pitfall for large organizations – one that hurts speed and inventiveness – is “one-size-fits-all” decision making. Some decisions are consequential and irreversible or nearly irreversible – one-way doors – and these decisions must be made methodically, carefully, slowly, with great deliberation and consultation. If you walk through and don’t like what you see on the other side, you can’t get back to where you were before. We can call these Type 1 decisions. But most decisions aren’t like that – they are changeable, reversible – they’re two-way doors. If you’ve made a suboptimal Type 2 decision, you don’t have to live with the consequences for that long. You can reopen the door and go back through. Type 2 decisions can and should be made quickly by high judgment individuals or small groups. As organizations get larger, there seems to be a tendency to use the heavy-weight Type 1 decision-making process on most decisions, including many Type 2 decisions. The end result of this is slowness, unthoughtful risk aversion, failure to experiment sufficiently, and consequently diminished invention.1 We’ll have to figure out how to fight that tendency. And one-size-fits-all thinking will turn out to be only one of the pitfalls. We’ll work hard to avoid it… and any other large organization maladies we can identify. /Sustainability and Social Invention/ Our growth has happened fast. Twenty years ago, I was driving boxes to the post office in my Chevy Blazer and dreaming of a forklift. In absolute numbers (as opposed to percentages), the past few years have been especially significant. We’ve grown from 30,000 employees in 2010 to more than 230,000 now. We’re a bit like parents who look around one day and realize their kids are grown – you blink and it happens. One thing that’s exciting about our current scale is that we can put our inventive culture to work on moving the needle on sustainability and social issues. Two years ago we set a long-term goal to use 100% renewable energy across our global AWS infrastructure. We’ve since announced four significant wind and solar farms that will deliver 1.6 million megawatt hours per The opposite situation is less interesting and there is undoubtedly some survivorship bias. Any companies that habitually use the light-weight Type 2 decision-making process to make Type 1 decisions go extinct before they get large. year of additional renewable energy into the electric grids that supply AWS data centers. Amazon Wind Farm Fowler Ridge has already come online. We reached 25% sustainable energy use across AWS last year, are on track to reach 40% this year, and are working on goals that will cover all of Amazon’s facilities around the world, including our fulfillment centers. We’ll keep expanding our efforts in areas like packaging, where our culture of invention led to a big winner – the Frustration-Free Packaging program. Seven years ago we introduced the initiative with 19 products. Today, there are more than 400,000 globally. In 2015, the program eliminated tens of millions of pounds of excess packaging material. Frustration-Free Packaging is a customer delighter because the packages are easier to open. It’s good for the planet because it creates less waste. And it’s good for shareholders because, with tighter packaging, we ship less “air” and save on transportation costs. We also continue to pioneer new programs for employees – like Career Choice, Leave Share, and Ramp Back. Career Choice pre-pays 95% of tuition for courses that teach in-demand skills, regardless of whether those skills are relevant to a career at Amazon. We’ll pay for nursing certifications, airplane mechanic courses, and many others. We’re building classrooms with glass walls right in our fulfillment centers as a way to encourage employees to participate in the program and to make it easy. We see the impact through stories like Sharie Warmack – a single mother of eight who worked in one of our Phoenix fulfillment centers. Career Choice paid for Sharie to get licensed to drive an 18-wheeler. Sharie worked hard, passed her tests, and she’s now a long-haul driver for Schneider Trucking – and loving it. This coming year, we’re launching a program to teach other interested companies the benefits of Career Choice and how to implement it. Leave Share and Ramp Back are programs that give new parents flexibility with their growing families. Leave Share lets employees share their Amazon paid leave with their spouse or domestic partner if their spouse’s employer doesn’t offer paid leave. Ramp Back gives birth mothers additional control over the pace at which they return to work. Just as with our health care plan, these benefits are egalitarian – they’re the same for our fulfillment center and customer service employees as they are for our most senior executives. Renewable energy, Frustration-Free Packaging, Career Choice, Leave Share, and Ramp Back are examples of a culture that embraces invention and long-term thinking. It’s very energizing to think that our scale provides opportunities to create impact in these areas. I can tell you it’s a great joy for me to get to work every day with a team of such smart, imaginative, and passionate people. On behalf of all of us at Amazon, thank you for your support as shareholders. As always, I attach a copy of our original 1997 letter. Our approach remains the same, and it’s still Day 1. Jeffrey P. Bezos ——Fact-check note—— Source: Amazon’s 2015 letter to shareowners (official public document). The original post ran a Chinese translation by Vinchent, a user of the Dedao app (as credited in the post). The English text above is Amazon’s official original, quoted directly rather than translated back from Chinese. Check: Compared against the official original (Amazon’s collected shareholder letters, cross-checked with SEC/official PDFs). Key figures and facts were spot-checked and match. Correction to the Chinese translation (1 item): “1.6 million megawatts per hour” should be 1.6 million megawatt hours per year.

These articles are personal notes, book excerpts, and collected reprints. Figures and events reflect the time of the original publication, are not current information, and do not constitute investment advice.

Share this article
LINEFacebook

Comments

What do you think? Share your experience or questions.

Loading comments…

My reading notes

These notes stay in this browser and are not published. Writing a note also saves the article. Back up or move your notes in My Reading.

Manage my bookmarks and notes →

STAY CURIOUS

Save a good question for your next visit.

Save this article, or follow the Facebook page and RSS feed for additions. Historical articles retain their original publication dates.

Browse All articlesSearch articles and full text →