Working Backwards: Insights, Stories, and Secrets from Inside Amazon

by Colin Bryar

“Long-term orientation interacts well with customer obsession. If we can identify a customer need and if we can further develop conviction that that need is meaningful and durable, our approach permits us to work patiently for multiple years to deliver a solution.”2 Key word: patiently. Many companies will give up on an initiative if it does not produce the kind of returns they are looking for within a handful of years. Amazon will stick with it for five, six, seven years—all the while keeping the investment manageable, constantly learning and improving—until it gains momentum and acceptance.

When we have invented, our long-term, patient approach—driven by customer need—has been fundamentally different from the more conventional “skills-forward” approach to invention, in which a company looks for new business opportunities that neatly fit with its existing skills and competencies. While this approach can be rewarding, there is a fundamental problem with it: the company will never be driven to master new skills and develop new competencies, hire new kinds of leaders, or create different types of organizations.

The point of the story, however, is that the process improves your odds of success but by no means guarantees it.

The magnitude of your inventions, and therefore your mistakes, needs to grow in lockstep with the growth of your organization. If it doesn’t, your inventions will likely not be big enough to move the needle.

In other words, his first action was not a “what” decision, it was a “who” and “how” decision. This is an incredibly important difference. Jeff did not jump straight to focusing on what product to build, which seems like the straightest line from A to B. Instead, the choices he made suggest he believed that the scale of the opportunity was large and that the scope of the work required to achieve success was equally large and complex. He focused first on how to organize the team and who was the right leader to achieve the right result.

At most companies, reducing a leader’s scope would be considered a demotion, and in fact there were many VPs and directors who saw each of these changes in that way. At Amazon, it was not a demotion. It was a signal that we were thinking big and investing in digital for the long term.

a typical company that wanted to grow would take stock of its existing capabilities and ask, “What can we do next with our skill set?” He emphasized that Amazon’s approach was always to start from the customer and work backwards. We would figure out what the customers’ needs were and then ask ourselves, “Do we have the skills necessary to build something that meets those needs? If not, how can we build or acquire them?” Once we determined what was necessary to create value for our customers and to differentiate ourselves from our competitors, we didn’t let our lack of ability deter us from achieving this important end result—our own device.

Furthermore, outsourcing in this context offers a classic example of short-term decisions with devastating long-term implications. Practically every day, Amazon could tweak its offering to make things a little better. And so practically every day, the distance between itself and its competitors widened. Outsourcing turned out to be the more expensive path.

“How much more money are you willing to invest in Kindle?” Jeff calmly turned to our CFO, Tom Szkutak, smiled, shrugged his shoulders, and asked the rhetorical question, “How much money do we have?” That was his way of signaling the strategic importance of Kindle and assuring the team that he was not putting the company at risk with the size of the investment. In Jeff’s view, it was way too early to give up on the project. Development continued.

Both of these features were ways for the Kindle to “get out of the way” so customers would forget they were reading on a machine.

We didn’t know if or when the cost per e-book would come down and make this a more profitable and sustainable business. We didn’t look at it through the short-term lens that the publishers did. We focused on what would make sense to customers and what it would take to get customers excited to buy a Kindle and load it up with their favorite books. We took a leap of faith, hoping that, over time, we would be able to reduce the cost of the device and the books themselves.

On October 24, 2008, she devoted an entire episode of her show to Kindle, gushing, “It’s absolutely my new favorite favorite thing in the world.”5 Because millions of viewers looked to Oprah, the “Queen of Reading,” for book recommendations, sales exploded.

Option one would be the skills-forward path—that is, using the existing skills and assets of the company to drive business opportunities. Leaders at most companies would likely be praised for choosing this path. The danger is that while they stand atop this local optimum, someone else will figure out how to scale a higher peak they couldn’t see at the time due to risk aversion.

those moments when you’re thinking, “But we don’t have a Jeff.” The good news is that you don’t need a Jeff to make this type of decision. You only need to ruthlessly stick to the simple-to-understand (but sometimes hard-to-follow) principles and process that insist on customer obsession, encourage thinking long term, value innovation, and stay connected to the details. None of us, including Jeff, knew exactly what we would end up building; it’s more like we stuck with the process and surrendered to where it was taking us.

That trend of decelerating growth was not good news for a relatively new company seeking to participate in a market so incredibly large that it could be considered virtually unlimited.

One action we did briefly consider was creating a national advertising campaign to build awareness for the Amazon brand. In 2002, we ran a long-term advertising test in Portland and Minneapolis. The campaign drove a bump-up in sales, but we ultimately decided not to fully move forward with it. The modest sales uplift was nowhere near enough to justify the $50 million per year we estimated we’d have to spend on an effective national marketing campaign. The better investment was to plow that money back into improving the customer experience.

I have to admit that in the late 1990s, a few years before the Prime discussions started, we had a few scenes that looked a bit like that as we grappled with our growth concerns. We tried a number of initiatives, including promotions (buy five books, get one free!) and online nudges for customers to buy across categories. Eventually we realized that such actions wouldn’t work, because they took precious resources away from improving the long-term customer experience.

In the end, as always, we looked to our Leadership Principles, two of which were particularly relevant at this time: Customer Obsession. Leaders start with the customer and work backwards. They work vigorously to earn and keep customer trust. Although leaders pay attention to competitors, they obsess over customers. 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.

Price. Is the price low enough? Selection. Does Amazon have a wide range of products—ideally everything? Convenience. Is the product in stock, and can I get it quickly? Can I easily find or discover the product?

These timelines were part of the DNA of a company whose very first employee’s job description, as you may recall, made clear that the candidate would have to accomplish large and complex tasks in “one-third the time that most competent people think possible.”

Our marketing, retail, and finance teams set three criteria that any new marketing initiative would have to meet to go forward: It had to be affordable (an eye-catching but financially unsustainable approach was out of the question). It had to drive the right customer behavior (that is, nudge customers to buy more from Amazon). It had to be a better use of funds than the obvious alternative, which was to invest those same funds into actions that would improve the customer experience, such as lowering prices even further or improving our in-stock rate.

problem, another software team reported on its work building a subscription platform for an Amazon DVD rental business that would soon launch. Charlie was intrigued. He asked, “Why couldn’t we have customers pay an annual subscription fee that would include free shipping for a year? That would be a big win for customers. And we could stop spending so much effort reconciling fees.” Kim Rachmeler, who ran the customer service department at the time, liked the idea. “You may have something there, Charlie,” she said. “Why don’t you run with it.”3

In other words, Charlie was, in Amazonian terms, a “strong general athlete” (SGA). These customer-obsessed, inventive, long-term thinkers take pride in operational excellence and embody the Amazon Leadership Principles. Amazon often puts SGAs like Charlie into leadership positions and gives them the tools to become subject matter experts. Kim Rachmeler too was an SGA. She held many leadership roles. In addition to running customer service, she led Amazon’s Supply Chain Systems and the Personalization department. She was also a member of the S-Team.

Even if our assumptions were correct, the program would take years to pay off; not a single person in the leadership team other than Jeff was pushing for Prime to launch in 2004; it was the holiday season, which meant we were already really busy! This very nearly became one of those scenarios when a company takes the seductive but ultimately wrong path of staying the course, making a serious error of omission as a result. The “institutional no” is a big reason why Amazon could have made an error of omission in this case. Jeff and other Amazon leaders often talk about the “institutional no” and its counterpart, the “institutional yes.” The institutional no refers to the tendency for well-meaning people within large organizations to say no to new ideas. The errors caused by the institutional no are typically errors of omission, that is, something a company doesn’t do versus something it does. Staying the current course offers managers comfort and certainty—even if the price of that short-term certainty is instability and value destruction later on.

Jeff would likely have been making just such an error of omission if his October 2004 email had read, “Let’s wait to introduce free shipping and just focus on making this 2004 holiday season our best ever!” If he had stopped pressing the teams for more free shipping ideas, there would no doubt have been a sigh of relief. We would have looked at each other and said thank goodness we made the right call to pause. Instead of marking a turning point in Amazon’s history, that mid-October day might have been remarkable in another way. It might have been the moment we made a disastrous mistake, even if we didn’t realize it until years later.

One spring 2004 walking-the-store email that Jeff wrote ended up making a direct contribution to the Prime conversation, though we didn’t know it at the time. It addressed an issue that was seemingly nontechnical in nature: making too much profit on an item. Jeff browsed our electronics and jewelry stores. The prices for flat-screen TVs and precious jewelry ran into the hundreds and sometimes thousands of dollars. We had little pricing flexibility on many of these items due to supplier relationships. Since we couldn’t offer lower prices, Jeff felt we should do the next best thing: offer free next-day shipping. We made quite a bit of gross profit from those items, compared to the gross profit from a $15 book or video, so we could afford to offer free expedited shipping for the same all-in price.

year. I think at some point he had decided that it wasn’t the idea that was flawed, but the decision-making process, a process encumbered by institutional risk-aversion. The “October surprise” email arose out of his realization that you simply could not prove a priori that free shipping would work. You just had to try it.

Cut to February 22, 2011: Amazon Prime users who visited the site discovered a new benefit of membership. We (Bill and team) had launched Prime Instant Video that morning.1 It offered streaming of five thousand movies and TV shows as part of the Amazon Prime membership, at no extra charge. Until that moment, the Amazon Prime brand had meant one thing to subscribers: fast, free shipping. Millions of customers had subscribed. Tens of millions of customers knew what Prime stood for. But now Amazon Prime meant streaming video too?

We knew from our research that this drove customers crazy, but the solution was not a technical one, because the problem was about rights. The studios earned 70 percent of the revenue on each purchase and considered every download to be a new sale. They did not think of themselves as being in the customer experience business—they simply wanted to extract as many royalty payments as they could from their distributors (e.g., Apple and Amazon) and their distributors’ customers.

In retrospect, it is easy to see the mistakes. We had rushed Unbox out the door before it was ready. In the weeks leading up to launch, rumors had been swirling around Hollywood and in the press that Apple was close to launching a digital video service. We didn’t want to come in second to Apple so we were in a frenzy to ship Unbox and ship it fast. This was directly antithetical to the notion of focusing on the customer, not the competitor. We had conducted an internal employee-only beta test, but we failed to use the results as an opportunity to slow down, carefully review the customer feedback, and take the time needed to make real changes to improve the quality of the customer experience. We were just focused on shipping. We had prioritized speed, press coverage, and competitor obsession over the customer experience. We had been decidedly un-Amazonian.

“Not enough,” he says. “Not enough. These rivets have to be completely flush. I want no air resistance on the fuselage. She’s got to be cleaner. Cleaner! You understand?” The team leader nods. Back to the drawing board. Jeff had told Steve that it was his job to be like Howard Hughes. From then on, Steve had to run his fingers over each new Amazon product, checking for anything that might reduce the quality, insisting that his team maintain the highest standards.

But I would quickly learn that this was not within our control. When I met with studio executives, I would explain that the TVOD business was going to take off and would one day be much more valuable to them than their pay-TV deals. The executives nodded, said they understood, agreed that this needed to change, but told me their bosses didn’t see it that way. As with all media companies, the decisions made by top Hollywood brass were then, and still are, all about achieving short-term financial goals.

Fast-forward ten years, and thanks to their short-term thinking, the studios were scrambling to launch their own streaming video services like Disney+, Warner’s HBO Max, and NBC’s Peacock in a bid to fight for their survival and compete with Amazon and Netflix.

the maximum base salary at Seattle headquarters was $160,000 per year, and there was no bonus system at all. Additional compensation was in Amazon stock. If you got a raise, it was completely in stock, which wouldn’t begin vesting for 18 to 24 months.

My incentives were very different from those of my counterparts at the movie studios and record companies. To benefit financially, I needed Amazon to grow over the long term. I can’t say that everyone at Amazon was always pleased with this compensation philosophy. We all have a need to be rewarded for an important accomplishment, and we want to receive our reward in a timely fashion. But for those who thought and acted long term, and hung around, it paid off.

But one implication of being at the earth’s most customer-centric company is that you don’t spend money on things that don’t benefit customers.

The two big, revolutionary breakthrough features of the Netflix service were subscription and streaming. Amazon and Apple were the leaders in premium movie and TV distribution, but we offered downloads only (and you had to purchase or rent each movie or show). We thought of streaming as a low-quality phenomenon—the domain of YouTube, with its videos of dancing cats that you watched on your PC for a couple of minutes between meetings. So when Netflix launched Watch Now, we took note and discussed the service in detail, but the prevailing wisdom inside our team and among others in the industry was that it was a half-baked test, not a serious offering.

In retrospect, it seems obvious that the Netflix launch was a significant threat, because streaming plus subscription would prove to be the magic combination in the digital video business. And they were smart and savvy about how they launched it by including the streaming titles as free/included for DVD subscribers.

They did not ban Netflix, however, because it was a subscription service, and they did not see it as directly competitive. We were seriously disadvantaged by our exclusion from those devices in those first two or three years. One alarming statistic shows why: we estimated that 95 percent of Netflix’s streams came either through their website, the three main game consoles (Xbox, PlayStation, and Nintendo Wii), or through the iPad and iPhone.

We also got pushback from some retailers. At the time, Walmart and Best Buy sold the most TVs to consumers, but Amazon’s retail business was a growing threat to them. Starting in 2007, they had employed a number of tactics to slow us down, such as declining to carry Amazon gift cards in their stores. Now their buyers warned some of the electronics manufacturers that any devices loaded with Amazon Video On Demand would not make it to their shelves. Walmart and Best Buy were such dominant players in the sale of electronic devices that many manufacturers wouldn’t work with us. When we convinced Sony to add Amazon’s streaming service to their Bravia TVs and Blu-ray players in September 2008, the logjam started to break. But it would take another four long years for us to convince the PlayStation team to play along with Amazon so we could finally tap into the 20 million–strong PlayStation user base.

Even our great asset, the Amazon website, wasn’t quite as important to the sale of digital media as it was to physical goods. Yes, it attracted lots of customers who were looking to buy media products, but applications on Macs, PCs, tablets, phones, and TVs were more and more important than our website for delivering a high-quality digital media experience. Apple, for example, sold all their digital media through an application running on Macs and PCs (iTunes) rather than on their website. And, unlike Amazon, Apple had control over their own devices. The combination of application and device delivered a high-quality streaming (or download) and playback experience and therefore delivered tremendous value to the consumer.

Sell in app not web

But as the Apple and HBO business models demonstrate, there was one important way that the world of digital media was the same as the old, analog media world: there was still a great advantage to be had in control. In the old media world, you could control one of two things: the method of distribution of the content or the content itself (or in some cases both). Broadcast networks like NBC and CBS controlled their networks and also developed exclusive content such as TV shows, sports events, and news broadcasts. Studios like Warner and Disney created movies and shows. In the new digital media world, broadcast networks and studios would lose their control on distribution, to be replaced by applications on internet-connected devices.

We were essentially a digital distribution system, with nothing unique or proprietary about it. No wonder we kept slamming into barriers on both ends of the value chain—content development and distribution on devices.

In 2010, we held a series of meetings with Jeff to discuss ideas and options for moving us toward the lucrative ends of the value chain. It was clear that consumers loved the all-you-can-eat, fee-based subscription model of Netflix, but we learned that Netflix was probably pouring between $30 and $40 million annually into licensing deals. While today, $40 million does not seem like a lot of money for Amazon to invest, believe me, this was not the case at the much smaller and leaner Amazon of 2010. That number shocked us. It seemed like a crazy big commitment. Not to Jeff. In one meeting he said something like, “In case it isn’t clear enough, I want to see the plan for how we are going to make a similar investment to get going in subscription video.” Jeff also made it clear that we should explore other big, digital-media ideas, including the creation of hardware devices.

Here’s another example of where a single-threaded leader and team helps. I was accountable for the financial performance and overall health of the affiliates business. Our team had virtually all the resources required to launch this feature: we had software engineers and product managers to build the feature; and we had our own customer service representatives armed with specialized knowledge and tools to field questions from affiliates. We knew our customers well, had conviction that the experiment was worth doing, and were willing to be misunderstood as we tried something new. We also had a rollback plan in the unlikely event that the experiment failed.

Jeff Barr joined Amazon a few weeks later and is still with the company, serving as VP and chief evangelist for AWS.

Andy Jassy had been Jeff’s technical advisor for the previous 18 months and was ready for a new role. He could have taken virtually any job in the company, including a leadership role in any of its largest businesses. Fortunately for us, he decided to start and lead a new team that would build on our experiment. Andy and his team envisioned and created a much more robust set of products that would usher in the era of cloud computing and become the massive hit that AWS is today. While the explosive growth of AWS and its sophisticated suite of product offerings all came after I had moved on to other projects, it serves as a vivid example of being Amazonian.

But Bias for Action does not obviate the need for the painstaking aspects of the Working Backwards process. We did not allow ourselves to be so driven by what our competitors might do that we would launch a product without first having thought very carefully about how our customers would use it and benefit from it. To put it another way, Working Backwards was the process that enabled us to put into action the principle of Customer Obsession.

The rewards, however, are clear and distinct for both the company and the person. Amazon is clear up front about seeking people who obsess over the customer experience and who value long-term success and continuous innovation over making a quick buck or earning a fancy title. It offers a context that supports risk-taking and openness to ideas from people at any level of the business; it also provides the fulfillment that comes from taking on difficult challenges under daunting time constraints and wrangling them to the best outcome possible. Most often, this brings superior results for the company.