At Amazon, we believe that focusing on solving customer problems enables us to set priorities and invest in activities that deliver the most value. We identify innovation opportunities by starting with the customer and working backwards.
Working Backwards is the main mechanism we use to innovate on behalf of our customers: We dive deep to understand customer goals and needs, and then we innovate to solve problems and deliver new products, services and experiences.
Working Backwards, for example, led AWS to design and produce our own computer chips, a decision that surprised our industry. We chose this new direction not because we wanted to manufacture processors but because our customers’ compute needs were expanding. By developing our own chips, we could create more compute options for customers and deliver greater performance at a lower cost. Since 2017, we’ve regularly introduced new custom chips, providing customers with increased performance and security while reducing energy consumption and costs.
Working Backwards is where every AWS product, service and program begins. And because customer focus shouldn’t be guesswork, data plays a crucial role. We’re constantly using the cloud to bring data together securely from different sources and analyze it to better understand what customers want and need. This helps ensure that we don’t invent in isolation or produce a solution in search of a customer.
Because Working Backwards focuses on the customer, it’s an approach that adapts to different industries, geographies and organizational structures. We’ve helped a variety of AWS customers use it to uncover new opportunities for innovation, including Yelloh, a U.S.-based, direct-to-consumer frozen food delivery company, and TC Energy, a major North American energy company based in Canada that operates in the natural gas, oil and power industries.
Working with AWS, Yelloh developed a data-driven, two-way messaging service that improves order tracking for customers, and an AI-powered “digital door knock” that informs customers when a Yelloh truck is in their neighborhood. And TC Energy, which manages 57,900 miles of natural gas pipelines, developed ORBIT, an operational business intelligence application that uses machine learning to help customers optimize gas throughput.
Although these two companies operate with different business models, each could apply Working Backwards to understand its customers’ challenges and needs and invent new ways to serve customers better.
Innovating successfully depends on rapidly determining where to innovate, which solution
to focus on and how to build it.
Work Backwards From Customers
As Jeff Bezos says, “To be innovative, you have to experiment. If you want to have more invention, you have to do more experiments per week, per month, per year, per decade. It’s that simple. You cannot invent without experimenting.”
An important practice that has helped AWS maintain a rapid pace of innovation is equipping our employees with data that helps them understand our customers and make faster, better decisions. This removes bottlenecks and speeds experimentation. Moreover, giving employees access to data inspires them to innovate because they can use that information to find and assess opportunities.
We’ve also helped AWS customers use data to improve the speed and quality of their decision making. One of these companies is VTEX, a global enterprise digital commerce platform. VTEX’s C-level dashboards, which gave leaders visibility into the company’s commercial performance, didn’t provide information on key indicators, such as gross merchandise volume (GMV) and Revenue and Sales Channel performance. Working with the AWS Data Lab, VTEX modernized its data platform to support advanced analytics and used Amazon QuickSight, a cloud business intelligence service, to develop a prototype for a new C-level dashboard.
Just two months later, VTEX launched its new data system, giving its leaders enhanced visibility into the company’s commercial performance and goal achievement. More importantly, the improved dashboard provided better analytics and made it easier for VTEX executives to make faster, better decisions to support its customers and partners.
To innovate more, you have to innovate faster.
As Mariano Gomide de Faria, the company’s founder and co-CEO, told us, “The AWS Data Lab empowered our team to not only build a prototype in a matter of days but to bring a solution to production that gives us enhanced visibility into our business performance and goal achievement. It strengthens even more VTEX's ability to support the success of our customers and partners in their digital commerce journeys globally.”
That building process, though, isn’t a straight line. Most great ideas don’t come out of the gate perfect. Instead, they require experimentation and iteration to get to where they need to be.
But in many organizations, employees are reluctant to experiment because trying something new carries a risk of failure. To help Amazon employees better assess risk and feel more confident with experimentation, we provide training and resources to help them evaluate whether a decision is a one-way or a two-way door. A one-way door decision is one that has significant and often irrevocable consequences, like building a new fulfillment center or retiring a popular product. It’s hard to undo and can impact customer trust. In contrast, a two-way door decision is one that has limited and reversible consequences, such as piloting a new feature on a mobile app.
We’ve found that most decisions are two-way doors. For these, we encourage our employees to move forward if they have enough evidence to believe that a decision could benefit our customers. “Enough evidence” tends to be about 70% of the data an employee wishes they had. Waiting for 90% or more will likely cause a decision to move too slowly and could lead to missing the chance to be first in providing a new experience for customers.
Another way to encourage experimentation is to reduce the cost—which lowers risk and can turn what might appear to be a one-way door into a two-way door. You can do this by providing digital tools, automation and agile development so your employees can build, test, and (when needed) abandon experiments quickly. Moreover, when those tools are cloud-based, you pay only for what you use rather than investing in capacity you may not need, and you can scale successful projects securely and reliably without drawn-out, costly rollout plans.
Perhaps the most important element in inspiring employees to innovate is having leaders embody the belief that risk-taking is critical. Amazon president and CEO Andy Jassy has said, "If you invent a lot, you will fail more often than you wish. Nobody likes this part, but it comes with the territory." And Adam Selipsky, CEO of AWS, encourages employees to focus more on an idea’s potential to solve customer problems than on that idea’s risk of failure.
Across the organization, Amazon leaders use the one-way/two-way door mental model in their own decision making, and they reward employees for taking chances and piloting new projects, whether or not those experiments are a “success.”
VTEX demonstrates that innovation is more than ideation—it’s building.
There are many ways to foster innovation, and each organization should develop an approach that reflects its own unique culture, values and business model. At AWS, we’ve found that putting customers at the center of our innovation efforts gives us focus, helps us set priorities and guides us as we iterate. Moreover, as we’ve helped AWS customers accelerate innovation, we’ve seen that Working Backwards adapts well to different types of organizations across a wide variety of industries.
To promote innovation in your own organization, experiment with mechanisms, like Working Backwards, to ensure that all creative thinking begins with a deep understanding of your customers. Explore providing your employees with data and tools that enable them to imagine, build and iterate faster—and at a lower cost. And, finally, celebrate the people who took a chance on an idea that could make life better for your customers. Whether it succeeds or fails, every experiment offers your organization the opportunity to learn.
How Can You Help Innovation Thrive In Your Organization?
Transforming
For Growth
Visit the series homepage for more stories of companies tapping into ingenuity and AWS technology to grow and change industries.
DISCOVER MORE
01
Make High-Velocity Decisions
02
Encourage Bold Invention And Experimentation
03
Operating a business at scale creates complexity, which often invites Day 2 thinking: focusing on short-term goals rather than customers, allowing bureaucracy to slow decision making and cultivating a fear of failure that prevents employees from envisioning what could be.
To be successful, businesses need to evolve in response to changing customer preferences and shifts in the external environment. Continuous innovation powers this
ability to adapt, so it’s imperative that business leaders take steps to resist Day 2 thinking. That almost always requires more than rewriting the company mission statement or hosting an annual hackathon.
At Amazon, we foster a culture of innovation through relentless customer focus and company-wide mechanisms. And we’re not alone. For example, Toyota Connected, which Toyota founded as an innovation center of excellence, takes a similar approach. Although it relies on different mechanisms that reflect its culture, like Investment Councils where engineers pitch ideas directly to the executive team, Toyota Connected uses the cloud to empower its employees “to be brave, bold, inspired, and to deliver memorable customer mobility experiences.”
There’s no surefire way to innovate, and it becomes more difficult as an organization grows.
Maintain A
Day 1 Mentality
04