Overview  

Amazon Payment Experience (APX) universal widgetized payment experience

• Analyzed/deconstructed all existing payments pipeline experiences.
• Created an encapsulated and extensible widget suite to accommodate those experiences.
• Translated component architecture into API based data models and workflows
• Lead role in negotiating voluntary adoption by 1000+ payment pipelines.
• Personally reviewed and approved all new payment features company-wide in a federated development model.
• Created or oversaw all new payment methods internationally with extensive exploratory and functional user testing.
• Participated in hundreds of interview loops and promotion reviews for both designers and other roles.
• Grew a team from 4 developers and one 1 designer to multiple teams employing hundreds to take ownership of Amazon payments globally.

Supplemental Axure Documents  

APX 101 – basic concepts and tenets

(some browsers may require plugin)

Problem and Solution  

     I was hired at Amazon by a four person R&D team to solve what was considered an impossible problem. Over 8 years we solved the problem, grew into multiple teams employing hundreds of people, and voluntarily migrated 99% of Amazon traffic to our product. In the process we became the owners of all payments experience across Amazon globally.

Problem

     There are over 1000 independent business pipelines within Amazon, each empowered to choose their own experience and technology stack. Each business links directly to dozens of backend payment services and owns their own workflows and UI. With no cross-pipeline mechanisms or oversight, experiences are highly variable with no ability to sync change in style or pattern. To add a new feature or enhancement, each business needs to do individual design, development, and testing. Likewise, each service is being used in countless unknown ways preventing the service from changing without breaking those experiences. Feature growth demand, the number of pipelines, and variation of experience are growing exponentially. There is no top-down mechanism to address this type of cross-vertical problem.

Solution

     Create an abstraction layer between pipelines and the backend services. Create a widget suite to replace all existing payment experiences in all pipelines that can accommodate all existing payment features in every region and on every surface. Voluntarily onboard every pipeline to the externally controlled widget. Every pipeline now has full access to the entire breadth of Amazon payment features and every subsequent feature with no additional development or effort. Backend services are now free to grow and change without unique pipeline dependencies.

Complications

• Broad browser support
• Progressive enhancement requires widgets to work end-to-end without javascript
• Needs to work on non-HTML based devices
• Major pipelines have their own payments design teams with their own roadmaps
• No top-level authority to resolve conflict or drive the project. Participation is entirely voluntary and based on product merit.
• Use of widget reduces individual pipeline control of their experience.

In Detail

     Over several years I analyzed and deconstructed every payments pipeline in Amazon and designed an experience extensible enough to accommodate the entire payments landscape. I unified payments patterns and styling architecture and made possible for the first time brand-wide change and growth. Previously only Amazon retail checkout had access to the full breadth of payment products available due the high cost of design and implementation. Now every pipeline has full access with no cost to the business. With a combination of plug-n-play nested components and detailed configuration, nearly any payment experience or business requirement can be accommodated.

     I played a lead role in negotiations with each client for onboarding and became a critical influencer of adoption. Once a critical mass of clients were onboarded, I created a Payments Experience Bar Raiser process to review and approve each new payment feature in a federated development model to maintain consistent patterns and reduce widget variation. I also used my position to seek out and champion similar products being developed by other teams, creating a new family of cross-vertical technology where none previously existed.

     As the lead payments designer, I created or oversaw all new payment method and feature design and provided consultation across Amazon as the subject matter expert within payments. Though my position was an individual contributor it contained many supervisory responsibilities.

     In addition, I was regularly included in hiring loops for both design and other positions. I participated in hundreds of interviews and promotion reviews and mentored both full time designers and Amazon design interns, writing intern proposals and managing their internships.

     Though, I feel my greatest accomplishment was shepherding an important product through 5 reorgs and VP level leadership changes with many subsequent team shuffles and constantly shifting priorities across enumerable individual roadmaps to achieve a scope of collaboration previously unknown at Amazon. And without it, the growth we now see would be unmanageable.

     Much of the actual design of APX is abstract. The goal of the product was to accommodate any number of visual styles and eventually transitioned to a universal styling architecture that controls much of the visual pieces which change over time. Though built in a visual language, the true core of the design is pattern, form, extensibility, and flexibility without proliferation. The beauty is in managing countless needs with a concise and well-designed tree of individual widget components. And creating a sound process for pattern creation that supersedes arbitrary design decisions (even my own). This unfortunately makes it difficult to represent effectively in a visual portfolio.