“Amazon Devices” are all the devices manufactured by Amazon, such as Kindle, Echo, and Fire TV. Over 200 million customers visit “Amazon Devices” product pages every year. Those customers come from different shopping stages: early-stage shoppers are brand new to the device and are looking for educational content; mid-stage shoppers want to compare different models; late-stage shoppers have already researched and are ready to buy, etc. Current product pages assume that all customers are ready to make a purchase. Unfortunately, this approach can alienate those who are in the early stages of their buying process (which comprises roughly 50% of the site's visitors) and make it challenging for mid-stage shoppers to locate the information they require.
To tackle this challenge, I organized design workshops to unite different teams and created the product page framework. I then worked with PM to pitch the solution to VPs, turning this project into our organization’s long-term goal. I defined mechanisms for teams to collaborate under limited resources when economic uncertainty hit. Despite the challenges, we made progress and achieved our goals.
My Role: UX Design
Frequent Collaborators/Tiger Team: Product Manager (1), Software Development Manager (1), Marketing Managers
Project Outcome: successfully pitched designs to VPs, turning the project into the organization’s long-term goal. Defined mechanisms for teams to moving towards the goal under limited resources.Company: Amazon
Previous Attempts
Initially, I received little support for this project as many teams at Amazon Devices had already made attempts to tackle this challenge. For instance, the Alexa team established an "Alexa Hub" that provided educational information for new customers. Unfortunately, the "Alexa Hub" was not integrated into customers' shopping experiences and was difficult to locate, thus having a minimal impact on new customers. Another example was from the Devices Accessory team. This team added extra educational content about complementary products on the product page. This content was helpful for shoppers ready to purchase, but it discouraged early and mid-stage shoppers, hurting overall sales of Amazon Devices.
Drawing from past lessons, I persevered with the belief that there were at least two different actions I could take to achieve a positive outcome:
Solve problems in collaboration. Different teams were fixated on their own domains (e.g., the Device Accessory team only cared about getting more late-stage customers to buy more accessories). I could unite teams to create a shared vision and prevent isolated problem-solving.
Focus on “effective content” instead of “more content.” Many teams did not consider customer behavior when creating content, leading to customers ignoring or finding the content irrelevant. I recognized an opportunity to create a framework that would improve the effectiveness of content.
REMOTE DESIGN WORKSHOP
Design Workshop Goals
I organized a series of remote design workshops, inviting key partner teams responsible for various aspects of the product page. These teams brought diverse backgrounds, such as product management, data science, development, marketing, branding, and extensive product knowledge, including product education, accessory, post-purchase, and product launches. I set up three goals for the design workshop.
Collect past learnings and product knowledge from each team.
Identify the most important customer problems on product pages.
Align on a shared product vision.
REMOTE DESIGN WORKSHOP
Preparation
Create a product feature Library: I created a comprehensive product feature library with detailed explanations so that participants are on the same page regarding the functionality of each feature. An example of this can be seen in the image below.
Provide templates and prompts: I developed research-based templates and prompts with examples to ensure effective conversations.
Assign responsibility: I organized the 30 participants into teams according to their areas of expertise. To ensure smooth operations in each team, I assigned a "timekeeper," "team leader," and "presenter"”
REMOTE DESIGN WORKSHOP
Literature Review
I gathered research studies and past A/B testing results. I held several literature review sessions with participants to familiarize everyone with what we already knew.
REMOTE DESIGN WORKSHOP
Gather Insights
I had teams write down customer problems in the context of the Device Customer Journey, using the prepared templates and prompts
REMOTE DESIGN WORKSHOP
Find Themes
We grouped customer problems into five themes using affinity diagramming. These themes summarized the common issues customers experience on Amazon Devices product pages.
REMOTE DESIGN WORKSHOP
Brainstorm Future-State Attributes
Participants brainstormed future-state attributes under each customer problem theme.
REMOTE DESIGN WORKSHOP
Identifying the Key Ingredients
By voting and categorizing the future attributes into "Must Have" and "Nice to Have," we identified the essential components and common denominators of each feature attribute. This allowed us to create a powerful and concise vision rooted in customer problems.
Vision Statement
Amazon Device product pages will evolve and adapt as customers progress in their shopping journeys, bringing relevant and useful information to customers.
FRAMEWORK CREATION
Organize Existing Features
With a clear and unified understanding of customer problems and product vision, I dove into creating the design framework. Since we had repeatedly discussed how disorganized the existing product pages were, I started off by organizing existing features. I used the customer problem themes identified during the design workshop as the foundation and created 8 “zones.” To illustrate, I established a “Category Evaluation” and “Glace Product” zone in response to “customer problem theme 1 - information overload” from the workshop. This theme was derived from previous research that Amazon Devices product pages were overwhelming, leading customers to overlook critical information and fail to obtain the necessary details.
One question multiple stakeholders asked was: “Why do we group different features into zones?”
By creating consistent and predictable zones on the product page, customers can effortlessly find the features they need. This is similar to a well-organized wardrobe with labeled sections, where everything has its designated place. This approach saves customers time that would otherwise be spent searching the entire page.
FRAMEWORK CREATION
Integrate Shopping Journey
As defined in the product vision, the product page will “evolve and adapt as customers progress in their shopping journeys.” I generated various ideas for how the organized product page meets customers’ needs throughout their shopping journey.
I shared my ideas with the core team, and we voted on the approach that was feasible, usable, valuable, and had the highest impact on customers.
The selected solution will expand zones that are relevant to customers' current stages of shopping journeys to provide them with more helpful information. At the same time, irrelevant zones will collapse but will still be accessible to customers when they need them.
FRAMEWORK TO DESIGN
Adding Signals and Content
I worked with data scientists to understand “Signals” that indicate what stage of the shopping journey customers are in. I also worked with marketing managers and UX researchers to determine which content is most meaningful at which stage.
FRAMEWORK TO DESIGN
Personalization based on Individual Interests
To prevent shipping our organizational chart to customers, I engaged in discussions with stakeholders outside of our organization. During these discussions, I carefully considered data and insights that differed from my initial assumptions. One important insight I gained was the significance of personalization based on individual interests. Through conversations with the Alexa team and the Amazon central shopping team, I realized that exposing customers to product features tailored to their interests not only increases the likelihood of them buying the product but also enhances their satisfaction with future Amazon Device purchases. Consequently, I incorporated personalization based on individual interests alongside personalization based on customer journeys.
FRAMEWORK TO DESIGN
Information Architecture
Combining the “Zone” concept and personalization, I created the IA (information architecture) of a product page that evolves and adapts as customers progress in their shopping journeys.
FRAMEWORK TO DESIGN
Putting Everything Together
I produced detailed designs to assist senior leaders who may not be familiar with the terminology used in the IA. The high-fidelity designs serve the purpose of providing a visual representation of the new framework.
IMPLEMENTATION
Create a PR/FAQ Document
At Amazon, a PR/FAQ document is at the core of every innovation. It is also an opportunity for me to get funding for this project. I co-created the PR/FAQ with my product manager. Below is my responsibility for the PR/FAQ:
Communicate product vision and strategy, especially the customer impact.
Create detailed designs to help leaders visualize the solution.
Identify an MLP backed by data (MLP stands for “Minimal Lovable Product,” which is the minimal tech effort required for customers to adore a product).
Demonstrate operation efficiency. Show how my product vision benefits customers and is easy to execute by internal teams.
IMPLEMENTATION
Turn Lemons into Lemonade
I pitched the design to our VPs, with the support of leaders from different teams (who were involved in the design workshop). The proposal was immediately accepted into my organization’s long-term goals.
Unfortunately, my pitch was happening around a sensitive time (2022 Q4) when Amazon cut down spending, and I didn’t get immediate funding for this project.
Luckily, the leaders were highly impressed with the design framework and expressed full confidence in it. Consequently, numerous teams have contacted me to inquire about the implementation timeline. This has allowed me to break down the framework into smaller components and delegate them to different teams.
IMPLEMENTATION
Generate Hypotheses
I scheduled two additional work sessions with colleagues and stakeholders who participated in the design workshop. Using the design vision as a guide, we established four main hypotheses and corresponding guidelines to direct our future designs. I emphasized that these hypotheses will be continuously refined as we gather new data and insights.
IMPLEMENTATION
Derive Experiments from Hypotheses
We devised experiments for each hypothesis and integrated those ideas into each team's 2023-2024 roadmap. Currently, teams are running A/B tests guided by the hypotheses and testing our ways into launching the full design framework.