iOS/ANDROID APP CASE STUDY
Walmart's Generative AI
Product Details Redesign
OVERVIEW
Walmart customers found the existing item detail experience frustrating—it was difficult to navigate, cluttered with dense content, and failed to highlight decision-making information.
Walmart's item detail page receives over 70 million views each day. It plays a critical role in providing customers with the information they need to make a confident purchase decision and add an item to their cart. However, customer feedback and user research revealed that shoppers often struggled to find key item details—information that's essential to evaluating and understanding the product. To address this, we set out to not only redesign the way item details are discovered and presented, but also explore the use of generative AI to enhance the item content for a better overall customer experience.
ROLE & DETAILS
Senior Product Designer
Competitive Research, Prototyping, Testing, Leadership Presentations, Documentation
Team
1 Product Designer
1 Product Owner
2 Engineers
Design Timeline
September 2023 – March 2025
Tools
Figma
THE CHALLENGE
"How might we enhance the item detail experience to make key information more discoverable and digestible—while leveraging generative AI to deliver the insights customers need to make confident purchase decisions?"
Mobile customers using the Walmart app faced a frustrating experience when trying to find key item information. Competing internal Walmart teams contributed to cluttered layouts, forcing customers to scroll extensively just to reach essential details. Additionally, third-party sellers had complete control over the content displayed on item pages, often resulting in overwhelming, repetitive, and poorly structured information. Our goal was to redesign this section of the item page to make product information more discoverable, easier to navigate, and user-friendly.
On load
First scroll
Second scroll
Third scroll
Amazon
Costco
Target
Wayfair
Best Buy
Temu
COMPETITIVE RESEARCH
I was surprised to find that many competitors offered lackluster approaches.
• Amazon highlights a few key features up top, followed by a short list of bullet points. At first, I thought the blue text was a bug—but they're actually clickable links. That interaction isn’t very intuitive.
• Target uses iconography initially, but quickly overwhelms the user with a dense block of text—roughly three full scrolls worth.
• Best Buy incorporates iconography and uses actionable arrows to signal expandable content. However, most of the other accordions are collapsed by default, requiring the user to actively click through to discover more.
• Costco seems more focused on promoting multiple purchases than on helping customers learn about the specific item they’re viewing.
• Wayfair puts some emphasis on key highlights using icons, but the overall layout lacks design and visual hierarchy.
• Temu includes a section for rich media, but it’s small and visually uninviting—it feels like an afterthought and could easily be overlooked by customers.
FIRST ITERATION
Leveraging generative AI, we transformed item detail content into a streamlined, all-in-one experience for the customer.
I began by organizing the experience into four main sections, each designed to simplify, elevate, and enhance how customers consume the information:
• Highlights: This section featured three key item highlights, each marked with a green checkmark to reinforce positive attributes.
• Quick description: This 400-character summary distilled lengthy, often overwhelming third-party item details into a single, easy-to-read paragraph. Blue highlights drew attention to the most important terms and phrases. A “View more” text link opened a bottom sheet containing the full item description provided by the third-party seller.
• At a glance: This section showcased six key specifications, each accompanied by intuitive icons to enhance visual appeal and customer understanding. A “View more” text link allowed users to explore the full list of item specifications in a bottom sheet.
• Reviews summary: This section condensed all customer reviews into a single, insightful headline and summary paragraph. Color-coded highlights—green for positive, red for negative, and gray for neutral sentiments—helped customers quickly scan sentiment and tap directly into the corresponding review.
I organized the experience into four streamlined, visually stimulating sections, each designed to make item information easier to find, understand, and act on. By leveraging generative AI and thoughtful visual design, I distilled complex content into digestible pieces and added interactive elements to guide customer attention and improve the decision-making process.
TESTING RESULTS
We tested the new generative AI item details experience on the native Walmart app and received both positive and negative results.
Early testing delivered promising results: add-to-cart rates rose by 1.10%, and bounce rates fell by 0.95%. However, the test also revealed a downside—engagement with Walmart+ and sponsored ads declined, due to those modules being pushed lower on the page. This drop translated to a potential annual loss of $4 million if the changes were fully launched, presenting a significant business concern. Additionally, other trade-offs were present, such as the duplication of information between this new, large module and the existing item details section further down the page. This redundancy made the page longer and feel more confusing. So, we decided to pause the rollout and focus on refining the experience further.
SECOND ITERATION
With learnings in hand, we decided to break up the experience—no hard feelings.
In the second iteration, we moved away from a single, oversized module to a more adaptive—and still somewhat complex—experience. Our goal was to recover lost ad revenue while keeping key product information visible higher up on the page. To strike this balance, we broke the modules into smaller components and enabled testing across different placements to determine the most effective layout. Additionally, we eliminated the redundant item detail information further down the page and consolidate.
TESTING RESULTS
This round delivered strong results across the board, but we knew there was still more to achieve.
The next round of testing continued to show positive results on our native apps. On iOS, we saw significant improvements in item page add-to-carts (+0.39%) and item page add-to-carts per visitor (+0.34%). Overall iOS order conversion rates had a significant increase (+0.30%). On Android, we saw similar gains in item page add-to-carts (+0.25%) and item page add-to-carts per visitor (+0.27%). Additionally, we recovered the negative impact on Walmart+ and sponsored ads from the earlier iteration. While the results were strong and led us to move forward with a full launch, we recognized the need to keep iterating—customers still had to scroll too far to access the key information they needed.
THIRD ITERATION
With every scroll carrying weight, how can we balance the visibility of key item information with the need to preserve sponsored ad revenue—all without displacing other critical page modules?
Our original goal was to make key item information concise, easy to understand, and readily discoverable. However, through testing and iteration, we encountered a new challenge: a noticeable decline in Walmart+ and sponsored ad revenue. Data revealed that most customers only scroll a few times, underscoring the importance of surfacing both critical item details and required ads within the same viewport. In our third iteration, we explored two layout options that reduced vertical space compared to the original iteration while still delivering the decision-making content customers need. As a result, Walmart+ and sponsored ads now appear 1,099px higher up on the page compared to the first iteration, while AI-generated content takes a more prominent position. Customers can also easily access full third-party item details via a bottom sheet or a convenient, scrollable jump link.
TESTING RESULTS & A FUTURE LAUNCH
This iteration is currently being tested on Walmart platforms. Check back for updates!
We believe this design strikes the right balance—surfacing key item details in a clear, accessible location without pushing other critical modules further down the page, all while preserving prime placement for Walmart+ and sponsored ads. You can see this test live in the Walmart app soon, and we’ll continue to share updates as testing progresses. The team is optimistic about receiving the green light to launch this new iteration in June 2025.