Same-Day Delivery Redesign
Same-Day Delivery Redesign
Same-Day Delivery Redesign
Redesigning the guest’s experience with same day delivery orders
Timeline
Timeline
April - September 2025
April - September 2025
April - September 2025
Status
Status
2026 Roadmap
2026 Roadmap
2026 Roadmap
Role
Role
UI/UX Design, Research, User flow, Wireframing, Prototyping, Usability testing
UI/UX Design, Research, User flow, Wireframing, Prototyping, Usability testing
UI/UX Design, Research, User flow, Wireframing, Prototyping, Usability testing
Enter password to view case study
Today, shoppers have a multitude of shipping options to receive their order, including same-day delivery. Guests are required to select a delivery window time before completing the checkout process. The flow works, but the UI has not been evaluated or updated in many years.
As order complexity increased, the time selection experience struggled to scale — particularly for multi-package orders. This project focused on simplifying delivery time selection while improving clarity, flexibility, and usability.
Today, shoppers have a multitude of shipping options to receive their order, including same-day delivery. Guests are required to select a delivery window time before completing the checkout process. The flow works, but the UI has not been evaluated or updated in many years.
As order complexity increased, the time selection experience struggled to scale — particularly for multi-package orders. This project focused on simplifying delivery time selection while improving clarity, flexibility, and usability.
Understanding the Problem
Understanding the Problem
The Never Ending List of Delivery Times
The Never Ending List of Delivery Times
Taking a heuristic evaluation of the current flow, it was obvious why the current existing user journey wasn't ideal. The flow technically worked in that guests are able to checkout successfully, but compared to other parts of the website, it was looking outdated.
The cramped nature of the UI discouraged future feature development, and was arguably costing the user time in having to look through all the available delivery options. More time and steps may risk the user deciding not to complete their purchase, especially since guests ordering SDD in particular are usually looking to receive their items fast and checkout fast.
Taking a heuristic evaluation of the current flow, it was obvious why the current existing user journey wasn't ideal. The flow technically worked in that guests are able to checkout successfully, but compared to other parts of the website, it was looking outdated.
The cramped nature of the UI discouraged future feature development, and was arguably costing the user time in having to look through all the available delivery options. More time and steps may risk the user deciding not to complete their purchase, especially since guests ordering SDD in particular are usually looking to receive their items fast and checkout fast.
I created 4 design solutions and iterated on them based on critique and dot voting results. The two solutions that performed best were eventually transformed into the designs for user testing.
I explored a variety of solutions, and after critique and dot voting, I landed on these two iterations for user testing.
Diving Deeper with Research
Diving Deeper with Research
What about 2 Hour Delivery Windows?
What about 2 Hour Delivery Windows?
One of my initial design considerations was listing delivery slots by 2 hour increments instead of 1 hour as this could reduce the height of the list by 50%.
Coincidentally, a sister company had recently conducted research around user behavior in regards to delivery windows. 6 Users were asked whether they preferred 1 or 2 hour delivery windows. Looking at this research, user’s strongly felt they liked the flexibility of delivery options but wanted control over when their order will be delivered.
One of my initial design considerations was listing delivery slots by 2 hour increments instead of 1 hour as this could reduce the height of the list by 50%.
Coincidentally, a sister company had recently conducted research around user behavior in regards to delivery windows. 6 Users were asked whether they preferred 1 or 2 hour delivery windows. Looking at this research, user’s strongly felt they liked the flexibility of delivery options but wanted control over when their order will be delivered.
6/6 participants
chose 1 hour over 2 hours windows.
Delivery windows <2 hours were considered “too long” and associated with other types of work (such as home repairs or cable installations).
After analyzing the research, I decided to prioritize solutions showing 1 hour window slots to help provide users with more of a sense of confidence in their orders.
After analyzing the research, I decided to prioritize solutions showing 1 hour window slots to help provide users with more of a sense of confidence in their orders.
Competitor Strategies
Competitor Strategies
Looking at competitors, companies like Walmart were taking advantage of their space and up-selling express delivery options. Amazon combines their same-day delivery option window with regular shipping, offering SDD potentially depending on when the order is placed or if a minimum dollar amount is reached.
Looking at competitors, companies like Walmart were taking advantage of their space and up-selling express delivery options. Amazon combines their same-day delivery option window with regular shipping, offering SDD potentially depending on when the order is placed or if a minimum dollar amount is reached.
How might we create a more streamlined same-day checkout process for guests so they can place their order with confidence, both now and in the future?
How might we create a more streamlined same-day checkout process for guests so they can place their order with confidence, both now and in the future?
Iterative Solutions in Figma
I created 4 design solutions and iterated on them based on critique and dot voting results. The two solutions that performed best were eventually transformed into the designs for user testing.
I explored a variety of solutions, and after critique and dot voting, I landed on these two iterations for user testing.
I created 4 design solutions and iterated on them based on critique and dot voting results. The two solutions that performed best were eventually transformed into the designs for user testing.
I explored a variety of solutions, and after critique and dot voting, I landed on these two iterations for user testing.
Hypothesis: By surfacing less delivery times, the user will be able to checkout more efficiently and more confidently.
Hypothesis: By surfacing less delivery times, the user will be able to checkout more efficiently and more confidently.
Idea 1: Accordion Structure
Idea 1: Accordion Structure
In order to attempt to lessen the cognitive load for the user at a glance, I created a accordion solution in which the user would be shown a 3 time slots per day, with the option to expand out and view more times. However, from an accessibility perspective this solution could be difficult for users utilizing voice control.
In order to attempt to lessen the cognitive load for the user at a glance, I created a accordion solution in which the user would be shown a 3 time slots per day, with the option to expand out and view more times. However, from an accessibility perspective this solution could be difficult for users utilizing voice control.
Idea 2: Default Delivery Time
Idea 2: Default Delivery Time
This solution selects the first delivery time available. As an alternative, the user could choose Schedule Delivery to display more times.
In an ideal scenario, the time for a guest to checkout would be significantly reduced. However, the tradeoff is that guests may fail to notice the preselected delivery time until after they've checked out.
This solution selects the first delivery time available. As an alternative, the user could choose Schedule Delivery to display more times.
In an ideal scenario, the time for a guest to checkout would be significantly reduced. However, the tradeoff is that guests may fail to notice the preselected delivery time until after they've checked out.
User Testing & Iteration
To ensure the checkout experience was both usable and confidence-inspiring, I ran three rounds of unmoderated testing . These rounds progressed from broad usability validation to targeted design decisions, culminating in an A/B test that informed the final interaction model.
To ensure the checkout experience was both usable and confidence-inspiring, I ran three rounds of unmoderated testing . These rounds progressed from broad usability validation to targeted design decisions, culminating in an A/B test that informed the final interaction model.
Round 1: Surface Less Delivery Times
Round 1: Surface Less Delivery Times
I created a prototype that would surface less times in chip format with a modal that would open to display more times to the user, while still pre-selecting the earliest delivery time.
Participants with recent Same Day Delivery experience were then asked to complete realistic checkout tasks, including selecting a delivery window, adjusting the delivery date, and editing delivery instructions.
I created a prototype that would surface less times in chip format with a modal that would open to display more times to the user, while still pre-selecting the earliest delivery time.
Participants with recent Same Day Delivery experience were then asked to complete realistic checkout tasks, including selecting a delivery window, adjusting the delivery date, and editing delivery instructions.
Overall, these tests performed well. Participants consistently described the experience as intuitive and easy to navigate. Delivery windows were clearly understood, “View more times” was discoverable, and users completed tasks quickly without hesitation.
Overall, these tests performed well. Participants consistently described the experience as intuitive and easy to navigate. Delivery windows were clearly understood, “View more times” was discoverable, and users completed tasks quickly without hesitation.
5/5
participants were able to open the modal to view more times ✅
2/5
wondered if a design with radio buttons would be easier to read. 🤔
After sharing these designs internally, the design systems team had concerns around inconsistencies such as the “View more times” not being consistent with other actionable items on the page. The ”Okay to delivery early?” checkbox also was distracting in a gray box, so I wondered if there was an alternative design for this as well.
After sharing these designs internally, the design systems team had concerns around inconsistencies such as the “View more times” not being consistent with other actionable items on the page. The ”Okay to delivery early?” checkbox also was distracting in a gray box, so I wondered if there was an alternative design for this as well.
Round 2: Mental Models & Instruction Persistence
Round 2: Mental Models & Instruction Persistence
In order to try and further cleanup the SDD section, moving delivery instructions to the address section could feel more natural and help users move through the flow quicker, which would be consistent with how the address settings under the user profile is structured today.
In order to try and further cleanup the SDD section, moving delivery instructions to the address section could feel more natural and help users move through the flow quicker, which would be consistent with how the address settings under the user profile is structured today.
Today, guest’s can save default delivery instructions in their settings.
Today, guest’s can save default delivery instructions in their settings
Participants consistently associated delivery instructions with the delivery address — not the delivery time. Instructions were perceived as location-specific and tied to physical context.
However, some participants pointed out that the instructions might be seen as a “global” instruction that would apply to all of their registered addressed. I refined the copy to “Save instructions to this address” to further clarify instructions applying to a specific address.
Participants consistently associated delivery instructions with the delivery address — not the delivery time. Instructions were perceived as location-specific and tied to physical context.
However, some participants pointed out that the instructions might be seen as a “global” instruction that would apply to all of their registered addressed. I refined the copy to “Save instructions to this address” to further clarify instructions applying to a specific address.
100%
participants successfully located and edited delivery instructions without assistance ✅
2/5
mentioned the risk of forgetting their instructions had been saved. 🤔
Round 3: A/B Testing the Delivery Time Selector
Round 3: A/B Testing the Delivery Time Selector
After many rounds of user testing and gathering data, I still felt the lingering question: chips vs list? After all, the list format was used in many areas of the website. Though, part of my hesitancy came the potential design issues regarding alignment with an embedded list format. In order to help clarify any doubts and optimize the delivery window experience, I ran an AB tests for the two interaction models to see if chips or embedded list would perform stronger.
After updating the designs to follow the design system more consistently, the final user testing focused on choosing the right component for the time selection: chips or list. The list format is used in many other areas of the website; however, it still felt clunky compared to the chips.
After many rounds of user testing and gathering data, I still felt the lingering question: chips vs list? After all, the list format was used in many areas of the website. Though, part of my hesitancy came the potential design issues regarding alignment with an embedded list format. In order to help clarify any doubts and optimize the delivery window experience, I ran an AB tests for the two interaction models to see if chips or embedded list would perform stronger.
After updating the designs to follow the design system more consistently, the final user testing focused on choosing the right component for the time selection: chips or list. The list format is used in many other areas of the website; however, it still felt clunky compared to the chips.
Both patterns were usable, and participants completed tasks successfully. However, behavioral differences emerged. Participants commented that the chip layout reduced visual density and enabled faster recognition, while the embedded layout felt more comprehensive but visually heavier.
Both patterns were usable, and participants completed tasks successfully. However, behavioral differences emerged. Participants commented that the chip layout reduced visual density and enabled faster recognition, while the embedded layout felt more comprehensive but visually heavier.
100%
task success rate for selecting and updating delivery windows ✅
Majority completed in under 10 seconds for both options
Conclusion
Final Design Decisions
Final Design Decisions
The final design incorporates a chip-based delivery selector, address-anchored delivery instructions, clarified persistence language, and improved confirmation states. These refinements reduced cognitive friction while reinforcing user confidence. Since desktop offers slightly more real estate, this version would be used to expand the chips while on mobile web they would stack vertically.
Due to the address feature being owned by another team, they felt delivery instructions was not quite ready to be moved into SDD due to the other trickle-down areas this could affect. For now, it will remain in SDD but with the potential and user research to make the best decision on it going forward. Since it would remain in SDD, I decided to keep it in a modal to stay aligned with the behavior with the other features in SDD.
The final design incorporates a chip-based delivery selector, address-anchored delivery instructions, clarified persistence language, and improved confirmation states. These refinements reduced cognitive friction while reinforcing user confidence. Since desktop offers slightly more real estate, this version would be used to expand the chips while on mobile web they would stack vertically.
Due to the address feature being owned by another team, they felt delivery instructions was not quite ready to be moved into SDD due to the other trickle-down areas this could affect. For now, it will remain in SDD but with the potential and user research to make the best decision on it going forward. Since it would remain in SDD, I decided to keep it in a modal to stay aligned with the behavior with the other features in SDD.
Key learnings & Emerging Design Principles
Across all rounds, several themes emerged. Users valued control over automation, particularly around saved delivery instructions. Trust was a recurring theme — even small ambiguities created anxiety in a high-stakes interaction like same-day delivery. Visual clarity and progressive disclosure reduced cognitive load and increased confidence.
These learnings could help shape design principles for any future design updates for the checkout page to ensure the best user experience.
Next Steps & Special Instructions
Key learnings & Emerging Design Principles
After meeting with the engineering team, this work is slated on the roadmap for 2026. The engineers were confident that data should be safe and available. This design will help provide options for adding more features as well, or upselling membership programs.
This project also continued the effort to remove "Special Instructions" from the SDD section. I still felt strongly that special instructions did not belong in SDD and moving it elsewhere could help provide clarity and create a faster checkout experience. Currently on mobile, this feature lives in cart so perhaps the most natural case is to move it there.
Additionally, since I proposed the feature “Special Instructions” be removed, designers mentioned the use cases in which special instructions is used. I still felt strongly that special instructions did not belong in SDD and moving it elsewhere could help provide clarity and create a faster checkout experience. Currently on mobile, this feature lives in cart so perhaps the most natural case is to move it there.
Across all rounds, several themes emerged. Users valued control over automation, particularly around saved delivery instructions. Trust was a recurring theme — even small ambiguities created anxiety in a high-stakes interaction like same-day delivery. Visual clarity and progressive disclosure reduced cognitive load and increased confidence.
These learnings could help shape design principles for any future design updates for the checkout page to ensure the best user experience.
Next Steps & Special Instructions
After meeting with the engineering team, this work is slated on the roadmap for 2026. The engineers were confident that data should be safe and available. This design will help provide options for adding more features as well, or upselling membership programs.
This project also continued the effort to remove "Special Instructions" from the SDD section. I still felt strongly that special instructions did not belong in SDD and moving it elsewhere could help provide clarity and create a faster checkout experience. Currently on mobile, this feature lives in cart so perhaps the most natural case is to move it there.
Additionally, since I proposed the feature “Special Instructions” be removed, designers mentioned the use cases in which special instructions is used. I still felt strongly that special instructions did not belong in SDD and moving it elsewhere could help provide clarity and create a faster checkout experience. Currently on mobile, this feature lives in cart so perhaps the most natural case is to move it there.
