how might we increase the adoption of UberEats' grocery shopping offerings?
context
sponsored school project @ Uber
a 5-month project @ Georgia Institute of Technology
master's thesis project @ Human-Computer Interaction program
team
a designer(me), a researcher,
guided by two designers from Uber
contribution
led design
completed all the design works
cooperate research & analysis
outcome
We designed Swipe It, a gamified feature to boost order volume and frequency.
It received positive feedback from target users, and notably, Uber later began testing a similar internal feature—highlighting the practical relevance of our design direction.
project context
While Uber is well established in ride-sharing, its other services—like food and grocery delivery—are still building traction. As part of Uber’s broader superapp vision, driving adoption of these offerings has become a key strategic goal.
In this concept project, we partnered with the Uber design team to focus specifically on the grocery delivery experience. We held bi-weekly sessions with Uber designers for feedback and guidance, aligning our work with real-world product considerations and user needs.
what is the challenge for Uber?
Uber already holds a significant share in the grocery delivery space with UberEats and offers a smooth user experience. However, growth in this area has plateaued. Despite a mature product, the service saw limited traction—reflected in slow market share growth, stagnant new user acquisition, and low retention.
challenge/ how might we help Uber gmrow its grocery shopping service?
what are the growth metrics?
Uber provided four key metrics to evaluate potential solutions:
Lifetime revenue
New users
Amount spent per order
Order frequency
success metrics/ These KPIs became our north star in defining and prioritizing problem spaces and design directions.
first, take a look of my final solution!
Swipe it!
and get your grocery delivered at your door.
Get inspiration.
Collaborate.
Get grocery item.
features
feature 1/
get inspirations
Simply swiping through a list generated based on user preference and order history.
feature 2/
get grocery by dish
In dish detail page, use ingredients as your shopping list and shop for your items from multiple stores.
feature 3/
collaborate
Invite your families or roommates to join your swiping. Know what they also like and get groceries for them.
what is the process…
01
understand the landscape
Goal/
❓ What strategies do other companies use to drive user adoption of a single service or feature—especially in relation to the four growth metrics: new users, lifetime revenue, order frequency, and amount spent per order?
❓ Are there existing studies or insights around Uber’s multi-service model and super app user behavior patterns?
Methods/
📘 Literature review
We reviewed academic papers and industry articles focused on digital platform growth, super apps, user behavior adoption, and TAM & SCoT frameworks. This helped us establish a foundation of behavioral and structural patterns across large-scale digital ecosystems.
📊 Competitive analysis
We conducted a strategic review of 10 platforms—including super apps (Alipay, WeChat, Amazon, Dazhong Dianping), ride-share services (Didi, Grab), and grocery delivery platforms (Walmart, Instacart, DoorDash, Gopuff). We identified their methods for promoting service adoption and increasing cross-service engagement.
Insights/
💡
New behaviors often emerge when services are paired with existing habits.
💡
Bulk-buying is more common than one-off purchases, often used to avoid fees.
💡
Key barriers to UberEats grocery adoption: high perceived cost and low trust in product quality.
💡
Gamification, social sharing, referrals, and predictive nudges are the most effective drivers of cross-service engagement.
But how does
literature
translate to the actual consumer?
02
reveal user behaviors
Goal/
❓ What are the key habits and mental models people follow during grocery shopping?
❓ What keeps users loyal to their current grocery methods (e.g., in-store vs. online)?
Methods/
📋 Survey - 48 participants, 20-52 years old, in-store+online shoppers
We conducted a survey to understand people's grocery shopping decision-making process. The goal was to identify which factors drive their choices (e.g., price, convenience, trust) and how these priorities differ across channels. It also helped us recruit suitable interviewees for the next phase.
🎙️ Interview - 9 participants, 23-40 years old, in-store+online shoppers
Through semi-structured interviews, we dove deeper into user motivations, habits, and emotional factors behind grocery shopping decisions. We aimed to uncover unexpected insights, especially around the emotional and situational contexts that shape preference for grocery experiences.
Insights/
💡
A typical grocery journey follows the flow: Need → Plan → Decide → Execute
💡
Users prioritize saving time across all stages—from planning to putting groceries away.
💡
Shopping needs shift based on life stage (e.g., affordability, nutrition, family routines).
💡
Many users distrust the quality of online groceries and are skeptical of customer support.
💡
In-store shopping fulfills emotional needs like exploration, habit, and family bonding.
with all these insights,
we further defined what we need to do with our design…
03
define design requirements
After analyzing our research findings—from market trends to user behaviors—we synthesized three key design requirements to guide our solution. These principles are grounded in user needs and strategic goals, and they helped us prioritize ideas and evaluate trade-offs during concept development.
We used these requirements as directional filters throughout the design process, ensuring that every interaction we proposed aligned with what would truly drive user adoption, engagement, and long-term value.
💡
Streamline multiple steps
Keep shopping flow simple and smooth
Provide improved efficiency at multiple parts in the journey
Increase user trust and control over quality of goods
💡
Make the value obvious
Clearly communicate convenience, flexibility, or other perks that may justify fees
Offer an experience that feels differentiated from competitors
💡
Bring the market online
Enable sharing, community or collaboration
Encourage exploration and discovery of different items
Foster alignment within a household
let's get into the design process from here!
04
first, ideation!
Using our design requirements and the success metrics as a guide, the researcher and designer independently generated ideas.
We then reconvened to discuss and apply the SCAMPER method, narrowing our concepts down to three that were both feasible and aligned with our requirements.
concept 1/
recipe inspiration
concept 2/
household profile
concept 3/
shopper live tracking
how do users and experts think about our ideas?
We then evaluated our concepts with 10 target users. And also we re-examined our problem statements through the lens of diverse brands—from similar to entirely different niches—to move away from Uber-centric thinking.
design challenge/ By deciding to move forward with concept 1 and 2, we also recognized the limitations of existing solutions in both spaces. Recipe apps are abundant. Household collaboration features often feel static or overly utilitarian. To differentiate our solution and enhance engagement, we brainstormed ways to bring a more playful, low-friction, and habit-forming experience to the table.
and it led us to our final idea:
What it keeps:
recipe inspiration - concept 1
invite and share with household members - concept 2
bundle-buy ingredients for selected dish - concept 1
What it adds:
gamified experience for browsing recipes inspired by dating app mechanics
swipe along with others asynchronously enabling mutual discovery
lightweight interaction, cut off the avatar household profile
This format not only supports our key design requirements—streamlining steps, bringing the market online, and making the value obvious and have the potential to fulfill all the success metrics — ✅ new users ✅ amount spend per order ✅ order frequency ✅ lifetime revenue, but also introduces a fresh, emotionally resonant interaction that aligns with Uber’s goal of making grocery shopping more engaging and frequent.
we then build our three key user flows
and have this idea validated with
5 users, 1 advisor, and 2 designers from Uber team
flow 1/
swipe through dishes
flow 2/
invite collaboration
flow 3/
shopping cart & check out
05
iteration / low-fi wireframes
i built out all wireframes
and have them tested with 5 target users
gathered 50+ feedbacks via user testing sessions and post-interview survey
5 major changes were made
problem 1 - lack of guidance/ During testing across age groups, we found that older users—an important segment for food delivery—struggled to understand the dating-app-style swipe interaction. Without guidance or context, they hesitated to explore the feature and distrusted the relevance of the recommendations.
solution - add an onboarding flow/ To reduce entry friction and build user trust, I designed an onboarding flow that explains the swipe mechanism and collects food preferences and dietary needs. This not only helped users get started more confidently, but also allowed us to deliver more personalized and meaningful recommendations from the first interaction.
problem 2 – lack of clarity in collaboration/ While the collaboration feature allowed users to swipe with household members, users couldn’t tell who was actively participating or see their dietary preferences (1 interviewee mentioned this needs)— limiting meaningful coordination and the value of the shared experience.
solution – add collaboration visibility/ To enhance transparency and decision-making, I added a live status bar showing which collaborators were active. Users can now view dietary preferences and restrictions, making it easier to communicate and choose meals that accommodate everyone.
problem 3 – lack of organization & utility/ Initially, we offered a single swipe list, but over time it became overwhelming and confusing. Users couldn’t distinguish between one-time likes and dishes they had actually shopped for. There was no way to filter, revisit previous bundle buys, or build a personalized collection of favorite meals.
solution – separated swipe history from personal dishes/ We introduced two lists that have different focuses, and both with filter, quick access to history, and "add new" function.
problem 4 – Ingredient shopping disconnect/ Our initial design didn’t account for the gap between recipe ingredients and real-world store listings. Ingredient names often didn’t match item names in stores, and users had to manually search and shop each item across different locations—creating friction and inefficiency.
solution – Integrated smart shopping workflow/ We redesigned the shopping experience to support real ingredient-to-product matching. Items can be searched, substituted, and added from multiple stores—all without leaving the page. This streamlined the shopping process and made bundle buying far more efficient and flexible.
problem 5 – users have different mental models for reviewing carts/ We thought some users prefer viewing cart items grouped by store, while others want to see them organized by dish—how they interacted with meals in the app. We need to test it out before moving forward.
balancing user preference and business constraints/ We tested three different versions of shopping cart. Our testing showed a clear preference for version 2 - two views shown as two tabs. However, during a design review with the Uber design team, they emphasized the need to control development complexity and engineering cost.
solution – one cart with toggleable views/ We introduced a lightweight filter toggle that lets users switch between viewing items by store or by dish—all within one cart. It maintained the clarity users liked, without adding development overhead.
are we ready for the next round of user testing?
YES!
06
iteration / mid-fi wireframes
i updated all the wireframes
and have them tested with 5 non-ux trained participants and 2 ux designers
gathered 45+ feedbacks via user testing sessions and post-interview survey
2 changes were made
problem 1 - new user has potential high drop-off due to lengthy onboarding/ The original four-step onboarding caused friction—especially for older users unfamiliar with swiping-based interfaces. Many users left before reaching the core experience.
solution - reduced steps, contextual guidance post- engagement/ We streamlined onboarding into two focused screens, letting users jump into swiping sooner. Hints for filtering and collaboration now appear after initial engagement, keeping the entry lightweight while still enabling personalization.
problem 2 – returning users faced friction before re-engaging / In the original design, returning users were required to reselect preferences and invite collaborators before they could access the core swiping experience. This interrupted their flow and added unnecessary steps that weren’t always relevant for repeat visits.
solution – prioritized core interaction, moved extras downstream/ We redesigned the flow so returning users now land directly on the swiping page, allowing them to jump back in quickly. Filters and co-swiping options remain accessible, but are surfaced contextually—only when users need them. This reduced friction and improved re-engagement.
let's see how we have improved our design!
To evaluate usability improvements across iterations, we distributed the System Usability Scale (SUS) questionnaire to participants after each testing session. In total, we collected 20 complete responses. All scores reported represent the average ratings from these participants. This approach allowed us to track how each round of design changes impacted perceived usability across key dimensions such as ease of use, complexity, and consistency.
concept 0.0
83.6%
overall SUS score
final design
30.6%
ease of use
21.9%
reduced complexity
45.2%
consistency
07
design system & UI portfolio
In designing this product, I adhered closely to Uber Eats’ design system to ensure consistency, scalability, and visual coherence. I utilized Uber Move typography across all screens with appropriate sizing, followed the official brand color palette, and built a system of reusable components.
Beyond visual fidelity, I maintained a modular structure aligned with atomic design principles, which allows for easy updates and consistent user experience. I also referenced official Uber design documentation and collaborated with teammates to ensure the system could scale efficiently across different features and states.
All referenced official Uber Eats guidelines.
looking back to our problem,
pure design is not our solution….
08
promotional strategy & visual assets
To support the adoption of our newly designed grocery shopping feature, we treated the feature itself not only as a tool but also as a strategic marketing touchpoint. Based on findings from our competitive analysis—especially a deep dive into super app strategies like Didi—we identified in-app promotion as a key driver for activation.
We proposed three in-app marketing strategies to complement the feature launch. These strategies focus on visibility, engagement, and timely nudges, helping users discover and return to the experience from various touchpoints within the Uber ecosystem.
strategy 1/
cross-functional entry point
Enable discovery by surfacing Swipe it! across multiple parts of the Uber ecosystem (e.g., Rides, Eats, Grocery), turning it into a familiar service touchpoint.
strategy 2/
in-app nudging banner
Show contextual banners to re-engage users who are passively browsing, nudging them to try grocery shopping or complete relevant actions.
strategy 3/
holiday & seasonal campaigns
Create seasonal banners (e.g. Easter, Halloween) to tap into festive shopping needs and inspire timely usage.
END
reflection
🤝 working with industry stakeholders/ We collaborated closely with Uber’s design and product teams. I learned how to align user needs with business priorities, navigate trade-offs, and clearly communicate design rationale.
🧩 solving real problems/ The project went beyond interface design—we tackled real adoption challenges through user research, feature design, and integrated in-app marketing strategies.
🔍 competitive analysis pays off/ Our in-depth study of super apps like Didi and Meituan directly inspired effective promotion strategies. It was time well spent.
🔄 designing flows in detail/ I worked across full user flows—from onboarding to checkout—balancing usability, functionality, and engineering feasibility while iterating through user testing.
✏️ designing within a system/ Adhering to Uber’s design system pushed me to be detail-oriented with layout, components, and branding. I strengthened my ability to design consistently at scale.