Discovery Flow
A personalized, swipe-based discovery flow that turns preference into meaningful recommendations.
Evaluation & Reviews
A clear, trustworthy evaluation system that helps users make confident decisions.
Booking & Social Coordination
A streamlined booking experience with integrated social features.
Overview
This project began with a familiar problem: tourists are often overwhelmed by endless restaurant options in unfamiliar places. Working as a team of four, we reimagined this moment of decision by turning restaurant discovery into a playful, swipe-based experience inspired by dating apps.
Across the term, I led major stages of the process: secondary research and competitor analysis, eight semi-structured interviews, persona creation, early task flows and sketches, and iterative prototyping from low-fidelity screens to the final interactive prototype.
Discovery & Research
I begin the research with 8 semi-structured interviews. The data revealed three core insights that guided the design direction:
Sometimes I don't understand the menu, so I just won't pick that restaurant.
I saw tons of people on TikTok saying it was amazing, but the reality was disappointing.
I spend so much time comparing reviews that by the time I decide, I'm not even hungry anymore.
Target Users
Based on our research, I created two personas to guide the design direction.
The Overwhelmed Planner. He needs structure, verified information, and efficiency to feel confident booking.
The Confused Communicator. She needs translation support and visual menus to overcome language barriers.
Initial Prototype
Before moving into high fidelity, I mapped the core flows with medium to low fidelity wireframes to test usability and interaction flows.
Usability Testing
With the initial prototype in place, we ran eight moderated, task-based usability sessions with participants who regularly use restaurant apps on their phones. Each session followed a realistic scenario: planning dinner for a small group, from setting preferences to completing a booking.
Using a “think aloud” protocol, we asked participants to verbalize what they were looking for, what they expected to happen next, and where they felt unsure while they moved through key tasks.
Here are some critical feedbacks from user testing:
Unclear feedback for “Like/Dislike” interaction
Many participants were confused about what happened after tapping the heart icon. They expected immediate feedback or a visible “saved” state, and were unsure where the saved restaurants would appear. This suggested the need for clearer affordances and feedback.
Insufficient decision support
Early versions of the prototype provided only minimal restaurant information. Participants hesitated to “Like” or “Pass” because they don't have enough information such as menus, hours, ratings, or location. This validated the need for a richer, more informative detail view before swiping decisions.
The need for integrated social features
Participants mentioned wanting to “check with friends first” before making a reservation. This suggested the importance of lightweight social tools such as sharing, sending recommendations within the app.
Demand for Predictive Recommendations
Instead of repeatedly searching with manually inputted preferences, participants expect the app to learn from their preferences and provides personalized recommendations.
The Pivot: Iterating Based on Feedback
Usability testing surfaced three critical gaps in the early prototype. Each one turned into a concrete design pivot that made Foodie Match clearer, more informative, and more aligned with how people actually decide where to eat.
Clear feedback after “Like”
Participants were unsure what happened after tapping the heart icon and whether a restaurant had truly been saved. I added a lightweight confirmation overlay (“Added! Saved to your list”) to provide immediate feedback and reduce uncertainty.
Richer detail page before swiping
Early versions did not show enough information for users to confidently “Like” or “Pass”. I introduced a scrollable detail view with hours, location, rating breakdown, tags, and a gallery so users can understand the restaurant before making a decision.
Lightweight social coordination
Many participants wanted to “check with friends first” before booking. I added a social layer with profile creation, friend selection, and a dedicated share flow (“Send and return”) so users can align choices with friends without leaving the app.
The Final Deliverable
Based on prior insights, this is the final high-fidelity prototype, incorporating three core values:
Personalized Recommendations
Smart suggestions based on location, preferences, dietary needs, and budget.
Trustworthy Review System
Verified reviews with rating breakdowns and translation support to reduce uncertainty.
Engaging Decision Flow
Swipe interaction that presents one restaurant at a time, making choosing feel fun and low-pressure.
Interactive Prototype
Explore the prototype below to experience the swipe and booking flow.
Reflection
My biggest takeaway from this project is understanding how thoughtful UX design can guide users through overwhelming choices and reduce the cognitive load behind decision-making. This project also sharpened my own design approach.
Interactions with participants revealed frustrations, motivations, and decision-making habits I hadn't previously considered. User testing made these insights tangible: direct feedback from users and moments of confusion pinpointed areas needing improvement. Thanks to these invaluable tests, each iteration felt purposeful and yielded tangible results.
In the future, I would love to explore core components such as the recommendation algorithm and further develop Foodie Match into a fully realized, real-world product.