top of page

Beli

Food has always been more than just a meal to me—it’s a love language, a way to connect, and a constant adventure. Growing up, my family was always on the hunt for new flavors and experiences, whether it was returning to our go-to comfort spots, experimenting in the kitchen, or discovering hidden gems in different cities. Some of my favorite memories are tied to food from family celebrations to late-night hot pots/shabu shabu at home. 

 

So when the opportunity came to redesign a food app, I knew I had to dive in. Enter Beli—a vibrant and dynamic food discovery app designed for people like me, who live to eat, explore, and share. Beli is all about building a community of food lovers, making it easy to rate, document, and recommend favorite restaurants. It keeps those dining memories alive while curating personalized recommendations, so the next best meal is always just around the corner.

New and Improved Rating System

Screenshot 2025-02-05 at 5.54.47 PM.png

Problem

The simplistic three-option rating (“I liked it,” “It was fine,” “I didn’t like it”) fails to capture the complexity of dining experiences, and the lack of clarity in the numerical ranking makes it difficult for users to understand how their reviews influence scores. This limits their ability to make informed decisions based on their preferences.

Solution

To solve the challenges we identified, we revamped Beli’s rating system to make it more accurate and reflective of real dining experiences. We introduced five key categories—taste, presentation, service, dietary accommodations, and ambiance—each rated on a 1-5 scale, allowing users to assess their meals more precisely. These individual scores are then averaged to create a comprehensive rating, which influences the rankings in “Your Lists.” Users can also filter restaurants by category, making it easier to find options that match their preferences. With this redesign, we created a more straightforward and engaging experience for Beli’s community.

Screenshot 2025-02-05 at 6.35.20 PM.png

Online Research

Competitive Analysis:
Google Maps
Screenshot 2025-02-05 at 7.10.18 PM.png
  • Target Users: Diners, event planners, and travelers

  • Features: Easy navigation, search and filtering, interactive map, real-time accuracy (hours, busy times), personalized recommendations

  • Interface: Clean, intuitive

  • Limitations:

    • Lacks integrated booking and ordering (relies on third-party sites)

    • Limited offline functionality

  • Overall: Great for discovering restaurants but could improve in streamlining the dining experience.

Door Dash 
Screenshot 2025-02-05 at 7.25.05 PM.png
  • Target Users: Consumers, families, college students, event hosts

  • Features:

    • Food delivery & pickup

    • Customer ratings & tailored recommendations

    • Order scheduling & personalization

  • Strengths:

    • Large platform with many options & intuitive app

    • Personalized recommendations

  • Limitations:

    • Markups & fees

    • Pressure on drivers/restaurants from ratings

    • Negative feedback due to uncontrollable factors (traffic, weather)

Screenshot 2025-02-05 at 7.16.49 PM.png
Yelp
  • Target Users: Consumers, business owners, travelers

  • Features:

    • Reviews & ratings

    • Search & filters

    • Business info & integrated booking/ordering

    • User-generated content (reviews, photos)

  • Strengths:

    • Extensive business database

    • Strong community engagement

    • Credibility & longevity

  • Limitations:

    • Unreliable or biased reviews

    • Cluttered interface with too much information

    • Lack of personalization

  • Competes with Google Maps & food delivery services

Screenshot 2025-02-05 at 11.49.06 PM.png

Digital Prototyping

Sketches & Wireframes 

circle rating.png
User Research

High Fidelity Prototypes - Redesign 1 & 2

Top errors we found when user testing: 
 

Error Prevention

Issue: Confusing Icons for Saving vs. Reviewing During task 1, users struggled to differentiate between the "plus" icon for reviews and the "bookmark" icon for saving restaurants. This caused confusion as users accidentally clicked the wrong icon when trying to save a restaurant for later.

​

Match Between the System and the Real World

Issue: Oversimplified Rating System Users found the three-option rating system ("I liked it," "It was fine," "I didn’t like it") limiting and unrealistic compared to how dining experiences are typically evaluated in the real world. They expressed a desire for more granular ratings, like evaluating taste, presentation, and ambiance.

Solution: We expanded the rating system to include five specific categories: taste, presentation, service, dietary accommodations, and ambiance, with a 1-5 rating scale. This allowed for a more accurate, detailed review process that aligned better with real-world dining assessments.

​

Flexibility and Efficiency of Use

Issue: Cluttered Feed Users found the endless scrolling feed overwhelming, especially when trying to engage with friends' rankings or bookmarks. The feed was cluttered with friends' bookmarks, which led to frustration when trying to find content they wanted to interact with.​​

Post User Testing Redesign 

redesign.png

Our redesign introduces category-specific ratings and a more intuitive filtering system, improving both usability and personalization. These changes align Beli’s ranking system with real-world dining preferences, making restaurant discovery more precise and engaging. With further iteration and testing, Beli can offer an even more seamless and insightful user experience.

Demo Video 

Looking Back

image.png

Reflection & Lessons Learned 

User Testing Drives Meaningful Iteration: The feedback from user testing revealed the need for a more nuanced rating system, including an "Ambience" category and a preference for a single, scrollable rating page. These insights led to critical design changes that better aligned with user expectations.

Collaborative Ideation Leads to Stronger Solutions: Our initial focus was on redesigning the restaurant menu page, but a team whiteboarding session helped us pivot towards improving the rating system, where we saw the most significant user feedback. Collaboration and brainstorming played a crucial role in refining our solution.

Don't be Quick to make Assumptions: We initially believed users would prefer a step-by-step approach to rating, but they favored a single-scroll option instead. This taught us the importance of challenging our assumptions to find better solutions that resonate with users.

next steps .png

Next Steps

While our redesign significantly improves the flexibility and accuracy of Beli’s rating system, there are still opportunities for further iteration:

​

  1. Tailored Rating Categories – Adapt rating criteria for different dining experiences, such as Wi-Fi access for cafés or wait times for food trucks, to provide more relevant insights.

  2. Enhanced Filtering Options – Improve search functionality by allowing users to set default preferences and apply multiple filters simultaneously for a more efficient experience.

  3. Optimized Rating Interface – Explore alternative UI elements like pre-set ratings or emoji-based scoring to further streamline the rating process.

  4. Broader User Testing – Conduct A/B testing with a larger user base to validate and refine design choices.

Acknowledgements :)

Thank you for joining me on this food adventure! Also a huge shoutout to my incredible and passionate team—you all made this journey unforgettable! <3

  • Kaitlyn Butcher

  • Gabriella Strudler

  • Jake Villasenor

  • Marisa Kawakami (yipee that's me :))

beliii.jpeg
bottom of page