Re-designing the Beli app for a more customized restaurant search
TIMELINE
Nov - Dec 2023
PROJECT
Passion Project
ROLE
Solo Designer
SKILLS
Product design, Interaction design, Prototyping, User research
OVERVIEW
PROBLEM
'Popular' doesn't mean 'Perfect for you'
Everyone has different standards and priorities for a restaurant. Since the Beli app's personalization features rely on holistic reviews, it sometimes fails to meet the specific preferences of different users.
TARGET USERS
Restaurant Enthusiasts
Foodies and culinary adventurers
Frequent diners seeking new experiences
Tourists exploring local cuisine
Date night planners
HERE'S HOW BELI WORKS
Meet Collin, a budget-conscious foodie navigating the Beli app for the perfect dining spot. As he explores, we'll uncover Beli's unique features and the challenges he faces:
Rate by Specific Categories
Allows users to rate restaurants based on specific criteria such as price, ambiance, food quality, and service.
Filter by Categories
Enables users to filter restaurant recommendations by specific criteria.
View Recommendations on Map
Provides a map view of tailored recommendations to help users visualize and select restaurants based on their preferences.
RESEARCH
USER INTERVIEW
To understand the needs and expectations of current Beli users, I distributed 20+ surveys and conducted 4 user interviews.
"I don't want to think too much to rate restaurants on Beli"
Yeri Won, 21
Occupation: Studying neuroscience
Sunjae Lee, 24
Occupation: Studying economics
Collin Kim, 22
Occupation: Studying data science
"Beli offers unfair restaurant comparison systems"
SJ Kim, 25
Occupation: Investment Banker
COMPETITIVE ANALYSIS
Unclear data source for recommendations
How trustworthy is the traveler ranking?
Excessive options to filter users' search
The data sources used for the recommendations are not clearly defined.
There are too many filter options, causing confusion for users.
The abundance of sponsored restaurants undermines the trustworthiness of the recommendations and rating system.
PAIN POINTS
One Algorithm Doesn't Fit All
Users distrust generic algorithms that don't reflect personal tastes
Unfair Rating System
Beli's one-click ratings struggle with unfair comparisons between different cuisines.
The Value-Seeking Generation
Core users are budget-conscious young adults seeking value
IDEATION
HMW
(How Might We)
provide a more tailored restaurant suggestions and build trust?
BRAINSTORMING
To kickstart ideation, I brainstormed solutions addressing three key issues: Generic algorithm distrust, unfair rating system, and lack of personalized recommendations.
LOW-FIDELITY WIREFRAMES
Exploring multi-category rating options
To explore the most user-friendly multi-category rating options, I developed three low-fidelity wireframes. I conducted a second round of interviews to determine which method users prefer.
Despite initial concerns about increased clicks, users favored the multi-category rating system, valuing more tailored recommendations over fewer interactions.
Users prefer dropdown menus for ease and speed of use.
FINAL SCREENS
RE-DESIGN #1: RATE BY SPECIFIC CATEGORIES
How can Beli refine its algorithm to better reflect each user's personal taste?
Before
Vague restaurant comparison without clear criteria
After
Detailed ratings across multiple categories for personalized comparisons
RE-DESIGN #2: FILTER BY SPECIFIC CATEGORIES
How can users search for restaurants based on their specific needs?
Before
Generic recommendations without transparent criteria
After
Customized search results based on user-selected rating categories
RE-DESIGN #3: VIEW RECOMMENDATIONS ON MAP
How can users more easily view information about their recommended restaurants on a map?
Before
Limited information in map view
After
Interactive map with easy-to-access restaurant details
RETROSPECTIVE
KEY LEARNINGS
Working within existing design systems is more complex than creating from scratch, requiring meticulous attention to detail.
Presenting lesser-known apps poses challenges in audience comprehension.
NEXT STEPS
Further improve alignment with the original design system.
Implement detailed micro-interactions, like map pin highlights.
VIEW MORE PROJECTS
