ADD A FEATURE
Yelp
ROLE
DURATION
TOOLKIT
Background
Yelp connects users to businesses such as restaurants, services, and salons. As a social platform, it allows users to share reviews, photos, ratings, and their experience with a local restaurant that gives the community insight on what to expect and prepare them for their visit. People can be more confident in their decisions and discover new spots because they have easy access to a business’ information. I chose to add a feature to Yelp because I am a foodie who understands the difficulty in agreeing on a place to eat at when planning with other people. I wanted to design a feature that makes the process a little easier and more fun for the group.
Problem
Yelp has limited search filters and currently presents recommendations in a way that may be overwhelming for groups trying to find restaurants to eat at, contributing to decision fatigue.
Solution
Reduce decision fatigue when selecting a restaurant in a group by making the process more efficient through collaborative decision-making, personalization, and AI-driven recommendations.
COMPETITOR ANALYSIS
I examined four restaurant finder services that use AI-driven recommendations to evaluate their strengths and weaknesses. This analysis provided insight into existing market solutions, highlighting a lack of a strong collaborative feature and opportunities for improvement.
USER INTERVIEWS
I conducted interviews with 5 individuals who have used Yelp to find restaurants, with two that have a preference towards using Google Maps. They all provided valuable insight on qualities they look for in restaurants, communication about dining options amongst a group, and their opinions about AI recommendations.
Key Findings
Many wanted Yelp’s current search features to be more intuitive: Disliked limited filter options, making it difficult to search for specific preferences, and they experienced scrolling fatigue when reviewing the suggested restaurants.
Main pain point when deciding in a group is a conflict in preferences: Requires compromise and it overall causes it to take longer to reach a decision.
Interest in a shared list where all users can add restaurant options: Typically collaborate by directly sending Yelp/Google links in a third party group chat.
Viewed the usefulness of AI recommendations on a spectrum: Certain factors such as how responsive the AI is, how often they interact with the platform, and the time commitment to learn if they like the suggested item or not can determine how much they trust or rely on AI.
Organizing common themes into an affinity map:
Need more advanced filters, Group preference conflicts, Ability to compare restaurant options, Interest in AI-driven recommendations
USER PERSONAS
The research yielded a diverse set of user needs when choosing a restaurant in a group that inspired me to focus my design around 3 user personas. This approach helps ensure an inclusive experience that accommodates different decision-making styles such as someone who has specific food preferences, enjoys taking the lead in planning, or prefers to go with the flow.
FLOWCHARTS
Participants expressed interest in AI-driven recommendations and found a shared list the most helpful. In response, I designed user flows to map out the possible pathways users take when creating a group collection and generating restaurant recommendations. The goal was to create a seamless, intuitive experience that simplifies the decision-making process.
KEY
Shared Collection: Create a Group Collection
Recommendations: Generate AI-driven recommendations
WIREFRAMES
I translated the user flows and sketches into digital wireframes to explore the layout and structure of the key group features. I conducted early user testing and received feedback on pain points such as confusion on how to invite members and more clarity on the shared list section’s functionality.
Building all low-fi wireframes.
WIREFRAMES
I recreated Yelp’s established style guide, researching their iconic branding. I applied Yelp’s UI elements to my lowfi wireframes to closely match their current aesthetics.
USABILITY TESTS
I conducted usability tests with 5 Yelp users who have experience deciding on a restaurant in a group or a pair. The study aimed to identify potential areas of improvements in the designs.
Performed 4 Tasks
#1
Create a Group Collection
#2
Invite members to created Collection
#3
Explore Group Collection with added restaurants
#4
Get Yelp Recommendations for the group
Key Findings
100% success rate in creating a Group Collection and inviting members due to clear labeling and iconography.
Many felt overwhelmed by how the comments were all listed below the restaurant, making them get lost in the comments and decreasing the readability of the page, especially when there was multiple comments.
Many wished there was a way to pin certain restaurant options and have better organization of the restaurant listings.
Most found the Recommendations interaction the most enjoyable and preferred an easy and fun interaction such as using the react buttons. Many found the AI-generated summaries to be helpful with the bolded keywords and concise details; not overloaded with information.
ITERATIONS
These iterations helped shape a more user-centered design, ensuring the app remains intuitive, engaging, visually-appealing and valuable for groups deciding on a place to eat.
FINAL PROTOTYPE
The final prototype integrates key improvements, polishing up the overall experience and interface.
REFLECTION
Lessons Learned
Designing within Brand Constraints
This project taught me to design within the constraints of an established brand. I compromised the interactivity of certain designs in order to stay consistent with Yelp’s already offered features. I learned the importance of understanding a company’s brand identity and getting familiar with its platform’s current features so that my designs and ideas seamlessly blend in.
Improved Prototyping
I gained more hands on experience and practice in building prototypes in Figma. I discovered new ways to make it more interactive and accurately portray the functionality or intentions of the features. This allowed participants to engage with the interface more realistically, leading to more insightful and meaningful feedback.
Next Steps
Group Match Percentage
Users needed more clarification on how the Group Match was calculated. Provide a visual breakdown of the factors that make up the percentage.
Refining AI Recs
Interest in AI-generated summaries of restaurants and what would be included such as dishes places is known for or key info in top reviews.
User Testing
Experiment with different Recommendation interactions such a swiping feature or side by side comparisons to see what resonates best with users.
Check out the next Case Study!
BACK TO TOP