Beli / Beyond Reviews, Between Friends

Beli / Beyond Reviews,
Between Friends

Beli / Beyond Reviews,
Between Friends

ROLE

UX Researcher

TIMELINE

8 weeks

TEAM

Catie R.

Valarie T.

Emma W.

Octrina Y.

TOOLS

Figma

INTRODUCTION

WHAT IS THE PRODUCT AREA AND HOW DO THESE PRODUCTS HELP PEOPLE?

The product area we are focusing on is Food, specifically review and restaurant discovery platforms. Products in this category enhance decision-making efficiency by aggregating ratings, reviews, and photos into a convenient format, so users can quickly compare options and choose where to dine. They often cater to a wide range of budget, dietary, and location needs, making it easier for users to find restaurants that match their preferences. Additionally, they can help people discover places they might not have found on their own.

PROTO PERSONAS

Early proto personas outlining our assumed key user types, goals, and pain points before formal user research.

USABILITY RESEARCH REPORT #1

How do users navigate a Beli to discover and review restaurants that match their preferences? What information influences their decisions?

WHAT IS BELI?

Beli is a social restaurant app for tracking, ranking, and sharing places you’ve eaten. It focuses more on personal taste and friend recommendations rather than traditional 5 star reviews.

Reviews are shared privately between your friends, and after leaving a review, you compare and rank that restaurant with other places you’ve been to, generating an overall score out of 10.

Objectives

  1. Evaluate how users navigate the Beli app to discover restaurants

  1. Identify what kind of information influences users’ restaurant selection

  1. Assess whether users are able to successfully complete their tasks in the app

  1. Identify problem areas in the UX flow to improve the app’s usability

Methodology

Each participant was completely new to Beli. We used a pre-existing account on one of our phones in a classroom setting. Pariticipants voiced their thoughts aloud as they went through the app.


Task 1: Find a food spot that you’d be interested going to. Add/save it to a list.
Task 2: Find a restaurant/food place that you recently went to and rate it.

Each participant was completely new to Beli. We used a pre-existing account on one of our phones in a classroom setting. Pariticipants were told to voice their thoughts aloud as they went through the app.

PRIOR TO THE TASKS

We asked questions regarding how participants go about finding food recommendations and their experience.

DURING THE TASKS

  • One moderator instructing the participant

  • Two observers/note takers

  • 1 Participant (4 total) testing the app

AFTER THE TASKS

We asked participants for their thoughts about their overall impressions, likes, dislikes, and suggestions for improvement.

Analyzing Qualitative Data

Afterwards, we analyzed the qualitative data we took through affinity diagramming:


  1. The clusters created by grouping our sticky notes into 4 categories: (initial questions, task 1, tast 2, post-task questions)

  2. Grouped highlights and next steps within those groups.

  3. Labeled clusters with code, hashtags, and stickers

  4. Summarized of our data into primary draft insights


Finally, we prioritized our draft insights using the 2x2 matrix, considering low/high cost and low/high user value.

Limitations

SAMPLE SIZE + SHORT RESEARCH TIME

Short research time resulted in some features not tested at all. Additionally, the small sample size resulted in limited insight about the features that were tested

FIRST TIME USERS

All of our participants were first time users, so feedback was more reflective of the initial impression of the app rather than long-term use.

FULL ACCESS

Participants were able to use an app that had full access rather than a user who first signs up.

Key Results

THE GOOD

→ Social Aspect

  • Ability to tag friends you ate with in the ratings

→ Colored Ratings

  • Color distinction made it easier to figure out the ratings

THE CONFUSING

→ Rating System

  • Participants were thrown off by the 10/10 rating

→ Icon Mix-Up

  • Participants struggled to find key features due to unclear icons

THE BAD

→ Feature Location

  • Figuring out where the features are caused great confusion

→ Visuals

  • Lack of photos made researching for restaurants harder

Reccomendations

HIGHER PRIORITY

  • Differentiating searches between restaurants and cuisines/dishes

  • Changing the review button to a more recognizable icon

LESSER PRIORITY

  • Toggle option to see photos when searching up restaurants

  • Toggle between a stars and numbers rating

  • Making the map interactive rather than a still photo

  • Make the view map button bigger for higher visual prominence

SURVEY RESULTS

Overview of Strategy

What is the overall behavior surrounding food review and its apps? What could we continue to research deeper on?

OBJECTIVES

  • Find out people’s experiences with food reviews

  • How (online, friends, factors) people find places to eat

  • What specific features in the app would be preferred to discover places

APPROACH

  • Sent out to our class, mainly college students

  • Asked questions with the fact that many people haven’t used Beli before in mind

Summary of Findings

  • 40% of responses trust online ratings

  • 60% of responses trust recommendations through friends/word of mouth

  • Most people find food reviews to be helpful

  • Top three factors when looking for place to eat: Price, Location, Overall Rating

  • Top Features used to discover restaurants: Search Bar, Trending

  • Most people never leave food reviews after visiting a restaurant

Suggestions

Since most people don’t leave reviews, Beli could add an incentive for users to write them. Rather than making review writing seem like a chore, Beli could highlight review writing as an fun experience to do with you and your friends—emphasizing on the fact reviews private and only shown to your friends, unlike Yelp.

Survey Improvements

  • The first question should have been multiple choice with numbers, making it a short answer reply makes the data hard to analyze and understand.

  • The last question “Elaborate why or why not” (that users do not write reviews) is very open ended and led to many unspecific/short answers, that are hard to support their yes/no.

USABILITY REPORT #2

Participant Persona

DISCOVERY:

  • Friends

  • TikTok

  • Instagram

  • Google reviews

  • Yelp

HABITS:

  • Rarely leaves reviews, only in some extreme chances

  • Prefer apps they are already familiar with

  • Ignore features that add additional task work

PAIN POINTS:

  • Don’t see the value in leaving reviews

  • Finds the process of leaving reviews too time consuming

  • No consistent incentive to rate places regularly

Methodology

We conducted competitive evaluations through usability tests with Yelp, one of the most widely recognized and commonly used review platforms. Was there something on Yelp that encourages people to leave reviews?


We asked our participants to perform the same task on each app.

Task: Find a restaurant/food place that you recently went to and rate it.

We conducted competitive evaluations through usability tests with Yelp, one of the most widely recognized and commonly used review platforms. Was there something on Yelp that encourages people to leave reviews? We asked our participants to perform the same task on each app. Task: Find a restaurant/food place that you recently went to and rate it.

  • 4 sessions, 2 starting with Yelp, 2 starting with Beli, to mitigate bias.

  • 1 moderator who asked the participant questions and explained instructions.

  • 2 observers/note takers, who silently watched the participant behind their shoulder and took notes.

  • Screen recording enabled to allow us to review participants’ actions and capture relevant screenshots afterward.

Limitations

USING A PRE-EXISTING ACCOUNT

  • We did not have time to create a new account on Beli for every session.

    • Many features aren’t available for new users.

    • Users had to compare/rank restaurants that they hadn’t been to.

    • Viewed friend rankings they had no personal connections to.

EVEN NUMBER OF SESSIONS

Sometimes we got splitting results (50/50 preferences) with only having an even number of sessions.

Analyzing Qualitative Data

Afterwards, we analyzed the qualitative data we took through affinity diagramming:


  1. The clusters created by grouping our sticky notes into 3 categories:

    Yelp, Beli, Post-task questions

  2. Labeled clusters with code, hashtags, and stickers

  3. Summarized of our data into primary draft insights


Finally, we prioritized our draft insights using the 2x2 matrix, considering low/high cost and low/high user value.

We conducted competitive evaluations through usability tests with Yelp, one of the most widely recognized and commonly used review platforms. Was there something on Yelp that encourages people to leave reviews? We asked our participants to perform the same task on each app. Task: Find a restaurant/food place that you recently went to and rate it.

Leaving Reviews on Yelp

Leaving Reviews on Beli

Key Findings

4/4 USERS MENTIONED MONETARY VALUE AS INCENTIVIZATION

  • Discounts and gift cards took biggest priority to incentivize users to leave reviews.

  • One user who rarely leaves reviews recalls the last review on Yelp was due to the restaurant offering a discount.

  • “Unless discounts or gift cards are awarded, what’s the point?”

3/4 USERS MENTIONED TOXICITY ON YELP REVIEWS

  • 2/4 users specifically mentioned prioritizing supporting small businesses—people use Yelp to bring businesses down.

  • “I’ve seen people bring restaurant ratings down because they have a personal vendetta.”

  • “People are just complaining on Yelp all the time.”

3/4 USERS SAID THEY ARE MORE DRAWN TO BELI OVER YELP

  • Beli feels “more personal” and “trustworthy.”

  • Social/friends feature incentivizes them.

  • “If I had a bunch of friends on here.”

  • “Sometimes I don’t care about randoms people’s opinions.”

“Friends score is a super good metric that I would pay attention to”

Actionable Reccomendations

Rebranding and shifting perspective of Beli from a food review app (like Yelp) into a social “food diary” or “restaurant log” app that emphasizes personal tracking, shared memories, and hangouts with friends. 

Rebranding and shifting perspective of Beli from a food review app (like Yelp) into a social “food diary” or “restaurant log” app that emphasizes personal tracking, shared memories, and hangouts with friends. 

  • Yelp reviews are sometimes incentivized by monetary rewards, making them less trustworthy; Beli’s friend-based model creates more authentic, reliable recommendations.

  • Yelp is built around stranger reviews and broad public discovery, while Beli’s strength is friend-only sharing, personal rankings, and memory tracking.

  • Beli should lean into personal restaurant history, social connection, and trusted recommendations from people you actually know.

    • Users are logging places for themselves and their friends, not writing formal reviews for strangers.

  • Yelp reviews are sometimes incentivized by monetary rewards, making them less trustworthy; Beli’s friend-based model creates more authentic, reliable recommendations.

  • Yelp is built around stranger reviews and broad public discovery, while Beli’s strength is friend-only sharing, personal rankings, and memory tracking.

  • Beli should lean into personal restaurant history, social connection, and trusted recommendations from people you actually know.

    • Users are logging places for themselves and their friends, not writing formal reviews for strangers.

REFLECTION

Our Research Journey

WHAT WE SET OUT TO LEARN

How can we improve Beli so people are more inclined to leave reviews over Yelp?

WHAT WE INITIALLY FOCUSED ON

Improving the review process (i.e icon change, rating system, feature locations) and overall app experience.

WHERE OUR THINKING SHIFTED

People weren’t avoiding leaving reviews because of the app, there was no clear incentive for leaving reviews in the long run.

“Why would I switch to Beli if Yelp gives me everything I want?”

The Real Insight

When we introduced Beli’s social feature, seeing where friends ate and their review, participants were far more willing to engage. Beli’s greatest asset is the trust that already exists between friends.

INTERESTED IN LEARNING MORE?

INTERESTED IN LEARNING MORE?

I'm happy to chat about my process, reach out through email :)

Let's be friends!

Or… lettuce… ketchup? :)

Let's be friends!

Or… lettuce…

ketchup? :)

© 2026 EMMA WEI

© 2026 EMMA WEI

BUILT WITH <3 IN FRAMER

BUILT WITH <3 IN FRAMER