Redesigning Yelp

February 2021 - July 2021

Overview

For my UX Design Certificate program at Cornell University. I was tasked to research a problem space and develop design solutions based off the information collected. Everything culminated into a redesign of Yelp.

My Role

This was a solo student project.

The Problem

There is no single trustworthy place to discover new and fun activities to do.

The Solution

Instead of creating a new app, redesign Yelp's existing trusted database of user-reviewed information to improve content discovery.

Trusted Discovery Feed

Browse and search content backed by real-life user reviews and ratings. Now with infinite scroll.

Something for Everyone

More categories, moods, filters and personalized recommendations to relate to and choose from.

Explore Anywhere

See what's happening in other places by simply changing the location.

Step 1: Finding a Problem

The Problem

My first task was to find a problem space. I recalled every time I've tried to lookup something new or fun to do, I've come up short or left somewhat dissatisfied. I decided to test this problem for my project and discovered...

I had to use multiple apps in conjunction to both find and facilitate a decision, which felt inefficient. Ultimately, there were no reliable app or website to discover new/fun things to do.

Step 2: Researching the Problem

Research Methods

Affinity Diagram

Step 3: Analyzing the Research Findings

What I Found

  1. The problem is real.
    No participant knew any direct tools to efficiently solve the problem. They all mainly used search-engine based apps, but felt this was slow as they had to search something first, rather than have those 'somethings' laid out in front of them.
  2. Multiple apps and websites were required to solve the problem.
    Participants verified the problem space by acknowledging the inefficiencies and frustrations in having to use certain applications to discover experiences, and others to verify their legitimacy in conjunction.
  3. Google was most used, Yelp was most trusted.
    All participants mainly used Google/Google Reviews to search for and verify results. Yelp was mainly used to verify food-related content, but was trusted more overall than Google.
  4. Ratings, reviews and pictures sourced the most trust.
    Participants did not fully trust the internet's results 100% of the time, and thus further searched for ratings, reviews and pictures to add credibility to their initial result(s).
User Persona
User Journey Map (Virtual Tour)

Setting User Requirements

After analyzing the data, I created a set of user requirements the solution should fulfill:

Step 4: Finding a Solution

Here's What I Noticed

  • Reviews were the most influential step in the user journey.
    After searching and discovering content, all participants had to search for complementing reviews/ratings via Google or Yelp to build trust in their initial searches to finalize decisions.
  • 66% of participants trusted Yelp more than Google Reviews.
    Although both were trusted, participants felt Google overall was a site designed to spread its focus further out across different areas, whereas Yelp was specialized in its ratings/reviews.

So...

  • Make the most influential step - the only step.
    The process of searching for something then searching again to verify the initial search is inefficient. Since everyone went to Google Reviews/Yelp anyway, these platforms should naturally promote content so it's the only place users would go in the first place.

Step 5: The Solution

Redesign parts of Yelp to promote content discovery.

Unlike Google, Yelp wasn't a strictly search-engine based application. With a varied database of real-world information backed by trusted ratings, reviews and pictures, I believed rearranging certain information within the app could solve the problem.

Also, participants heavily vocalized their superior trust in Yelp's ratings over others as well as its potential to solve the actual problem. Before designing, I took inspiration from other apps with successful discovery feeds such as Instagram and Postmates.
Storyboarding & Sketching
Lo-Fi Paper Prototype
UI Sketches
Medium-Fidelity Prototype

Final Changes for Hi-Fi Prototype

  • Focus on redesigning the homepage.
    Due to timeframe and feedback, I opted to focus the redesign mainly to the homepage, as the homepage would be where users spend the most time discovering/interacting.
  • Discover and search simultaneously.
    I ran a heuristic evaluation and found the two functions should be combined as the discovery feed already has a search function within it.
  • More pictures.
    Feedback suggested to keep the three-picture layout Yelp used by default, but stressed how they enjoyed the minimalistic visual aesthetics of apps like Instagram.
Main Homepage, Sub-categories
Promoting Events, Infinite Scroll Feed
Changing Location, Updated Feed
Categories, Filters

Step 6: Usability Testing

Key Findings

I recruited a former participant and gave them two contextual scenarios to complete the test (i.e. you're on lunch break trying to find something fun to do for the weekend). My key findings were:

Next Steps

Step 7: Lessons Learned

Contact