Yelp Collections

Improving Yelp’s Collections to help users navigate maps more easily and rediscover businesses in categories that drive a third of Yelp’s ad revenue

Note: This project is not officially affiliated with Yelp. Thoughts and proposed modifications of the app are my own.

Project Deliverable
6 new sub-features and improving UI of Collections on Yelp’s iOS mobile app

Timeline
4 weeks

Role/Skills
Sole Designer:
UX/UI Design, Prototyping, Testing, Research, Working Within Existing Design Systems, Product Strategy, Feature Prioritization

Jump to Section

The Problem

It’s Difficult to Rediscover Businesses After Saving to Yelp Collections

Users save restaurants and businesses to Collections but rarely visit them because the current map interface makes it difficult to rediscover and choose from saved places.

This leads to decreased engagement and lost ad revenue.

A current screenshot shows this issue: it currently does not provide any open/closed status nor does it show any custom user notes.

Why is this important?

As of December 2024, roughly one-third of Yelp’s ad revenue comes from “Restaurants, retail, and other.” While they’ve recently released AI features to help with personalization, a community-driven platform will always benefit from a greater sense of human interaction.

After all, Yelp’s mission is to “connect people with great local businesses.”

The Solution

The Key: Put Information Where It’s Expected

Solution #1: Increase Visibility of Information

All of the attributes of a business that are important to a user already exist on Yelp but are all in different parts of the app. Our solution brings these disjointed features together for a seamless and personalized user experience. The major features and how they stack up against the current app are shown below:

The key screens of the solution are above, each with information added that improves the usability of the app!

Solution #2: Enhance the Map View

By enhancing the map view, users are able to get details of places they want to try in neighborhoods they want to be in.

Overview of improvements & new features on the map view:

Map and List view together
Seeing All Collections/Bookmarks altogether
Custom map pins
Filtering and searching
“Share Business” button
Check-in history
Today’s Open hours
The Collections a business belongs to

Now that we’ve seen the biggest improvements to the app, let’s walk back to the beginning of the process and see how we got here…

How do our users browse?

Conducting Research

Competitive Analysis

Competitive Landscape

What are competitors in the restaurant bookmarking space doing?

User Interviews

Business Discovery

How are users finding spots to eat, and what makes them save a place?

Bookmark Interaction

How do users visit saved spots and how do they organize their lists?

Pain Points

What issues and roadblocks do users face when it comes to using bookmarks?

Yelp is Great for Community Photos, But Lags Behind with the Map

Popular products that offer ways for users to connect with local businesses and restaurants were evaluated in the competitive analysis. While the biggest competitor to Yelp is/was Google Maps, it was still important to look at other types of products to understand their unique value propositions and see where else Yelp could improve.

Where Yelp Shines

Word of mouth and feeling of community
More foodie-type critiques and reviews
High quality photos of food with descriptions
Reservations and communicating with businesses

Opportunities for Improvement

Map and list view on the same mobile screen
Customization of notes and different types of lists
Directions and navigation

Users Always Start Searches With a Map

Nine participants were interviewed about the way they use saved lists to browse and interact with local restaurants and businesses. Participants were of all genders, in their late 20s to early 30s, lived in major metropolitan areas, and used Yelp and/or Google Maps to browse new restaurants 1-3+ times per week.

Questions that users were asked centered around Saving (searching, browsing, saving) new businesses, primarily restaurants, and Visiting saved businesses (choosing from their list, navigating, using the app once there).

The map view is immensely helpful when searching for places to go
Notes are helpful for people to use and remember why they wanted to go places
In an age of paid social media influencers and reviews written by bots, word of mouth and recommendations from friends have the biggest influence on whether to visit a business

Confirmed Assumptions

Even though navigating is the last step when going to the restaurant, users always started their search with the map view, regardless of the platform they used (Yelp, Google, or Apple Maps)
Google users tended to use both the Want to Go and Favorites saved lists, Yelp users tended to add to My Bookmarks but not revisit the list
Many users were overwhelmed by the number of businesses they had saved and would want to stay organized, but also did not know the Notes feature (on either Google and Yelp) existed

New Insights

Our Main Persona, A Big, BIG Foodie

Using the information gathered in user interviews, the main persona chosen for this project was someone who loves knowing all the local spots and is the go-to person in her community when it comes to recommendations. This person wants to use Yelp as a tool to keep exploring the city and try new spots, but she gets frustrated when she can’t easily access places she’s previously saved on the app.

Defining the opportunity

How Might We…

make it easier for Yelp users to rediscover and visit restaurants they've saved in their Collections?

Gathering Ideas and Making it Real

Developing a Solution

Usability Audit of the Current App: Too Many Taps!

I looked at a common task to understand the flow of the current Yelp app across several different screens.

The goal: Move a saved business from My Bookmarks (the default collection from Yelp) into Favorites (a newly created Collection) after verifying it’s the right business they want to move.

All in all, it takes 5 steps to confirm the business is the right one they are looking for and move it to a new collection.

Furthermore, the amount of information that is missing about a business is inconsistent between the list and map views, causing confusion to the user.

Design Inspiration From Travel Apps

After identifying inconsistencies and areas for improvement in the earlier task flow, I looked at what some of the competitors from the earlier analysis were doing differently and see what might inspire some feature improvements to the Yelp app. Two apps that had features I could learn from were Google Maps and Tripadvisor.

For both apps, the map was a key player in helping a user organize their saved lists. On Google Maps, which originated as a wayfinding and navigation tool, a user can easily get directions to a saved destination. On Tripadvisor, someone planning out their travels only has a limited amount of time at their destination and will likely want to understand what’s closest to them by distance.

Both Google Maps and Tripadvisor showed the map and a scrollable item view together on the same screen, and with different pin icons, prevents the user from getting fatigued from too many of the same pins to look at.

Focusing on the User Needs

Based on user feedback, it was essential to design expected features for using the map and incorporate a social/word of mouth element into the user flow.

Overcoming Challenges During Development

New features drew upon many existing features, which led to a desire to redo increasing parts of the app UI

Challenge #1: Feature Creep

To overcome feature creep, it was important to just focus on the main user task flows as well as using a prioritization matrix throughout the design process.

Creating visual hierarchy in limited real estate was difficult, with so much information competing for attention.

Challenge #2: Information Overload

To overcome information overload, it was essential to use icons and graphics when possible to guide and inform the user without cluttering the screen.

Prioritizing Features

All potential product features were sorted using an effort/impact prioritization matrix. The color coding in the feature matrix shows the status of each feature, including those that were evaluated but not built out due to the level of effort required.

Prototyping

Leveraging the Existing App to Speed Up Workflow

There were four main task flows to build out in this project that centered around the main use case of interacting with Collections and saved businesses.

The main user flow: A user is interacting with their bookmarked businesses and either deciding where to go next or reorganizing their Collections

Since these tasks built on what was already existing in the app, such as opening up the Collections tab from the home page or going to a business page to move it to a different Collection, I used my limited time and resources to focus on where the most impact would be gained: the map.

To speed up workflow, screenshots of the existing app were used whenever possible. I was able to get quick feedback on several map options just due to collaging app screenshots together and making slight tweaks to the UI. They did not need to be pixel perfect but already looked higher-fidelity just due to the fact that they were taken from the existing app. One of the constraints on this project was to maintain branding and UI, and the best way to do that was by using components that already existed. For new components that were created, I referenced Yelp’s style guide (or as they call it, the Yelp Cookbook) and looked through every screen on the app to ensure colors, typefaces, and icons were consistent with the current UI.

Testing & Revisions

Observing Natural Behaviors in Unmoderated Tests

Usability Testing

To evaluate the flows, I ran an unmoderated usability test with 10 participants via Maze. The setup was made for a desktop experience (to view the prototype and instructions side-by-side) but the prototype experience was designed for mobile.

Users were given 4 main tasks to complete, and I tracked number of clicks (rather than time to completion) for measuring success.

Using Maze’s generated click heatmaps, I was able to track how users interacted with the prototype, including in areas that were not part clickable or part of the task flows. This helped understand users’ expectations as well as expose limitations of an unmoderated test with a static prototype.

This also lowered the success metric score (67%) for the second screen of the Filtering by Collection on Map view task.

Overall, there were six main new features tested. These did not previously exist on the Yelp app at the time of the project.

At the end of the test, were asked to rank the usefulness of the features. The top 3 were:

C. Seeing the map view and list view on the same screen
A. Viewing all bookmarks/multiple Collections at once
E. Details of past activity (at that business) by you and/or your friends (Social Feed)

The Social Feed also consistently ranked as the most confusing feature — what it was, where to find it, and what it meant.

The top 3 rated features were also the same screens in the task flows that got <75% Pass in the success metrics. This made it straightforward to focus on what needed iterating.

Final Revisions

Before & After: Drawing Attention to the Map View

Before & After: Anticipating User Behavior for Filtering

Before & After: Labeling UI Elements to Reduce Confusion

Before & After: Getting People to the Social Feed

Overall Improvements

Overall: A Big Upgrade in UX with Subtle Visual Changes

The final solution was one that seamlessly fit into the existing user flows on the Yelp app.

The biggest changes (and in my opinion, the biggest improvements!) are noted under green in the final prototype key screens below.

Key Screens: Current Yelp App

Key Screens: Final Prototype

To see all the improved key screens in action, take a look at the interactive prototype and try it for yourself below!

A Look back on the Journey

Takeaways & Next Steps

Reflections

What Went Well

A lot went well in this project. First off, I had a lot of fun doing it. I use Yelp several times a week and I’ve had so many moments where I wished these features existed, so it was enjoyable to create a prototype where they finally did exist. I felt validated when testing because all of the users agreed that the features added brought value. I also enjoyed getting to use screenshots from the existing Yelp app. Their branding is great — it’s punchy, exciting, friendly, and bold. It encourages users to be part of the community and get to know their local businesses. I felt like I could start at a very high fidelity when wireframing, which really helped me move fast and get targeted and helpful feedback.

What I Learned: Users Want Only Relevant Information to Them

It’s important to give the user key information...

Seeing new concepts (“Collections” and social feed) on multiple screens can reinforce its importance
Key attributes about a business should be visible at early stages in user flow

...without overwhelming them

Too many links/buttons can cause analysis paralysis
Too much text can cause eye fatigue and glossing over entire sections
Graphics and color, even used subtly, can draw the user’s attention

What I Learned: Unmoderated Tests are Tricky

Metrics might not be completely accurate

Drop off due to unique-device testing (no restarts)
Increased duration from inactivity with test open
Increased misclicks due to unfamiliarity with prototype

Potential different user behavior due to test device

Tests completed on desktop so users can see side by side of instructions and mobile screen with prototype
People don’t like reading instructions, but without a moderator they can get stuck using the platform

Next Steps

Post-Launch Metrics for User Experience

After these features launch, I would want to collect data regarding the following:

Most common path to the social feed (Your Feed)
Do users use the badge, personal check-ins, friend check-ins, or the More menu to get to the social feed?
How often are users leaving comments?
“Want to Go” & “Favorites” usage
By creating two default Collections rather than the single default My Bookmarks, do users actually use both on Yelp?
Filter & search usage
Do users have the same behavior while searching/filter within Collections, compared to while in the Search tab?
Product abandon
At which screens do users abandon the app while browsing from the Collections tab?
When users open the app, how often are they returning to a previously viewed business vs. starting a new search?

Post-Launch Metrics for Business Impact

To validate the design of these new features, I would track:

Increased number of visits to the Collections tab
Increased percentage of time spent exploring the map on Collections
Increased percentage of business page views from Collections
Increased percentage of businesses added to Collections

Gathering these metrics would allow Yelp to confirm increased ad revenue is driven from Collections, which would in turn open up the value of continuing to improve the Collections user experience.

With More Time, Let’s Improve More Than Just Collections

With more time, I would look into implementing the following improvements beyond just Collections for a more holistic user experience:

Suggest businesses to add to Collections
These could be popular business based off of Collections, and another opportunity for sponsored results to appear (only if relevant)
Prioritize and highlight bookmarked/saved businesses
Rank them higher on Search, renamed the Collections tab to Bookmarked or Saved (A/B test the naming)
Clean up the UI (for both the existing app and new features)
Check the grid spacing and icon consistency
Explore new features
Such as toggling the privacy of notes, and improving the accuracy of food menus

Thanks for perusing!

If you’ve made it this far, I appreciate the time you spent reading this case study. If you have any questions about it I’d be more than happy to chat about the process and any of the findings.

If you like what you saw, send me an email and let’s talk about working together! (If you didn’t, you can still send me an email because I appreciate getting all feedback.)