AutoTrader.ca is Canada’s go-to auto classifieds app for buying and selling cars. The app features 400,000+ vehicle listings, 1 million+ users and 2 million+ downloads across Android and iOS platforms. (Keep in mind Canada only has 37 million people!)
In this project, the team and I (3 UX/UI designers + product manager) is tasked with redesigning the entirety AutoTrader.ca's app so serve our users' needs better.
To create an experience that aligns with both the user and AutoTrader, we have to understand how their business model works. It’s completely free for regular users to browse, buy, and sell cars on AutoTrader. AutoTrader’s revenue stream comes from two offerings:
AutoTrader’s VP of marketing Roger Dunbar has made it exceedingly clear that the company’s number one priority is to increase the “quality of the traffic and the quality of the conversion.” Meeting user needs is the best way to achieve this.
We understand that buying, selling, and browsing are the 3 primary factors why users use AutoTrader, but what are the specific in-app goals?
To better understand what exactly users wish to achieve with AutoTrader, we conducted user studies with 14 respondents who are users of either AutoTrader’s app or an equivalent competitor (Kijiji, cargurus, etc). We asked why each respondent uses car buy/sell apps and these were the most common responses:
The responses reveal that the 3 main functions AutoTrader must provide is a good searching, browsing, and saving experience.
We turned our findings into 3 demographic-free personas and created user flow maps for their AutoTrader experience.
The car buyer uses AutoTrader for a simple purpose – finding and buying the car they want. Their user flow revolves around searching for their ideal car and taking action.
The car browser uses AutoTrader as a tool to keep up with the car market and bookmark vehicles for future reference. Their user flow revolves around searching, saving, and revisiting results of interest.
The car seller uses AutoTrader to sell their car by creating a private listing or selling directly to a dealer. The seller also uses the app to manage their listings.
Based on the proposed solutions for the above pain points, I created a low-fidelity wireframe to present it to the team.
After getting positive feedback regarding the new and improved user flow, I was confident to move onto the next step. I converted my low-fidelity wireframe to a high-fidelity wireframe to iron out the finer details. Survey results revealed that users enjoyed the design language of the current AutoTrader app, so I kept it the same. Here is the first iteration of my high-fidelity wireframe.
I conducted live moderated studies for each participant focusing on 3 main questions:
Feedback from my study group lead to several iterations of improvements. Here are the main changes I made:
Landing Screen:
I attempted to address the pain point of users not knowing the landing page was scrollable by shrinking the search element and shifting the ‘Get Your Quote’ ad up. However, 2 of my study group participants indicated that it was still unclear whether they could scroll down. This is important because both of the ‘Browser’ testers considered the ‘Recently Viewed’ and ‘Recommended Deals’ sections to be a relevant part of their user flow. I revised my strategy by further reducing the search element and setting it to expand back to full size upon interaction.
The result is a partially covered ‘Recommended Deals’ element that is revealed enough to display all relevant information, but covered enough that the user knows to scroll down. This was achieved without significantly harming the usability of the search feature.
Make Selector:
My decision to change the make selector from a large scrolling list to a search was well-received by all 4 testers. However, an oversight I made was not putting a list of all vehicles. This became frustrating for users because if they did not know exactly which make they wanted to search (or perhaps forgot the name), there was no way look through a list. I reimplemented the list and added 2 quality of life features. The first was a ‘Popular Makes’ section which contains the most selected makes. Next was adding a small logo for each make for easier recognition.
The result is a more intuitive make selector that makes it faster and easier for users to find and select their desired brands.
Saving a Vehicle:
One of my ‘The Browser’ testers pointed out that when saving a vehicle for the first time, they expected a prompt or button that would take them to the ‘Saved Vehicles’ page. I decided to add a small pop-up message to instruct first time users how they can access their saved vehicles.
Notification Setting:
The current AutoTrader app has a feature that will notify the user if any of their saved cars gets a price drop. This feature can be enabled in a dedicated notification settings page. I initially followed this design but quickly realized it was unintuitive to have an entire page dedicated to a single setting. I changed this setting into a toggle and moved it to the main ‘Saved’ page for easier access.
Saved Searches:
I discovered during the usability tests that users did not consider the ability to edit searches a crucial feature. I took a step back to reconsider why users use the ‘Saved Searches’ function. Feedback from the testers revealed that a substantial reason to go in this section is to see the new listings. I moved the edit feature to the menu and added a new feature that aggregates all new listings for each saved search (since the user’s last search). This ‘New Vehicles’ feature streamlines the browsing and buying user flow by allowing users to effortlessly keep up to date with vehicles they are interested in.
Listing Details:
The final revision was a small one. My testers pointed out that the buttons on the listing page was too small and hard to interact with. I made each button and contained text 50% larger for improved usability.
The latest revision of the AutoTrader app is displayed below. The redesign eliminates the existing pain points listed above and augments the flow of the users. All in all, candid feedback from the users helped create a truly user-centric experience.
How can we improve the AutoTrader app even more? My first suggestion is to develop new features that will augment the user experience. Conducting more open-ended usability tests can provide valuable insight on features that can make app experience even better. My second suggestion is to cater the app to an even broader audience. For example, a feature that suggests cars to users based on a questionnaire will make AutoTrader a better experience for people who don’t know much about cars.