Stocklist
How my research and redesign boosted
inventory visibility and user interactions up to 159%.
Company
AutoScout24
Context
AutoScout24 is the largest car marketplace in Europe. With over 30 million users per month, and more than 43,000 dealer partners. Our users for this project, were dealership managers and sales people. With heavy users having an average of 40 cars in their stock.
What I did
As a B2B Sr. Product Designer, I led an end-to-end UX process, discussing problems, solutions, and implementation with the team.
The tool
The dealer cockpit is our tool for dealerships. Its most crucial feature is the stocklist (see below), which shows the vehicles and their statuses (with +2.6m visits in 2022).
Old design
Opportunities
Internal requirements
Dealers had an average of 54 vehicles, and with the previous design we showed 2 vehicles before the fold. Creating an endless scroll, so we needed to condense information. And to enhance engagement, we wanted to implement some AI models to provide them with dynamic insights.
Dealers' needs
With attitudinal data (from 700 feedbacks), and behavioural data (from their paths on Google Analytics), we learned that dealers wanted to sort and print their list of vehicles, as well as edit, activate, and access more information about their vehicles.
Benchmarks
Function
I investigated 7 companies from the same industry around the world (including direct competitors), and how they were approaching the challenge.
Form
I listed products like Google Flights, Stripe, Spotify, and others to learn how they were solving similar problems in terms of UI and, more specifically, table views. Besides that, I started a short collection of articles about table views.
Function
+
Form
To create our final solution, I organized a workshop with the tech, data, and business teams. At this workshop, we all used Miro to collaboratively design our ideas, starting from user and business required elements.
Designing solutions
Check how I created a design that addresses the main user and business needs.
More details
A first layer of data is presented on the table view, a deeper one would be accessible from our drawer.
Printing
Being the most requested feature, with printing dealers can generate printable PDF’s of their whole stock to work offline as well.
Dynamic info
Born as an internal requirement, dynamic infos are a result of our AI models to support dealership decisions and predict results while maintaining then engaged with us.
Sorting
Inspired by spreadsheet tools, we included sorting on the header of our table view.
Activity and edits
Dealers can edit their vehicles data and enable/disable their ads with AutoScout24 from our new feature.
Interviews
and tests
Since we had a good amount of data to work with, we decided to test our hypothesis with users once we had a prototype. To do that, I partnered with a UX researcher, and we conducted 8 sessions with users from the Netherlands, Germany, Belgium, and Italy. In these sessions, we asked questions and evaluated how they performed their tasks using it. In the end, we didn't have any major usability issues, but we saw that feature localization would be needed to address all their different needs. So, we made those iterations to prepare for our next step.
Developing
and shipping
With the design fine-tuned, we discussed the implementation details and started building it. My team decided to push it to 100% of our base (I would do it differently), and our users would be able to try it out using a switch to enable this new version or go back to the previous one.
Final results
These results show that we achieved our goals to present simplified and meaningful content:
-65%
Average inventory pagination
-44%
Use of extensive filters
+159%
Price changes with our insights
84,6%
Interactions / Per pageview
Did you liked it?
