Subaru is one of the world largest automotive manufacturers. In 2019 we were engaged to assist with the clean-up of the analytics and set-up an experimentation program.
Our experimentation program was focused on creating a framework to help the team make data led decisions.
The showroom page of the Subaru website is one of the most visited pages on the site.
The goal of the page was to educate users on the Subaru range, provide bite sized information on drive the users into the model details page to get more information.
This however was not always the case. 44% of users who came to the page would immediately bounce and over 21% would exit the site permanently. Of the users who stayed on the page 31% would not progress to the models page.
Within our research we identified the original page was long and lacked details. This allowed visitors to compare models before deciding which model to explore further, resulting in 31.2% (8,926 per month) not progressing through to a models page.
Given the importance of the page and impending media activity, we decided to create multi-clustered experiment.
As part of the experiment design we focused on a number of test factors grouped together to create a new showroom template.
Within the test we looked to reduce page size, making page content more scannable, reducing the page header, decluttering the page and show pricing on the page.
The new page experience proved to be a great success. The shortening of the page and content delivered a 9.7% uplift in page activity and an increase in flow through to the models page.
uplift in page activity