In October 2011, I joined the e-commerce company ATG Stores (a Lowe's company) as their very first UX designer. I worked closely with the marketing team, business analysts, project managers, and developers. I redesigned the website and built the style guide from the ground up.Established a UX process for my team.
I later became a lead designer and hired two excellent team members. My design team was able to convert the entire e-commerce website to responsive design in 6 months with using CSS and Javascript.
One of my favorite parts of the UX process is using research to support design. We used users' feedback, usertesting.com, heat map, and Google Analytics to help find pain points and test out the new design. When proposing this monumental change to the team, we listened to stakeholders' suggestions and concerns and tested the "big ideas" using Optimizely A/B Testing.
Some of the design proposals did raise concerns for stakeholders - such as the navigation bar redesign. From user testing videos we could see that the cluttered four layer top navigation made it hard for the user to find anything quickly. I proposed to the team that we should try to simplify the navigation by grouping information together appropriately (e.g. No need to list all categories, place them under Departments). However, the business analysts were worried that customers would not be able to find their desired items and would not realize the wide variety of of available products at a glance, which was a valid concern. To remedy this uncertainty, I set up an A/B testing on optimizely.com across all browsers. I changed the top navigation simply using CSS and pushed it live. Half of our users would still shop with old navigation while the other half would experience the new navigation.
Was the design proposal good or bad? Will the cleaner design impact the sales positively or negatively?
The Answer: new design wins!
The new navigation started to show its advantage in 11 days among almost 500k unique visitors, improving the order completion rate by 3%. We ran the test for 60 days among nearly 3 million users across all browsers, and the new design performed so well that the test result entirely eliminated all concerns.