L'Oréal x Yves Saint Laurent
Leverage AI technology to enhance the in-store experience, building trust and credibility through exceptional service
The current manual process which requires our engineering team to facilitate the implementation of feedback, is slowing the iteration process, resulting in significant delays and inefficiencies
I conducted interviews and shadowed beauty advisors (BAs) during their skin-tone analysis sessions with in-store customers, observing their process of recommending the most suitable shades.
BAs need a better way to prove to customers that their recommendations are correct without just saying, "trust me".
They need an app that supplements their credibility without taking away the personal aspect of their role.
Not all customers trust an AI to recommend them their shade, they want to see that their recommendation was personal.
Personalization was key to help build the trust between BAs and customers. This was visualized through the app in real-time.
Complete look - Product upsell
Perfect complexion - Product upsell
With the initial requirements for an MVP from the engineering team, I proceeded to develop wireframe sketches based on the established user flow.
The goal was to craft an interactive experience enabling users to engage with the live scan visualization. This allowed users to see how each scan affected their results in real-time. From there, they could personalize their recommendations, test them out, and provide feedback on the AI's suggestions for further enhancement.
Our competitors took a more straightforward approach by simply presenting the final recommended shades and products without much explanation. However, our research showed that customers were eager to delve deeper into the "how" and "why" behind their shade recommendations.
The wireframes converted into a simple grayscale prototype to start the usability testing.
Task 1
Achieve 3 passing scans
Task 2
Test and choose the most accurate shade
Task 3
Rate the recommendation accuracy
Scan results
7/8
Users need a more comprehensive breakdown of the live scan results.
8/8
Users mentioned that they test the best match first, and then device if they need to try others.
6/8
Users suggested a visual diagram to show customers where to scan and why.
6/8
Users rarely utilized the side nav, and some did not even know it was clickable.
Real-time scan visualization
AI shade recommendation
Facial zones visual aid
Active user feedback
Improved feedback loop time by automatically delivering customer feedback on shade match quality to the engineering team, resulting in a 30% increase in workflow speed.
Increased Beauty Advisors' credibility with a comprehensive visual breakdown of their customers' shade recommendation.
No need to reinvent the wheel. Few users used the left sidebar, and it was not obvious it could be clicked. moving it to the top and making it more recognizable significantly increased its usage and opened up more space for the main content.
