The original work optimized one shoppable content experience for everyone, and it was a success. Today, I'd use machine learning to understand shopper behavior at a deeper level, building specific customer profiles from signals like browsing patterns and purchase history to surface the product categories, styles, and colors each shopper actually favors. Generative AI would then create the copy, imagery, and layout for those selected products, tailored to each profile, rather than relying on a fixed set of pre-made content. My role shifts from picking the "best" static design to defining which customer signals the system should respond to, and setting the taste and guardrails so every personalized version still feels on-brand.