AI Personalization: Recommendations & Checkout

Delivering personalized experiences is especially important in healthcare and prescription-based ecommerce, where regulatory requirements often disrupt the traditional user flow. At National Vision, I led the strategy and design for two key personalization initiatives: a behavioral product recommendation engine and a dynamic Prescription Checkout step.

To help users find the right product faster, we built a personalized recommendation engine that analyzed browsing behavior, saved preferences, past purchases, and known prescription data (like pupillary distance). This feature surfaced high-relevance frames on PLP and PDP views, improving product discovery and reducing decision fatigue. The logic prioritized material, size, and color patterns based on user intent signals and favorited items.

We originally surfaced the output of this data to returning, authenticated users to measure engagement and streamline reording flows utilizing implicit data. In order to improve unauthenticated experiences, we later built a small recommendation wizard - gathering simple pereferences requested in a conversational tone and educating our users on how to best achieve measurement fit.