Shipped · AI/AR · Mobile
A mobile-first virtual try-on that uses AI facial detection to take customers from "what's my brow shape?" to a curated shopping list. It's live today.
The brief
MAC customers wanted a way to understand and shop for brow products tailored to their unique features, without booking an in-store consultation. The goal was a hyper-personalized virtual try-on that could detect a customer's brow attributes, recommend a tailored look, and convert that recommendation into a sale, end to end, on a phone.
Goal
Detect: scan a user's unique brow attributes through the camera.
Recommend: suggest tailored brow goals: shape, thickness, length, definition.
Educate: walk through step-by-step tips so the recommendation actually makes sense.
Convert: connect directly to a curated shopping list for instant purchase.
Process
Identified user needs around brow education, try-on, and product selection before any screen was sketched.
Defined the technical flows for AI detection, camera consent, and mobile performance constraints up front, so the design never outran what the tech could actually do.
Built interactive wireframes and full journey maps covering consent, scanning, analysis, and goal selection.
Designed makeup tips and personalized guides tied directly to each customer's detected brow attributes.
Designed "Shop Your Brow" overlays and product bundles that link straight into checkout.
Added snapshot sharing, post-experience surveys, and follow-up email touchpoints to close the loop after purchase.
Key features
AI facial detection, Golden Ratio overlay guidance, customizable goals (shape, length, thickness, color), before/after comparisons, step-by-step education, a shoppable product list with direct add-to-bag, and a snapshot-and-share moment that turns the result into something worth posting.
Outcome
The Brow Recommender created a seamless path from discovery to purchase: customers gained confidence in brow styling, experienced an immersive mobile-optimized try-on, and converted at a higher rate thanks to the personalization layer.