Ulta Beauty Deploys AI Recommendation Platform to Personalize Product Combinations for 46M+ Members

Background

Ulta Beauty operates at the intersection of mass and prestige beauty, with a loyalty membership base exceeding 46 million customers. The scale of its product catalog — spanning skincare, makeup, haircare, and fragrance across thousands of brands — creates both a challenge and opportunity: customers benefit from personalized discovery, but the catalog complexity makes generic recommendations ineffective. Ulta invested in building Quazi, a proprietary AI personalization platform, to address this at scale — delivering individualized product recommendations that reflect each customer's evolving preferences and purchase behavior.

What Was Implemented

• Quazi: Ulta Beauty's proprietary AI personalization platform, updating recommendations in real time based on customer interactions across marketing, social, and sales channels
• "Ulta AI" shopping assistant: powered by Google Gemini Enterprise for Customer Experience; available on Ulta.com and rolling out to the Ulta Beauty App
• Agentic commerce integration with Google AI Mode in Search and Gemini app: Ulta's assortment shoppable within Google's conversational interfaces
• Individual replenishment reminders, personalized offer recommendations, and product combination suggestions across all channels
• Insights from 46M+ loyalty members powering personalization engine

Results

No specific outcome metrics (conversion rate, AOV, revenue lift) were confirmed in primary Ulta Beauty sources fetched. The Quazi platform and Gemini AI assistant are confirmed operational capabilities as of 2026. A secondary aggregator article attributes a 25% AOV increase to Ulta's AI recommendations (unverified — not confirmed in a primary source fetched). The book's characterization of Ulta using AI to recommend personalized product combinations is accurate.

Lessons

• A proprietary personalization platform (Quazi) allows a retailer to control the full recommendation stack and integrate loyalty data directly — a competitive advantage not available with third-party tools
• Agentic commerce (making inventory shoppable within AI search interfaces like Google Gemini) represents the next frontier beyond on-site recommendations
• Loyalty program scale (46M+ members) provides a significant data advantage for training personalization models
• No publicly disclosed outcome metrics were located for Ulta's AI bundling and recommendation work at time of research — underscoring that even major retailers often do not publish granular performance data