Ulta Beauty Deploys Gemini-Enabled Agentic Shopping

Ulta Beauty is deploying agentic commerce with Google across search and the Gemini app and launching a first-party AI assistant. The retailer will make its assortment shoppable within AI Mode in Search and the Gemini app over the next month, while debuting Ulta AI, an on-site assistant built on Gemini Enterprise that draws on Ulta's 46 million loyalty members to personalize discovery and checkout.

Ulta and Google are using the Universal Commerce Protocol to enable agentic flows where a conversational agent can recommend, compare, and complete purchases inside a single interaction. Key capabilities called out in the launch include:

• •Personalized product recommendations surfaced inside conversational threads and AI-assisted search results
• •Side-by-side comparisons and guidance that prioritize trust and discovery for beauty categories
• •Streamlined checkout for eligible items completed within Google surfaces or handed off to Ulta channels

The partnership explicitly uses Gemini Enterprise for customer experience signals and integrates Ulta product catalogs into Google conversational surfaces. Ulta positioned this as an early collaboration with the Universal Commerce Protocol, an open standard designed to let merchants and agents exchange intent, product metadata, pricing, and cart state in a machine-friendly way.

This is a commercially meaningful deployment at the intersection of search, recommendation, and transaction. Beauty is a high-friction vertical where visual signals, brand trust, and micro-personalization matter, so agentic commerce is a natural fit. Ulta's loyalty dataset and curated assortments give it a defensible personalization advantage versus general marketplaces that win on assortment and logistics.

For practitioners the launch highlights three trends to watch: first, the migration of discovery from catalog search to intent-driven, agentic conversations; second, the emergence of protocol-level interoperability with Universal Commerce Protocol that reduces friction for merchants to be transactible inside third-party agents; third, operational demands around catalog metadata, real-time inventory, pricing integrity, and consented personalization when agents can complete checkout.

Track adoption metrics, conversion lift, and attribution changes away from traditional search and marketplace channels. Pay attention to how Ulta maps inventory, handles promotions, and secures payment flows inside third-party conversational interfaces. Privacy, fraud controls, and error handling in cross-surface checkouts will reveal whether agentic commerce scales without adding merchant risk.

This is not only a marketing play. Implementing agentic commerce requires engineering investments in structured product metadata, cart and order APIs, and real-time signals for availability and pricing. It also creates new instrumentation needs for measuring intent-to-conversion inside conversational sessions and demands tightened data governance when loyalty data is used to personalize agent responses. Expect other vertical retailers to accelerate early integrations or risk ceding discovery control to the platform agents that customers increasingly use.