Google's UCP Gets Its First Major Update — Multi-Item Carts, Identity Linking, and Simplified Onboarding Signal Agentic Commerce Is Accelerating

Google published a major update to its Universal Commerce Protocol on March 19, three months after unveiling the open standard at NRF 2026 in January. The update adds three capabilities that shift UCP from a product-discovery layer to something closer to a full transaction stack: multi-item cart operations, real-time catalog queries against retailer backends, and identity linking that connects shoppers' loyalty programs to UCP-integrated platforms. Google also announced simplified onboarding through Merchant Center, lowering the integration barrier for mid-market and smaller retailers who lack dedicated engineering teams.

UCP was co-developed with Shopify, Etsy, Wayfair, Target, and Walmart, and endorsed by more than 20 global partners including Stripe, Visa, Mastercard, Best Buy, and Macy's (Google, January 2026). The March update is the first concrete evidence that this coalition is shipping features, not just signing press releases.

What exactly changed in the March 19 update?

Cart operations are the most significant addition. Before this update, UCP handled single-product interactions — a shopper asked an AI agent about one item, and the agent returned results for that item. Multi-item cart support means agents can now save several products in a single session, matching how people actually buy groceries, household supplies, or back-to-school lists. The difference between adding items one at a time and building a real cart is the difference between a product search tool and a shopping assistant.

Catalog queries let AI agents pull real-time pricing and inventory data directly from retailer backends. Static product feeds have been the default in e-commerce for decades — merchants export a snapshot of their catalog, and platforms ingest it on a schedule. Real-time queries eliminate the lag. When an agent tells a shopper that a product is in stock at a specific price, that information reflects the retailer's live system, not a feed that was generated six hours ago.

Identity linking connects existing loyalty programs to UCP-integrated platforms. A shopper with a Target Circle or Macy's Star Rewards account gets those benefits applied automatically when purchasing through an AI agent that speaks UCP. This solves a friction point that has slowed AI commerce adoption: consumers who have accumulated loyalty benefits are reluctant to shift purchasing to a new channel if it means losing those perks. Identity linking removes that trade-off.

Why does this matter beyond Google's own products?

UCP is designed as an open standard compatible with Anthropic's Model Context Protocol, Google's Agent-to-Agent protocol, and Apple's AP2 framework. That interoperability claim is what separates it from competing approaches. OpenAI's Agentic Commerce Protocol, built with Stripe for Instant Checkout, operates as a more closed system tied to ChatGPT's transaction flow. Shopify's Agentic Storefronts take a merchant-first approach, distributing product data across AI platforms but keeping the commerce logic within Shopify's infrastructure.

The coalition backing UCP gives it structural weight. EMARKETER projects AI platforms will account for 1.5% of total retail e-commerce sales in 2026 — roughly $20.9 billion — nearly quadruple 2025 figures (EMARKETER, December 2025). Adobe reported a 693% increase in AI-driven traffic to US retail sites during the 2025 holiday season (Adobe, 2025). Twenty-two percent of shoppers now incorporate AI tools into their buying journeys (Salsify, 2026). The infrastructure handling that traffic needs shared standards, and UCP is the most broadly supported candidate.

What comes next for UCP adoption?

Google is actively onboarding retailers through the simplified Merchant Center process. Commerce Inc, Salesforce, and Stripe have committed to implementing UCP on their platforms, which extends the protocol's reach to merchants who may never interact with Google's tools directly. Broader rollout is planned across AI Mode in Search and the Gemini app, meaning UCP-enabled product data will surface wherever Google's AI interfaces appear.

The speed of iteration matters as much as the features themselves. Three months from announcement to a substantive capability update suggests Google is treating UCP as a living protocol, not a specification published once and left to gather dust.