Want a self serve tool to track AI Visibility? Checkout Passionfruit Labs

Learn More

Want a self serve tool to track AI Visibility? Checkout Passionfruit Labs

Learn More

Want a self serve tool to track AI Visibility? Checkout Passionfruit Labs

Learn More

SEO

What is Google's UCP Update? Carts, Catalogs, and Loyalty in AI Shopping

What is Google's UCP Update? Carts, Catalogs, and Loyalty in AI Shopping

What is Google's UCP Update? Carts, Catalogs, and Loyalty in AI Shopping

Summarize this article with

Summarize this article with

Table of Contents

Don’t Just Read About SEO & GEO Experience The Future.

Don’t Just Read About SEO & GEO Experience The Future.

Join 500+ brands growing with Passionfruit! 

Google updated the Universal Commerce Protocol (UCP) on March 19, 2026, two months after launching it at NRF. The update added three new capabilities, Cart, Catalog, and Identity Linking, plus a simpler onboarding path through Merchant Center. Together, the changes turn UCP from a single-item checkout spec into something closer to a full commerce protocol for AI agents.

The short version: agentic commerce just got more usable, and retailers who treat their product data as machine-readable infrastructure get a head start.

What is the Universal Commerce Protocol (UCP)?

Universal Commerce Protocol is an open standard, co-developed by Google and Shopify, that lets AI agents transact with retailers across the shopping journey. Agents can discover products, build carts, complete checkouts, and handle post-purchase actions through a common interface. UCP launched at NRF 2026 in January with backing from Walmart, Target, Best Buy, Visa, Mastercard, and over twenty other ecosystem partners.

The protocol is decentralized. Merchants host JSON profiles at /.well-known/ucp on their own domains, similar to how robots.txt or sitemap.xml work. Any compliant AI agent can read these profiles to learn what a retailer sells, what capabilities they support, and how to interact with their commerce APIs.

How UCP fits into agentic commerce

Agentic commerce describes the shift where AI agents like Gemini, ChatGPT, and Perplexity act on behalf of shoppers across discovery, comparison, and checkout. Most existing search and SEO infrastructure assumes a human is browsing. Agentic flows assume an AI is parsing structured data, negotiating capabilities, and executing transactions inside its own surface.

For brands building visibility across both Google and AI search, the rules are converging. The same machine-readable product data that powers UCP transactions also feeds AI Overviews, Gemini, and other answer engines. If you want a primer on how the surfaces overlap, our breakdown of how SEO, GEO, and AEO strategies fit together covers the ground.

What changed in the March 2026 UCP update

Google's March 19 update introduced three new optional capabilities and announced a simplified Merchant Center onboarding path. Cart and Catalog are published as draft specifications. Identity Linking is already in UCP's stable release. Each addresses a specific limitation that had been holding adoption back.

Cart: multi-item AI shopping baskets

Before the update, UCP could only handle one item per checkout session. A shopper asking Gemini for "running shoes, socks, and a water bottle" would have hit three separate transactions. Cart lets AI agents add multiple items from one retailer in a single operation, and supports pre-purchase exploration: agents can build baskets before the shopper commits, then convert the basket into a checkout when ready.

Catalog: real-time product data for agents

Catalog gives AI agents a live read on a retailer's inventory, including variants, pricing, and stock levels. Existing Google Shopping product feeds are static snapshots that update on a schedule. Catalog is queried at the moment an agent needs to surface a product, so the agent can confirm a specific size or color is in stock before recommending it to the shopper.

Identity Linking: loyalty pricing inside AI surfaces

Identity Linking uses OAuth 2.0 to connect a shopper's retailer account to UCP-integrated platforms. A Nike member buying through Google AI Mode can carry over the member pricing, free shipping, and other benefits available on the brand's own site. Without Identity Linking, shopping through an AI agent meant losing those benefits. Among the three capabilities, Identity Linking is the only one already in UCP's stable release.

Simplified Merchant Center onboarding

Google is building UCP onboarding directly into Merchant Center, the dashboard most retailers already use for Google Shopping feeds. Products using the native_commerce product attribute will show a checkout button in Google AI Mode and the Gemini app. Salesforce, Stripe, and Commerce Inc. are slated to implement UCP support on their platforms, so retailers on those stacks will get the protocol layer handled for them.

UCP vs ACP: how the two agentic commerce protocols compare

UCP is one of two major open standards competing to power AI shopping. The other is the Agentic Commerce Protocol (ACP), launched by OpenAI and Stripe in September 2025. The two protocols solve similar problems with different architectures, and most retailers planning for agentic commerce will end up supporting both.

Dimension

UCP (Google + Shopify)

ACP (OpenAI + Stripe)

Launch

January 2026 (NRF)

September 2025

Architecture

Decentralized; merchants host JSON profiles

Centralized; merchants submit catalogs to OpenAI

AI surfaces

Google AI Mode, Gemini, multi-agent ecosystem

ChatGPT, OpenAI-powered surfaces

Payment layer

Open; works with AP2 and multiple processors

Stripe Shared Payment Token

Loyalty support

Identity Linking carries over member benefits

No equivalent capability

Best for

Multi-surface visibility, retailer control

Conversational checkout inside ChatGPT

UCP follows the open-web pattern. ACP follows a managed marketplace pattern. Retailers serious about full agentic shopping coverage typically implement both, since each connects to a different AI agent ecosystem.

What the UCP update means for retailers

The action items have not changed much. The pressure to act on them has. AI-driven traffic to U.S. retail sites grew 693.4% year over year in November and December 2025, according to Adobe Analytics, and the protocols are catching up to that demand. Three priorities matter most.

Get product data structured for machines

UCP, ACP, and AI Overviews all draw on the same underlying signal: clean, structured, machine-readable product data. Schema markup, accurate inventory, complete attributes, and consistent fulfillment information are no longer just SEO hygiene. Structured product data is the prerequisite for being included in agent-driven shopping flows. The patterns we cover in our guide to measuring AI search visibility for ecommerce apply directly here, since the same data feeds both citation tracking and agentic transactions.

Audit what AI surfaces currently say about your brand

Before optimizing for transactions, audit visibility. Run prompts across ChatGPT, Perplexity, Gemini, and Google AI Mode that match how your customers describe your category, and log which competitors get cited. Pages that already rank in Google's top results have a strong head start in AI citations, but the relationship is not automatic. The audit is the diagnostic step that tells you what to fix next, and how to invest budget across traditional search and AI search visibility without spreading thin.

Decide your protocol coverage

For most retailers, the answer is "both UCP and ACP," but the sequencing matters. UCP integration is dropping in cost as Merchant Center handles more of the onboarding. ACP integration is faster on the technical side but pricier in fees. Retailers selling through a major platform such as Shopify, Salesforce Commerce Cloud, or BigCommerce should check what their platform is doing for them before scoping in-house engineering work. Smaller catalogs can usually start with UCP via Merchant Center; larger catalogs with custom commerce stacks need to scope both protocols carefully.

For brands looking at where this fits inside a broader visibility plan, generative engine optimization services cover the structured-data and citation work the protocols rely on. The technical SEO foundation those flows depend on sits inside search engine optimization services.

Bring your brand into the answer

Agentic commerce is no longer a forecast. UCP, ACP, and AI Overviews are reshaping how shoppers find, compare, and buy products right now. The brands that show up inside those answers are the ones treating product data, schema, and citation visibility as core infrastructure, not seasonal projects.

If your category is moving and your team is not, the gap compounds quarter over quarter. The fastest way to close it is a baseline audit that shows where you stand today across Google and AI surfaces, then a focused plan that fixes the biggest visibility leaks first. Look at how Passionfruit's GEO services approach the audit, see what citation tracking looks like inside Passionfruit Labs, and talk to the team when you are ready to move.

Frequently asked questions

The questions below come up most often when retailers, ecommerce operators, and marketing teams start mapping UCP into their stack.

What does UCP stand for?

UCP stands for Universal Commerce Protocol. The open standard, co-developed by Google and Shopify, lets AI agents discover, transact with, and manage post-purchase actions for retailers across consumer surfaces such as Google AI Mode and the Gemini app.

When did Google update UCP?

Google announced the latest UCP update on March 19, 2026, two months after the protocol's initial launch at NRF in January 2026. The update introduced Cart, Catalog, and Identity Linking capabilities, plus a simplified Merchant Center onboarding path.

Is UCP the same as ACP?

No. UCP is a Google and Shopify open standard covering the full shopping journey. ACP, or Agentic Commerce Protocol, is co-developed by OpenAI and Stripe and focuses on checkout flows inside ChatGPT. The two protocols are compatible at the merchant infrastructure layer, and many retailers support both.

Do I need engineering work to enable UCP?

Not always. Retailers managing product feeds through Google Merchant Center can enable UCP through a simplified onboarding flow that is rolling out over the coming months. Retailers on Salesforce, Stripe, Commerce Inc., or Shopify benefit from platform-level UCP support that abstracts the protocol layer.

What is Identity Linking in UCP?

Identity Linking uses OAuth 2.0 to connect a shopper's retailer account to UCP-integrated AI platforms. When a shopper buys through Google AI Mode or Gemini, Identity Linking carries over loyalty pricing, member discounts, and free shipping the shopper would normally receive on the retailer's own website.

How does UCP affect AI search visibility?

UCP shares its data foundation with AI Overviews, Gemini, and other AI search surfaces: structured product information, accurate inventory, and machine-readable schema. Brands investing in UCP infrastructure end up improving their broader AI search visibility at the same time, since the same signals feed both flows.

grayscale photography of man smiling

Dewang Mishra

Content Writer

grayscale photography of man smiling

Dewang Mishra

Content Writer

grayscale photography of man smiling

Dewang Mishra

Content Writer

Trusted by teams at high growth companies

Ready to win search?

End to End, managed experience to drive growth from Google and AI search

Passionfruit

Trusted by teams at high growth companies

Ready to win search?

End to End, managed experience to drive growth from Google and AI search

Passionfruit

Trusted by teams at high growth companies

Ready to win search?

End to End, managed experience to drive growth from Google and AI search

Passionfruit