AI shopping agents need catalogs they can parse: price, variants, compatibility, stock. UCP (Universal Commerce Protocol) is Shopify and Google's 2026 spec for declaring those capabilities once and selling through ChatGPT, Google AI Mode, Gemini, Copilot, and similar surfaces.
This post explains what that means in practice, what Shopify and Square are shipping, and how to clean up product data now—even before every AI channel is live.
What Is Agentic Commerce?
Agentic commerce is shopping where an AI agent acts on behalf of a user: discovering products, answering pre-purchase questions, building a cart, and completing checkout. Instead of a human browsing ten tabs, a user might tell ChatGPT "find me a waterproof hiking backpack under $80 with free shipping" and expect a purchasable result.
For this to work, agents need structured, unambiguous product data—not marketing fluff or duplicate manufacturer copy. Ambiguous catalogs get skipped; clear catalogs get recommended.
Google's 2026 agentic commerce announcements describe dozens of new Merchant Center attributes designed for conversational discovery—beyond traditional keywords to include FAQ answers, compatible accessories, and substitute products. Shopify's UCP standard format lets merchants declare capabilities once and sell across multiple AI channels from admin.
The sellers who prepare now will surface in agent results when traffic arrives—not scramble to retrofit data later.
What Is the Universal Commerce Protocol (UCP)?
UCP is an open standard for AI agents to connect and transact with any merchant. Co-developed by Shopify and Google, it is supported by major retailers (Target, Walmart, Wayfair, Etsy) and millions of Shopify merchants. The specification is public on GitHub.
Core design principles
| Principle | What it means for sellers |
|---|---|
| Discovery and negotiation | Merchants declare supported capabilities; agents negotiate what they can handle per transaction |
| Composable architecture | Core checkout + optional extensions (discounts, loyalty, fulfillment) |
| Transport agnostic | Same business logic across REST, GraphQL, MCP, and other transports |
| Payment flexibility | Agents and merchants negotiate payment handlers per cart—not one fixed gateway |
| Graceful handoff | Human-in-the-loop escalation when automation cannot complete |
UCP models the full commerce journey: product discovery → cart → checkout → order → post-purchase. Agents do not guess your return policy or shipping rules—you declare them in machine-readable profiles.
UCP vs. other protocols
The agentic commerce landscape includes several standards:
| Protocol | Led by | Primary use |
|---|---|---|
| UCP | Shopify + Google | End-to-end commerce across agents and merchants |
| ACP (Agentic Commerce Protocol) | OpenAI + Stripe | ChatGPT Instant Checkout flows |
| MCP (Model Context Protocol) | Anthropic | Tool and data access for AI models |
Shopify's Agentic Storefronts abstracts multiple protocols—merchants toggle AI sales channels from admin without implementing each standard separately. UCP is the broadest open standard for cross-platform agent commerce as of 2026.
Shopify UCP: Agentic Storefronts
Shopify merchants can enable agentic selling through Agentic Storefronts in Shopify admin. Products listed in Shopify Catalog—Shopify's LLM-enriched product index—become discoverable in:
- ChatGPT (via OpenAI partnership)
- Google AI Mode in Search and Gemini app
- Microsoft Copilot (embedded checkout update, 2026)
- Shop app and future catalog partners
What Shopify handles for you
If you sell on Shopify, much of the UCP infrastructure is managed automatically:
- Product data flows from your admin into Shopify Catalog
- LLMs categorize, enrich, and standardize attributes
- Agentic Storefronts exposes UCP-compliant endpoints to partner agents
- Checkout can complete natively inside the agent surface (Embedded Commerce Protocol)
Merchants on Shopify's Agentic plan can list products in Shopify Catalog even if their primary storefront is not Shopify—opening AI channels to brands on other platforms.
What you still must optimize
Shopify's infrastructure does not fix vague copy. Garbage in, garbage out:
- Duplicate supplier descriptions produce duplicate agent recommendations (or none)
- Missing dimensions cause agents to exclude products from constraint queries ("fits 15-inch laptop")
- No FAQ content means agents cannot answer "is this dishwasher safe?"
UCP-ready product data starts with human-readable copy that doubles as machine-readable entities.
Square and ChatGPT Integration
Square sellers enter agentic commerce through Square's integrations with AI platforms and Google's UCP ecosystem. Square's point-of-sale and online store data—catalog, inventory, pricing—feeds into discovery surfaces when merchants enable compatible channels.
For Square merchants, the preparation steps mirror Shopify:
- Complete catalog fields — every item should have category, description, price, and image
- Unique descriptions per SKU — not "assorted widget" repeated 50 times
- Explicit attributes — size, color, material, dietary flags (for food), compatibility
- Accurate inventory — agents that recommend out-of-stock items destroy trust
Square's strength is unified online + in-person catalog. Ensure Square Online item descriptions match the detail level of your best e-commerce listings.
The UCP-Ready Product Data Checklist
Use this checklist to audit any catalog—Shopify, Square, WooCommerce, or marketplace feeds destined for Google Merchant Center.
Product identity (required for discovery)
- Unique product title per SKU — product type + benefit + key differentiator
- Stable product identifier — SKU, GTIN, or MPN consistently mapped
- Accurate product category — Google product category or platform equivalent
- Primary image URL — high resolution, product fills 85%+ of frame
- Price and currency — current, including variant-level pricing
- Availability / inventory status — in stock, out of stock, preorder with dates
Attributes agents use for filtering
- Dimensions and weight — shipping agents and "fits my space" queries
- Color, size, material — variant attributes explicitly set, not buried in title only
- Compatibility lists — devices, models, accessories ("works with iPhone 15")
- Age range / safety — baby, food, and regulated categories
- Condition — new, refurbished, used (critical for marketplaces)
Natural-language content agents can quote
- Who is this for? — opening description sentence naming audience
- What problem does it solve? — outcome-focused, not feature-dump
- FAQ block — 3–5 Q&A pairs in plain language
- What's included — box contents, accessories, warranty length
- Care and limitations — dishwasher safe, indoor only, not medical advice
Structured data and feeds
-
Productschema on owned storefronts — name, description, image, Offer, availability -
FAQPageschema where FAQ blocks exist - Google Merchant Center feed — all required attributes populated; no blank description fields
- Consistent data across channels — same title semantics on Shopify, Amazon, and feeds (agents cross-reference)
Policy and fulfillment transparency
- Return policy — days, condition, who pays shipping
- Shipping regions and timelines — agents quote delivery expectations
- Tax behavior — configured correctly per jurisdiction
How to Upgrade Existing Listings for UCP
Phase 1: Fix your top 20% (week 1)
Rank SKUs by revenue. For each:
- Rewrite title using product type + benefit + differentiator (first 80 characters matter on mobile)
- Add FAQ block to description
- Fill all variant attributes in admin
- Set unique alt text on every image
An AI product page optimizer cuts rewrite time from 30–60 minutes to 30–60 seconds per SKU.
Phase 2: Catalog-wide consistency (weeks 2–4)
Define a data dictionary for your store:
| Field | Standard | Example |
|---|---|---|
| Title pattern | [Type] — [Benefit], [Spec] |
Insulated Mug — 12hr Hot, 16oz Stainless |
| FAQ minimum | 3 questions per SKU | Shipping, compatibility, care |
| Image alt text | Product + attribute | 16oz stainless insulated travel mug with leak-proof lid |
Batch-optimize remaining SKUs. Shopify sellers can use our Shopify optimizer; Amazon sellers use Amazon optimizer.
Phase 3: Enable agent channels (when available)
- Shopify: enable Agentic Storefronts in admin; review which AI channels are active
- Google: monitor Merchant Center for new conversational attributes as they roll out
- Square: follow Square dashboard announcements for AI channel opt-in
Toggle channels on only when your top SKUs pass the checklist—agents surface your best data first.
UCP and GEO: Two Sides of AI Discovery
GEO (Generative Engine Optimization) helps your content appear in Google AI Overviews and Perplexity answers. UCP helps agents complete transactions. Both reward the same underlying asset: clear, structured, factual product data.
Practices that serve both:
- Self-contained definition paragraphs ("This is a …")
- Statistics and specs with units (not "large" but "16 oz / 473 ml")
- Dated content refreshes (show
lastUpdatedon guides and key pages) - FAQ sections in natural language
Read our future-proof ecommerce product catalog guide for the combined GEO + UCP + visual search strategy.
Frequently Asked Questions
What does UCP-ready mean for my product catalog?
Your product data is structured and complete enough that AI shopping agents can find items, answer questions, and start checkout through UCP without a custom integration per agent.
Do I need to be on Shopify to use UCP?
No. UCP is an open standard. Shopify has the most turnkey path (Agentic Storefronts, Catalog). Other platforms and Google Merchant Center are adding compatible fields. You can adopt UCP-ready practices on any catalog today.
How is UCP different from Google Shopping feeds?
Shopping feeds feed ads and organic Shopping. UCP adds a negotiation layer so agents can run full checkout (cart, payment, order status) across AI surfaces. Feeds are one input; UCP is the transaction protocol.
When will ChatGPT shopping work for my store?
Shopify merchants can sell through ChatGPT via Agentic Storefronts (2026 rollout). Non-Shopify sellers may list through Shopify's Agentic plan and Catalog. Square and others are on their own timelines. Clean up catalog data now regardless.
What product fields matter most for AI agents?
Title, description, price, stock, image, category, dimensions, compatibility, and FAQ content. Agents filter on structured fields and quote FAQ answers. Missing any of these hurts discoverability.
Can AI tools help me become UCP-ready?
Yes. Listing optimizers produce FAQ-friendly, intent-clear copy agents can quote. AI Product Page Optimizer runs on Amazon, Shopify, eBay, and Etsy admin pages. New users get 5 free credits.
Is UCP the same as OpenAI's ACP?
No. UCP (Shopify + Google) is a broad open commerce protocol. ACP (OpenAI + Stripe) targets ChatGPT Instant Checkout. Shopify's Agentic Storefronts abstract both—you don't implement each separately.
How often should I refresh product data for agentic commerce?
Audit top SKUs quarterly. Update within 90 days of material changes (variants, pricing, compatibility). Stale specs hurt both agents and Google.