Google and Shopify just rewired how AI agents buy products. In January 2026, they launched the Universal Commerce Protocol (UCP) at the National Retail Federation conference — three months after OpenAI and Stripe debuted the Agentic Commerce Protocol (ACP) powering ChatGPT's Instant Checkout. McKinsey projects agentic commerce will orchestrate up to $1 trillion in U.S. retail revenue by 2030. These two protocols are the rails it will run on — and Shopify merchants are caught in the middle.
What Are UCP and ACP?
Both protocols solve the same fundamental problem: AI agents need a standardized way to find products, present them to shoppers, and complete transactions. But they approach it from different directions.
UCP (Universal Commerce Protocol) is Google and Shopify's open standard for agentic commerce. It standardizes the entire shopping journey — discovery, checkout, order management, and post-purchase support — across any AI surface. Think of it as a universal language that lets any AI agent (Gemini, Copilot, or even third-party bots) interact with any merchant's catalog, checkout flows, loyalty programs, and fulfillment systems.
ACP (Agentic Commerce Protocol) is OpenAI and Stripe's infrastructure for enabling purchases inside conversational AI. It's laser-focused on one thing: making checkout seamless within chat. When a ChatGPT user says "buy that moisturizer," ACP handles the payment flow using Stripe's Shared Payment Tokens — single-use, time-bound credentials that never expose the buyer's actual payment details.
The distinction matters for Shopify merchants because each protocol determines how and where your products get discovered and sold. If your product data isn't optimized for AI discovery, neither protocol can help you.
Head-to-Head: UCP vs ACP Comparison
Here's how the two protocols stack up across the dimensions that matter most to merchants:
| Feature | UCP (Google + Shopify) | ACP (OpenAI + Stripe) |
|---|---|---|
| Launched | January 2026 | September 2025 |
| Scope | Full commerce lifecycle | Checkout-focused |
| Discovery | Decentralized (/.well-known/ucp) + Google Merchant Center | Centralized (OpenAI approval required) |
| AI Surfaces | Gemini, Google AI Mode, Copilot, third-party agents | ChatGPT (800M+ weekly users) |
| Payments | Google Pay, PayPal, Visa, Mastercard, Amex | Stripe Shared Payment Tokens, PayPal |
| Transport | REST, MCP, A2A | REST API |
| Shopify Integration | Native via Agentic Storefronts | Via Shopify Catalog (rolling out) |
| License | Open standard (industry coalition) | Apache 2.0 (open source) |
| Key Partners | Walmart, Target, Etsy, Wayfair, Best Buy, Visa | Etsy, Salesforce, Shopify (coming) |
The critical takeaway: UCP is broader, ACP is deeper. UCP wants to be the universal language for all agentic commerce. ACP wants to make buying inside a conversation as frictionless as tapping "Buy Now." As checkout.com's analysis puts it, "UCP makes commerce accessible, and ACP makes commerce autonomous."
How Each Protocol Handles Discovery
Discovery is where Generative Engine Optimization becomes critical — and where the two protocols diverge sharply.
UCP's approach is architecturally decentralized. Merchants can publish their commerce capabilities at /.well-known/ucp on their own domain, similar to how robots.txt works for search crawlers. AI agents can theoretically discover merchants without going through any central authority. In practice, most merchants selling on Gemini still go through Google Merchant Center, but the open discovery mechanism exists for any agent to use.
ACP's approach is centralized and curated. Merchants who want to sell in ChatGPT apply through OpenAI. OpenAI approves them. Products appear in ChatGPT's shopping interface. This creates quality control but limits access — especially for smaller Shopify stores that haven't been approved yet.
For Shopify merchants, this means your optimization strategy must account for both models. Your structured data and schema markup need to be machine-readable for UCP's open crawlers, while your product feed data needs to meet the curated standards that get you into ChatGPT's Instant Checkout.
The stores winning AI visibility today are the ones treating product data as their most valuable asset — not an afterthought. Google AI Overviews already pull from structured product data, and UCP takes this dependency even further.
What Shopify Merchants Get for Free
Here's the good news: if you're on Shopify, you're already closer to UCP readiness than you think.
Shopify's Agentic Storefronts, launched as part of the Winter '26 Edition, give every merchant a one-toggle path to selling on AI platforms. Configure once in your Shopify Admin, and you're live on ChatGPT, Perplexity, Google AI Mode, Gemini, and Microsoft Copilot.
Behind the scenes, Shopify Catalog — the infrastructure powering Agentic Storefronts — uses specialized LLMs to automatically categorize and enrich your product data so AI agents can understand it. This is the same backbone that feeds both UCP (for Gemini/Google AI Mode) and ACP (for ChatGPT Instant Checkout).
What merchants still need to do:
- Enable Agentic Storefronts in Shopify Admin and select which AI platforms to sell on
- Optimize product data with specific, factual descriptions (not marketing fluff) — the same principles that drive ChatGPT shopping recommendations
- Set up the Knowledge Base App to control FAQs, policies, and brand voice that agents relay to customers
- Review your schema markup to ensure structured data is comprehensive and accurate
The merchants who take the time to optimize their product data for AI consumption will be the ones AI agents recommend. The ones relying on default catalog data will be invisible — just like in traditional AI search.
The Walmart Playbook: Why "Both" Is the Right Answer
If you're wondering whether to focus on UCP or ACP, look at what the largest retailers are doing. Walmart isn't picking sides — and neither should you.
In October 2025, Walmart integrated with ChatGPT's Instant Checkout via ACP, making apparel, entertainment, packaged foods, and marketplace items available for purchase directly in chat. By August 2025, ChatGPT already accounted for 20% of Walmart's referral traffic, according to Similarweb data.
Then in January 2026, Walmart announced a Gemini integration through UCP, letting Google's AI automatically surface Walmart and Sam's Club products when relevant.
PayPal is taking the same dual-protocol approach. In a January 2026 blog post from PayPal's newsroom, the company framed the landscape into three layers: commerce protocols (ACP, UCP), payment and trust protocols (Google AP2, Visa's Trusted Agent Protocol), and infrastructure protocols (A2A, MCP).
The lesson for Shopify merchants: this is not VHS vs Betamax. It's iOS vs Android — two ecosystems, both successful, and the smart play is building for both. Your product data quality is the shared foundation that makes both protocols work. As we've written about in our 2026 predictions piece, the merchants who invest in AI-readable product content now will capture disproportionate share as agentic commerce scales.
The B2B Angle: Why This Goes Beyond DTC
While the headlines focus on consumer shopping, the B2B implications are even more staggering. Gartner predicts that by 2028, 90% of B2B buying will be AI agent-intermediated, pushing over $15 trillion in B2B spend through AI agent exchanges.
Both UCP and ACP are designed with extensibility that supports B2B workflows — procurement, subscription management, and complex fulfillment scenarios. UCP's Extensions Layer already handles domain-specific features like split shipments, pre-orders, and delivery windows.
For Shopify merchants selling wholesale or operating in B2B, this means your digital catalogs need to be designed for machines as much as humans. Link building and authority signals become even more critical when AI procurement agents are evaluating supplier credibility programmatically.
Your 5-Step Agentic Commerce Action Plan
Here's what to do right now to prepare your Shopify store for both protocols:
1. Audit your product data quality. AI agents extract specific, factual content — not marketing copy. Every product needs exact specifications, ingredients, dimensions, use cases, and pricing. A comprehensive AI search audit can identify exactly where your data falls short.
2. Enable Shopify Agentic Storefronts. Go to Shopify Admin, toggle on the AI platforms you want to sell on. This connects you to both UCP (Google/Gemini) and ACP (ChatGPT) through Shopify's infrastructure.
3. Implement comprehensive schema markup. Both protocols rely on structured data to understand your products. Product schema, FAQ schema, review schema, and organization schema all feed into how AI agents evaluate and present your store.
4. Build your Knowledge Base. Use Shopify's Knowledge Base App to define shipping policies, return windows, brand voice, and FAQs. This is the data AI agents relay to customers when they ask follow-up questions during checkout.
5. Monitor AI-driven traffic and conversions. Adobe data shows AI-referred visitors have 32% longer sessions and 27% lower bounce rates than non-AI traffic. By October 2025, they were converting 16% higher. Track this channel separately — it's growing fast and behaves differently from traditional search.
Tools like PageX can help with steps 1 and 3 — auditing your store's AI visibility and identifying exactly which product data gaps are keeping you out of AI agent recommendations across platforms like ChatGPT, Perplexity, and Gemini.
Is Your Store Ready for Agentic Commerce?
Get a free AI visibility audit to see how your products appear to AI shopping agents — and what's missing from your structured data.
Frequently Asked Questions
Do I need to choose between UCP and ACP for my Shopify store?
No. The industry consensus — echoed by Walmart, PayPal, and every major retailer — is to support both. UCP powers Google AI Mode and Gemini shopping, while ACP powers ChatGPT's Instant Checkout. As a Shopify merchant, Agentic Storefronts let you connect to both ecosystems from a single setup. The real work isn't choosing a protocol; it's ensuring your product data is complete, specific, and machine-readable enough for either system to use.
Is UCP available to all Shopify merchants right now?
Yes. Shopify's Agentic Storefronts are available to every Shopify merchant through the admin panel. Enable it, select your AI platforms, and Shopify Catalog handles the data syndication. For ChatGPT specifically, Shopify is rolling out Instant Checkout access to over a million merchants, though some features are still in phased rollout. Google AI Mode checkout via UCP is also launching on Google surfaces soon.
How does agentic commerce affect my existing SEO strategy?
Agentic commerce amplifies the importance of product data quality and structured markup — the same foundations that drive AI search visibility. If your products already appear in AI Overviews and ChatGPT recommendations, you're well-positioned. The new protocols add a transactional layer on top of discovery, meaning AI agents can now complete purchases — not just recommend products. This makes Generative Engine Optimization even more directly tied to revenue.
What payment methods do UCP and ACP support?
UCP supports Google Pay at launch, with PayPal integration coming soon, and is designed to work with Visa, Mastercard, American Express, and other providers through Google's Agent Payments Protocol (AP2). ACP uses Stripe's Shared Payment Tokens and also supports PayPal. Both protocols are payment-provider agnostic by design, meaning more options will be added as the ecosystem matures.
Will Amazon adopt either protocol?
Amazon is currently taking the opposite approach. Rather than integrating with external AI shopping agents, Amazon is blocking AI crawlers from platforms like ChatGPT and Google while building its own closed ecosystem through Rufus, its in-house AI assistant. ChatGPT referrals to Amazon dropped 18% month-over-month as of August 2025, sitting at sub-3%. This creates a significant opportunity for Shopify merchants on open protocols to capture AI-driven shopping demand that Amazon is actively rejecting.