Five different protocols now control how AI agents find, recommend, and sell your products. If that sentence made your eyes glaze over, you’re not alone.
The agentic commerce protocol landscape is moving fast. Google launched UCP in January 2026. Anthropic donated MCP to the Linux Foundation. OpenAI and Stripe built ACP. Google added AP2 for payments. And A2A has been quietly connecting AI agents since 2025.
Here’s the problem: 45% of consumers already use AI for some part of their buying journey (IBM, 2026). These protocols determine whether your products show up when those consumers ask an AI to find something. If you don’t know how they work, you can’t prepare for where ecommerce is heading.
This guide breaks down each protocol in plain terms: what it does, who built it, and which ones actually require action from Shopify merchants right now.
What Are Ecommerce Protocols and Why Do They Matter?
Ecommerce protocols are standardized rules that allow AI agents to interact with online stores. They define how agents discover products, communicate with other agents, process transactions, and handle payments.
Think of it like this: when you visit a website, your browser uses HTTP to load pages and HTTPS to keep data secure. You never think about those protocols. They just work in the background. Ecommerce protocols do the same thing for AI shopping agents.
Without them, every AI platform would need a custom integration with every store. That doesn’t scale. Protocols create a common language so that ChatGPT, Gemini, Perplexity, and any future AI assistant can all interact with your Shopify store using the same standardized methods.
The $3-5 trillion global agentic commerce opportunity projected by 2030 depends entirely on these protocols working together (McKinsey, 2025).
Here’s the landscape in 2026:
| Protocol | Created By | Purpose | Layer |
|---|---|---|---|
| UCP | Google + Shopify | End-to-end shopping transactions | Commerce |
| MCP | Anthropic (Linux Foundation) | AI agent access to data and tools | Data Integration |
| A2A | Google (Linux Foundation) | Agent-to-agent communication | Coordination |
| ACP | OpenAI + Stripe | ChatGPT checkout | Commerce (ChatGPT) |
| AP2 | Secure agent-initiated payments | Payments |
Each protocol handles a different part of the AI shopping experience. Let’s break them down.
What Is UCP (Universal Commerce Protocol)?
UCP is an open commerce protocol that manages the full shopping lifecycle, from product discovery through checkout and fulfillment, across any AI platform.
Google and Shopify co-developed UCP and launched it in January 2026. It’s the broadest commerce protocol available, covering everything from browsing products to processing returns.
How UCP Works
When an AI agent needs to help someone shop, UCP provides the structure for:
- Discovery: Agent queries product catalogs to find relevant items
- Comparison: Agent retrieves pricing, availability, reviews, and specifications
- Checkout: Agent initiates secure payment on behalf of the shopper
- Fulfillment: Agent tracks shipping and delivery status
- Post-purchase: Agent handles returns, exchanges, and support
UCP is designed to be protocol-agnostic. It has built-in support for MCP (data access), A2A (agent communication), and AP2 (payments). This makes it more of an orchestration layer than a standalone protocol.
UCP and Shopify
Shopify has native UCP integration through the Shopify Admin. When you enable Agentic Storefronts, your product catalog becomes queryable through UCP-compatible AI agents.
20+ companies endorsed UCP at launch, including Visa, Mastercard, Stripe, Target, Walmart, and Wayfair (Google Developers Blog, 2026). That’s not a small list. It signals that major retailers and payment processors are building around this standard.
What Is MCP (Model Context Protocol)?
MCP is a standardized interface that lets AI agents access external data, tools, and APIs. Think of it as the “USB-C port” for AI. It provides a universal connection between AI models and the systems they need to work with.
Anthropic created MCP and later donated it to the Linux Foundation’s Agentic AI Foundation. Unlike UCP, MCP isn’t commerce-specific. It’s a general-purpose data access layer used across industries.
How MCP Works
MCP provides three core functions for AI agents:
- Resources: Access to external data (product catalogs, inventory feeds, CRM records)
- Tools: Ability to call functions and APIs (place orders, update records, generate reports)
- Prompts: Pre-built templates for common interactions
For ecommerce, this means an AI agent can use MCP to read your Shopify product data, check inventory levels, and pull customer order history, all through a standardized connection.
MCP Adoption
MCP is the most widely adopted protocol on this list. The numbers tell the story:
- 10,000+ active MCP servers deployed across industries (MCP Manager, 2026)
- 97 million+ monthly SDK downloads (Pento, 2025)
- Supported by Claude, ChatGPT, Gemini, VS Code, Cursor, and GitHub Copilot
MCP and Shopify
Shopify built MCP support directly into every store. Your store has a /api/mcp endpoint that AI agents can use to access your product data. When an agent connects through MCP, it can read your catalog, check stock levels, and retrieve product details in a format it understands.
This is different from UCP. MCP handles the data access. UCP handles the transaction. They work together: MCP lets the agent see your products, and UCP lets the agent sell them.
What Is A2A (Agent2Agent Protocol)?
A2A is a protocol that lets AI agents discover, communicate with, and delegate tasks to other AI agents. While MCP connects agents to data and UCP handles transactions, A2A connects agents to each other.
Google created A2A and donated it to the Linux Foundation. It has 50+ partners in its ecosystem (Google, 2025).
How A2A Works in Shopping
Here’s a real scenario:
- You ask your personal AI assistant to find a birthday gift under $100
- Your agent uses A2A to contact a shopping specialist agent
- The shopping agent uses MCP to query product catalogs from multiple stores
- The shopping agent uses A2A to contact a payment agent for secure checkout
- The payment agent uses UCP to complete the transaction
No single agent does everything. A2A allows specialized agents to collaborate, each handling what they’re best at. It’s like having a team of specialists instead of one generalist.
A2A Agent Cards
Every A2A-compatible agent publishes an “Agent Card” that describes:
- What the agent can do (capabilities)
- What data formats it accepts (input schema)
- How to communicate with it (endpoint)
This is how agents find and evaluate each other. A shopping agent can discover a price comparison agent, check its capabilities, and decide whether to delegate work to it.
What About ACP and AP2?
Two other protocols round out the picture. They’re narrower in scope but important for specific use cases.
ACP (Agentic Commerce Protocol)
ACP is OpenAI and Stripe’s commerce protocol, designed specifically for purchases inside ChatGPT. It’s the protocol behind ChatGPT’s Instant Checkout feature.
ACP is open source (Apache 2.0 license) and currently powers transactions for Etsy sellers inside ChatGPT. Over 1 million Shopify merchants are announced for the next phase of rollout (OpenAI, 2025).
ACP vs UCP: Both handle commerce, but they serve different platforms. UCP is platform-agnostic (works with any AI). ACP is optimized for ChatGPT specifically. Given ChatGPT’s 700 million+ weekly users (OpenAI, 2025), ACP represents a massive distribution channel even with its narrower scope.
| Feature | UCP | ACP |
|---|---|---|
| Creator | Google + Shopify | OpenAI + Stripe |
| Platform | Any AI platform | ChatGPT primarily |
| Scope | Full lifecycle | Checkout focused |
| Payment | Google Pay, AP2 | Stripe |
| Status | Launched Jan 2026 | Beta, expanding |
| Shopify support | Native via Admin | Coming (1M+ merchants) |
AP2 (Agent Payments Protocol)
AP2 is Google’s open protocol for secure agent-initiated payment transactions. It solves a specific problem: how do you let an AI agent spend your money safely?
AP2 introduces “Mandates,” which are cryptographically signed payment contracts. When you authorize an AI agent to make a purchase, the Mandate defines exactly what it can buy, how much it can spend, and for how long. The agent can’t exceed those boundaries.
60+ organizations are collaborating on AP2, including major payment processors and financial institutions (Google Cloud Blog, 2026).
AP2 works within UCP as the payment security layer. When a UCP transaction reaches the payment step, AP2 handles the authorization and fraud protection.
How All Five Protocols Work Together
These protocols aren’t competing. They’re layers in a stack, each handling a different part of the AI shopping experience.
Here’s how they interact during a single AI-powered purchase:
- MCP reads product data from your Shopify store (data access layer)
- A2A coordinates between your personal agent and specialized shopping agents (coordination layer)
- UCP manages the shopping transaction from discovery to checkout (commerce layer)
- AP2 secures the payment with cryptographic authorization (payment layer)
- ACP provides an alternative commerce path for ChatGPT-specific purchases (ChatGPT layer)
The good news: as a Shopify merchant, you don’t need to implement each protocol separately. UCP has built-in support for MCP, A2A, and AP2. Shopify’s native integrations handle most of the technical work.
| Layer | Protocol | What It Does | Merchant Action |
|---|---|---|---|
| Data Access | MCP | Agents read your product data | Enable via Shopify’s /api/mcp |
| Coordination | A2A | Agents talk to other agents | No direct merchant action |
| Commerce | UCP | Full shopping lifecycle | Enable Agentic Storefronts |
| Commerce (ChatGPT) | ACP | ChatGPT Instant Checkout | Connect Stripe account |
| Payments | AP2 | Secure agent payments | Handled by payment processor |
Which Protocols Do Shopify Merchants Need to Prepare For?
Not all five protocols require action from you. Here’s a practical breakdown of what actually needs your attention.
High Priority: UCP and MCP
These two protocols directly depend on your product data quality. If your data is poorly structured, AI agents won’t find your products regardless of which protocol they use.
What to do now:
- Enable Agentic Storefronts in Shopify Admin (covers UCP)
- Verify your /api/mcp endpoint is active (covers MCP)
- Optimize product data for AI readability:
- Write descriptive titles (include material, size, use case)
- Complete all product metafields
- Add structured data (JSON-LD schema)
- Ensure accurate inventory and pricing
- Set up your Shopify Catalog so AI agents can query your full product range
Your product data quality is the single biggest factor in whether AI agents recommend your products. Protocols are the roads. Your product data is the vehicle.
Medium Priority: ACP
If you use Stripe for payments (most Shopify merchants do), ACP activation should be straightforward when Shopify rolls out support. With 700 million+ weekly ChatGPT users, this is a sales channel worth preparing for.
What to do now:
- Ensure your Stripe account is in good standing
- Watch for Shopify’s ACP announcement
- Review your product titles and descriptions (ChatGPT pulls from the same structured data)
Low Priority: A2A and AP2
These protocols work behind the scenes. You don’t configure them directly. A2A handles agent-to-agent communication, and AP2 handles payment security. Both are managed by the platforms and payment processors, not by merchants.
What to do: Nothing specific. These protocols benefit you automatically when you’ve set up UCP and MCP correctly.
How These Protocols Affect Your Ad Attribution and ROAS
Here’s something most protocol explainers miss: when AI agents handle purchases through UCP or ACP, your traditional ad attribution breaks.
A customer might see your Google Shopping ad on Monday, then ask ChatGPT to “buy that moisturizer I saw” on Wednesday. ChatGPT processes the purchase through ACP. Your Google Ads dashboard shows a click with no conversion. Your analytics show a sale with no source.
$20.9 billion in retail spending already flows through AI platforms, representing 1.5% of total retail spending (eMarketer, 2025). As that number grows, the gap between your ad spend and your tracked conversions will widen.
UCP is designed to help with this. It includes attribution data in the transaction payload, so merchants can trace which AI platform and which agent initiated the sale. But this requires:
- Server-side tracking set up on your store
- UCP attribution fields mapped in your analytics
- Cross-platform reporting that stitches together ad impressions and agent-initiated sales
If you’re running paid ads on Shopify, understanding how these protocols handle attribution isn’t optional. It’s how you protect your ROAS as agentic commerce grows.
What’s Coming Next for Ecommerce Protocols
The protocol landscape is evolving quickly. Here’s what to watch:
Convergence around UCP: Google designed UCP to incorporate MCP, A2A, and AP2. As more platforms adopt UCP, it’s likely to become the default commerce protocol. Shopify’s early investment here gives merchants an advantage.
ACP expansion beyond ChatGPT: OpenAI’s protocol is open source, which means other platforms could adopt it. But with UCP backed by Google, Shopify, Visa, Mastercard, and 20+ other companies, ACP may remain ChatGPT-focused.
AI agent adoption accelerating: 40% of enterprise apps will include AI agents by end of 2026 (Gartner, 2026). That means more agents querying your products through these protocols, more frequently.
39% of consumers already use AI for product discovery (Ekamoira, 2026). By the time most merchants start paying attention to protocols, the early movers will have already captured the AI-driven demand.
Frequently Asked Questions
What is the difference between UCP and MCP for ecommerce?
UCP handles the full shopping transaction lifecycle (discovery, checkout, fulfillment) while MCP handles data access (reading product catalogs, checking inventory). They work together: MCP lets AI agents see your products, UCP lets them sell your products.
Do I need to implement all five protocols on my Shopify store?
No. Shopify handles most protocol integration natively. Focus on enabling Agentic Storefronts (UCP) and verifying your /api/mcp endpoint (MCP). A2A and AP2 operate behind the scenes. ACP support through Shopify is rolling out.
Which protocol does ChatGPT use for shopping?
ChatGPT uses ACP (Agentic Commerce Protocol) developed by OpenAI and Stripe. It powers the Instant Checkout feature where users buy products without leaving the chat. Over 1 million Shopify merchants are in the ACP rollout pipeline.
How does A2A differ from MCP?
MCP connects AI agents to data and tools (vertical integration). A2A connects AI agents to other AI agents (horizontal coordination). An agent uses MCP to read your product catalog and A2A to delegate shopping tasks to a specialist agent.
Will UCP replace MCP?
No. UCP incorporates MCP rather than replacing it. UCP uses MCP for data access, A2A for agent coordination, and AP2 for payment security. They’re complementary layers, not competitors.
What is AP2 and why does it matter for Shopify merchants?
AP2 (Agent Payments Protocol) secures AI-initiated payments using cryptographic Mandates that limit what an agent can purchase, spend, and for how long. It protects both merchants and consumers from unauthorized agent transactions.
How do ecommerce protocols affect my Google Ads performance?
When AI agents process purchases through UCP or ACP, traditional pixel-based tracking breaks. Sales happen without pageviews, so your ad dashboard may show clicks without conversions. UCP includes server-side attribution data to help bridge this gap.
Are these protocols only for large Shopify stores?
No. Protocol support is built into every Shopify plan through Agentic Storefronts and the /api/mcp endpoint. Small stores benefit equally because AI agents query product data based on quality and relevance, not store size.
What should I do first to prepare for agentic commerce protocols?
Start with your product data. Optimize titles, complete metafields, add structured data, and ensure accurate inventory. Then enable Agentic Storefronts in Shopify Admin. Protocol readiness starts with data readiness.
Is UCP compatible with non-Shopify platforms?
Yes. UCP is an open protocol. While Shopify has native integration, any ecommerce platform can implement UCP. Google designed it to be platform-agnostic and compatible with MCP, A2A, and AP2.
Take Action Before the Protocols Take Over
The protocol landscape looks complicated from the outside, but for Shopify merchants, it comes down to three actions:
-
Get your product data right. Every protocol depends on structured, complete, accurate product information. This is the foundation that makes everything else work.
-
Enable Agentic Storefronts. This activates UCP support and makes your catalog visible to AI shopping agents across platforms.
-
Watch for ACP rollout. When Shopify announces ACP support for your store, connect your Stripe account to tap into ChatGPT’s 700 million weekly users.
The merchants who prepare now will show up in AI recommendations when $20.9 billion in AI-driven retail spending grows to tens of billions more. The merchants who wait will wonder why their products stopped appearing in search results that no longer look like search results.
Protocols are the infrastructure. Your product data is the competitive advantage. Start there.
About Akash Radadiya
