UCP vs ACP: The Two Protocols That Decide Which Stores AI Agents Recommend
In January 2026, Google and Shopify announced the Universal Commerce Protocol (UCP). Months earlier, OpenAI and Stripe launched the Agentic Commerce Protocol (ACP). These two protocols now control which online stores AI agents can discover, recommend, and transact with. If your store does not support at least one of them, AI shopping agents cannot see you. Period.
This is not a theoretical shift. Google AI Mode is already live. ChatGPT Instant Checkout is processing real transactions. Walmart checks out inside Gemini. JD Sports sells through ChatGPT. The infrastructure for AI-mediated commerce is built and running. The question is no longer whether this matters. It is whether your store is visible to it.
For competitive intelligence teams, these protocols open a new dimension of analysis. UCP manifests are designed to be public and crawlable, which makes them rich with strategic data as adoption grows. This article breaks down both protocols, explains what they reveal about your competitors, and shows how to use that intelligence.
What Is UCP (Universal Commerce Protocol)?
UCP is the robots.txt of AI commerce. Co-developed by Google and Shopify with partners including Target, Walmart, Etsy, and Wayfair, UCP provides a standardized way for merchants to declare their commerce capabilities to any AI agent.
The protocol works through a publicly accessible manifest file at /.well-known/ucp on a merchant's domain. Any AI agent can fetch this file to discover what the merchant supports: checkout, catalog search, order management, payment methods, fulfillment options, and more.
What Lives Inside a UCP Manifest
- Services: Which commerce APIs the merchant exposes (REST, MCP, Agent-to-Agent)
- Capabilities: Checkout, discounts, fulfillment, loyalty programs, order management, catalog search
- Payment handlers: Google Pay, Shop Pay, Apple Pay, PayPal, Stripe, Adyen
- Signing keys: Cryptographic identity verification for secure transactions
- Extension support: Delivery windows, subscription schedules, split shipments, pre-orders
Think of it as a machine-readable business card that tells AI agents exactly what the store can do and how to interact with it.
The critical detail: UCP manifests are public by design. They must be accessible so that AI agents can discover merchants without pre-configured integrations. This creates a massive competitive intelligence opportunity that most brands have not realized yet.
What Is ACP (Agentic Commerce Protocol)?
ACP is OpenAI and Stripe's answer to the same problem, but with a different philosophy. Where UCP is decentralized (any agent can crawl any merchant), ACP is centralized (merchants apply through ChatGPT's commerce onboarding and configure through Stripe's dashboard).
UCP vs ACP: Key Differences
| Dimension | UCP (Google + Shopify) | ACP (OpenAI + Stripe) |
|---|---|---|
| Discovery | Decentralized (/.well-known/ucp) | Centralized (ChatGPT onboarding) |
| Primary surface | Google AI Mode, Gemini, Copilot | ChatGPT Instant Checkout |
| Payment | Modular (Google Pay, Shop Pay, etc.) | Stripe required (4% fee) |
| Cart support | Multi-item carts | Single-item only (currently) |
| Geography | Global from launch | US only for instant checkout |
| Product catalog | Google Shopping Graph (50B+ listings) | Merchant feeds only |
| Manifest visibility | Public and crawlable | Private (Stripe dashboard) |
The industry consensus is clear: implement both. Google launched UCP at NRF 2026 with 25+ partners including Shopify, Walmart, Target, Visa, and Mastercard. UCP captures top-of-funnel discovery intent (people searching via Google and Gemini). ACP captures deep-funnel conversational intent (people buying inside ChatGPT). Supporting both protocols captures the widest agentic traffic surface.
Why This Matters for Competitive Intelligence
Here is where it gets strategic for eCommerce teams. Because UCP manifests are designed to be public and discoverable, they turn into crawlable competitive intelligence as the standard settles. That opens a view into which capabilities your competitors have enabled, which payment methods they support, which transport protocols they expose, and how current their implementation is.
What You Can Learn From Competitor UCP Manifests
- 1. Agentic commerce readiness: Does the competitor have a UCP manifest at all? If yes, they are visible to AI shopping agents. If not, they are invisible.
- 2. Capability depth: Do they support checkout only, or do they also expose catalog search, order management, loyalty programs, and fulfillment options?
- 3. Payment infrastructure: Which payment handlers do they list? Google Pay, Shop Pay, Apple Pay, PayPal? More payment options means more ways AI agents can complete transactions.
- 4. Transport protocols: REST only, REST + MCP, or REST + MCP + Agent-to-Agent? More protocols means more types of AI agents can interact with them.
- 5. Fulfillment sophistication: Delivery windows, subscription schedules, split shipments, in-store pickup. Richer fulfillment data gets preferential treatment from AI agents.
- 6. Version freshness: Are they on the latest spec version or lagging behind? Outdated versions signal abandoned implementations.
This is the kind of intelligence that was previously impossible to extract at scale. As UCP standardizes, it becomes structured and public by design - waiting to be analyzed the moment the discovery mechanism settles.
Where RivalSweeper Fits in UCP Adoption
RivalSweeper already monitors 30+ competitive dimensions for eCommerce brands and grades every store on its Agent Readiness Score. As UCP matures into a stable, reliably discoverable standard, manifest analysis is a natural next layer to sit alongside that score. Here is how we think about it.
What This Looks Like Alongside Agent Readiness
- Agent Readiness today: RivalSweeper grades AI-shopping-agent readiness across six signals - structured data, technical access, content quality, trust signals, AI-agent files, and brand-entity recognition - for you and your competitors. That is the live foundation any UCP tracking would build on. See the Agent Readiness Score.
- UCP manifest presence (as the standard stabilizes): Once UCP settles on a stable, publicly discoverable manifest location, presence detection across monitored competitor domains becomes a signal worth surfacing - so you would know when a competitor turns on agent visibility and you have not.
- Capability and payment comparison: Where merchants publish machine-readable manifests, a side-by-side view of the capabilities they support (checkout, discounts, fulfillment, loyalty, orders, catalog) and the payment handlers they expose is exactly the kind of structured intelligence that layer would add.
- Change detection: Alerts when competitors add new capabilities, upgrade protocol versions, or enable new payment methods - the same change-detection engine RivalSweeper already runs on other competitive dimensions, extended to agentic-commerce signals as they become reliably observable.
What You Should Do Right Now
1. Understand Where UCP Discovery Is Heading
UCP is designed around a publicly discoverable manifest so any AI agent can find a merchant without a pre-built integration. The exact discovery location is still being standardized, so treat any single path as provisional for now. The takeaway that matters: agent-visibility is becoming a public, checkable signal, and you want to be on the right side of it before your competitors are.
2. Audit Your Structured Data Foundation
UCP builds on top of structured data. If your Schema.org markup is incomplete, AI agents cannot understand your products even before UCP enters the picture. Product, Offer, AggregateRating, BreadcrumbList, MerchantReturnPolicy, and ShippingDetails are the minimum.
3. Implement Both Protocols
If you are on Shopify, UCP is available through Agentic Storefronts. For custom platforms, you need to implement the REST endpoints, publish the manifest, and configure Google Merchant Center. For ACP, integrate with Stripe and apply through ChatGPT commerce onboarding. Dual implementation captures the widest agentic traffic.
4. Monitor Competitor Adoption Continuously
This is not a one-time check. Agentic-commerce adoption is moving every week. RivalSweeper already monitors competitors in real time across 30+ dimensions and alerts you when the landscape shifts, and agentic-commerce signals join that monitoring as they become reliably observable. For a broader framework covering all 10 essential Competitive Intelligence capabilities for 2026, read our complete eCommerce competitive intelligence guide.
The Bottom Line
UCP and ACP are not optional upgrades. They are the new gatekeepers of AI-mediated commerce. Reddit now reaches 121 million daily active users and 471 million weekly, with AI search users surging 30% year-over-year to 80 million. Reddit launched AI-powered shopping carousels in February 2026, while the competitive intelligence tools market is projected to reach $96 billion by 2030 (Research and Markets). If your store is not structured for these protocols, you are invisible to the fastest-growing distribution channel in eCommerce.
The brands that move first do not just get a technology advantage. They get a compounding data advantage as AI agents learn to prefer merchants with richer, more complete commerce capabilities. Every week you wait, your competitors' agentic commerce moat gets deeper.
Want to see where you stand? Book a demo and we will show you your Agent Readiness Score and where you stand on agentic-commerce readiness. Or start with our Agent Readiness feature to see the full picture.