A2A. MCP. WebMCP. UCP. AP2. What are all of these protocols? Do they actually matter for SEO?
A2A (Agent-to-Agent)
- Purpose: Helps different AI agents communicate and work with each other
- What it does: A2A connects agents to other agents. The A2A protocol defines how agents can discover each other via JSON Agent Cards. The Cards list capabilities, skills, and basic information. This protocol also helps agents manage and coordinate tasks.
Model Context Protocol (MCP)
- Purpose: Connect AI agents to external tools, data, and APIs
- What it does: MCP connects an agent to tools and data. Website owners set up an MCP server that defines what data and actions are available on their website (like searching products) using JSON schema. The MCP server translates the AI agent’s standardized requests into the site’s backend format and returns structured responses.
Web Model Context Protocol (WebMCP)
- Purpose: Allow websites to present structured actions directly to AI agents through the browser
- What it does: WebMCP is a new client-side browser API that helps AI agents interact with web pages more reliably. Instead of scraping screenshots or parsing raw HTML, agents can invoke defined actions directly. There are two APIs:
- Declarative adds new attributes to existing HTML form elements.
- Imperative uses JavaScript to register complex tools.
Learn how to get started with WebMCP.
Agent Payments Protocol (AP2)
- Purpose: Enable AI agents to securely make payments on behalf of users
- What it does: AP2 is an open protocol designed to allow AI agents to initiate payments with verifiable authorization and accountability. In other words, it helps the agent make sure it has permission before actually completing a transaction. AP2 supports “human-present” and “human-not-present” shopping scenarios.
Universal Commerce Protocol (UCP)
- Purpose: Standardize AI-agent-driven commerce across the entire shopping journey
- What it does: UCP provides a unified standard for AI agents to handle product discovery, checkout, payment, and post-purchase support across participating merchants. Google plans to use UCP with Merchant Center in AI Mode and the Gemini apps.
Agentic Commerce Protocol (ACP)
- Purpose: Facilitate transactions between buyers, AI agents, and merchants
- What it does: ACP defines how merchants provide product feeds for discovery, maintain accurate pricing and inventory, and enable agent-initiated checkout within AI interfaces. ACP currently powers Instant Checkout in ChatGPT.
Bottom line: Agentic AI shifts the focus from information discovery to action discovery. These protocols are the tools we have available to make a website’s actions and data accessible to AI agents.