Discover our GTM Flywheel: Content, Ads & Outbound working as oneLearn more
Skip to main content
MCP Server

Customer.io MCP

Jun 21, 2026
About

The Customer.io MCP Server enables seamless AI-powered integration with Customer.io, supporting segment creation, campaign insights, and SDK troubleshooting to enhance your app's messaging and user engagement.

How to Use

How to Install and Use Customer.io MCP

Customer.io MCP (Model Context Protocol) Server lets you connect your AI tools, like ChatGPT or Claude, with your Customer.io workspace. This helps you manage your customer data, segments, and campaigns through AI with ease. Here's a simple guide on how to install and use Customer.io MCP.

Step 1: Enable MCP Support for Your Account

Before using Customer.io MCP, you need to turn it on in your account settings. Only account admins have permission to do this.

To enable MCP support:

  1. Go to Settings > Privacy, Data, & AI in your Customer.io dashboard.
  2. Turn on Customer.io AI.
  3. Then, turn on Customer.io MCP.

You must enable both Customer.io AI and Customer.io MCP for everyone on your team who wants to use it.

Step 2: Set Up Your MCP Client

Now that MCP support is enabled, you need to connect your favorite AI tool or IDE (like ChatGPT, Claude Desktop, or Cursor) to the MCP server.

For ChatGPT:

  • Make sure you have a Pro, Plus, or Enterprise plan.

  • Enable Developer Mode under ChatGPT Settings > Connectors.

  • Add the Customer.io MCP server connector:

    • The MCP Server URL is https://mcp.customer.io/mcp for US accounts or https://mcp-eu.customer.io/mcp for EU accounts.
    • Use OAuth to authenticate.
  • Select the workspaces you want to connect and allow access.

Example:

{
  "name": "Customer.io",
  "description": "Access Customer.io workspace data and tools",
  "url": "https://mcp.customer.io/mcp",
  "authentication": "OAuth"
}

For Claude Desktop or Similar Tools:

  1. Go to Settings > Personal Settings in Customer.io.
  2. Click Claude Desktop tab and enable support.
  3. Add a custom connector with the MCP server URL (https://mcp.customer.io/mcp or EU URL).
  4. Choose the workspaces to use.

For Cursor or Other IDEs:

Add this JSON to your MCP config file to connect:

{
  "mcpServers": {
    "CustomerIO": {
      "type": "http",
      "url": "https://mcp.customer.io/mcp"
    }
  }
}

After adding, login to authenticate and pick your workspace.

Step 3: Find and Use Your Workspace

When using the MCP server, you need your workspace info to perform actions. You can find your workspace name in the top-left of the Customer.io dashboard. The workspace ID is visible in the URL on the dashboard page.

Use this workspace info when prompting your AI tools.

Step 4: Use MCP Tools and Actions

Customer.io MCP offers various tools to work with your data. Here is a brief overview:

  • create: Create segments (create_segment).
  • get: Get info about profiles, campaigns, segments, and newsletters.
  • integration: Help install and troubleshoot SDKs (mobile or web).
  • list: List attributes, segments, events, campaigns, newsletters.
  • search: Search across docs and workspace data.
  • metrics: Fetch engagement and deliverability metrics.

You can ask your AI tool, for example:

  • “Create a new segment of active users.”
  • “Help me troubleshoot push notifications on iOS.”
  • “List all campaigns in workspace XYZ.”

Step 5: Manage MCP Sessions

You can view and revoke active MCP sessions to keep your workspace secure:

  1. Go to Settings > Personal Settings.
  2. Under Connected clients, click View sessions.
  3. Choose and revoke sessions if needed.

Summary

Customer.io MCP connects your workspace with AI tools to simplify your customer data work. Enable MCP support in your account, connect your AI tool using the MCP server URL, pick your workspace, and start interacting with the tools it offers. Use the simple JSON examples above to set up IDEs or add connectors. Manage sessions anytime to keep control.

Try it out and see how AI can help you work faster with Customer.io data!

Details

TypeMCP Server
UpdatedJun 21, 2026
CreatedMar 16, 2026
DocumentationView docs
View Documentation