Ramp Data MCP Server

Public MCP server for querying Ramp's anonymized economic data across 50,000+ U.S. businesses, including vendor benchmarks (Ramp Rate) and AI adoption metrics (AI Index).

Data & Enrichment by Ramp None active

Connect (Remote Server) Docs

Overview

Ramp Data is a provider-hosted MCP server that exposes Ramp's public economic indices to AI assistants. The data is aggregated and anonymized from billions of dollars in real corporate spend processed through Ramp's corporate card and bill pay platform, covering 50,000+ American businesses. Unlike the customer-facing Ramp MCP (which uses your own Ramp account data), Ramp Data is a public dataset and requires no Ramp account or API key to use.

The server bundles two datasets behind a single connection. Ramp Rate provides software category and vendor benchmarks: adoption rates, year-over-year and month-over-month changes, vendor leaderboards, share of spend, and segmentation by company size. The AI Index tracks how businesses are adopting and spending on AI tools, with breakdowns by vendor, industry sector, and company size, queryable over rolling 1 to 120 month windows.

This makes Ramp Data useful for competitive intelligence, market sizing, vendor research, and AI trend analysis directly from inside Claude, ChatGPT, Cursor, or any other MCP-compatible client. It is one connection that answers questions like "what are companies our size paying for Salesforce" or "which AI vendors are growing fastest in financial services."

Tools

Tool	Description
`ramp_rate_list_categories`	List all software categories. Call this first to discover valid category names before querying summaries or leaderboards.
`ramp_rate_get_category_summary`	Get adoption statistics for a specific category, including adoption rate, YoY/MoM change, growth rank, vendor count, and size segmentation.
`ramp_rate_get_category_vendors`	Get a ranked vendor leaderboard within a category.
`ramp_rate_resolve_vendor`	Resolve a human-readable vendor name (e.g. Slack) to its canonical vendor_slug and list the categories it appears in.
`ramp_rate_get_vendor_profile`	Get a vendor's full Ramp Rate profile across all categories it appears in.
`ramp_rate_compare_vendors`	Compare adoption metrics across multiple vendors side by side.
`ai_index_get_adoption`	Get AI adoption history with vendor breakdown.
`ai_index_get_adoption_by_sector`	Get AI adoption history broken down by industry sector.
`ai_index_get_adoption_by_size`	Get AI adoption history broken down by company size (FTE segment).

Setup Guide

Ramp Data is a remote MCP server hosted by Ramp. There is nothing to install, no API key, and no Ramp account required.

Server URL

https://mcp.ramp.com/ramp-data/mcp

Claude (web, desktop, mobile)

Add it as a connector: Settings, Connectors, then either use the built-in Ramp Data connector at claude.com/connectors/ramp-data or add a custom connector with the URL above.

Claude Desktop (manual config)

Edit ~/Library/Application Support/Claude/claude_desktop_config.json on macOS or %APPDATA%\Claude\claude_desktop_config.json on Windows:

{
  "mcpServers": {
    "ramp-data": {
      "url": "https://mcp.ramp.com/ramp-data/mcp"
    }
  }
}

Cursor / VS Code / other MCP clients

Point the client at https://mcp.ramp.com/ramp-data/mcp as a remote MCP server. Custom or third-party clients must have their redirect URI allowlisted by Ramp first if they use OAuth.

ChatGPT

Settings, Connectors, Add connector, paste the server URL.

No environment variables or credentials are needed.

Use Cases

Vendor research and benchmarking: pull a category leaderboard for "Data warehouses" and compare adoption trends between Snowflake, Databricks, and BigQuery.
Competitive procurement intel: see what share of spend the top vendors hold in your software category, segmented by company size.
AI market tracking: chart how AI vendor adoption has shifted over the last 12 to 24 months across the whole economy.
Sector trend analysis: break down AI adoption by industry (e.g. financial services vs. healthcare) to inform GTM targeting or investment theses.
Inline analyst answers in Claude or ChatGPT: ask natural-language questions like "which CRMs are growing fastest at companies with 200-1,000 FTEs" without leaving your chat.

Example Prompts

"List the top 10 vendors in the Observability category by share of spend at companies with 500 to 2,000 FTEs."
"Compare Slack, Microsoft Teams, and Zoom adoption over the last 24 months."
"Show me AI adoption broken down by industry sector for the last 12 months."
"Which AI vendors are growing fastest at companies with under 100 employees?"
"Give me a full Ramp Rate profile for Datadog across every category it appears in."

Pros

Officially hosted by Ramp, no install, no API key, and no Ramp account required.
Single connection exposes both Ramp Rate (vendor benchmarks) and AI Index (AI adoption) datasets.
Backed by real anonymized spend from 50,000+ U.S. businesses, not survey data.
Read-only by design, so it is safe to expose to autonomous agents.

Limitations

Read-only public data only. It does not access your own Ramp account, transactions, or budgets (use the separate Ramp MCP at mcp.ramp.com/mcp for that).
Coverage is U.S. corporate spend on Ramp's platform, so it skews toward U.S. SMB and mid-market software buyers.
Limited surface area: 9 tools focused on two indices, not a general financial data API.

Alternatives

Ramp MCP (https://mcp.ramp.com/mcp): the customer-facing companion server for querying your own Ramp transactions, cards, bills, and approvals.
ramp-public/ramp_mcp on GitHub: open-source self-hosted MCP server over the Ramp Developer API for customer-specific data.
Bloomberg, FactSet, or other market data MCP integrations: broader financial datasets, but none with Ramp's vendor-level anonymized spend granularity.