← All articles

How to give Claude access to your HubSpot CRM data

June 18, 2026

Giving Claude access to your HubSpot data means connecting your HubSpot account once through a synced data layer, so Claude can query your contacts, deals, and pipeline directly and accurately — instead of you exporting a CSV from HubSpot every time you want to ask a question about it.

Why exporting a report first defeats the point

The most common workaround is exporting a HubSpot report and pasting it into the chat, or uploading it as a file. It works, but it puts the manual step right back in: you're still the one running the export, deciding which fields to include, and remembering to do it again next week. If a deal moves stage five minutes after you export, Claude doesn't know.

A real connection removes that step. HubSpot stays the system you work in day to day; Claude (and any other AI tool you use) reads from a synced, current copy of the same data, automatically.

Connecting HubSpot to Claude

  1. Connect HubSpot. In your Synquil dashboard, go to /dashboard/connect/new and select HubSpot. You'll be sent through HubSpot's OAuth flow to authorize read access to your portal.
  2. Choose your objects. HubSpot's fixed CRM objects — contacts, companies, deals, tickets, and so on — appear as selectable items. Toggle on what you want synced, then "Save and Deploy."
  3. Review the schema. Synquil proposes a unified schema for your selected HubSpot objects under /dashboard/schema/versions. Confirm it to apply.
  4. Connect Claude. In /dashboard/mcp, open the Claude tab and follow the link to claude.ai/settings/connectors to add Synquil as a custom connector. Paste in your MCP URL; Claude redirects you to Synquil to authorize. No separate HubSpot credentials are ever shared with Claude — Claude only ever talks to Synquil's synced copy.
  5. Ask Claude about your pipeline. Try "which deals have been in the same stage for more than 14 days" or "what's our average deal size this quarter, broken down by source." Claude runs these as read-only queries against your synced HubSpot data.

HubSpot's OAuth tokens expire every 30 minutes; Synquil refreshes them automatically in the background, so the sync keeps running without you re-authorizing.

One CRM connection, every AI tool

Because the sync happens once into a shared schema, the same setup works for ChatGPT (see connecting HubSpot to ChatGPT) or any other MCP-compatible tool without reconnecting HubSpot again per tool. Everyone asking about your pipeline — in whichever AI tool they prefer — gets the same answer, because they're reading the same synced data.

Connect your HubSpot account and give Claude a real view into your pipeline.