Ever wanted to ask 10 questions of your HubSpot portal and get the answers in minutes? Here’s me and Claude doing exactly that.

The typical RevOps rhythm looks like this: someone on the team has a question about the pipeline. You open HubSpot, build a filter, wait for the view to load, check the result, and then — because the answer raised another question — you start the process again. Ten questions means ten separate interface sessions. It can take the better part of an afternoon.

What I’m showing in this video is a different model entirely. Daeda MCP connects Claude directly to a local mirror of your HubSpot data. Every question runs against that local database, instantly, in plain language. The ten questions below took under 10 minutes to answer in total.

The 10 Questions We Asked

These are real questions I ran in the session. All answers came back in the same Claude conversation:

1. How many open deals are in each pipeline stage right now? A full stage-by-stage breakdown with deal counts and total values.

2. Which deals have been sitting in the same stage for more than 30 days? Named each deal, showed the stage, and calculated how many days it had been stuck.

3. Are there deals closing this month with no associated contact? Surfaced the association gaps before they became a problem at month-end.

4. Which companies have open deals but no active contact engagement in 60 days? A clean list of at-risk accounts that were not on anyone’s radar.

5. What is the average deal velocity for closed-won deals in the last quarter? Days-to-close by stage, with a median and range.

6. Which deal owners have the highest and lowest close rates over the last 6 months? Performance comparison across the team with total deal counts for context.

7. Are there any deals where the deal amount has never been set? A data quality check — found several deals sitting at £0 that needed updating.

8. Which contact roles appear most often in our closed-won deals? Mapped the buying committee patterns from actual closed revenue.

9. How many contacts are associated with more than one open deal right now? Identified contacts who might be getting conflicting outreach from different reps.

10. Which source leads convert to closed-won most often? Attribution breakdown by lead source field across the last 12 months.

Why the Speed Matters

The answers to these questions already exist in your HubSpot data. The bottleneck is not the insight — it is the interface. Filtering, exporting, cross-referencing, and repeating takes time, and that time adds up every single week.

Daeda MCP moves the query layer out of the HubSpot UI and into a local database that Claude can read directly. The result is that operational questions become conversational — you ask, Claude queries, and you get an answer immediately.

This is not about replacing your HubSpot views. It is about being able to ask follow-up questions in the same session without rebuilding a filter each time. When the tenth question is as fast as the first, the whole operating rhythm changes.

What Makes This Work

Daeda MCP runs a sync that pulls your HubSpot CRM data — deals, contacts, companies, associations, properties — into a local encrypted SQLite database. Claude connects to that database through the MCP protocol.

Every question in the session above ran locally. No rate limits. No waiting on API calls. No switching between browser tabs.

The setup process takes about 10 minutes. After that, the portal is queryable from Claude whenever you need it.

Getting Started

Install Daeda MCP from npm: npmjs.com/package/daeda-mcp

You will need a Claude desktop client with MCP support and a connected HubSpot account. The free version gives you full local read access and the ability to run exactly this kind of question session against your portal.

When you are ready for plan mode, multi-portal access, and operational execution — that is what Daeda MCP Pro is built for.

Join the Daeda MCP Pro waitlist