How to Connect HubSpot to ChatGPT in 1 Minute (CRM & Cross-Channel Analysis)

HubSpot holds all the data you need, but not the answers.

Your deals, contacts, emails, sequences, and activities are all there, connected. Yet most reports show them in isolation, making it hard to see what actually drives revenues, what converts, what stalls, and why.

So teams export CSVs, stitch data in spreadsheets, and lose hours on manual analysis.

Windsor MCP changes that. It connects your entire HubSpot account to ChatGPT as a single, live dataset, so you can ask complex questions about diversified areas of your business and get actionable insights in just seconds.

🚀 Connect your HubSpot data to ChatGPT with Windsor MCP. Try it free for 30 days → onboard.windsor.ai/app/hubspot.

With your HubSpot data live in ChatGPT, you can instantly see:

• Which lead sources produce deals that close fastest?
• Which sales reps have the highest close rates, and how does their deal velocity compare?
• Which accounts have no activity in 30+ days but still have high-value open deals?
• What patterns appear in closed-lost deals across sizes or industries?

Below, we’ll walk you through how to connect HubSpot to ChatGPT using the Windsor.ai native app in under a minute.

Connecting HubSpot to ChatGPT via Windsor MCP: quick setup, live data

Windsor.ai connects via its native ChatGPT app. Install it once and then access and query your HubSpot and other data sources without any additional setup.

What you need to get started

• A HubSpot account with the required permissions
• A Windsor.ai account (free trial or paid plan)
• A ChatGPT account 

Step 1. Link your HubSpot account to Windsor

Go to onboard.windsor.ai/app/hubspot and click Grant HubSpot Access.

Authorize via HubSpot’s OAuth and select the account(s) you want to connect to ChatGPT.

Windsor connects directly to HubSpot’s API, instantly pulling contacts, companies, deals, activities, email data, and tickets (covering 1,000+ data fields).

Custom properties and custom objects are included automatically. Once connected, Windsor detects and maps them directly from your HubSpot portal — no manual setup required.

Step 2. Install Windsor’s native app in ChatGPT

Open Windsor’s native ChatGPT app and click Connect. Grant access to your Windsor account.

Confirm your HubSpot data is available in ChatGPT using this test prompt:

What HubSpot accounts are connected to Windsor.ai?
Show me total open deals right now.

Pipeline data confirmed? You’re ready to dive deeper.

What ChatGPT can see across your HubSpot account

Windsor gives ChatGPT access to your HubSpot data across all major objects, and, importantly, the relationships between them. You can view the full list of supported fields here.

Below are the key metrics and dimensions Windsor streams into ChatGPT, and what type of analysis you can do with them.

The questions worth asking: HubSpot prompt ideas by use case

The quality of your insights depends on the questions you ask.

To help you get started faster, we’ve compiled a set of ready-to-use prompts for HubSpot data analysis, tailored to common use cases and pain points.

Pipeline health: what’s real and what’s noise

A HubSpot pipeline report shows deal value by stage. But it doesn’t tell you which of those deals have been sitting untouched for three weeks, which are in a stage they’ve been in twice before, or which reps have inflated pipelines padded with deals that will never close.

ChatGPT can answer all of this.

Find stalled deals before they become lost deals

Which open deals have had no activity logged — no calls, emails, or meetings — in the last 21 days?

Sort by deal_amount descending.

Which pipeline stage are they in, and how long have they been there?

Pipeline by stage with weighted forecast

Show all open deals grouped by deal stage.

For each stage: total deal value, number of deals, and average deal_hs_deal_stage_probability.

Calculate the weighted pipeline value per stage.

Which stage holds the most risk — high value but low probability?

Close date accuracy: are reps’ predictions reliable?

For deals closed in the last 90 days, compare the original expected close date vs. the actual close date.

Which reps had the smallest gap (most accurate forecasters)?

Which had deals slip by more than 30 days most frequently?

Win and loss analysis: learning from every outcome

HubSpot tracks every closed-won and closed-lost deal, but the patterns inside that data rarely surface.

Which industries convert best? Which objections kill deals most often? Which lead sources bring real buyers vs. window shoppers?

These insights are far more valuable than any sales training, yet they usually stay hidden.

With ChatGPT, you can uncover these patterns instantly and turn every deal into a learning opportunity.

What’s actually killing deals

Analyze all closed-lost deals from the last 6 months.

What are the most common deal closed lost reasons?

Do loss reasons differ by deal size (small under $5k vs. mid $5-20k vs. enterprise $20k+)?

Which pipeline stage do most deals exit from?

Win rate by lead source

For all deals created in the last 6 months, compare win rate by contact lead source.

Which sources produce the highest percentage of closed-won deals?

Which produce the most pipeline volume but lowest conversion?

Rep performance: close rate and deal velocity

Compare sales rep performance for the last quarter.

For each deal hubspot owner id show: total deals created, win rate, average deal amount, and average days from creation to close.

Who is closing fastest? Who has the highest average deal size?

Lead funnel: from first touch to pipeline

HubSpot tracks every contact’s journey across lifecycle stages, but most teams analyze each stage in isolation.

The real insight comes from connecting those stages and linking them to actual deal outcomes.

With ChatGPT, this deep, end-to-end funnel analysis happens automatically.

Funnel conversion rates end to end

How many contacts entered each lifecycle stage in the last 90 days:
subscriber → lead → MQL → SQL → opportunity → customer?

What is the conversion rate between each consecutive stage?

Where is the biggest drop-off?

Lead quality by source: which channels send buyers?

For contacts created in the last 6 months by contact lead source:

What % reached SQL stage?

What % converted to a closed-won deal?

What was the average deal amount for converted contacts per source?

Contacts going cold — catch them before they’re gone

Which contacts with a lifecycle stage of MQL or SQL have had no activity in the last 30 days?

List them with their company, lead source, and last activity date.

These are leads that need immediate re-engagement. Generate an email to re-activate these leads.

Email and sequences: what’s actually moving deals forward

HubSpot sequences are one of the most powerful sales tools, and one of the least analyzed.

Open rates are tracked. But what’s rarely measured is real impact: which sequences drive stage progression, booked demos, and closed-won deals.

With ChatGPT, this analysis happens in seconds, turning engagement data into clear revenue insights.

Which sequences generate the most replies

Compare all active email sequences by reply rate over the last 60 days.

Which sequence has the highest reply rate?

Which has the most sends but lowest engagement?

Are there sequences with high open rate but near-zero replies — a subject line working but content not landing?

Sequence to pipeline: the attribution question

For contacts who received an email sequence and then moved from MQL to SQL in the same 30-day window:

Which sequences appear most frequently before that stage transition?

This is our strongest sequence-to-pipeline signal.

Form and landing page performance

Which HubSpot forms had the highest submission rate in the last 30 days?

For each form show: views, submissions, conversion rate.

Which forms are generating the most new contacts?

Rep briefing: everything about an account before a call

This is where ChatGPT’s natural language capability makes the most immediate practical difference.

Before an important call or meeting, instead of clicking through multiple HubSpot records, you just ask for a complete summary.

Full account briefing in seconds

Give me a briefing on the account [company name].

Include: company details, all associated contacts and their lifecycle stages, all open deals with current stage and last activity date, the last 5 interactions logged, and any open tickets.

Which accounts are about to go dark

Which companies with open deals above $10k have had no calls, emails, or meetings logged in the last 14 days?

Include the deal owner and deal stage for each.

These accounts need outreach today. Help me generate an outreach email.

Revenue forecasting: stop guessing

Most pipeline reviews happen weekly, in a meeting, from a slide that’s already out of date.

With HubSpot data live in ChatGPT, a weighted revenue forecast is a single prompt away — at any point, for any time period, cut by any dimension.

Weighted revenue forecast by close month

Using open deals with a close date in the next 90 days:

For each month, calculate total expected revenue using deal amount × deal_hs_deal_stage_probability.

Which month has the most weighted pipeline?

Which deals are moving the needle most significantly?

ARR and MRR snapshot

What is the total deal_hs_arr from all deals closed in the last 12 months?

How does this month's new ARR compare to the previous 3 months?

Which deal stage currently holds the most potential new ARR?

Bonus: Connect HubSpot to your full stack, and see what actually drives revenue

HubSpot shows what happens inside your CRM. But to understand where revenue really comes from, including marketing spend, website behavior, and actual payments, you need to connect it with the rest of your business data.

🔗 Windsor brings it all together, connecting 325+ platforms (marketing, e-commerce, accounting, HR, analytics) into a single ChatGPT conversation.

Here are some high-impact ways to combine your data for deeper insights:

• HubSpot + ad platforms (Google, LinkedIn, Meta): Which campaigns generate leads that actually convert into revenue, not just MQLs? Identify which channels drive deals that close, and which bring a low-quality pipeline.
• HubSpot + GA4: What do high-value leads do before they convert? Discover which pages, content, and journeys correlate with faster deal progression and higher close rates.
• HubSpot + Stripe or other billing systems: Which closed deals actually turn into revenue? Reconcile CRM pipeline with real payments to uncover revenue gaps, churn risks, and your true LTV.
• HubSpot + Google Ads (deep dive): Which keywords drive deals, not just clicks? Identify search terms that consistently lead to SQLs and closed-won revenue vs those that stall at MQL.
• HubSpot + Shopify (for B2C / hybrid models): Do marketing-qualified leads actually spend more? Compare customer purchase behavior with HubSpot engagement to understand what drives repeat purchases and higher AOV.

HubSpot already has a native ChatGPT connector — so what does Windsor add?

HubSpot launched its official ChatGPT connector in 2025, and it’s a solid integration.

It’s easy to set up and works well for everyday CRM questions:

• pipeline summaries
• contact and deal lookups
• ticket trends
• account briefings

For anything inside HubSpot, it does the job.

But that’s also its limitation. It only sees your CRM.

The moment your questions involve data outside HubSpot, including ad spend, website behavior, billing, or product usage, the native connector reaches the wall.

And those are the questions that actually define revenue but are left unanswered:

• Which campaigns generate deals that close, not just leads?
• Which channels drive high-value customers, not just conversions?
• Which closed-won deals turn into real revenue and which don’t?

Where Windsor MCP goes further

Windsor MCP connects HubSpot to the rest of your stack and brings all of it into a single conversation.

Instead of analyzing one system at a time, you can ask questions across all of them at once.

So, HubSpot answers: what’s happening in your CRM.
Windsor answers: how your entire revenue system works.

Quick comparison:

Conclusion

HubSpot holds the answers to your most important revenue questions all along:

• Which campaigns actually drive the pipeline?
• Where do deals stall, and why?
• Which reps consistently close, and which need support?
• What does your forecast really look like right now?

But getting these answers meant building custom reports, exporting data, or relying on BI tools.

Until now.

Windsor MCP removes that barrier. It connects your HubSpot data to ChatGPT, so you can analyze your entire pipeline, uncover patterns, and get actionable answers using the power of AI in just seconds.

🚀 Connect HubSpot to ChatGPT with Windsor MCP to turn your raw CRM data into clear revenue insights! Start your 30-day free trial now.

FAQs

What are the ways to connect HubSpot to ChatGPT?

Three main options exist:

• Windsor.ai native app (both for CRM and cross-channel analysis): Connects HubSpot alongside 325+ other data sources — Google Ads, LinkedIn Ads, GA4, Shopify, and more. All sources are normalized and queryable in the same ChatGPT conversation. Read-only. The right choice when questions require connecting CRM data to ad spend, website analytics, or revenue platforms.
• HubSpot native connector (only for CRM-only analysis): Official connector. Available to all HubSpot tiers with a paid ChatGPT plan (Plus, Pro, Team, or Enterprise). A Super Admin sets it up once via ChatGPT Settings. Supports regular chat and Deep Research mode. Can create and update CRM records and log engagements. Limited to HubSpot data only.
• Manual CSV export: Export HubSpot object lists and upload to ChatGPT. Works for one-off isolated questions; loses all object relationships on export and is immediately stale.

What is the fastest way to connect HubSpot to ChatGPT?

Windsor’s native app provides the fastest HubSpot to ChatGPT integration method. Connect your HubSpot account at onboard.windsor.ai/app/hubspot, then open Windsor’s ChatGPT app and click Connect. Your HubSpot data, along with any other sources connected to Windsor, will be live in every ChatGPT conversation.

Is there a native HubSpot connector for ChatGPT?

Yes. HubSpot launched an official ChatGPT connector, which is available to all HubSpot customers on any plan, as long as you have a paid ChatGPT subscription (Plus, Pro, Team, Enterprise, or Edu). The connector supports contacts, companies, deals, tickets, standard, and custom properties. It also includes a Deep Research mode for more complex, multi-variable analysis, and as of early 2026, bidirectional access to create records and log engagements. The one limitation: it is HubSpot-only. It can’t blend CRM data with your ad platforms, website analytics, or other tools.

What HubSpot data does Windsor pull into ChatGPT?

Windsor connects to HubSpot’s API and gives ChatGPT access to over 1,000 metrics and properties across all major objects: contacts (name, email, lifecycle stage, lead source, last activity, custom properties), companies (name, industry, revenue, employee count), deals (name, amount, stage, pipeline, close date, probability, ARR/MRR/ACV, lost reason, owner), activities (calls, emails, meetings, notes, tasks with timestamps), email and sequence metrics (sent, opened, clicked, replied, bounced, CTR), and tickets (status, priority, SLA, stage history). Custom properties and custom objects are also available once your HubSpot portal is connected.

Does Windsor preserve the relationships between HubSpot objects?

Yes, and this is the key advantage over manual CSV exports. Windsor streams HubSpot’s object relationships into ChatGPT: contacts are associated with their deals, deals are associated with their activity history, and companies are linked to both contacts and deals. This means ChatGPT can answer cross-object questions, like which lead source produces deals that close fastest, or which accounts have open deals with no recent activity, that HubSpot’s native reports can’t answer without the Custom Report Builder.

Do I need HubSpot Professional or Enterprise to use this?

No. Windsor connects to HubSpot’s API on any tier, including Starter. Standard objects (contacts, companies, deals, activities, email data, and tickets) are available regardless of the HubSpot plan. Custom properties are available on all paid plans. The Custom Report Builder feature in HubSpot itself requires Professional or Enterprise, but connecting your data to ChatGPT via Windsor doesn’t depend on that feature at all.

Can ChatGPT see custom properties and custom objects from my HubSpot portal?

Yes. Windsor connects to your specific HubSpot portal and discovers your custom properties and custom pipeline stages automatically. If your team has created custom deal properties (like preferred_budget, technology_stack, or deal_source_detail), those are available for ChatGPT to filter, group, and analyse alongside standard properties.

Can I use this to generate pipeline forecasts?

Yes. Windsor pulls each deal’s amount, stage probability, and expected close date. ChatGPT can calculate weighted pipeline (amount × probability) by stage, by owner, by close month, or by any other dimension you specify. You can ask for a forecast at any time, not just at the Monday pipeline review, and the answer reflects the live state of your CRM.

Can I combine HubSpot with my ad platforms to understand true marketing ROI?

Yes, and this is one of the most powerful uses of Windsor’s cross-source capability for B2B teams. Connect HubSpot alongside LinkedIn Ads, Google Ads, Meta Ads, or any of Windsor’s 325+ supported sources. Ask ChatGPT to join the contact lead source with the deal outcome to calculate which campaigns produced a pipeline that actually closed — the marketing attribution question that usually requires a dedicated BI tool and a significant data engineering effort.

Can I make any changes to my HubSpot account via Windsor MCP?

No, Windsor connects to HubSpot with read-only API access. ChatGPT can analyze, summarise, and surface insights from your CRM data, but it can’t create contacts, update deal stages, log activities, send emails, or make any changes to your HubSpot portal. All actions remain in your hands inside HubSpot.

How fresh is the HubSpot data Windsor sends to ChatGPT?

Windsor queries HubSpot’s API in real time when ChatGPT needs data to answer a question. If a deal was updated an hour ago, that change is visible when you ask about it. There is no fixed sync schedule or stale cache; every question reflects the current state of your CRM.