HubSpot and ChatGPT Deep Research – Practical Guide + 24 Prompts to Start With
Discover how to connect ChatGPT with HubSpot and use 24 expert prompts to drive sales, marketing, and service insights. Download the library
In June 2025, HubSpot announced integration with OpenAI’s ChatGPT, launching the first CRM-native deep research connector. This move allows over 250,000 HubSpot customers to query their CRM data directly through ChatGPT, enabling advanced, natural language-driven analysis without the need for technical expertise or complex report builders. The integration is designed to let users ask plain-English questions about their contacts, companies, deals, and tickets, and receive actionable insights instantly. HubSpot’s CTO, Dharmesh Shah , described this as a major milestone, emphasizing how the connector transforms ChatGPT into a real-time business intelligence analyst for marketing, sales, and service teams.
This guide will walk you through how to set up the connection, what to watch out for, and how to start using your CRM data more effectively today.
How to Connect Your HubSpot to ChatGPT
Connecting HubSpot to ChatGPT doesn’t require any coding skills. Just follow these simple steps:
Enable Deep Research in ChatGPT Navigate to Settings & Beta → Features and toggle on “Enable Deep Research.”
Add HubSpot as a Source Go to Tools → Run Deep Research → Sources and select HubSpot.
Authenticate Your Account Log in with your HubSpot credentials and choose the account you want to connect.
Admin Permissions Required A HubSpot Super Admin (or someone with App Marketplace permissions) needs to complete the initial setup. After that, individual team members can connect on their own.
Start Exploring You’re now ready to run natural language prompts that query your HubSpot data inside ChatGPT. (Note: Deep Research works only in the web version of ChatGPT.)
ChatGPT Account Requirements and Privacy Implications
To access this integration, you’ll need:
A ChatGPT Plus or Team account with Deep Research enabled.
Access to the web version of ChatGPT, as Deep Research isn’t supported in mobile apps yet.
As with any data-driven tool, privacy matters. When you connect your HubSpot account, your data is used only for your session and is not stored by OpenAI or used to train future models. Still, always confirm with your compliance or data privacy team before connecting sensitive information.
Pitfalls: What You Need to Get Right First
Before diving into prompts and insights, it’s crucial to ensure your HubSpot instance is set up for success. Here’s what most teams get wrong:
No data, no insight: If your CRM data is sparse or inconsistent, even the most advanced AI can’t help you. Make sure your deals, pipelines, contact fields, and marketing campaigns are in active use.
Poor feature usage: AI works best when HubSpot’s native features - lifecycle stages, lead scores, personas, and campaign tagging - are fully leveraged.
Think of this as preparing your data environment for AI. Garbage in, garbage out still applies.
Use Cases and Prompts to Explore Your Data
Sales
Review At-Risk Deals & Re-Engage
Review my open deals in HubSpot and identify which ones are at risk based on stage duration, last activity, and buyer engagement. Recommend specific next steps or messaging to re-engage each account.
Nudge Top Deals to Close
List the 20 open deals most likely to close this month (based on close date, last-modified date, and deal amount). Draft a one-sentence follow-up nudge for each, then send follow-ups directly from HubSpot.
Compare Win Rates & Remove Friction
Compare win rate and average sales-cycle length for deals > $50,000 this quarter vs. last. Highlight the three slowest stages and recommend one friction-busting tactic per stage, then update stage guidance in HubSpot.
Define ICP & Refine Based on Deals
Analyze my top 50 closed-won deals from HubSpot. Identify common firmographics, buyer roles, and deal behaviors. Recommend a refined ICP and suggest segments for targeting.
Identify Best-Converting Personas & Build Lookalike Campaign
Identify the personas that converted fastest from lead to customer in the last 90 days. Recommend key attributes to target in future campaigns, then use the insights to build a lookalike audience or launch a persona-based sequence in HubSpot.
Generate Multi-Step Email Sequence
Generate a 3-step email sequence for {Name} at {Company} promoting {Product} using recent activity from HubSpot like page views, email clicks, or previous conversations.
Review MQL Handoff & Improve Process
Analyze the last 25 MQLs that didn’t convert. Use form data and sales notes to identify handoff issues. Output: Lead | Issue | Fix.
Map 3-Step Upsell Playbook Post-Pilot
Map a 3-step upsell playbook for a client who completed a successful pilot. Input: pilot value is proven. I want messaging, visual proof of ROI, and the right executive sequencing to land a larger deal. Bonus: ideas to increase urgency.
Summarize Sales Performance by Rep
Summarize sales performance by rep for the past quarter including number of deals closed, total value, and average deal size. Highlight top performer and low performer.
Follow-Up to Reduce Risk in Final Stage
Write a follow-up note for a prospect who’s in final-stage consideration. Assume decision maker is aligned, but internal team has concerns. The email should reduce perceived risk, reaffirm alignment with strategic goals, and use social proof to close.
Diagnose Pipeline Bottlenecks & Optimize
Analyse our HubSpot sales pipelines and find bottlenecks and challenges. Make suggestions for improvements. Main goals are increase revenue and optimise current processes.
Draft Personalized Renewal Email
Draft a personalized renewal email for {Name} at {Company}. Use metrics from last year’s engagement (e.g., efficiency gains, time saved) and position an upgraded package aligned with their fiscal goals. Frame around risk reduction + ROI, not features.
Re-Engage Inactive Companies
Find inactive companies with growth potential and generate re-engagement plans.
Marketing
Analyze Patterns & Create Buyer Personas
Analyse our HubSpot data by identifying patterns and trends within the data to understand customer behavior, needs, and pain points. Create buyer personas.
Review Leads That Didn’t Become MQLs
Analyze the last 25 leads that didn’t become MQLs. Identify key reasons and recommend improvements to qualification criteria or nurturing process. Output: Lead | Issue | Fix.
Refine Lead Scoring for Revenue Prediction
Examine our Ideal Customer Profile’s historical data—seniority levels, industry segments, campaign sources, and qualitative sales notes—to identify the top signals that correlate with upsell and renewal revenue (not just initial MQLs). For each signal, assign a relative weight (e.g., 1–10) based on its predictive power, and explain how you’d combine them into a single lead score. Output a ranked list of signals with weights, an example scoring formula, and a brief rationale for why each signal predicts buying intent beyond initial qualification.
Segment & Scale High-Performing Sources
Segment contacts by source (e.g., LinkedIn Ads, organic search, webinars, ebook downloads, paid social). Compare conversion rates to customers and identify the best-performing cohort, then scale high-performing sources through nurture workflows in HubSpot.
Outline Executive Webinar Plan
Outline a 30-min webinar for mid-level to executive buyers in a key industry vertical. Title: {Webinar Title}. Include common pain points, bold POV slides, and one anonymized client success story. CTA: diagnostic session or ROI consult.
Write LinkedIn Transformation Posts
You’re a LinkedIn content strategist for senior B2B audiences. Draft three post outlines that spotlight our role in a client’s business transformation: the first should zero in on the client’s pain points before partnering with us, the second should illustrate how our solution unlocked new avenues of growth, and the third should highlight the hard metrics that prove our impact. For each outline, include a compelling hook, three concise talking points, and a clear call to action—tailored to resonate with senior leaders at similar firms.
Build 90-Day Content Roadmap by Vertical
You’re a content strategist. Given our ICP — Mid-market SaaS healthcare providers ($5–50M ARR) — and Persona — VP of Marketing — create a 90-day content roadmap across these verticals: Product Adoption, Customer Retention, Revenue Expansion. Output as a Markdown table.
Prepare Persona-Based Campaign
Review our last 90 days of engagement and conversion data to pinpoint the top 3 fastest-converting personas—naming each, summarizing their key traits and what drove their conversions. Then, for those personas, define HubSpot list filters to build a lookalike audience and draft a five-step email nurture sequence for each, complete with subject lines, core messages, and CTAs.
Service
Analyze Support Tickets & Generate Churn Brief
Analyze recent support ticket language and satisfaction scores for signs of risk. Identify companies with open deals and negative service interactions, then generate a churn-risk brief for the CSM in HubSpot.
Analyze Renewal Risk from Tickets & Renewals
Analyze ticket volume, resolution times, escalation frequency, and upcoming deal renewal dates over the last quarter. Flag the top 10 accounts at highest churn risk, then use the insights to guide intervention strategies in HubSpot.
Audit Top Ticket Issues & Recommend Automations
Audit our top support tickets (e.g., login issues, billing confusion, password resets). What’s the pattern? What workflows can we automate or improve? Give 3 ways to turn these friction points into retention and expansion moments.
Download the prompt library for free
Final Words: A New Chapter for CRM Intelligence
This is just the beginning. HubSpot has made a bold, forward-thinking move by allowing users to connect their live data to generative AI tools like ChatGPT. But what’s even more exciting is what this enables: teams can now perform deep analysis, uncover insights, and trigger actions—all with natural language.
If you're a business leader, marketer, or sales manager, this shift turns your CRM from a static database into a living assistant. And it’s up to you to take advantage of it.
The better your data, the smarter your AI. Explore. Experiment. And elevate your CRM.