Most ChatGPT connectors pull data from systems of record — CRM platforms, marketing databases, analytics tools. That gives ChatGPT access to what companies already know about their customers. User Intuition’s ChatGPT connector does something different: it gives ChatGPT the ability to ask customers directly.
The distinction matters because most of the questions that drive product, marketing, and strategy decisions cannot be answered by CRM data alone. Why did a customer churn? Which message will resonate with a new segment? Does this concept actually solve the problem? CRM systems log what happened; research connectors surface why — and what to do next.
This guide covers the connector setup, explains how it differs from CRM-style connectors, and walks through four research workflows you can run from a ChatGPT conversation. For the underlying platform, see the User Intuition agentic research platform.
What a ChatGPT customer research connector actually does
ChatGPT connectors work through the Model Context Protocol (MCP) over Streamable HTTP and OAuth. When you configure a connector in ChatGPT (Settings → Connectors → Create), ChatGPT fetches the tool manifest from the endpoint URL you provide. From that point forward, the tools appear natively in the ChatGPT conversation — the model can call them from natural-language prompts, chain them across a multi-step workflow, and return results without the user opening a separate application.
A customer research connector specifically exposes research workflow operations as callable tools: defining a study, setting up an AI moderator, recruiting participants, launching interviews, retrieving transcripts, and querying accumulated findings. The connector bridges the gap between ChatGPT’s reasoning capability and real participant evidence from research participants.
The practical result: you can brief ChatGPT on a product question, ask it to run a study, wait for participants to complete interviews, and then ask ChatGPT to analyze the findings — all within the conversation, without logging into a separate research platform.
The two data-source split: CRM data vs. research data
The ChatGPT connector ecosystem currently has two dominant data types, and it is worth understanding the split clearly before configuring your workflow.
CRM data connectors (HubSpot’s Deep Research Connector for ChatGPT and similar tools from Salesforce and other CRM platforms) pull structured record fields from your CRM database — contact properties, deal stages, account history, logged email threads. They are useful for answering questions that live inside your CRM: “How many deals closed from this segment last quarter?” or “What did the last 20 calls with enterprise prospects have in common?”
CRM connectors are limited to what was previously logged. If the insight does not exist in the CRM — because it requires asking customers directly — a CRM connector cannot surface it.
Research data connectors supply primary qualitative signal: moderated interviews with recruited participants, panel responses to specific research questions, AI-analyzed themes across a transcript set, and synthesized findings over accumulated past studies. Research connectors are necessary for questions that require direct participant input: “Why are customers churning before the 90-day mark?” or “Which of these three positioning messages resonates most with enterprise buyers?”
User Intuition ships a research data connector. The two connector types are complementary — CRM context + research signal is a stronger combination than either alone — but they are not substitutes.
Setup: ChatGPT Connector for User Intuition
Connect an OAuth-capable ChatGPT client to https://mcp.userintuition.ai/mcp and complete the discovered authorization flow. Follow the current client setup instructions for account and interface requirements. Start a new conversation after connecting, and use list_studies to verify a read.
Hosted OAuth does not require you to paste a User Intuition API key. API keys remain available for local stdio, the CLI, and direct REST integrations.
Four example workflows
1. Validate a tagline from ChatGPT
Provide the tagline options and audience. Create a metadata draft with create_study, then send the research brief through customize_study. Relay any planning questions to the user. Retrieve the persisted plan with get_study and obtain approval before recruitment. The user must choose panel or BYOP explicitly.
2. Run a churn study from chat
Use customers with direct experience of the cancellation. For a BYOP study, use create_participants with 1–100 unique participant emails per batch after the saved plan and invitations are approved. Invitations send by default; set silent: true on individual participant records when invitations should not send. Source customer lists through your own authorized export or integration; MCP has no direct CRM segment-sync tools.
3. Review past research without new fieldwork
Agents can search findings and participant responses across authorized studies, then retrieve the underlying reports and interviews. Results preserve study context and source links; the calling agent interprets the evidence. Search coverage is explicit, and retrieving evidence does not launch research.
4. Brief a product launch with real customer voice
Select relevant studies, retrieve their reports, and ask ChatGPT to draft a launch brief with supporting sources and uncertainty. Study results expose findings, participant responses, sample profiles, recommendations, and source references in JSON. Use generate_report when analysis is needed, and get_interview to verify supporting messages and recording links. Preference shares, credibility scores, and ranked themes are not guaranteed typed fields in this response.
How does User Intuition handle ChatGPT-driven research?
MCP exposes study planning, recruitment, interviews, results, evidence search, and supporting configuration operations. The CLI provides shell access to research workflows. Use the current tool catalog and API reference for exact names and arguments; tool counts vary by release.
For a panel study, use launch_panel with dry_run: true to obtain the recruitment estimate. Show the country, language, cost, and timeline; launch with the same settings after approval. Each launch specifies one country. Audiences below 10% incidence require a feasibility request.
The integration supports creating new research and retrieving existing study evidence. Plan approval, invitation authorization, and estimate approval remain part of the conversation. ChatGPT can synthesize the evidence it retrieves, while cross-study evidence search is available through the research API.
Getting started
Create an account, connect with OAuth, and start with a read. Then describe one research decision, choose recruitment, and review the saved plan before any invitations or paid launch. See agentic research and MCP setup.