← Reference Deep-Dives Reference Deep-Dive May 20, 2026 · 6 min read

ChatGPT Customer Research Connector: Real Interviews, Not Just CRM Data

By Kevin, Founder & CEO

TL;DR

ChatGPT connectors extend the model's context window with live external data by connecting external services via the Model Context Protocol over Streamable HTTP and OAuth. The connector category has two distinct data types: CRM connectors, which pull structured record fields from HubSpot, Salesforce, or similar platforms, and research connectors, which supply primary qualitative signal from moderated interviews and participant panels. CRM connectors surface what is already recorded about a customer; research connectors are necessary when the insight does not exist and must be commissioned from real participants. The distinction matters because AI models reasoning over CRM data can only work with what sales and support teams have logged. Research connectors give the model access to direct customer voice, stated preferences, and validated messaging outcomes that CRM systems never capture. User Intuition's ChatGPT connector runs on the Streamable HTTP endpoint at https://mcp.userintuition.ai/mcp, exposes all 72 research tools, and lets ChatGPT users launch studies, recruit from a 4M+ panel, and query accumulated findings without leaving the conversation.

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 human signal 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

The User Intuition connector uses the Streamable HTTP transport, which is the correct path for ChatGPT integrations. The stdio transport (local subprocess) is not used here.

Step 1: Get your API key. Log in to app.userintuition.ai and navigate to Settings → API Keys. Copy your ui_sk_ key.

Step 2: Open ChatGPT connector settings. In ChatGPT, go to Settings → Connectors → Create (this UI may also appear as Settings → Plugins → Add Custom Connector, depending on your ChatGPT tier).

Step 3: Enter the endpoint. In the connector URL field, enter https://mcp.userintuition.ai/mcp. ChatGPT will auto-discover the available tools from the MCP manifest at this endpoint.

Step 4: Complete OAuth. ChatGPT will redirect you through User Intuition’s OAuth flow to authorize the connection. Your ui_sk_ key is bound at this step. Approve the authorization.

Step 5: Verify. Once set up, ask ChatGPT: “What research tools are available from User Intuition?” It should list the capability groups and tools from the 72-tool manifest. The connection is ready.

Full docs are at docs.userintuition.ai/mcp-server/overview.

Four example workflows

1. Validate a tagline from ChatGPT

You have two tagline candidates and want to know which lands better with a target segment before the campaign goes live.

Ask ChatGPT: “Set up a concept-testing study for these two taglines with 20 SaaS product managers. Use voice interviews.” ChatGPT calls create_assistant to configure the AI moderator with your tagline framing, then create_invite to recruit 20 participants matching the profile from the 4M+ panel. When results arrive, ask ChatGPT to analyze_transcripts and surface which tagline generated stronger resonance signals and why.

2. Run a churn study from chat

You have a spike in 30-day churn and need qualitative signal on the driver before next week’s board meeting.

Ask ChatGPT: “Launch a 15-person exit interview study targeting customers who churned in the last 30 days. Focus on the moment they decided to leave.” ChatGPT uses create_assistant to set the AI moderator’s research goals, then bulk_create_invites_from_segment to reach the churned-customer segment. When interviews complete, analyze_transcripts returns AI-scored themes across the full participant set. The whole workflow — from study creation to findings — runs from the conversation.

3. Query past research without new fieldwork

Your team has run dozens of studies over the past year. You want to know whether any of them covered a specific product decision you are now revisiting.

Ask ChatGPT: “Has any of our past research addressed why customers don’t use the collaboration feature?” ChatGPT calls query_intelligence against the Intelligence Hub, which searches across all accumulated study transcripts and findings. If a relevant study exists, it returns a synthesized answer grounded in the historical data. No new fieldwork, no manual search through old reports.

4. Brief a product launch with real customer voice

A new feature is shipping next week and you want a launch brief grounded in what customers have said about the problem it solves.

Ask ChatGPT: “Pull together a launch brief for the new reporting feature based on what our research shows about how customers think about reporting.” ChatGPT calls query_intelligence to retrieve relevant past findings, then generate_report to produce a structured output. The brief is grounded in actual interview data, not synthesized from general AI knowledge.

How does User Intuition handle ChatGPT-driven research?

User Intuition’s Streamable HTTP endpoint at https://mcp.userintuition.ai/mcp is purpose-built for cloud agent integrations, including ChatGPT. OAuth handles authentication automatically — no manual header configuration, no local key management.

The 72-tool surface gives ChatGPT full research workflow capability, not just read access. Three proof points specific to ChatGPT workflows: query_intelligence in the Intelligence Hub group lets ChatGPT answer questions over accumulated findings without commissioning new fieldwork first; bulk_create_invites_from_segment lets ChatGPT recruit targeted participant cohorts in a single call; and analyze_transcripts returns AI-scored themes across a full study set, giving ChatGPT structured findings to reason over rather than raw transcript text.

The 4M+ panel means ChatGPT can recruit any standard research profile — consumer segments, B2B job functions, behavioral criteria — without the team sourcing participants separately. Studies in 50+ languages are available from the same tool surface. Most studies complete within 24-48 hours of the create_invite call, so a ChatGPT-initiated study launched in the morning can return analyzed findings by the next session.

The connector is fully documented at docs.userintuition.ai/mcp-server/overview. The product context is at the agentic research platform page.

Getting started

Create a free account at app.userintuition.ai — 3 free interviews, no card required.
Get your ui_sk_ key from Settings → API Keys.
In ChatGPT: Settings → Connectors → Create → enter https://mcp.userintuition.ai/mcp → complete OAuth.
Ask ChatGPT to list available research tools to verify the connection.
Run a quick test: ask_humans with a single research question to see the panel response flow.

For the full tool reference and config options, start at docs.userintuition.ai/mcp-server/overview. For teams evaluating whether agent-driven research fits their workflow, the agentic research platform page is the right place to start.

Note from the User Intuition Team

Your research informs million-dollar decisions — we built User Intuition so you never have to choose between rigor and affordability. We price at $20/interview not because the research is worth less, but because we want to enable you to run studies continuously, not once a year. Ongoing research compounds into a competitive moat that episodic studies can never build.

Don't take our word for it — see an actual study output before you spend a dollar. No other platform in this industry lets you evaluate the work before you buy it. Already convinced? Sign up and try today with 3 free interviews.

Frequently Asked Questions

How is a ChatGPT customer research connector different from a CRM connector?

A CRM connector (such as HubSpot's Deep Research Connector for ChatGPT) pulls structured fields from CRM records — deal status, contact properties, pipeline stage, email history. The data already exists in the CRM and the connector makes it accessible in the ChatGPT context. A customer research connector commissions new primary data — moderated interviews with recruited participants, panel responses to specific questions, findings synthesis over accumulated studies. CRM connectors surface what was previously logged; research connectors surface what participants say when asked directly, which CRM systems do not capture.

What endpoint does the User Intuition ChatGPT connector use?

The User Intuition connector uses the Streamable HTTP endpoint at https://mcp.userintuition.ai/mcp with OAuth authentication. ChatGPT's connector setup (Settings → Connectors → Create) auto-discovers the available tools from this endpoint. The stdio transport (npx -y @userintuition-ai/mcp) is not used for ChatGPT — it requires a local subprocess, which is not available in the ChatGPT environment.

How many tools does ChatGPT gain access to through the User Intuition connector?

72 tools across 9 capability groups — the full MCP surface. This includes Human Signal tools for rapid panel polls, Studies tools for study and assistant management, Invites tools for participant recruitment, Calls tools for transcript retrieval and analysis, Intelligence Hub tools including query_intelligence and generate_report, and Account tools. ChatGPT gains read and write access to the full research workflow, not just read access to past data.

Do I need a paid plan to use the User Intuition ChatGPT connector?

No — you can start with a free account at app.userintuition.ai (3 free interviews, no card required). The connector itself is free to set up. Costs accrue per interview when you launch studies: Starter plan credits are $25 each (audio), Professional plan is $20/credit with 50 free credits included monthly. The connector is available on all plans.

Can ChatGPT both launch new studies and query past research through the connector?

Yes. The connector exposes both creation tools (create_assistant, create_invite) and retrieval tools (query_intelligence, list_calls, analyze_transcripts). In a single ChatGPT session, you can ask it to check whether past research covered a topic (via query_intelligence), then commission a new study on a gap (via create_assistant + create_invite), and later retrieve and analyze the results (via list_calls + analyze_transcripts).

What authentication does the ChatGPT connector setup require?

OAuth, resolved automatically during the connector setup flow in ChatGPT (Settings → Connectors → Create → enter the endpoint URL). You will be prompted to authorize User Intuition during the OAuth flow. Your ui_sk_ API key is bound at that step. No manual header configuration is required — OAuth handles authentication for the Streamable HTTP transport.

Is the ChatGPT connector read-only or does it support full study creation?

Full study creation. The 72-tool surface includes write tools: create_assistant (define study parameters and AI moderator), create_invite (recruit participants from the panel), update_assistant, and more. ChatGPT can create, configure, launch, and retrieve research entirely within the conversation. This is different from connectors that only expose read access to already-completed data.