One builds your survey while you sit inside the platform. The other lets your AI assistant run the platform from wherever you already work.
Two ways to put AI to work in QuestionPro
You want AI to handle the tedious part of research. The real question is where you want that to happen.
Some days you are inside QuestionPro building a survey, and you would rather describe what you need than click it together. Other days you are already deep in Claude or ChatGPT, halfway through a report, and you want live data without opening a single tab.
QuestionPro built for both.
One option lives inside the platform and builds surveys for you. The other lets your AI assistant reach into the platform from outside. They are not rivals. They are two answers to the same instinct, and knowing which is which saves you a lot of guessing.
Learn More: See how QuestionPro goes headless and how Survey Agent builds surveys from a plain-language prompt.
What Survey Agent does
Survey Agent is an autonomous AI agent built into QuestionPro. You describe what you want to learn in plain language, and it plans, builds, and configures the survey for you. Questions, blocks, flow, and logic, all handled inside the app.
It lives where you already build. Open Research Suite, start a new survey, and the agent is right there. It also sits in the corner of the survey builder in draft mode, so you can ask for changes on a survey you already started.
Three things it is good at. It handles the fiddly skip logic and branching: criteria, show or hide questions, and multi-condition compound branching, the parts that are easiest to get wrong under time pressure. It turns a Word doc or PDF into a structured survey, so a brief sitting in someone’s inbox becomes a real questionnaire without manual re-entry. And it can summarize a survey you did not build, mapping the flow and topics in seconds.
The short version: Survey Agent is AI that comes to your survey.
What headless (MCP) does
Headless is the other direction. Instead of you coming to QuestionPro, the platform comes to you, inside the AI assistant you already use.
It works through the Model Context Protocol, an open standard that lets an assistant like Claude, ChatGPT, Cursor, or Windsurf talk to QuestionPro directly. You connect once through a secure sign-in, with no API keys, and your assistant can then act on your account with the same permissions you already have.
What changes is reach. Survey Agent is about surveys. Headless spans the platform. The registry keeps growing, with servers live today across surveys, BI dashboards, CX, InsightsHub, Audience, Communities, and more. So you can ask your assistant to build a survey, pull an NPS trend into a dashboard, then summarize findings from your research repository, all in one conversation.
The short version: headless is your AI assistant coming to the whole platform.
How they differ, point by point
Where you work
Survey Agent runs inside QuestionPro. You are in the builder, talking to the agent, watching the survey take shape on the same screen. Headless runs inside your assistant. You never open the dashboard unless you want to.
What they reach
Survey Agent focuses on one job and does it well: building and editing surveys, with logic and document import. Headless reaches across the platform, so the same prompt can move from a survey to a dashboard to a repository without you switching tools.
What it takes to start
Survey Agent needs nothing. If your account includes it, it is already there. Headless takes about five minutes to connect: open your assistant’s settings, add the QuestionPro connector, sign in, and approve. The help center walks through it for Claude and ChatGPT.
One team, both tools
Picture a UK insurer running a customer experience program.
On Monday, a CX manager needs a post-claim satisfaction survey with follow-up questions only for unhappy customers. She opens QuestionPro, describes the flow to Survey Agent, and the branching is built for her. No rule wiring, no research team ticket.
On Thursday, the same team is in a review meeting, working in ChatGPT. Someone asks how detractor sentiment shifted last quarter. Through the headless connection, the assistant pulls the data, segments it, and surfaces the answer while the meeting is still running. Nobody exports a CSV.
Same platform. Same data. Two tools, picked by where the work was already happening.
Which one fits your task
Reach for Survey Agent when the job is a survey. You are building or editing one, you want logic set up correctly, you have a document to import, or you simply want to stay inside QuestionPro while you work.
Reach for headless when the job is bigger than a survey, or when you are already living in your AI assistant. You want to drive dashboards, repositories, audiences, or CX workflows by prompt, and you would rather not leave the tool you are in.
Often the honest answer is both. One builds inside the app. The other connects the app to everywhere else you think. The aim is simple: put AI where it actually helps you, not where a roadmap decided it should live. If you want the wider view on that shift, the MCP survey tool breakdown goes deeper.
Get started
You do not have to choose on day one. Try Survey Agent the next time you build a survey, and connect headless the next time you are working in Claude or ChatGPT. See which one fits the way you already work. Create a free QuestionPro account and start with whichever door is closer.
Frequently asked questions
Survey Agent is an AI agent inside QuestionPro that builds and edits surveys for you, including logic and document import. Headless is a set of connections that let an outside AI assistant, like Claude or ChatGPT, operate QuestionPro directly. The simplest way to hold it: Survey Agent brings AI to your survey, while headless brings the whole platform to your AI assistant. Many teams use both.
No. Survey Agent works from plain-language prompts inside the survey builder, so there is nothing to configure. Headless takes about five minutes to set up: you open your AI assistant’s settings, add the QuestionPro connector, sign in with your account, and approve the connection. After that, you ask for what you need in plain English. No code, no API keys.
Yes, and they often do. They cover different jobs. You might use Survey Agent inside QuestionPro to build a survey with branching, then use a headless connection from your AI assistant to pull the responses into a dashboard and summarize them later. Same account, same data, and the same permissions apply across both.
Any assistant that supports the Model Context Protocol can connect. Today that includes Claude, ChatGPT, Cursor, Windsurf, and VS Code, among others. Because the protocol is an open standard rather than a private integration, new assistants that adopt it can connect without QuestionPro building a separate link for each one.
Yes, within the controls you already have. Connections use OAuth, so there are no shared keys, and every session is bound to your real account. Your assistant can only touch what your role allows, and you can revoke access at any time from your settings. Actions are scoped and logged, so a view-only user stays view-only through the connection.
Yes. Survey Agent supports importing questions from Word and PDF files. Open the chat window, attach your document, tell it what you want, for example to import every question, and it parses and organizes the questions into a structured survey. That turns an inherited brief or client questionnaire into a working draft without retyping.
Survey Agent is available on Team Edition, Research Edition, and University Research Edition, with support for the Essential License noted as coming soon. If you are unsure which edition your account uses, your account admin can confirm, or you can request a walkthrough from the QuestionPro team.
No. Surveys are one part of it. The headless registry keeps expanding across the platform, with servers for BI dashboards, CX, InsightsHub, Audience, Communities, employee experience, and more. That is the main way it differs from Survey Agent, which stays focused on survey building and editing.
