UX Research Interview Questions Prompt for ChatGPT
Turns a product and a research goal into a structured set of user interview questions warm-up, behavioral, pain-point, and wrap-up written to avoid leading the participant. Most AI-generated interview question lists have the same problem: they’re leading. “Don’t you find this feature frustrating?” isn’t a research question, it’s a suggestion wearing a question mark and...
Act as a senior UX researcher preparing for a round of user interviews. The
product is: "{product description}". The research goal is: "{research goal}.
Generate a structured interview guide with 10-12 questions, grouped into
four sections:
1. Warm-up (2 questions) build rapport, ease into the topic
2. Behavioral (4-5 questions) what the user actually does today, not
what they think or want
3. Pain-point exploration (3-4 questions) tied directly to the research
goal above
4. Wrap-up (1-2 questions) anything missed, final thoughts
Rules:
- No leading questions (nothing that implies a "correct" or expected answer)
- No double-barreled questions (asking two things at once)
- Favor open-ended phrasing ("Walk me through...", "Tell me about a time...")
over yes/no questions
Present as a numbered list under each section heading. After the list, add
2-3 optional follow-up probes a researcher could use if an answer is too
brief.Variables
| Placeholder | Example | What changes if you alter it |
|---|---|---|
| {Product description} | a mobile app that helps freelancers track invoices and get paid faster | What the product is shapes which behaviors and pain points the questions probe. |
| {Research goal} | understand why users stop using the app after the first week | The specific thing you're trying to learn this is what the pain-point section targets directly. |
Actual output, by model
Real output this exact prompt produced — not a description of what it might do.
[full 12-question guide as pasted, verbatim, sections 1-4 plus 3 optional probes]
Builds a ready-to-use interview guide questions grouped by stage, written to avoid leading the participant.
How To Use
- Replace {product_description} with a sentence about what you're building
- Replace {research_goal} with the specific thing you're trying to learn
- Paste into ChatGPT
- Print or copy the output into your interview notes before the session
Pro Tips
- ★Run this once per research goal, not once for your whole product a guide built for "why do users churn" and one built for "how do users discover feature X" should not be the same document
- ★Read the output out loud before your first interview questions that look fine on paper sometimes sound leading when actually spoken
- ★Use the optional follow-up probes only when an answer is thin; don't read them as required questions, or the interview turns into an interrogation
Example output
1. Warm-up Tell me about the kinds of freelance work you do and how you typically handle getting paid. Walk me through the most recent time you sent an invoice to a client. 2. Behavioral Walk me through what you usually do after completing work for a client and before receiving payment. Tell me about how you currently keep track of invoices that are still unpaid. Think about the last time you used an app or tool to manage an invoice. What did you do with it? Walk me through what you did during your first week using this app. Tell me about the last time you stopped using the app. What happened around that time? 3. Pain-point exploration Tell me about anything that made using the app difficult during your first week. Walk me through a moment when you decided not to use the app for something you had previously used it for. Tell me about anything that happened in the app that made it less useful to you over time. Think about the point when you stopped opening or using the app regularly. What was happening at that point? 4. Wrap-up Looking back at your experience, what should I understand about why your use of the app changed after the first week? 4. Optional follow-up probes “What happened next?” “Can you give me a specific example?” “What did you do instead?”
FAQs
Sources & further reading
Articles, websites, and tools this prompt draws on or recommends.
No Perfect Match?
Generate a custom prompt built around your profession, AI model and goal.
Generate a Prompt →UX Research Interview Questions Prompt for ChatGPT Description⌄
Turns a product and a research goal into a structured set of user interview questions warm-up, behavioral, pain-point, and wrap-up written to avoid leading the participant.
Most AI-generated interview question lists have the same problem: they’re leading. “Don’t you find this feature frustrating?” isn’t a research question, it’s a suggestion wearing a question mark and it quietly wrecks the data before the interview even starts. This prompt is built to catch that failure mode by naming it directly in the instructions, not by hoping the model avoids it on its own.
It’s aimed at product managers, designers, and researchers who need a solid first draft of an interview guide fast before a round of user interviews, a usability test, or early discovery work without needing formal research training to know what a good question looks like.
What makes it hold up:
- Questions are explicitly grouped by interview stage (warm-up, behavioral, pain-point, wrap-up), which mirrors how real interview guides are structured, not just a flat list of ten questions in no particular order
- The “avoid leading or double-barreled questions” instruction is stated outright, because that’s the single most common way DIY research questions go wrong
- {product_description} and {research_goal} are separate variables, so the same skeleton produces a genuinely different question set for “increase trial-to-paid conversion” versus “understand why users abandon onboarding” on the same product
One thing it won’t do: pick your participants or your research method for you see the Pro tips for what to pair it with.
Version history
- September 26, 2026Baseline entry first test of this prompt with product description="a mobile
Send me your AI wisdom
One email a week: new prompts worth using, and what's changed across the models.
Leave a reply