{"site":{"name":"Koji","description":"AI-native customer research platform that helps teams conduct, analyze, and synthesize customer interviews at scale.","url":"https://www.koji.so","contentTypes":["blog","documentation"],"lastUpdated":"2026-05-18T13:33:30.763Z"},"content":[{"type":"documentation","id":"7692e72d-c4f7-4977-a315-e91a0939886a","slug":"writing-interview-questions","title":"How to Write Great Interview Questions","url":"https://www.koji.so/docs/writing-interview-questions","summary":"This guide covers the principles of effective interview question design, including the distinction between open and closed questions, how to avoid leading questions, five probing strategies, and how to structure an interview guide. It includes before-and-after examples of question rewrites.","content":"# How to Write Great Interview Questions\n\nThe questions you ask in a user interview determine the quality of insights you get out. Ask the wrong questions and you'll collect an hour of data that tells you nothing useful — or worse, data that misleads you.\n\nThis guide covers the principles, techniques, and specific examples you need to write interview questions that surface genuine, actionable insights.\n\n## The Golden Rule: Ask About Behavior, Not Opinion\n\nThe single most important principle in interview question design is this: **ask people what they do, not what they think they would do.**\n\nHuman beings are remarkably bad at predicting their own future behavior. Research from the Journal of Consumer Research shows that stated purchase intentions correlate with actual purchasing behavior only about 30% of the time. When you ask \"Would you use a feature that does X?\" you're asking someone to simulate a future scenario — and their answer is essentially worthless.\n\nInstead, ground your questions in past behavior and concrete experiences:\n\n- **Bad:** \"Would you find it useful to have a dashboard that shows team progress?\"\n- **Good:** \"Tell me about the last time you needed to understand what your team had accomplished. What did you do?\"\n\nThe good version reveals actual behavior, real frustrations, and genuine workarounds — none of which the opinion-based version would surface.\n\n## Open vs. Closed Questions\n\nOpen-ended questions invite elaboration. Closed questions invite a single word or phrase.\n\n| Question Type | Example | What You Get |\n|---|---|---|\n| **Closed** | \"Do you use project management software?\" | \"Yes\" or \"No\" — a dead end |\n| **Open** | \"Walk me through how you keep track of your projects.\" | A rich description of actual workflow |\n| **Closed** | \"Is collaboration important to you?\" | \"Yes\" (obviously) — no insight |\n| **Open** | \"Tell me about a recent project where you had to coordinate with others. How did that go?\" | Specific challenges, tools, and pain points |\n\nAs a rule, **80% or more of your interview questions should be open-ended.** You can use a few closed questions for factual setup (\"How many people are on your team?\"), but the substance of your interview should invite stories and explanations.\n\n## Avoiding Leading Questions\n\nA leading question pushes the participant toward a particular answer. They're insidious because they often feel natural to the interviewer but contaminate the data.\n\nResearch by Elizabeth Loftus demonstrated that even subtle wording changes in questions can significantly alter participant responses and memory recall. This applies directly to user interviews.\n\n### Examples of Leading Questions and Rewrites\n\n| Leading Question | What's Wrong | Neutral Rewrite |\n|---|---|---|\n| \"How frustrated do you get when the page loads slowly?\" | Assumes frustration exists | \"How do you feel about the page load times?\" or better: \"Walk me through what happens when you open the app.\" |\n| \"Don't you think it would be better if we added search?\" | Leads toward agreement | \"If you can't find what you're looking for, what do you do?\" |\n| \"Most people love this feature — what do you think?\" | Social pressure to agree | \"What has your experience been with this feature?\" |\n| \"How much time does this clunky process waste?\" | \"Clunky\" and \"waste\" embed a judgment | \"Walk me through this process. How does it fit into your day?\" |\n| \"Would you pay for a premium version with advanced analytics?\" | Suggests they should want it | \"How do you currently measure results? What works and what doesn't?\" |\n\n## Probing Techniques\n\nYour initial questions open a topic. Probing questions dig deeper into it. Great interviewers spend more time probing than asking new questions.\n\n### The Five Probing Strategies\n\n**1. Elaboration probes** — Ask for more detail.\n- \"Can you tell me more about that?\"\n- \"What do you mean when you say it was 'complicated'?\"\n\n**2. Clarification probes** — Ensure you understand correctly.\n- \"When you say 'the team,' who specifically are you referring to?\"\n- \"You mentioned switching tools — which tools were those?\"\n\n**3. Example probes** — Ground abstract statements in concrete instances.\n- \"Can you give me a specific example of when that happened?\"\n- \"Think about the last time you did that — walk me through it step by step.\"\n\n**4. Contrast probes** — Explore differences and preferences.\n- \"You mentioned you used to do it differently. How does the current way compare?\"\n- \"What would the ideal version of that look like?\"\n\n**5. Emotional probes** — Understand feelings and motivations.\n- \"How did that make you feel?\"\n- \"What was going through your mind at that point?\"\n\n## Structuring Your Interview Guide\n\nA well-structured guide follows a natural conversational arc:\n\n### 1. Warm-Up (2–3 questions)\nPurpose: Build rapport, gather context, ease into the conversation.\n\n- \"Tell me a bit about your role. What does a typical day look like?\"\n- \"How long have you been working in [domain]?\"\n\n### 2. Broad Context (2–3 questions)\nPurpose: Understand the participant's world before zooming into your topic.\n\n- \"Walk me through how your team currently handles [process].\"\n- \"What tools or methods do you use for [activity]?\"\n\n### 3. Core Topic (5–8 questions)\nPurpose: Deep dive into the specific area you're researching.\n\nThis is where you spend most of your time. Frame questions around key research themes, and be prepared to abandon your planned order to follow interesting threads.\n\n### 4. Reflection (1–2 questions)\nPurpose: Let participants step back and reflect on what they've shared.\n\n- \"If you could change one thing about how you [do X], what would it be?\"\n- \"What would need to be true for [process] to work really well for you?\"\n\n### 5. Wrap-Up (1–2 questions)\nPurpose: Capture anything you missed.\n\n- \"Is there anything about [topic] that I should have asked about but didn't?\"\n- \"Anything else you'd like to share before we wrap up?\"\n\n## The Mom Test Connection\n\nRob Fitzpatrick's *The Mom Test* offers a complementary lens on question design. His core insight is that even your mom will lie to you if you ask bad questions — not out of malice, but because humans naturally want to be encouraging and supportive.\n\nThe antidote is to never ask about the future or opinions, and instead focus on specifics from the past. For a deeper dive, see our [Mom Test methodology guide](/docs/mom-test-methodology).\n\n## Question Design for Different Research Goals\n\n| Research Goal | Question Strategy | Example |\n|---|---|---|\n| **Explore a problem** | Ask about current behavior and pain points | \"Walk me through the last time you had to [task]. What was that like?\" |\n| **Understand motivation** | Ask about triggers and decision-making | \"What prompted you to start looking for a new solution?\" |\n| **Evaluate a concept** | Show, don't tell; ask for reactions | \"Here's a rough sketch of an idea. What's your first reaction?\" |\n| **Map a journey** | Ask for step-by-step walkthroughs | \"Take me through the entire process from when you first realized you needed X to today.\" |\n| **Assess value** | Ask about current costs and workarounds | \"How do you handle this today? What does that cost you in time or effort?\" |\n\n## Advanced Technique: The Critical Incident Method\n\nDeveloped by John Flanagan in the 1950s, the critical incident technique asks participants to recall a specific, vivid experience — positive or negative — related to your research topic.\n\nResearch in the Journal of Service Research found that critical incident narratives yield richer behavioral and emotional data than general questions about the same topic.\n\n**Template:** \"Think about a time when [situation] went particularly [well/badly]. Can you walk me through exactly what happened, from the beginning?\"\n\nThis technique produces vivid, story-rich data that's highly quotable in reports and highly useful for identifying specific design opportunities.\n\n## Practice Exercise\n\nTake any one of these research questions and write 5 interview questions for it:\n\n1. \"How do product managers decide which features to build next?\"\n2. \"What motivates people to switch from one email client to another?\"\n3. \"How do remote teams build trust?\"\n\nFor each question, check: Is it open-ended? Is it neutral? Does it ask about behavior rather than opinion? Could it be more specific?\n\nTools like Koji can help you brainstorm and refine interview questions before your study, ensuring your guide is balanced and free of common biases.\n\n## Further Reading\n\n- [The Definitive Guide to User Interviews](/docs/user-interview-guide) — the complete interview process\n- [Mom Test Methodology](/docs/mom-test-methodology) — complementary question design principles\n- [Choosing a Methodology](/docs/choosing-a-methodology) — pick the right approach for your research goals\n\n## Further reading on the blog\n\n- [Customer Interview Questions: 50+ Templates for Discovery, Churn, and Win/Loss (2026)](/blog/customer-interview-questions-templates) — The template is not the bottleneck — conducting the interview at scale is. Here are 50+ customer interview questions organized by use case, \n- [How to Run AI-Powered Customer Interviews at Scale](/blog/how-to-run-ai-powered-customer-interviews-at-scale) — Learn how to conduct effective customer interviews at scale using AI. This comprehensive guide covers everything from planning and question \n- [The Mom Test for Customer Interviews: How to Ask Questions That Get Real Answers (2026)](/blog/mom-test-customer-interviews-2026) — The Mom Test is the founder's guide to customer interviews that actually work. This complete guide covers Rob Fitzpatrick's three rules, rea\n\n<!-- further-reading:blog -->\n","category":"Research Methods","lastModified":"2026-05-16T03:27:38.00446+00:00","metaTitle":"Writing Interview Questions — Koji Docs","metaDescription":"Learn to write open-ended, neutral interview questions that produce genuine user insights. Includes examples, rewrites, and probing techniques.","keywords":["interview questions","question design","user research questions","open-ended questions","research methodology","probing techniques","leading questions"],"aiSummary":"This guide covers the principles of effective interview question design, including the distinction between open and closed questions, how to avoid leading questions, five probing strategies, and how to structure an interview guide. It includes before-and-after examples of question rewrites.","aiPrerequisites":["user-interview-guide"],"aiLearningOutcomes":["Write open-ended questions that elicit rich behavioral data","Identify and rewrite leading questions","Apply five probing techniques to deepen responses","Structure an interview guide with a natural conversational arc","Adapt question strategies to different research goals"],"aiDifficulty":"beginner","aiEstimatedTime":"10 min read"}],"pagination":{"total":1,"returned":1,"offset":0}}