{"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":"2026-09-21T04:03:45.269Z"},"content":[{"type":"blog","id":"46dae7ec-ad9e-41af-9114-e92d3f497bf8","slug":"the-strategic-guide-to-planning-user-interviews-from-hypothesis-to-insight","title":"The Strategic Guide to Planning User Interviews: From Hypothesis to Insight","url":"https://www.koji.so/blog/the-strategic-guide-to-planning-user-interviews-from-hypothesis-to-insight","summary":"Comprehensive guide to planning user interviews. Covers defining problem statements, targeting behaviors over demographics, choosing methodologies (JTBD vs. Exploratory), and scripting questions for AI moderators like Koji.","content":"\r\n---\r\n\r\n## 1. PROBLEM — Define the Core Objective\r\n\r\nThe most common mistake in user research is starting with a solution (\"Do they like this feature?\") rather than a problem (\"Why do they struggle?\"). Before writing a single question, articulate exactly what you are trying to solve.\r\n\r\n### A. Problem Statement\r\n**What specific problem or question are you trying to answer?**\r\nAvoid vague goals like \"learn about users.\" Be precise. A strong problem statement anchors the entire project.\r\n* **Weak:** \"We want to know if people like our new checkout flow.\"\r\n* **Strong:** \"Users are abandoning the checkout process at the payment method selection screen. We need to understand the specific friction points or trust issues causing this drop-off.\"\r\n\r\n### B. Decision to Inform\r\n**What decision will this research help you make?**\r\nResearch without a decision is just trivia. Connect your study to a specific business outcome.\r\n* **Example:** \"This research will determine whether we prioritize integrating a 'Buy Now, Pay Later' option or focus on simplifying the existing credit card entry form.\"\r\n\r\n### C. Current Hypothesis\r\n**What do you currently believe to be true?**\r\nState your assumptions clearly so they can be validated or debunked. This prevents confirmation bias.\r\n* **Example:** \"We believe users are overwhelmed by too many shipping choices, rather than a lack of trust in the platform.\"\r\n\r\n### D. Success Criteria\r\n**How will you know the research was successful?**\r\nSuccess isn't just finishing the interviews; it's getting the data needed to move forward.\r\n* **Example:** \"We will have identified the top 3 specific anxieties users feel when entering payment details.\"\r\n\r\n---\r\n\r\n## 2. AUDIENCE — Target Behaviors, Not Demographics\r\n\r\nIn modern product discovery, **psychographics and behaviors** are infinitely more valuable than demographics. A 20-year-old student and a 50-year-old executive might both struggle with the same productivity issue for the same reason.\r\n\r\n### A. Required Experience\r\n**What experience must they have had to be relevant?**\r\nYou need participants who have recently encountered the problem you are solving. Memory fades quickly, so recency is key.\r\n* **Criteria:** \"Must have purchased a flight online within the last 3 months.\"\r\n\r\n### B. Behavior of Interest\r\n**What specific behavior are you interested in understanding?**\r\nLook for actions that signal intent or friction.\r\n* **Behavior:** \"Users who added items to their cart but abandoned the session before purchase.\"\r\n\r\n### C. Screening Question\r\n**The \"Knockout\" Questions**\r\nUse binary (Yes/No) or multiple-choice questions to strictly verify the fit. Do not ask leading questions in the screener.\r\n* **Bad:** \"Do you like travel apps?\" (Too easy to say yes).\r\n* **Good:** \"When was the last time you booked a flight online?\" (Options: Last week, Last month, >6 months ago, Never).\n\nWhere the participants come from is the same decision. If you already have them, users in your product or a CRM segment, import the list into Koji so each person gets a personalised link (you see who responded and the report is segmented by whatever you imported), or share the generic link with the lead form off when anonymity will get you franker answers. If you do not, open Panel recruit, describe the audience in plain language, and Koji finds them across 130+ countries at a price per completed interview in credits you approve first, on paid plans; your knockout questions run as study-level screening questions before the interview starts.\r\n\r\n---\r\n\r\n## 3. APPROACH — Choosing the Right Methodology\r\n\r\nYour approach dictates the \"vibe\" of the interview and the type of data you collect. Choose the methodology that matches your product stage.\r\n\r\n### Option A: The \"Jobs to be Done\" (JTBD) Approach\r\n**Use Case:** *Understanding why users switch products or determining true competitors.*\r\nThe JTBD framework focuses on the \"job\" the user is hiring your product to do. It ignores features and focuses on outcomes.\r\n* **Focus:** The timeline of events. What triggered the search for a solution? What was the \"struggle\"?\r\n* **The Approach:** Treat the interview like a documentary of their decision-making process.\r\n* **Key Question:** \"Take me back to the moment you realized your old solution wasn't working anymore. What happened that day?\"\r\n\r\n### Option B: The Exploratory (Generative) Approach\r\n**Use Case:** *Early-stage discovery, finding new opportunities, or \"Greenfield\" projects.*\r\nThis is broad and open-ended. You are looking for patterns in behavior, not feedback on a specific solution.\r\n* **Focus:** Understanding the user's worldview, daily routines, and hidden pain points.\r\n* **The Approach:** Adopt a \"Master/Apprentice\" model where the user teaches you about their life/work.\r\n* **Key Question:** \"Walk me through your morning routine from the moment you wake up.\"\r\n\r\n### Option C: The Evaluative (Usability) Approach\r\n**Use Case:** *Testing a prototype, beta feature, or live product iteration.*\r\nThis is specific and task-oriented. You are testing if a solution actually solves the problem identified earlier.\r\n* **Focus:** Friction, confusion, and success rates.\r\n* **The Approach:** Observation over conversation. Ask users to perform a task and \"think aloud.\"\r\n* **Key Question:** \"I'd like you to try to book a flight for next Tuesday. Please speak your thoughts out loud as you go.\"\r\n\r\n### Option D: The Critical Incident Technique\r\n**Use Case:** *Debugging churn, analyzing major failures, or understanding \"Power User\" moments.*\r\nThis focuses on extreme highs or lows to uncover deep insights.\r\n* **Focus:** A specific event that stands out in memory (e.g., a service outage or a moment of delight).\r\n* **The Approach:** Forensic detailed analysis of one specific event.\r\n* **Key Question:** \"Tell me about the last time this software completely failed you. What were the consequences?\"\r\n\r\n### Option E: Continuous Discovery (Scaled Asynchronous)\r\n**Use Case:** *Ongoing product health, keeping a pulse on the user base without scheduling bottlenecks.*\r\nInstead of one massive study, you run frequent, lightweight touchpoints.\r\n* **Focus:** Speed and volume of qualitative data.\r\n* **The Approach:** Using AI tools to conduct interviews 24/7. This is where platforms like **[Koji](https://koji.so)** excel. You can deploy a research link to hundreds of users simultaneously, or let Koji recruit the audience from its built-in panel. Koji acts as the moderator, asking personalized follow-up questions based on the user's previous answers, allowing you to gather deep qualitative insights at a quantitative scale.\r\n\r\n---\r\n\r\n## 4. QUESTIONS — Scripting the Conversation\r\n\r\nGreat interviews feel like conversations, but they are built on a skeleton of rigorous questions.\r\n\r\n### A. Key Questions (The \"Spine\")\r\nThese are your non-negotiables. Regardless of where the conversation goes, these must be answered.\r\n1.  **The Context Setter:** \"Walk me through the last time you [performed the behavior].\"\r\n2.  **The Trigger:** \"What was happening in your life that made you realize you needed to solve this problem?\"\r\n3.  **The Evaluation:** \"How did you go about looking for a solution?\"\r\n4.  **The Comparison:** \"What other tools did you try, and why didn't they work?\"\r\n\r\n### B. Topics to Explore\r\nBeyond specific questions, list themes you want to uncover. This is useful for \"semistructured\" interviews where you want to follow the flow but hit specific nodes.\r\n* *Perceived Value vs. Cost*\r\n* *Trust & Credibility*\r\n* *Workflow Integration*\r\n\r\n### C. Guardrails (Essential for AI Interviewers)\r\nWhen using AI tools like **Koji** to scale your research, you don't just provide questions; you provide boundaries. This ensures the AI digs deeper without leading the witness.\r\n\r\n**Examples of Guardrails:**\r\n* **\"Do not mention our product name:** Keep the conversation focused on the user's problem, not our solution, until the final question.\"\r\n* **\"Probe on Price:** If the user mentions 'budget' or 'expensive', immediately ask follow-up questions about their spending limits.\"\r\n* **\"Avoid Yes/No Questions:** If a user gives a one-word answer, ask 'Can you tell me more about that?'\"\r\n\r\n*Note: One of Koji’s distinct advantages is its ability to understand these guardrails and ask relevant, context-aware questions to understand the user better, rather than just reading off a static list.*","category":"Research","lastModified":"2026-09-20T15:27:10.965171+00:00","metaTitle":"The Strategic Guide to Planning User Interviews (with AI Templates)","metaDescription":"Master the art of user research planning. A step-by-step guide covering problem definition, audience targeting, and scripting for AI tools like Koji.so.","keywords":["User Research Plan","UX Interview Script","Continuous Discovery","Jobs to be Done Framework","AI User Research Tools","Product Discovery Process","Koji User Interviews"],"aiSummary":"Comprehensive guide to planning user interviews. Covers defining problem statements, targeting behaviors over demographics, choosing methodologies (JTBD vs. Exploratory), and scripting questions for AI moderators like Koji.","aiKeywords":["user interview plan","research hypothesis","recruiting participants","interview script template","AI user research"],"aiContentType":"guide","faqItems":[{"answer":"Start by defining the core problem statement, identify your target audience based on behavior, choose a methodology (like JTBD), and script your questions with clear guardrails.","question":"How do I plan a user interview?"},{"answer":"Exploratory research (generative) discovers new opportunities and user needs, while evaluative research (usability) tests if a specific solution works.","question":"What is the difference between evaluative and exploratory research?"},{"answer":"AI tools like Koji can help structure interview scripts, ensure guardrails are followed during moderation, and scale qualitative interviews to hundreds of users.","question":"How can AI help with user research planning?"}],"relatedTopics":["ai-user-interviews","product-discovery","continuous-discovery"]}],"pagination":{"total":1,"returned":1,"offset":0}}