How to Conduct a JTBD Interview: Questions, Process, and Template

Why Most Customer Interviews Produce Useless Data

Your team runs customer interviews. You hear phrases like “it would be great if…” and “we really need a way to…” and “our biggest pain point is…” You diligently record these insights, synthesize them into themes, and feed them into your product roadmap.

And then you build the features, launch them, and discover that adoption is lukewarm.

The problem is not that you talked to the wrong customers. The problem is that you asked questions designed to extract feature requests when you should have been asking questions designed to extract desired outcomes. Traditional customer interviews are solution-oriented — they probe what customers want in a product. JTBD interviews are outcome-oriented — they probe what customers are trying to accomplish and how they measure success.

This distinction changes the questions you ask, the way you listen, and the analysis you produce. This guide walks through the complete JTBD interview process — from preparation through analysis — with specific questions, practical techniques, and a framework you can apply immediately.

For the broader JTBD methodology, see our Complete Guide to Jobs to Be Done.

Before the Interview: Preparation That Matters

Define the Job First

Before conducting a single interview, your team needs a working hypothesis of the job. This does not need to be perfect — the interviews will refine it — but you need a starting point.

Use the standard syntax: [verb] + [object of the verb] + [contextual clarifier]

Examples:

• “Monitor patient hemodynamic status during cardiac surgery”
• “Transport materials from ground level to elevated positions on a construction site”
• “Manage the nutritional intake of tube-fed patients in a clinical setting”

If you are unsure about the right level of abstraction, err on the broader side. It is easier to narrow a job that is too broad than to expand one that is too narrow.

Select the Right Participants

JTBD interviews target the job executor — the person who actually performs the job. This is not always the buyer or the decision-maker. For a surgical instrument, the job executor is the surgeon, not the procurement manager. For a loader crane, the job executor is the operator, not the fleet manager who authorized the purchase.

Aim for 15-30 interviews to achieve outcome saturation — the point where new interviews are no longer revealing new desired outcomes. In our experience:

• 8-12 interviews: You have captured 70-80% of the outcomes. Good enough for exploratory work.
• 15-20 interviews: You have captured 85-95% of the outcomes. Sufficient for most projects.
• 25-30 interviews: Near-complete coverage (95%+). Necessary for high-stakes product decisions or complex jobs.

Recruit participants who represent variety in how they accomplish the job, not variety in demographics. You want people who use different tools, work in different contexts, and have different levels of experience with the job — because these variations surface different outcomes.

Set Up the Interview

• Duration: 60-90 minutes. JTBD interviews go deeper than typical user interviews and cannot be rushed.
• Format: One-on-one, either in person or via video call. Avoid group settings — they suppress individual variation.
• Recording: Always record (with consent). You will need to review the transcript for precise outcome language.
• Team: One interviewer plus one note-taker. The interviewer focuses on the conversation; the note-taker captures candidate outcome statements in real time.

The Interview: A Stage-by-Stage Question Framework

The interview follows the structure of the Universal Job Map. For each stage of the job, you are trying to identify: (1) what the person does at this stage, (2) what outcomes they care about, and (3) where current solutions fall short.

Opening: Establishing the Job Context (10 minutes)

Start broad. You are establishing the context in which the job arises and confirming that your job definition is accurate.

Questions:

• “Walk me through the last time you [performed the job]. Start from the very beginning — what triggered it?”
• “What were you ultimately trying to accomplish?”
• “How often do you do this? Has the frequency changed over the past few years?”
• “Who else is involved when you do this? What are their roles?”
• “What tools, equipment, or systems do you use?”

What to listen for: The circumstance that triggers the job, the language the participant uses to describe the goal, and any stakeholders who influence how the job gets done.

Stage 1: Define (5-10 minutes)

The Define stage is where the person determines what needs to be done and plans their approach.

Questions:

• “Before you start, how do you determine exactly what needs to be done?”
• “What information do you need to figure out the right approach?”
• “What makes it difficult to determine the best approach?”
• “How do you know when you have enough information to proceed?”

What to listen for: Desired outcomes related to planning accuracy, information completeness, and approach selection. Example outcome: “Minimize the time it takes to determine the optimal configuration for the specific procedure.”

Stage 2: Locate (5-10 minutes)

The Locate stage is where the person gathers the inputs, materials, information, and resources needed.

Questions:

• “What inputs or materials do you need to gather before you can begin?”
• “Where do these come from? How do you find them?”
• “What happens when an input is missing or unavailable?”
• “How much time do you spend gathering what you need versus actually doing the core task?”

What to listen for: Outcomes related to input availability, search time, and completeness. Example outcome: “Minimize the time required to locate all necessary components before beginning the procedure.”

Stage 3: Prepare (5-10 minutes)

The Prepare stage is where the person sets up the environment and inputs for execution.

Questions:

• “What setup do you need to do before the core task?”
• “What is the most time-consuming part of preparation?”
• “What mistakes can happen during setup, and what are the consequences?”
• “How do you know when everything is properly prepared?”

What to listen for: Outcomes related to setup speed, error prevention, and readiness confirmation. In industrial contexts, this stage often contains the richest set of underserved outcomes — teams optimize execution and neglect preparation.

Stage 4: Confirm (5-10 minutes)

The Confirm stage is where the person verifies that everything is ready to proceed.

Questions:

• “Before you execute, how do you verify that everything is correct?”
• “What checks do you perform? Are they formal or informal?”
• “What happens if you discover something is wrong at this stage?”
• “How confident are you typically that everything is ready? What would increase that confidence?”

What to listen for: Outcomes related to verification speed, confidence level, and error detection. Example outcome: “Minimize the likelihood of discovering an error after execution has begun.”

Stage 5: Execute (10-15 minutes)

The Execute stage is the core of the job — the main task being performed.

Questions:

• “Walk me through the actual execution step by step.”
• “What requires the most concentration or skill?”
• “What can go wrong during execution, and how do you handle it?”
• “What does ‘perfect execution’ look like? How close do you typically get?”
• “What is the most frustrating aspect of the execution?”

What to listen for: Outcomes related to precision, speed, consistency, and error avoidance during the core task. This stage typically generates the most outcomes, but they are also the outcomes most likely to be already addressed by current solutions.

Stage 6: Monitor (5-10 minutes)

The Monitor stage is where the person tracks whether the job is being done successfully.

Questions:

• “During execution, how do you know things are going well?”
• “What feedback do you get? Is it sufficient?”
• “What would you want to know in real time that you currently cannot see?”
• “How do you detect problems before they become serious?”

What to listen for: Outcomes related to visibility, feedback quality, and early warning. Example outcome: “Minimize the time between a deviation occurring and becoming aware of it.”

Stage 7: Modify (5-10 minutes)

The Modify stage is where the person makes adjustments when something is not going as planned.

Questions:

• “When things go wrong, what adjustments do you make?”
• “How quickly can you course-correct?”
• “What is the cost of a late correction versus an early one?”
• “What modifications are easy to make? Which are difficult?”

What to listen for: Outcomes related to adjustment speed, correction accuracy, and recovery from errors.

Stage 8: Conclude (5-10 minutes)

The Conclude stage is where the person finishes the job, cleans up, and prepares for the next cycle.

Questions:

• “How do you know the job is complete?”
• “What happens after the core task is done? What cleanup or documentation is needed?”
• “How do you evaluate whether it was done well?”
• “What do you need to do to prepare for the next time you perform this job?”

What to listen for: Outcomes related to completion verification, documentation burden, and cycle time. In regulated industries (MedTech, pharma), this stage often contains significant unmet needs related to compliance documentation.

Closing: Emotional and Social Dimensions (10 minutes)

After walking through the job map, probe the emotional and social dimensions. These are harder to surface during the step-by-step walkthrough because they cut across the entire job.

Questions:

• “When you are performing this job, what worries you most?”
• “What would make you feel more confident during this process?”
• “How does your performance on this job affect how others perceive your competence?”
• “If a colleague were watching you do this, what would you want them to notice?”
• “What aspect of this job, if you could improve it, would most affect your professional reputation?”

What to listen for: Emotional outcomes (feel confident, avoid anxiety, reduce stress) and social outcomes (be perceived as competent, demonstrate expertise, earn trust). See our deep dive on functional, emotional, and social jobs.

The emotional and social questions are where most interviewers get uncomfortable. B2B product managers are used to discussing specifications and performance metrics. Asking a surgeon what worries her or asking an engineer what he wants his boss to notice feels personal. But these questions consistently reveal the outcomes that drive purchase decisions. Push through the discomfort.

What to Listen For: Recognizing Outcome Statements

The most important skill in a JTBD interview is recognizing when a participant is describing a desired outcome — even when they do not express it in outcome statement format.

Signals That an Outcome Is Being Expressed

• Complaints: “I hate how long it takes to…” → Desired outcome: “Minimize the time it takes to…”
• Workarounds: “I ended up building a spreadsheet to track…” → Desired outcome: “Minimize the effort required to track…”
• Wishes: “It would be great if I could see…” → Desired outcome: “Minimize the likelihood of missing…”
• Comparisons: “The old system was better at…” → Desired outcome related to the dimension being compared
• Frustrations: “The worst part is when…” → Desired outcome: “Minimize the frequency of…”

Converting to Proper Outcome Statement Format

Every desired outcome should be expressed as: [Direction of improvement] + [unit of measure] + [object of control] + [contextual clarifier]

The direction of improvement is almost always “minimize” — minimize the time, minimize the likelihood, minimize the effort, minimize the number of, minimize the variability of.

Raw interview data: “I spend forever trying to figure out whether the settings are right for this particular patient.”

Converted outcome: “Minimize the time it takes to determine the optimal device settings for a specific patient profile.”

This conversion happens during analysis, not during the interview. In the interview, capture the raw language. In analysis, convert to proper outcome format.

After the Interview: Analysis Framework

Step 1: Extract Candidate Outcomes

Review each interview transcript and extract every statement that expresses a desired outcome. Aim for 15-25 candidate outcomes per interview. Over 20 interviews, you will generate 300-500 raw candidate outcomes.

Step 2: Deduplicate and Standardize

Many candidate outcomes will be expressed differently but mean the same thing. Group duplicates and near-duplicates. Standardize the language using the outcome statement syntax. Your 300-500 raw candidates should consolidate to 50–150 unique desired outcomes.

Step 3: Map to the Job Map

Assign each outcome to a stage of the Universal Job Map. This reveals the distribution of outcomes across stages. If you find that 80% of your outcomes are in the Execute stage, you may have under-probed the other stages — consider running a few additional interviews focused on preparation, confirmation, and conclusion.

Step 4: Classify by Dimension

Tag each outcome as functional, emotional, or social. A typical distribution is 60-70% functional, 20-30% emotional, and 5-10% social. If your emotional and social outcomes are below 15% combined, you likely under-probed these dimensions.

Step 5: Prepare for Quantitative Validation

The qualitative outcomes from your interviews become the items in your quantitative survey instrument. Each outcome becomes a survey question with two scales: “How important is this to you?” and “How satisfied are you with how current solutions address this?” This quantitative step — surveying 200-600 respondents — is what transforms qualitative insights into prioritized opportunity scores.

For how to translate these outcomes into actionable product specifications, see our article on outcome statements and product requirements.

The JTBD Interview Template

Here is a condensed template you can use as a reference during interviews:

Pre-Interview Checklist

• Job statement defined (working hypothesis)
• Participant represents a job executor
• Recording consent obtained
• Interview scheduled for 60-90 minutes
• Note-taker assigned

Interview Flow

1. Context (10 min): Trigger, frequency, stakeholders, tools
2. Define (5-10 min): Planning, information needs, approach selection
3. Locate (5-10 min): Input gathering, resource availability
4. Prepare (5-10 min): Setup, configuration, readiness
5. Confirm (5-10 min): Verification, checks, confidence
6. Execute (10-15 min): Core task, precision, errors, frustrations
7. Monitor (5-10 min): Feedback, visibility, early warning
8. Modify (5-10 min): Adjustments, corrections, recovery
9. Conclude (5-10 min): Completion, documentation, next cycle
10. Emotional/Social (10 min): Worries, confidence, reputation, perception

Post-Interview

• Extract candidate outcomes (15-25 per interview)
• Flag key quotes and contextual insights
• Note any job map stages that were under-explored
• Update running outcome list

Common Interview Mistakes

Mistake 1: Asking “What Do You Need?”

This question invites feature requests, not outcomes. Replace it with “What are you trying to accomplish?” or “How do you measure whether this went well?”

Mistake 2: Leading the Witness

“Would it be helpful if the system could automatically calibrate?” is a leading question that embeds a solution. Instead: “What is involved in calibration, and what would make that part of the process go better?”

Mistake 3: Accepting Vague Answers

“I need it to be faster” is not an outcome. Probe deeper: “Faster at which specific step? What would faster mean — how fast is fast enough? What are the consequences of the current speed?” The goal is to get to a specific, measurable outcome statement.

Mistake 4: Spending All Your Time on the Core Task

Inexperienced JTBD interviewers spend 70% of the interview on the Execute stage because it feels like the most important part. Resist this. Some of the richest underserved outcomes live in Prepare, Confirm, and Conclude. Budget your time across all eight stages.

Mistake 5: Skipping Emotional and Social Probes

Particularly in B2B technical interviews, interviewers (and participants) are more comfortable discussing functional performance than emotional experience. The emotional and social questions at the end of the interview consistently produce insights that no other research method captures. Do not skip them.

Frequently Asked Questions