How to run a Christensen JTBD interview: a practical guide with AI prompts

The Christensen JTBD Interview is a qualitative discovery method rooted in Clayton Christensen’s Jobs Theory, designed to uncover the “job” a customer hires a product to do by combining structured observation with exploratory conversation. Unlike the Switch Interview (Moesta), which reconstructs one purchasing timeline, the Christensen approach starts from the theory that people buy products to make progress in specific circumstances and probes three dimensions of every job — functional, emotional, and social. The method is especially suited for new market creation, innovation strategy, and situations where the real competition includes non-consumption — people choosing to do nothing at all.

Where this fits among JTBD approaches

Jobs to be Done is a family of approaches, not a single method. Four distinct schools have emerged, each with its own research design, philosophy, and output:

Christensen JTBD Interview (this guide) — Clayton Christensen. Qualitative, theory-driven. Discovers jobs through observation and open-ended conversation, probing functional, emotional, and social dimensions. Outputs: job definitions, circumstance maps, non-consumption analysis. Best for: innovation strategy, new market creation, product-market fit discovery.
JTBD Switch Interview — Bob Moesta. Qualitative, retrospective. Reconstructs the full purchasing timeline and maps four forces (push, pull, anxiety, habit). Outputs: force diagrams, job stories, switching timelines. Best for: positioning, go-to-market, churn reduction.
Outcome-Driven Innovation (ODI) — Tony Ulwick. Quantitative. Surveys 100-150 desired outcomes, calculates opportunity scores. Outputs: job maps, outcome-based segments. Best for: systematic feature prioritization, large-scale market segmentation.
JTBD Canvas Workshop — Jim Kalbach. Mixed methods. Practical synthesis of all schools using a canvas format. Outputs: JTBD Hypothesis Canvas, job stories. Best for: agile teams, cross-functional alignment.

Christensen and Moesta developed Jobs Theory together, but their interview approaches serve different purposes. The Christensen Interview is broader and more exploratory — it aims to discover what the job is. The Switch Interview is narrower and more structured — it reconstructs how a specific switching decision happened. In practice, teams often run Christensen-style discovery interviews first to identify the job, then Switch Interviews to understand the purchase dynamics around that job.

What question does it answer?

What “job” are customers hiring this product (or category of products) to do, expressed in terms of the progress they seek in their life or work?
What are the functional, emotional, and social dimensions of that job — and which dimension dominates the hiring decision?
Who are the non-consumers — people who need this progress but aren’t hiring any solution — and what circumstances prevent them from acting?
What are the real competitors for this job, including workarounds, manual processes, and “do nothing”?
What circumstances trigger the job to arise — and what makes some circumstances more urgent than others?
Is there a gap between the “Big Hire” (deciding to buy) and the “Little Hire” (deciding to use) — and what causes that gap?

When to use

When you are creating a new product or entering a new market and need to discover what job exists before you can design a solution. The Christensen Interview finds the job; other methods (Switch, ODI) help you optimize around it.
When existing research describes demographics and preferences but not motivation — you know who buys, but not why. The three-dimension framework (functional, emotional, social) unpacks what demographic segments cannot.
When you suspect significant non-consumption — a large group of potential customers who need progress but are not hiring any solution. Christensen’s approach specifically looks for these people, unlike the Switch Interview which requires a completed purchase.
When your product is “hired” for a job you did not design for, and you need to discover what that job is. The observation phase reveals patterns that interviews alone miss.
When making innovation strategy decisions — should we pursue sustaining innovation (improve existing job performance) or disruptive innovation (serve non-consumers or overserved customers)?
When building a business case for a new category and you need to define the job in terms that executives, engineers, and marketers all understand.

Not the right method when you already know the job and need to understand the purchasing decision dynamics (use Switch Interview), when you need to prioritize across a large set of outcome statements (use ODI), when you need statistically representative data (use a survey), or when you are optimizing a known workflow (use usability testing).

What you get (deliverables)

Job definition: a structured statement of the progress the customer seeks, including the circumstances that make the job arise
Three-dimension job map: the functional requirements, emotional needs, and social needs of the job
Non-consumption analysis: a description of who is not hiring any solution, why, and what would need to change
Circumstance inventory: the situations, triggers, and contexts that cause the job to arise, ranked by frequency and urgency
Competitive set by job: the actual solutions customers hire for the same job — typically spanning multiple product categories
Big Hire / Little Hire gap analysis: where the decision to buy diverges from the decision to use, and what causes abandonment between the two
Experience map: the full set of experiences (purchase, setup, use, maintenance, disposal) that the product must deliver to satisfy the job

Participants and timing

Interview 10-15 people in three groups: (1) recent buyers who are actively using the product, (2) recent buyers who stopped using it or use it less than expected (Big Hire without Little Hire), and (3) non-consumers who experience the same circumstances but have not hired any solution.

Each session runs 45-60 minutes. The observation phase (if conducted) adds 1-2 hours per participant. Plan for 3-5 weeks total: 3-7 days recruitment, 1-3 days observation, 5-10 days interviews (2-3 per day), 3-5 days analysis.

How to run a Christensen JTBD interview

1. Define the job hypothesis

Start with a rough hypothesis about what job your product (or product category) is being hired to do. Frame it as a progress statement, not a feature list: “Help me [verb] [object] in [circumstance] so I can [outcome].” This hypothesis will be tested and refined through observation and interviews — it is a starting point, not a conclusion. Write down what you assume the functional, emotional, and social dimensions are. These assumptions will be challenged.

2. Observe before you ask

Before any interview, spend time watching how people behave in the circumstances where the job arises. This is Christensen’s “milkshake method”: the team that studied McDonald’s milkshake sales stood in the restaurant at 6:30 AM and watched who bought milkshakes, when, with what, and what they did next — before asking anyone a single question. Observation reveals patterns that people cannot articulate: the commuter who buys a milkshake to make a boring drive less tedious is not going to volunteer that information in a focus group. Record what you see: who, when, where, what else they bought, how they behaved, what surprised you.

3. Recruit across three groups

Recruit from three populations: active users (people currently hiring the product for the job), lapsed users (people who bought but stopped using — the “Big Hire without Little Hire” group), and non-consumers (people in the same circumstances who are not hiring any product). Non-consumers are the hardest to find and the most valuable. Screen for circumstances, not demographics: “Have you experienced [the circumstance] in the last 90 days?” matters more than age, income, or job title.

4. Open with the circumstance, not the product

Begin each interview by asking about the circumstance, not the product. “Tell me about the last time you needed to [progress statement]. Where were you? What was happening?” The goal is documentary-style recall of a specific moment, not abstract opinions. If the participant starts describing features or preferences, redirect: “Before we talk about what you chose, let’s go back to the moment the need appeared. What was going on in your life or work that made this matter?“

5. Probe three dimensions for each job

For every job that surfaces, probe all three dimensions:

Functional: “What did you actually need to get done? What would count as success?” Look for specific, measurable outcomes.
Emotional: “How did you want to feel during and after? What feelings were you trying to avoid?” Emotional dimensions often dominate decisions but are rarely volunteered.
Social: “How did you want to be seen by others? Whose opinion mattered?” Social jobs explain why people hire premium products, branded alternatives, or visible solutions.

Many jobs are primarily emotional or social with a thin functional wrapper. The milkshake’s functional job (fill me up during a commute) is less interesting than its emotional job (make a boring drive less tedious). If you only probe functional, you miss the real hiring criteria.

6. Map the competitive set by job

Ask: “Before you found [product], how were you handling this? What else did you try or consider?” The answers will rarely match your assumed competitor set. Christensen’s key insight: competition is defined by the job, not the product category. A milkshake competes with a banana, a bagel, and boredom — not with other milkshakes. Map every alternative the participant mentions, including “did nothing” and manual workarounds.

7. Investigate non-consumption

For non-consumer participants, focus on: “You experience [circumstance] but you haven’t hired any solution. Walk me through why.” Common barriers include: they do not realize a solution exists, they cannot access available solutions, they lack the skills or resources to use available solutions, or the available solutions demand too much time, money, or complexity. Non-consumption is the largest market opportunity in Christensen’s framework — it represents unserved demand.

8. Identify the Big Hire / Little Hire gap

Ask users and lapsed users: “Walk me through the moment you decided to buy. Now walk me through the first time you actually used it. What happened between those two moments?” Many products get bought but not used. The condo developer in Competing Against Luck discovered that retirees would buy a condo but delay moving for months because they could not face sorting through decades of possessions. The “Big Hire” (buying) was complete, but the “Little Hire” (moving in) was stalled. Understanding this gap reveals design opportunities in onboarding, setup, and first-use experience.

9. Synthesize: define the job and its dimensions

After 10-15 interviews, cluster the findings by circumstance similarity, not demographic similarity. For each cluster, write a full job definition that includes the circumstance that triggers the job, the functional, emotional, and social progress sought, the current competitive set, the barriers preventing non-consumers from hiring, and the Big Hire / Little Hire gap. Test each job definition against every interview transcript. A good job definition should explain at least 80% of the behaviors you observed and heard about.

10. Translate into innovation decisions

Job definitions become the input for product strategy. Ask: “Are we solving this job well enough for current consumers? Are we solving it at all for non-consumers? Is there a simpler, cheaper way to address this job that would reach people who currently cannot access existing solutions?” These questions separate sustaining innovation (better solutions for existing consumers) from disruptive innovation (adequate solutions for non-consumers).

How AI changes this method

AI compatibility: partial — AI can accelerate analysis, transcription, and pattern detection, but cannot replace observation or the exploratory interview itself. The Christensen Interview depends on noticing unexpected behaviors (observation phase) and following unpredictable conversational threads (interview phase) — both require human judgment.

What AI can do

Transcription and coding: Otter.ai or Grain transcribes interviews; an LLM tags each segment by job dimension (functional, emotional, social) and participant group (active, lapsed, non-consumer)
Competitive set extraction: After 10-15 transcripts, an LLM extracts all mentioned alternatives and clusters them by job, producing a competitive map
Non-consumption pattern detection: AI identifies recurring barriers across non-consumer interviews and ranks them by frequency
Job definition drafting: LLM generates candidate job definitions from coded transcripts, which the researcher validates and refines
Cross-interview pattern analysis: AI surfaces circumstance clusters, repeated emotional triggers, and common Big Hire / Little Hire gaps across the full dataset
Literature and market scanning: Before observation, AI scans reviews, forums, and support tickets for early signals of the job and competitive alternatives

What requires a human researcher

Observation: Watching customers in context requires physical or video presence, attention to non-verbal behavior, and the ability to notice anomalies
Exploratory interviewing: The Christensen Interview follows threads that emerge from the conversation — this requires real-time judgment, not a script
Emotional and social dimension probing: Participants rarely volunteer emotional or social motivations — drawing these out requires rapport, indirect questioning, and sensitivity to tone
Strategic synthesis: Deciding which jobs are strategically important and which represent disruptive opportunities requires business judgment

AI-enhanced workflow

Before the Christensen Interview, preparation meant manual review of customer data, support tickets, and market reports to form a job hypothesis. An LLM can now scan thousands of reviews and tickets in minutes, surfacing recurring circumstances and unmet needs. This compresses hypothesis formation from days to hours and produces a richer starting hypothesis with real language from customers.

After interviews, the most time-consuming step is coding transcripts by dimension and synthesizing across participants. Without AI, a team spends 2-3 hours per transcript on coding alone — 20-45 hours for a full study. With AI transcription and dimension-tagging, coding drops to 20-30 minutes of review per transcript. Pattern detection across 10-15 interviews, which previously required a full day of wall-mapping with sticky notes, now takes an hour of reviewing and refining AI-generated clusters.

The interview itself remains entirely human. The value of the Christensen approach is precisely in moments that cannot be scripted: the pause when a participant realizes they never thought about why they made a choice, the offhand comment about a workaround nobody on the product team knew about, the non-consumer who describes a circumstance the team had not considered.

Tools

Observation: Lookback (remote observation and screen recording), Hotjar / FullStory (session recordings for digital product observation), field notebook for physical observation.

Recording and transcription: Zoom / Google Meet for remote sessions, Otter.ai for real-time transcription, Grain for highlights and clips.

Recruitment: User Interviews for screened participant recruitment by circumstance, Respondent.io for B2B participants, Calendly for scheduling.

Analysis and synthesis: Dovetail for tagging and cross-interview pattern detection, Miro for job dimension mapping and competitive set visualization, Notion / Google Docs for collaborative analysis templates.

AI-assisted analysis: Insight7 for automated JTBD coding, Speak AI for transcript analysis, Qualz.ai for JTBD lens analysis, Claude / ChatGPT for transcript coding and job definition drafting.

Works well with

JTBD Switch Interview (Js): The Christensen Interview discovers the job; the Switch Interview reconstructs the purchase decision dynamics around that job. Run Christensen first to find the job, then Switch to understand how people hire and fire solutions for it.
In-depth Interview (Di): Broader exploratory conversations fill in life context and reveal adjacent jobs. The Christensen Interview is focused on progress and circumstances; the in-depth interview captures the person’s full situation.
Desk Research (Dr): Market reports, patent filings, and competitor analysis provide the raw material for the initial job hypothesis. Desk research before observation saves time and sharpens the hypothesis.
Survey (Sv): After discovering the job through interviews, a survey quantifies how many people experience the same circumstance and which job dimensions matter most. Surveys turn qualitative job definitions into market-sizing evidence.
Persona Building (Ps): Christensen’s circumstance-based clusters can be combined with demographic and behavioral data to build richer, job-driven personas that describe both who the customer is and what progress they seek.

Example from practice

A developer of retirement condominiums spent millions on high-end finishes — granite countertops, stainless steel appliances, three-car garages — and ran print ads in newspapers targeting affluent retirees. Units were not selling. The developer’s assumption was that the job was “move into a nicer, more manageable home.” Focus groups confirmed this: retirees said they wanted single-floor living, less maintenance, and proximity to their grandchildren.

The team applied the Christensen approach: they interviewed 30 people who had actually moved in, plus 20 who had visited a model unit, expressed interest, and then disappeared. The interviews focused on the circumstance, not the product. What emerged was unexpected: the functional job (smaller, easier home) was real but not the driver. The emotional job was managing a life transition without regret — retirees were paralyzed by the thought of sorting through thirty years of possessions and deciding what to keep. The social job was maintaining the family identity that their dining room table represented (Thanksgiving dinners, homework sessions, holiday meals). People who did not buy were stuck at the emotional and social dimensions, not the functional one.

The developer responded: they built storage units across the street and offered two years of free storage. They assigned a moving coordinator to help with sorting. They redesigned smaller units to accommodate a full-size dining room table. Sales went from one unit per month to three. The insight did not come from asking about features or running an A/B test on pricing — it came from probing the emotional and social dimensions of the job through circumstance-focused interviews.

Beginner mistakes

Treating Jobs Theory as the Switch Interview

Newcomers often read Competing Against Luck and then run a standard Switch Interview. The two approaches serve different purposes. The Christensen Interview is exploratory — its goal is to discover what the job is. The Switch Interview is retrospective — it reconstructs one purchasing decision. If you skip the discovery phase and go straight to switching timelines, you may reconstruct the purchase of a product hired for a job you never identified.

Defining jobs at the wrong abstraction level

“Help people be more productive” is too broad — it competes with coffee, meditation apps, and a good night’s sleep. “Help Denver accountants reconcile Q3 reports on Tuesdays” is too narrow — the market is three people. The right level has a circumstance specific enough to identify real competitors but broad enough to include thousands or millions of people.

Product teams default to functional job dimensions because they are concrete and measurable. However, the decision between two functionally equivalent alternatives is almost always resolved by an emotional or social factor. If you probe only function and skip emotion and social, you will produce job definitions that explain what people do but not why they choose one solution over another.

Skipping the observation phase

Going straight to interviews without observation means your questions are shaped by your assumptions, not by customer behavior. Observation reveals patterns that participants cannot articulate or would never think to mention.

Confusing demographics with circumstances

“Health-conscious millennials who shop at Whole Foods” is a demographic segment, not a job. A 55-year-old buying the same product for a completely different reason occupies the same demographic segment but is hiring for a different job. Christensen’s framework segments by circumstance, not by demographic.

AI prompts for this method

4 ready-to-use AI prompts with placeholders — copy-paste and fill in with your context. See all prompts for Christensen JTBD interviews →.