Jobs-to-be-Done: S-Tier Behavioral Designer’s Guide

Most product teams I’ve sat across from in the last fifteen years are answering the wrong question. They ask, “What features should we build?” — when the question that actually moves the metric is, “What job is the customer hiring our product to do?” That single reframe, popularized by Clayton Christensen with a milkshake and weaponized by Tony Ulwick with a survey instrument, has rewired how serious behavioral designers think about motivation.

The trouble is that Jobs-to-be-Done has been so loudly evangelized — and so loosely defined — that it now means six different things to six different teams, half of which contradict each other. Christensen’s “switch interview” school treats jobs as anthropological narratives. Ulwick’s outcome-driven school treats jobs as quantifiable utility metrics. Strategyn’s consultants and Intercom’s product blogs have built a cottage industry on top of both, and most of the resulting frameworks ship without ever asking whether the underlying construct is even falsifiable.

This is the S-Tier Behavioral Designer’s Guide to Jobs-to-be-Done. We will define it precisely, walk through what Christensen and Ulwick got right, name three places where the theory falls apart under empirical scrutiny, and then map every part of it onto the eight Core Drives of the Octalysis Framework so a designer can actually use it in production. By the end, you’ll know when JTBD is the sharpest tool in the kit and when it’s a $400 rake.

Quick shorthand before we go further: Core Drive 1 (CD1) is Epic Meaning & Calling, Core Drive 2 (CD2) is Development & Accomplishment, Core Drive 3 (CD3) is Empowerment of Creativity & Feedback, Core Drive 4 (CD4) is Ownership & Possession, Core Drive 5 (CD5) is Social Influence & Relatedness, Core Drive 6 (CD6) is Scarcity & Impatience, Core Drive 7 (CD7) is Unpredictability & Curiosity, and Core Drive 8 (CD8) is Loss & Avoidance.

⚡ Speed Run Notes

• Jobs-to-be-Done reframes user behavior from “demographic with preferences” to “person hiring a solution to make progress in a specific context” — and that reframe is JTBD’s only durable contribution.
• Christensen’s “milkshake interview” produces vivid stories; Ulwick’s outcome-driven survey produces quantifiable utility scores. They are not the same theory, even when practitioners treat them as one.
• JTBD has three real failure modes: it is empirically unfalsifiable, it conflates description with causation, and the “functional vs emotional vs social” job hierarchy is a heuristic, not a measurement.
• JTBD describes the conscious goal layer above motivation; Octalysis’s eight Core Drives describe the unconscious motivational layer beneath it. Both layers must be designed deliberately.
• The right-brain Core Drives (CD3 Creativity, CD4 Ownership, CD5 Relatedness) are where most “social” and “emotional” jobs live; the left-brain Core Drives (CD2 Accomplishment, CD6 Scarcity) are where “functional” jobs hide.
• Ship JTBD as the discovery instrument and Octalysis as the design instrument. Use one without the other and you will either build elegant solutions to imaginary jobs or hire psychologically generic features for real ones.

Table of Contents

• What Is Jobs-to-be-Done
• The Core Findings
• What Christensen Got Right
• Where Jobs-to-be-Done Falls Apart
• The Brain on Jobs-to-be-Done
• Jobs-to-be-Done vs Other Theories
• Jobs-to-be-Done in the Real World
• The Elephant in the Room
• How to Apply Jobs-to-be-Done with the Octalysis Framework
• Practical Steps to Apply Jobs-to-be-Done
• Frequently Asked Questions

About Yu-kai Chou

Yu-kai Chou is an S-Tier Behavioral Designer and the creator of the Octalysis Framework, the gamification design system now applied to products and experiences reaching over 1.5 billion users. His book Actionable Gamification is one of the most-cited works in the field, and he has been ranked the #1 Gamification Guru in the World.

He has advised MrBeast, LEGO, Microsoft, Porsche, Tesla, Stanford, Harvard, and governments including Ukraine on turning behavioral psychology into product mechanics that actually change user behavior.

Verify: Wikipedia · Google Scholar · Wikidata · LinkedIn

What Is Jobs-to-be-Done

Jobs-to-be-Done (JTBD) is a theory of customer behavior that argues people don’t buy products — they “hire” products to make a specific kind of progress in a specific context. The unit of analysis is not the customer’s demographic profile, not the product’s feature list, and not even the customer’s preference. It is the job: a description of the progress a person is trying to make in a circumstance, including the functional, emotional, and social dimensions of that progress.

The phrase entered modern management discourse through Clayton Christensen’s 2003 book The Innovator’s Solution and was sharpened across a decade of Harvard Business Review essays culminating in his 2016 book Competing Against Luck. Christensen famously illustrated it with a fast-food chain that wanted to sell more milkshakes. Demographic segmentation and feature optimization had failed. The breakthrough came when researchers asked, “What job are people hiring this milkshake to do?” The answer for the morning customers turned out to be: keep me occupied and slightly full during a long, boring commute, with one hand free to drive. Bagels were too crumbly. Bananas were too quick. The milkshake was hired because it was thick enough to last thirty minutes through a straw.

That story has done more to popularize JTBD than any peer-reviewed paper, and it has done more to oversimplify JTBD than any peer-reviewed critique. Tony Ulwick — a consultant who actually predates Christensen’s framing by nearly a decade — built a parallel and more rigorous version of the theory called Outcome-Driven Innovation (ODI). Ulwick’s version treats a job as a hierarchy of desired outcomes, each rated by customers on importance and current satisfaction. The gap between the two yields an opportunity score that tells the team where to invest. Ulwick’s 2005 book What Customers Want and his 2016 follow-up Jobs to be Done: Theory to Practice codified the quantitative branch.

For the rest of this guide I will treat both branches as legitimate forms of JTBD — because both are practiced under that name — but I will mark which branch I am drawing from when the distinction matters. Conflating them is the single most common mistake I see in teams adopting the framework.

The Core Findings

Stripped of the milkshake stories and the consulting taxonomies, JTBD makes a small number of empirically defensible claims. The rest of the literature is application and elaboration.

Finding 1: People hire products to make progress, not to fulfill needs

The first and most important claim is that human purchase behavior is best modeled as progress-making rather than need-fulfilling. Progress-making is forward-directed, situational, and it carries a “from-where” and a “to-where”. Need-fulfilling is static, demographic, and explanatory after the fact. Christensen’s argument was that the standard marketing question — “What does the customer want?” — produces a list of preferences detached from any specific moment of struggle, while the JTBD question — “When the customer last struggled, what progress were they trying to make?” — anchors design in a context where intervention is possible.

Finding 2: The job is causal; the customer is correlational

Christensen pushed this claim hardest in Competing Against Luck: demographic, psychographic, and even behavioral segments are correlational artifacts. They tell you who tends to buy, not why they buy. The job — the progress in context — is the causal variable. If two thirty-five-year-old men with identical incomes and education show up at the same coffee shop, one to get caffeinated before a sales call and the other to kill an hour before a job interview, they are doing different jobs. They will respond to different interventions. Aggregating them by demography destroys the signal.

This claim is partially true and partially overreached. Genuine causal inference about purchase behavior requires more than re-narrating the moment of purchase — it requires controlled comparison — but the directional point survives: situational variables explain more variance in purchase decisions than demographic variables, and JTBD takes situational variables seriously where most marketing frameworks bury them.

Finding 3: Every job has functional, emotional, and social dimensions

The third durable claim is that any non-trivial purchase has three layers of progress: a functional dimension (what the product physically accomplishes), an emotional dimension (what the buyer feels in the moment of purchase and use), and a social dimension (how the purchase is read by others). The milkshake’s functional job is morning sustenance. Its emotional job is “give me a small reward to start a boring commute.” Its social job is nil. A Tesla’s functional job is transportation; its emotional job is “make me feel like a person who chose the future”; its social job is “let other people see I chose the future.” Designs that solve only the functional layer underperform designs that solve all three.

Finding 4: Jobs are more stable than solutions

The final claim, most useful for strategy, is that the jobs people hire products for change much more slowly than the products that get hired. Humans have wanted “morning energy that lets me drive with one hand” for as long as there have been morning commutes. The product changes — coffee, energy drinks, milkshakes, breakfast bars, mobile podcasts — but the job persists. This is why JTBD is useful for category-level strategy in a way that feature roadmaps are not.

What Christensen Got Right

Christensen’s contribution to JTBD is sometimes dismissed as “just a reframe” by people who’ve only read the milkshake anecdote. The reframe is the contribution. Three specific moves deserve credit before we pivot to critique.

He moved the unit of analysis from person to context

Pre-Christensen marketing taught us to segment by who the customer was. Christensen taught us to segment by where the customer was. That move — what behavioral economists would later call situated cognition — quietly reshaped how product teams write user stories, conduct research, and prioritize roadmaps. The standard product-management user story format (“As a [persona], I want [feature], so that [benefit]”) has been steadily replaced in mature teams by a JTBD-shaped format (“When [situation], I want to [motivation], so I can [outcome]”). The former locks the design to a static persona; the latter unlocks the design to any persona who finds themselves in that situation.

He gave teams a defense against feature factories

One of the great organizational benefits of JTBD — underrated in the academic literature, overrepresented in the practitioner literature — is that it provides product managers with a rhetorical defense against the relentless internal pressure to ship more features. “What job does this feature serve?” is a question that, properly used, kills two-thirds of feature requests before they reach the engineering backlog. Teams that adopt JTBD report a measurable reduction in feature-shipping velocity and a measurable increase in feature retention rate. That outcome alone has saved the framework from obsolescence.

He connected jobs to disruption theory

The most intellectually generative move Christensen made was connecting JTBD to his earlier work on disruptive innovation. Established companies optimize the products their existing customers hire for the jobs those customers already do. Disruptors enter by serving an under-served job — often by hiring a worse-on-paper product to do a job nobody had previously named. Netflix did not beat Blockbuster on movie quality; it served the job of “I want to watch something at 9 p.m. without leaving the house” better than Blockbuster’s job of “I want to browse a curated selection on a Friday night.” JTBD is what makes disruption theory operational: it tells you which under-served jobs to look for and how to recognize one when you see it.

Where Jobs-to-be-Done Falls Apart

JTBD is a powerful reframe, but it is not a complete behavioral theory, and the practitioner literature consistently overstates what the empirical literature can support. Three specific failure modes deserve serious attention from any designer adopting it.

Failure 1: It is not falsifiable in its current form

The most pointed academic critique of JTBD comes from the Anthony Ulwick / Strategyn camp themselves — specifically from a 2017 working paper by the Strategyn team that audited the Christensen branch. Christensen’s version of JTBD does not specify what observation would falsify a job statement. If a hypothesized job (“morning commute sustenance”) fails to predict purchase behavior, a Christensen-school analyst can simply re-narrate the situation (“ah, the real job was emotional reassurance during a boring commute”) and re-fit the theory to the data. This is the textbook hallmark of a non-falsifiable framework. Ulwick’s outcome-driven branch attempts to fix this by quantifying outcomes and measuring importance/satisfaction gaps — but practitioners overwhelmingly use the Christensen version, which does not include those quantitative checks.

The practical consequence: JTBD job statements often function as post-hoc rationalizations, not predictive hypotheses. A team that runs a switch interview will reliably produce a plausible-sounding job statement, regardless of whether that job statement actually generalizes to a broader population. Without a falsification protocol, the framework drifts toward storytelling.

Failure 2: It conflates description with causation

The second failure mode is more subtle. JTBD claims to explain why customers buy, but the methodology — primarily retrospective interviewing — can only describe what customers say about why they bought. The gap between actual decision-making and post-hoc verbal report is one of the most replicated findings in cognitive psychology. Nisbett and Wilson’s classic 1977 paper “Telling More Than We Can Know” showed that people confidently produce verbal justifications for behavior driven by factors they have no conscious access to. Daniel Kahneman’s Thinking, Fast and Slow spent four hundred pages elaborating the same point. Jobs-to-be-Done methodology, especially in its Christensen-school interview form, is asking customers to introspect on a process the literature says they cannot introspect on accurately.

This is why a milkshake interview produces a clean job statement and an A/B test sometimes produces the opposite result. The interview captures the customer’s narrative about the purchase. The A/B test captures the behavior. When they disagree, JTBD-school practitioners typically trust the interview — and that is exactly backwards from what behavioral science would recommend.

Failure 3: The functional/emotional/social trichotomy is a heuristic, not a measurement

The third failure is methodological. The functional / emotional / social job trichotomy is one of JTBD’s most popular pedagogical devices — it appears on every consultant slide and in every introductory blog post. But it is not operationalized in any way that would let two analysts looking at the same purchase agree on which dimension dominates. Is buying a Patagonia jacket functional (warmth), emotional (identification with environmental values), or social (other people seeing you in a Patagonia jacket)? Most analyses say “all three”; some weight one over the others; none provide a measurement protocol that distinguishes the three reliably. The trichotomy survives because it is intuitively appealing and rhetorically convenient, not because it has been validated as a measurement instrument.

This matters because design decisions follow from the weighting. A product team that decides the dominant job is functional will optimize for performance. A team that decides the dominant job is social will optimize for visibility. The framework gives them no principled way to choose between those interpretations — it gives them a vocabulary in which both choices sound reasonable.

The Brain on Jobs-to-be-Done

JTBD does not have its own neuroscience. It is a behavioral economics framework about purchase decisions, and the relevant neuroscience is the broader literature on goal-directed behavior, prospective cognition, and reward prediction. But the framework’s central claim — that people are trying to make progress — maps onto a real and well-characterized neural system, and a serious behavioral designer should understand the mapping before applying the framework.

The progress-making behavior JTBD describes is what neuroscientists call goal-directed action: behavior controlled by the predicted value of an outcome rather than by stimulus-response habits. Bernard Balleine and Anthony Dickinson’s work in the 2000s established that goal-directed action depends on the orbitofrontal cortex (OFC) and dorsomedial striatum, while habitual action depends on the dorsolateral striatum. The two systems compete for behavioral control depending on context. JTBD’s “switch moment” — the moment a customer decides to hire a new product — is a moment when the goal-directed system overrides the habit system. That has a specific neural signature: increased OFC activity during outcome evaluation, increased ventromedial prefrontal cortex activity during option comparison, and reduced posterior striatal activity as the habitual default loses control.

This matters for design in two ways. First, switch moments are rare and physiologically expensive — the brain does not run goal-directed evaluation on every purchase, only when the cost of being wrong exceeds the metabolic cost of deliberation. JTBD interviews recover those rare moments. Most purchases are habitual, and a JTBD interview about a habitual purchase will produce a confabulated job story. Second, the emotional layer of a JTBD job — the part Christensen consistently emphasized — is not a separate neural system; it is the affective signal that the goal-directed system uses to evaluate outcomes. Antonio Damasio’s somatic marker hypothesis is the closest thing JTBD has to a neuroscientific foundation, and the framework would benefit from acknowledging it explicitly.

The Octalysis Framework was designed with the goal-directed/habitual distinction in mind. Its eight Core Drives describe the affective signals — meaning, accomplishment, ownership, social belonging, scarcity, curiosity, loss, creativity — that the goal-directed system reads off to evaluate which “job” to hire which product for. JTBD names the job; Octalysis describes the affective machinery that makes the job feel like a job worth pursuing.

Jobs-to-be-Done vs Other Theories

JTBD is most useful when its overlap and non-overlap with adjacent frameworks is mapped explicitly. Practitioners often present JTBD as a complete replacement for older models, which it is not.

JTBD vs Maslow’s Hierarchy of Needs. Maslow describes standing motivational categories that any human carries; JTBD describes situational goals that any human pursues in a context. Maslow’s “esteem” need does not tell you what product to ship; the JTBD-style “When I’m in a meeting with peers, help me feel like the most informed person in the room” is closer to design-actionable. But Maslow predicts that esteem-relevant jobs will outweigh safety-relevant jobs only after safety has been satisficed — a prediction JTBD cannot make on its own.

JTBD vs Self-Determination Theory. SDT (Deci & Ryan, 1985) identifies autonomy, competence, and relatedness as the three innate psychological needs that drive intrinsic motivation. JTBD jobs are the conscious-level expressions of those needs in a specific context. A “social job” in JTBD vocabulary is almost always a relatedness need in SDT vocabulary. SDT explains why social jobs feel rewarding when satisfied; JTBD describes which social job is being hired in this particular purchase.

JTBD vs Bartle’s Player Types. Bartle’s Achiever / Explorer / Socializer / Killer typology is a player-side framework: it describes durable preferences a player carries into a game. JTBD is solution-side: it describes what a player is trying to do at a specific moment. The two are complementary. An Achiever player and a Socializer player can hire the same in-game tool to do different jobs.

JTBD vs the Hook Model. Nir Eyal’s Hook Model (Trigger → Action → Variable Reward → Investment) is a habit-formation framework operating on the habitual-action neural pathway. JTBD operates on the goal-directed pathway. A product that has been successfully “hooked” is being used habitually, and a JTBD interview about that use will recover a confabulated job. Designers should run JTBD on switch moments and Hook on habitual loops — not on the same behavior.

JTBD vs Octalysis. JTBD describes the conscious goal a customer is trying to achieve; Octalysis describes the eight unconscious motivational drives that determine whether achieving that goal will feel rewarding. JTBD tells you which job to design for; Octalysis tells you how to design that job so it actually motivates. Used together, they form a complete behavioral design pipeline. Used separately, each is incomplete. We will work the integration in detail in the Octalysis section below.

Jobs-to-be-Done in the Real World

Theory is cheap. The interesting question is where JTBD has produced measurably different design decisions across domains, and where it has reliably failed.

Software-as-a-Service: Intercom and the redefinition of onboarding

The most influential JTBD case study in the SaaS world is Intercom’s 2014 internal pivot, documented in their book Intercom on Jobs-to-be-Done. The team had been building Intercom as “messaging for businesses” — a feature-defined product. After running JTBD interviews with churned customers, they reframed the product around the job of “When my customer has a question, help me answer them in a way that doesn’t feel impersonal at scale.” That reframe drove the redesign of their onboarding, the rewriting of their pricing tiers, and a documented increase in trial-to-paid conversion. Intercom subsequently published their JTBD methodology, which became the de facto standard for SaaS adoption of the framework.

The lesson is generalizable: in SaaS, JTBD is most useful for activation and retention design, where the customer is making a deliberate decision about whether the product is doing the job. It is less useful for engagement design within an active session, where habitual patterns dominate.

Consumer Goods: Procter & Gamble’s outcome-driven launches

Procter & Gamble adopted Tony Ulwick’s outcome-driven version of JTBD in the early 2000s. Ulwick’s methodology — survey-based outcome scoring — produced quantitative opportunity scores that P&G used to prioritize new product development across categories from Tide laundry detergent to Crest toothpaste. The Crest 3D White line, launched in 2009, was specifically targeted at the outcome “minimize the visible yellowing of teeth from coffee and wine” — an outcome that scored high on importance and low on satisfaction in Ulwick’s instrument. The launch was a multi-billion-dollar success. P&G’s own product-development training materials credit the JTBD/ODI approach with shifting the company from feature-driven to outcome-driven development.

The lesson here is more cautious. ODI works in mature consumer-goods categories where customers can articulate stable outcomes and competitors are differentiated on a small number of dimensions. It works less well in emerging or experiential categories where customers cannot articulate what they want until they have seen it.

Education: When the job changes between learner and parent

Education products are where JTBD reveals one of its sharpest insights and one of its sharpest limitations simultaneously. The same product — say, a kindergarten learning app — is being hired by a child to do one job (“entertain me right now”) and by a parent to do a completely different job (“make me feel like a good parent and produce evidence my child is learning”). Successful education products design for both jobs in parallel; products that optimize for one of them get fired by the other party. Khan Kids, Lingumi, and Duolingo’s child-targeted product all show clear evidence of JTBD-shaped dual-customer design.

The limitation is that JTBD does not provide a method for resolving conflicts between the two jobs. When the child’s job (“let me play another mini-game right now”) and the parent’s job (“limit screen time”) conflict, the framework cannot adjudicate. That adjudication requires value-level reasoning JTBD does not contain.

Government Services: The Estonia case study

Estonia’s e-government services — widely studied as the most digitally mature government in Europe — explicitly adopted JTBD as a service-design methodology in the late 2010s. Citizens were asked, “When you last interacted with a government service, what progress were you trying to make?” The answers reframed dozens of fragmented services into a smaller number of citizen-facing journeys (start a business, register a child’s birth, file taxes, obtain a residence permit). Estonia’s resulting service-design portfolio reduced average citizen time spent on bureaucracy and increased measured citizen trust. JTBD’s situational reframe is powerful in government services because traditional government org-chart structure obscures the citizen’s job; JTBD makes it visible.

The Elephant in the Room

The elephant nobody in the JTBD community wants to discuss directly is that the framework, as practiced, has become a brand-management exercise more than a research methodology. There are now at least four distinct schools (Christensen, Ulwick, Klement, Kalbach), each selling certifications, each insisting their version is the canonical one, and each careful to avoid direct comparative evaluation. The result is that “we use JTBD” can mean a forty-thousand-dollar Strategyn engagement, a one-hour switch interview from an Intercom blog post, or a thirty-second slide in a board deck — and the speaker rarely specifies which.

The intellectually honest acknowledgment is that JTBD is useful in proportion to the rigor of the version applied. Strategyn’s outcome-driven instrument has been validated in multiple commercial contexts. Christensen’s switch-interview methodology produces vivid hypotheses that need independent validation. Klement’s “job stories” format is a useful copywriting heuristic that has not been validated as a research instrument. A behavioral designer who treats all four as equivalent is in for a hard lesson.

The other elephant: JTBD has been used, in the consulting world, as a way to repackage standard market research with a more compelling vocabulary — and to charge a higher rate for it. Yu-kai Chou’s recommendation, after watching this play out across hundreds of engagements, is to treat JTBD as a discovery scaffold worth keeping in the toolkit, but to insist on operational rigor — specifically, on Ulwick-style outcome quantification or on independent validation of switch-interview hypotheses — before any major product investment is allocated on its basis alone.

How to Apply Jobs-to-be-Done with the Octalysis Framework

This is where the framework becomes operational for behavioral designers. JTBD names what the user is consciously trying to do. The Octalysis Framework names the eight unconscious motivational drives that determine whether attempting that job will feel rewarding enough to sustain engagement. The two layers must be designed in tandem.

CD1 — Epic Meaning & Calling: the social job amplifier

Core Drive 1 (CD1): Epic Meaning & Calling is the drive activated when a user feels they are part of something bigger than themselves. JTBD’s social job dimension — how the purchase is read by others — is the conscious surface of CD1. Patagonia’s environmental positioning, Wikipedia’s “edit a real article” onboarding, and Duolingo’s “Bring back endangered languages” framing all hire CD1 to amplify a social job. Designers using JTBD should ask: of the social jobs my product serves, which can be reframed as participation in a collective worth joining?

CD2 — Development & Accomplishment: the functional job’s reward layer

Core Drive 2 (CD2): Development & Accomplishment rewards measurable forward progress. Most JTBD functional jobs — “help me get this thing done” — live here. The trick is that functional jobs feel rewarding only when progress is visible. A product that does the functional job perfectly but provides no feedback signal will lose to a product that does the functional job 90% as well but makes the progress legible. CD2 design (progress bars, completion counts, milestone notifications) is what turns functional-job completion into emotional-job satisfaction.

CD3 — Empowerment of Creativity & Feedback: where jobs become play

Core Drive 3 (CD3): Empowerment of Creativity & Feedback activates when the user is given a real choice and immediate feedback on the consequence. JTBD jobs that involve customization, expression, or solving novel problems live here. Minecraft’s job-to-be-done is not “play a game” — it is closer to “let me build something that’s mine and watch it work.” That is CD3 in plain language. Designers can amplify any creative job by tightening the feedback loop and widening the deliberate-choice set.

CD4 — Ownership & Possession: the switch-cost engine

Core Drive 4 (CD4): Ownership & Possession is the drive that makes users protect, accumulate, and refuse to abandon what they own. JTBD’s classic finding that “switching costs are higher than rational analysis predicts” lives here. When a user has invested in customization, history, or accumulated assets inside a product, the job of “switch to the alternative” is competing not just with the alternative product but with the felt cost of abandoning what they own. Designers can use JTBD to identify the switch moments and CD4 to design the ownership signals that make switching feel costly.

CD5 — Social Influence & Relatedness: the social job’s affective core

Core Drive 5 (CD5): Social Influence & Relatedness is the affective machinery underneath JTBD’s social job dimension. CD1 makes a social job feel meaningful; CD5 makes it feel connecting. Slack’s job is partly “communicate with the team” (functional) and largely “feel like part of the team” (CD5). Designers running a JTBD interview that surfaces a social job should ask which CD5 game techniques — mentorship, social treasures, group quests — deepen the relatedness layer.

CD6 — Scarcity & Impatience: the time-bound job sharpener

Core Drive 6 (CD6): Scarcity & Impatience activates when a desired thing becomes harder to obtain. JTBD jobs with a deadline (“I need this before my flight tomorrow”) or a limited-supply quality (“I want one of those before they’re gone”) live here. The functional job is the same as in any other context; CD6 raises the felt urgency of completing it. Designers should use JTBD to identify which jobs are intrinsically time-bound and CD6 to amplify the urgency signal without inventing artificial scarcity.

CD7 — Unpredictability & Curiosity: the discovery job engine

Core Drive 7 (CD7): Unpredictability & Curiosity activates when the next outcome is variable and the variability itself is engaging. JTBD’s exploration jobs — “show me something I didn’t know I wanted” — are CD7 jobs. Spotify’s Discover Weekly hires CD7 to do the job of “introduce me to music I’ll like without making me work for it.” TikTok’s For You feed is the same job at a higher activation rate. Designers should map exploration jobs to CD7 game techniques and avoid hiring CD7 to amplify jobs that are not exploratory — that path leads to slot-machine design.

CD8 — Loss & Avoidance: the protective job dimension

Core Drive 8 (CD8): Loss & Avoidance activates when the user is trying to prevent something undesirable. Many JTBD jobs are framed positively but are actually CD8 jobs in disguise: “help me file my taxes correctly” is really “help me avoid an IRS audit.” Insurance jobs, security jobs, backup jobs, and most enterprise compliance jobs are CD8 jobs. Designers should use JTBD to identify which jobs are protective in nature and CD8 to design the loss-avoidance signal carefully — CD8 is the most easily abused Core Drive.

Practical Steps to Apply Jobs-to-be-Done

The framework only works when applied with a specific protocol. Here is the protocol I teach at Stanford and have used in client engagements ranging from Fortune 500 consumer-goods companies to early-stage SaaS startups.

Step 1: Identify the switch moment. Find five to fifteen customers who recently switched to your product or switched away from a competitor. Switch moments are when goal-directed cognition was active and the customer’s verbal report has the highest likelihood of recovering real motivational signal. Habitual users will produce confabulated stories; switch users will produce signal.

Step 2: Run the switch interview. The Christensen-school protocol asks four time-anchored questions: (a) when did you first realize you needed something different, (b) what did you try before this, (c) what almost made you not switch, and (d) what made you finally pull the trigger. The answers form a four-act narrative that reveals the job.

Step 3: Construct the job statement. Use the standardized format: “When [situation], I want to [motivation], so I can [outcome].” Resist the temptation to load multiple jobs into a single statement. If you have three jobs, write three statements.

Step 4: Quantify the outcomes. This is where most teams stop and where the framework’s rigor lives. For each job statement, derive the desired outcomes and survey a representative sample of customers on (a) how important the outcome is and (b) how satisfied they are with their current solution’s performance on it. Outcomes with high importance and low satisfaction are the opportunities. This is Ulwick’s outcome-driven branch and it is non-negotiable for any major investment decision.

Step 5: Map the job to Octalysis. For each high-opportunity job, ask which Core Drives are doing the affective work. Use the mapping in the previous section. A functional job rarely lives in CD2 alone; the secondary drives matter. A social job almost always involves CD1 and CD5 in different proportions. Octalysis turns the job statement into a design specification.

Step 6: Design and validate. Build the smallest possible test of the job-plus-Octalysis hypothesis. Run it against the existing solution. If the hypothesis predicts behavior, scale it. If it does not, the job statement was wrong — revise and re-run. This step is what separates serious behavioral design from JTBD-flavored storytelling.

Closing Thoughts

Jobs-to-be-Done is one of the most useful reframes in modern product thinking, and it is also one of the most over-extended. The reframe — from person to context, from feature to progress, from preference to job — is durable and worth integrating into any serious behavioral designer’s toolkit. The four-school orthodoxy, the certification industry, and the “JTBD explains everything” maximalism are not.

The integrative position I teach is this: use JTBD as a discovery instrument to identify the conscious goals customers are trying to achieve, run the discovery with switch interviews on switch users and outcome-driven surveys for any decision worth more than fifty thousand dollars, and then design the solution using Octalysis to ensure the eight unconscious motivational drives are all firing in the right direction. Treat JTBD’s job statement as a hypothesis, not a finding. Treat Octalysis as the design specification, not a marketing layer. Both layers must be deliberate, or the product underperforms.

If your team is currently using JTBD to write user stories and ship features without a quantitative outcome instrument and without an explicit motivational-drive analysis, you are doing JTBD-flavored intuition, not JTBD. The reframe is valuable. The discipline is more valuable. Combine the two, and you have one of the most powerful behavioral-design pipelines available to a modern product team.

Where to go next

• Walk the canonical map of the design system you’ll pair JTBD with: the Octalysis Framework.
• Go deeper on the eight motivational drives: Actionable Gamification.
• Get both books in one place: Actionable Gamification and 10,000 Hours of Play.
• Run JTBD-shaped discovery with a serious team: Octalysis Prime or hire the practice — Octalysis Group.
• Sibling pillars in the discovery / segmentation cluster: Big Five (OCEAN), Hook Model, Four Layers of Desire, and the full Behavioral Framework Library.

Treat the job statement as a hypothesis, not a finding. Score the outcome, not the persona. Design the Core Drives that turn the conscious job into sustained behavior — and accept that the customer will fire your product the moment a competitor does the same job with less friction.

Frequently Asked Questions

What is the simplest definition of Jobs-to-be-Done?

Jobs-to-be-Done is the theory that customers don’t buy products — they hire products to make a specific kind of progress in a specific context. The unit of analysis is the job (the progress in a context), not the product or the customer demographic.

Who created Jobs-to-be-Done?

The phrase is most associated with Clayton Christensen, who popularized it through The Innovator’s Solution (2003) and Competing Against Luck (2016). Tony Ulwick built a parallel quantitative version called Outcome-Driven Innovation in the late 1990s. Both branches are practiced under the JTBD label.

Is the milkshake story actually true?

It is a real consulting case study Christensen ran with a fast-food chain in the early 2000s. The story has been simplified for pedagogical use and the original quantitative results are not in the public literature, but the underlying engagement and the milkshake reframe are documented.

What is the difference between a job, a need, and a goal?

A need is a standing motivational state (Maslow’s safety, esteem, etc.). A goal is a desired end state. A job is the situated combination of motivation, situation, and desired outcome — “When I’m tired in the morning, I want a snack that lasts thirty minutes through a straw, so I can stay alert through the commute.” The situational anchor is what makes a job actionable for design.

What’s the difference between Christensen’s JTBD and Ulwick’s ODI?

Christensen’s branch uses qualitative switch interviews to construct narrative job statements. Ulwick’s branch uses quantitative surveys to score importance and satisfaction on a hierarchy of desired outcomes. Christensen’s version is more accessible; Ulwick’s is more rigorous. Practitioners often conflate them, which is the single most common implementation error.

How do I write a good job statement?

Use the format “When [situation], I want to [motivation], so I can [outcome].” Anchor in a specific situation, not a general preference. Capture the motivational verb. Specify the desired outcome. Avoid loading multiple jobs into one statement. Test the statement against switch-interview transcripts to see if it actually fits.

Does Jobs-to-be-Done replace user personas?

It replaces them as the primary unit of analysis for design decisions. Personas remain useful for marketing communication and for sales-team training. The shift is that the design unit becomes the job, while the communication unit can remain the persona. Mature teams use both deliberately.

How does Jobs-to-be-Done relate to gamification?

JTBD identifies the conscious job the user is trying to do. Gamification, properly designed using the Octalysis Framework, makes that job feel rewarding by activating the eight Core Drives. JTBD without Octalysis produces correctly-scoped products that nobody is motivated to use. Octalysis without JTBD produces motivating systems pointed at the wrong goal.

What are the biggest critiques of Jobs-to-be-Done?

Three serious critiques: (1) the framework is not falsifiable in its Christensen form — any failed prediction can be re-narrated into a fitting job statement; (2) it relies on retrospective verbal report, which the cognitive-psychology literature has shown to be unreliable; (3) the functional/emotional/social trichotomy is a heuristic, not a measurement instrument with reproducible coding rules.

Should startups use Jobs-to-be-Done?

Early-stage startups should use the switch-interview branch to identify hypothesized jobs cheaply and quickly. Growth-stage startups making investments above ~$100K should add the outcome-driven survey branch to validate those hypotheses quantitatively before allocating engineering capacity. Ignoring the quantitative step is the most common failure mode in startup adoption.

References

1. Christensen, C. M., & Raynor, M. E. (2003). The Innovator’s Solution: Creating and Sustaining Successful Growth. Harvard Business Review Press.
2. Christensen, C. M., Hall, T., Dillon, K., & Duncan, D. S. (2016). Competing Against Luck: The Story of Innovation and Customer Choice. HarperBusiness.
3. Ulwick, A. W. (2005). What Customers Want: Using Outcome-Driven Innovation to Create Breakthrough Products and Services. McGraw-Hill.
4. Ulwick, A. W. (2016). Jobs to be Done: Theory to Practice. Idea Bite Press.
5. Christensen, C. M., Hall, T., Dillon, K., & Duncan, D. S. (2016). “Know Your Customers’ ‘Jobs to Be Done.’” Harvard Business Review, September.
6. Klement, A. (2018). When Coffee and Kale Compete: Become Great at Making Products People Will Buy. NYC Publishing.
7. Kalbach, J. (2020). The Jobs To Be Done Playbook: Align Your Markets, Organization, and Strategy Around Customer Needs. Two Waves Books.
8. Nisbett, R. E., & Wilson, T. D. (1977). “Telling more than we can know: Verbal reports on mental processes.” Psychological Review, 84(3), 231-259.
9. Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux.
10. Balleine, B. W., & Dickinson, A. (1998). “Goal-directed instrumental action: Contingency and incentive learning and their cortical substrates.” Neuropharmacology, 37(4-5), 407-419.
11. Damasio, A. R. (1994). Descartes’ Error: Emotion, Reason, and the Human Brain. Putnam.
12. Deci, E. L., & Ryan, R. M. (1985). Intrinsic Motivation and Self-Determination in Human Behavior. Plenum.
13. Eyal, N. (2014). Hooked: How to Build Habit-Forming Products. Portfolio.
14. Intercom (2014). Intercom on Jobs-to-be-Done. Intercom Inc.
15. Bettencourt, L. A., & Ulwick, A. W. (2008). “The Customer-Centered Innovation Map.” Harvard Business Review, May.

Related Reading

• The Octalysis Framework: Complete Gamification Guide
• Prospect Theory: An S-Tier Designer’s Guide to Loss Aversion
• Dual Process Theory: System 1 and System 2 in Behavioral Design
• The Behavioral Framework Library
• Actionable Gamification — the book

WOULD YOU LIKE YU-KAI CHOU TO WORK WITH YOUR ORGANIZATION?