JTBD vs Design Thinking: Which Product Framework to Use
JTBD uncovers why customers hire or switch products; Design Thinking prototypes and tests solutions. When to use each, how to pair them, and Duolingo's 2026 case.
Teams reach for Jobs to be Done (JTBD) and Design Thinking for the same reason: neither starts with a feature checklist, and both put the customer at the center of product work. That overlap hides a sharp division of labor. JTBD explains why a buyer switches and what progress they are hiring a product to deliver; Design Thinking runs an iterative process—Empathize, Define, Ideate, Prototype, Test—to turn a human problem into a tested interface or service.
Pick the wrong one and the failure mode is predictable. Running Design Thinking before you understand the job yields polished prototypes for a demand that does not exist. Stopping at JTBD gives you sharp research slides with no tested product experience.
At a glance
| Dimension | Jobs to be Done (JTBD) | Design Thinking |
|---|---|---|
| Primary question | "What progress is the customer hiring this product to make?" | "How might we prototype and test a better solution for this user?" |
| Core unit | The job (functional, emotional, and social) + switching triggers | The human workflow across 5 phases (Empathize, Define, Ideate, Prototype, Test) |
| Competitive lens | Category-agnostic (includes spreadsheets, habits, non-consumption) | Typically scoped around the user's experience and journey |
| Primary output | Job statements, desired outcomes, push/pull switching map | Point-of-View statements, sketches, clickable prototypes, test notes |
| Origin | Tony Ulwick (ODI, 1990s) & Clayton Christensen (HBR, 2016) | IDEO & Stanford d.school |
| Main risk | Stopping at interview synthesis without testing real artifacts | Falling in love with clever prototypes that miss why users actually switch |
| Best for | Positioning, market definition, diagnosing stalled retention | New UX flows, service redesign, rapid concept validation |
What JTBD is best for
Jobs to be Done separates the enduring customer motive from the current product form. As Clayton Christensen and co-authors wrote in Harvard Business Review, circumstances predict behavior better than demographics—two buyers with identical profiles hire completely different products when their situation changes. Meanwhile, Tony Ulwick's Outcome-Driven Innovation breaks the job into measurable outcomes so teams know which steps are underserved before writing specs.
JTBD wins when the ambiguity sits in market demand rather than interface mechanics:
- Uncovering non-obvious competitors. A meal-kit startup does not just compete with rival meal kits; it competes with takeout, frozen dinners, and skipping dinner prep.
- Fixing retention when usage diverges from stated intent. In our JTBD examples breakdown, Duolingo's Q1 2026 earnings call reported daily active users up 21% year over year to 56.5 million and revenue up 27% to $292 million. Retention follows the daily streak—the emotional job of feeling productive in idle minutes without breaking a chain—far more closely than full language fluency.
- Killing features before engineering starts. When an idea does not serve the functional, emotional, or social job buyers actually hire for, it gets cut before a single prototype is drawn.
What JTBD does not give you is a structured creative loop for exploring ten UI directions, building a cheap mockup in an afternoon, or testing whether users understand a screen.
What Design Thinking is best for
Design Thinking is built for solution discovery. Formalized by IDEO and taught as a five-phase loop by the Stanford d.school—Empathize, Define, Ideate, Prototype, Test—it forces teams to observe real users, reframe assumptions into "How Might We" questions, diverge across multiple concepts, and put cheap, falsifiable prototypes in front of people early.
Design Thinking earns its keep when:
- The job is known, but the workflow is broken. Users already want to file an expense report, book a clinic visit, or complete checkout—they drop off because of confusion, trust gaps, or cognitive load.
- Multiple teams disagree on how to solve an agreed problem. Rapid prototyping replaces weeks of abstract slide debates with three clickable flows tested with five users in two days.
- You need to discover unarticulated UX friction. Shadowing users in the Empathize phase catches workarounds—sticky notes on monitors, manual copy-pasting—that people never mention in a questionnaire.
What Design Thinking does not guarantee is that you picked the right market bet in the first place. Empathy inside a flawed product premise can lead a team to optimize an onboarding flow for users who never had a strong reason to switch.
The Job-Before-Prototype Gate
Instead of treating JTBD and Design Thinking as competing religions, sequence them through The Job-Before-Prototype Gate—two checks a product team should pass before spending a week on prototypes:
| Gate question | Pass signal | If it fails, run… | What it prevents |
|---|---|---|---|
| 1. Switching Evidence — Can you name the specific situation that triggers buyers to hire you, and what alternative they fire? | 5–8 recent-buyer interviews converge on one repeatable push/pull trigger and job statement | JTBD switching interviews first | Prototyping features for an invented persona or wrong competitive set |
| 2. Delivery Bottleneck — Is growth capped because the promise is unclear, or because the current flow breaks trust/usability? | Users try to complete the job but stall on friction, unclear steps, or UI anxiety | Design Thinking (Ideate → Prototype → Test) | Rewriting positioning endlessly when the real problem is a broken workflow |
Look at how that gate separates two real product levers at Duolingo. Third-party retention analysis estimates Duolingo's Streak Freeze cut churn by roughly 21% for at-risk users, and learners hitting a 7-day streak are roughly 3.6× more likely to stay engaged (Trophy, 2026). JTBD explains why those mechanics work: users hire five-minute sessions to protect daily momentum and avoid guilt, not just to pass CEFR grammar exams. Once that job is clear, Design Thinking takes over—prototyping streak-repair prompts, leaderboard nudges, and widget layouts to test which interface keeps the habit intact without exhausting learners.
The decision rule
Use JTBD when you need to define what job customers are hiring you for and why they switch. Use Design Thinking when the job is clear and you need to ideate, prototype, and test how the product delivers it.
To test the rule on a concrete choice—imagine Duolingo's product team deciding next quarter whether to overhaul intermediate grammar lessons or test a new social accountability mechanic:
- Run JTBD first if telemetry shows learners churning after week two despite high completion scores: you need to know whether users are firing the app because another tool fits their commute better, or because the emotional reward stopped working.
- Run Design Thinking once the job is confirmed ("help me keep a 3-minute daily habit with low friction"): sketch eight ways to make re-entry effortless after a missed day, prototype two in Figma, and test them with lapsed learners before writing production code.
When NOT to use either
- When you already have a backlog of validated bets and need a shipping order. Neither JTBD nor Design Thinking ranks effort against impact—use RICE or WSJF once options are defined.
- When the bottleneck is industry structure or distribution economics, not product fit. If supplier power or switching costs block adoption regardless of UX, run Porter's Five Forces first.
- When you are choosing between delivery-level artifacts. If the team already knows the job and is debating user profiles vs sprint tickets, see JTBD vs Personas and JTBD vs User Stories.
In 30 seconds
JTBD defines the demand—what progress customers hire your product to make and what alternatives they fire. Design Thinking builds and tests the delivery—empathizing, prototyping, and iterating until the experience works cleanly in human hands. Validate the job first with JTBD, then prototype the solution with Design Thinking.
Want both templates ready on your phone? Framework for iPhone & iPad includes interactive JTBD and Design Thinking workflows with AI-assisted prompts.
Related
- Jobs-to-be-Done framework explained — the interview method and the JTBD → Personas → User Stories sequencing rule
- Jobs-to-be-Done examples: 8 real companies — Duolingo, Slack, Uber, Spotify, and Notion mapped to functional, emotional, and social jobs
- JTBD framework catalog — copyable template, provenance chain, and free online tool
- Design Thinking catalog — the 5 phases, worked example, and free online template
- JTBD vs Personas — when to use job statements vs user profiles
- JTBD vs User Stories — turning a validated job into sprint increments
Sources
- Clayton M. Christensen, Taddy Hall, Karen Dillon, and David S. Duncan — "Know Your Customers' Jobs to Be Done", Harvard Business Review (Sep 2016)
- Tony Ulwick / Strategyn — "Tony Ulwick and Outcome-Driven Innovation"
- IDEO — "Design Thinking"
- Stanford d.school — "Get Started with Design Thinking"
- The Motley Fool — "Duolingo (DUOL) Q1 2026 Earnings Call Transcript" (4 May 2026)
- Trophy — "Duolingo Gamification Case Study" (2026)
Frequently asked questions
What is the main difference between Jobs to be Done and Design Thinking?
Jobs to be Done (JTBD) defines demand—the underlying progress a customer wants in a specific situation and what alternatives they fire to hire your product. Design Thinking is an iterative problem-solving process (Empathize, Define, Ideate, Prototype, Test) focused on generating, prototyping, and testing human-centered solutions.
Should you run JTBD or Design Thinking first?
Run JTBD first when you are entering a new market, repositioning a stalled product, or unsure why customers buy or churn. Once switching interviews confirm the real job, run Design Thinking's Ideate, Prototype, and Test phases to build and validate the experience that fulfills that job.
Can Design Thinking replace Jobs to be Done?
Not reliably. Design Thinking's Empathize phase observes user pain inside an existing workflow, which is excellent for fixing UX friction but can miss the broader competitive set—such as non-consumption or a completely different category hired for the same job. JTBD supplies that strategic boundary before prototyping begins.
How do JTBD and Design Thinking work together in practice?
Use JTBD to replace or anchor the Define stage with a falsifiable job statement ('When [situation], I want to [motivation], so I can [outcome]'). Then run Design Thinking's Ideate, Prototype, and Test stages to explore multiple solution concepts and measure which one actually resolves the job with the lowest friction.