Jobs To Be Done (JTBD)
Jobs To Be Done is a framework that views customers as 'hiring' products to make progress on a specific goal in a given context, focusing on the underlying job rather than demographics or features.
Key takeaways
- A purchase follows a struggling moment shaped by push, pull, habit and switching anxiety.
- Interviews rebuild the switch in sequence, from first noticing to the forcing event.
- Reducing switching anxiety usually moves more stalled deals than adding another feature.
- One job statement can order a landing page, a discovery call and an onboarding flow.
- JTBD explains why buying happens; it does not size a market or name a segment.
In depth
The mechanism sits in a comparison the buyer runs without your involvement. Before any purchase there is a struggling moment, a push away from the current approach and a pull toward a new one, both held back by habit and by anxiety about switching. A jobs-to-be-done interview reconstructs that sequence in order: when the buyer first noticed the problem, what they tried on their own, what event finally forced action, and what worried them at the moment of committing to a change.
What raises the odds of being hired is reducing the two forces holding buyers in place, not adding capability. Anxiety about switching falls with migration help, a trial that runs on the buyer's own data, and evidence from someone in an identical situation. Habit weakens when the current approach visibly fails at a moment the buyer already recognises. Adding features addresses the pull, which in most stalled deals was already the strongest of the four forces.
In practice a job statement can order the first line of a landing page, the questions in a discovery call and the sequence of onboarding, because all three should follow the progress the buyer is making rather than the structure of your feature set. A quiz funnel fits this naturally: asking where the visitor currently stands in the job, what they have already tried and what forced them to look now yields a qualification signal and the exact language for the result page.
JTBD explains why a purchase happens, not who will purchase or how many, so it does not replace segmentation or market sizing. It is weak where switching is rare and gathering enough recent switchers would take years, and where the buyer and the user are different people pursuing different jobs. Applied loosely, almost any sentence can be dressed up as a job statement, at which point the framework stops constraining decisions and becomes a vocabulary.
Example in practice
How to measure it
The signal that a job statement is right is that it changes behaviour when used verbatim. Run it as a headline or as the opening discovery question and watch whether people recognise themselves faster, measured through scroll depth, reply rate, or how early in a call the buyer starts describing their own situation unprompted. A statement that produces no change in recognition is not the job.
On the qualification side, group leads by the job they arrived with rather than by source or industry, then compare win rate and cycle length across those groups. Where a quiz asks what forced the visitor to look now, that single answer usually separates buyers who close from researchers who do not, and it does so more cleanly than firmographic scoring.
Common mistakes
The most common failure is writing job statements in the company's own language, so the job becomes the product restated. 'Help me centralise my reporting' is a feature described from the buyer's chair rather than a job. A real statement survives the disappearance of your entire product category. Test it by asking whether a competitor with a completely different approach could also be hired for it; if not, you have written a solution.
The second is interviewing satisfied customers about preferences instead of recent switchers about a timeline. Preference questions produce tidy, agreeable answers that predict nothing. Ask about one specific purchase: the week it happened, what the buyer did first, who else got pulled in, what almost stopped them. Recency matters here, because memory of the struggling moment fades within a few months.
Frequently asked questions
How do you uncover a customer's job to be done?
The most reliable method is interviewing recent buyers about the moment they decided to switch and what they hoped would change. These switch stories reveal the functional and emotional job far better than feature surveys.
Is JTBD a replacement for buyer personas?
No, they answer different questions. A persona describes who is in the room and what each of them is accountable for, while a job describes what progress is being attempted and why now. Teams that keep both use the job to shape the offer and the message, and the persona map to decide who hears which part of it.
How do I run a jobs-to-be-done interview?
Pick buyers who switched recently and rebuild the timeline backwards from the purchase: what happened the day they decided, what they tried before that, when the problem first became noticeable, and what almost stopped them. Ask about events rather than opinions. Avoid asking what they want, since that turns the conversation into a feature request within minutes.
What does a good job statement look like?
It names a situation, the progress wanted and a constraint, and it holds up even if your product category disappeared. Something like 'when a new rep joins mid-quarter, help me get them selling without pulling my best people off their own deals' works because it describes the buyer's world and can be quoted in their own voice.
How many jobs should one product serve?
Usually one primary job with adjacent ones served secondarily. Products claiming several equal jobs tend to be describing features instead, and the messaging drifts back into a feature list because no single progress anchors it. When two jobs genuinely have different buyers and different urgency, that is often a signal to build separate offers.
Does JTBD apply to low-consideration purchases?
Yes, though the interview is harder because the struggling moment is short and poorly remembered. The push and pull still exist, compressed into minutes rather than months. For those products, behavioural evidence and short in-context surveys usually beat retrospective interviews, since buyers cannot reconstruct a decision they barely made consciously in the first place.
How do I use JTBD in a quiz funnel?
Build the questions around the visitor's current situation rather than your feature list: what they do today, what has already failed, and what changed recently. Those answers score intent more reliably than firmographics, and they hand the result page the exact words needed to describe the visitor's position before recommending any path forward.