Jobs To Be Done (JTBD)

The central metaphor of Jobs To Be Done, popularized by Clayton Christensen and developed in parallel by practitioners like Tony Ulwick and Bob Moesta, is that people do not buy products — they hire them to make progress in a particular circumstance. When a job arises, a customer pulls a solution into their life to get it done, and they fire it when something does better. The unit of analysis is the job, not the customer's age, role, or industry.

Christensen's canonical example is the morning milkshake: a fast-food chain found people were buying milkshakes alone, early, to-go. The job was not "I want a tasty drink" but "make my boring commute more interesting and keep me full until lunch." The shake was competing with bananas, bagels, and boredom — not with other milkshakes. Defining the job that way changes what you build and who you actually compete against.

A well-formed job has three layers. The functional dimension is the practical task to accomplish — transfer money, schedule a meeting, ship a release. The emotional dimension is how the person wants to feel while doing it — confident, in control, unburdened. The social dimension is how they want to be perceived by others — seen as competent, responsible, or current. A product that nails the functional job but ignores the emotional and social layers often loses to one that feels better to use, even when it does less.

A common practice is to write the job as a stable statement of progress rather than a feature: "When I onboard a new engineer, I want to get them productive in the codebase quickly, so I can keep the team's velocity up." The situation-motivation-outcome shape keeps the job durable — solutions and technologies change, but the underlying job a person is trying to get done tends to persist for years.

Personas describe who a customer is — a fictional "Marketing Mary, 34, mid-market SaaS." JTBD describes what a customer is trying to accomplish and why. The two are not mutually exclusive, but JTBD argues that demographic attributes are weak predictors of behavior: two people with identical profiles can hire wildly different products because they are in different situations. The job, not the persona, predicts the purchase.

The failure mode JTBD guards against is building for an idealized user instead of a real moment of need. In discovery, this means interviewing for the timeline of a real decision — what triggered the search, what the person tried first, what made them switch — rather than asking what features they want. The output is a map of jobs and unmet outcomes, which feeds directly into opportunity mapping and prioritization. It pairs naturally with continuous discovery and the opportunity solution tree, where jobs become the opportunities a team scores and bets on.

A JTBD framework is only as good as the evidence behind it. Jobs articulated in a workshop and then filed in a slide deck quietly drift from reality; the strongest job statements are continuously re-grounded in how customers actually behave, what they pay, and what they ask for. The hard part is keeping the job, the account, and the work it informs in the same line of sight.

A connected product operating system helps here by joining feedback, customer behaviour, and revenue on one shared spine. An insight captured against a job can be read alongside the same customer's usage and subscription, and a roadmap item framed as a job carries the accounts and revenue that hired it — so a job statement stays anchored to evidence rather than becoming a static artifact divorced from the customers it claims to describe.