← Back to concept explainers

Concept Fable

Tasks to be completed

Conclusion first:Jobs to Be Done (JTBD) focuses on the progress users want to make in a specific situation, rather than understanding needs only by identity, industry, or feature preference. When implementing enterprise Agents, JTBD can break "the customer wants AI" into real tasks: reducing customer service backlogs, accelerating pre-sales material production, lowering compliance review costs, shortening ticket handling time. It requires teams to probe the triggering scenario, existing alternatives, cost of failure, acceptance criteria, and budget source, thereby avoiding building the Agent into a showy feature.

Jobs to Be Done, JTBD · Product strategy / research

Read the story first. There is no rush to give a definition here; let the question grow on its own first.

Fable

The miller no longer asks what wheel people want to buy, but why the grain jams at night

The riverside mill had three kinds of waterwheels: the wooden wheel was cheap, the iron wheel was durable, and the new wheel spun fastest. The miller asked each customer which one they liked, and the answers changed every day.

The farmer said he wanted it cheap, the caravan said it wanted it fast, the winery said it wanted it stable. The miller adjusted the wheel according to these words, the warehouse filled with parts, and at night people were still queuing outside the door.

He thought he had not asked in enough detail, so he gave the customer a long form: how many spokes they liked, what color, what speed, and whether a name could be engraved. The longer the form, the livelier the answers, but the slower the delivery became.

After a rainstorm, the distillery owner was anxious to finish milling the wet wheat, or the whole batch would sour the next day. He did not care about the color of the wheel; he only asked whether it could finish milling tonight, who would be responsible if it broke, and how the payment terms would be calculated.

Only then did the miller realize that customers weren't buying a preference for waterwheels; they were seeking progress in a certain situation: turning grain that would spoil into flour that could be sold, and turning loss into a controllable cost.

He first tried making all the wheels faster. But with no one on duty at night, wet grain still queued up; with no compensation for spoiled material, the distillery still didn't dare hand over its entire batch of wheat. Speed was only one small piece of it.

He stopped most of the gimmicks and switched to a dedicated night-milling line, priority for wet grain, compensation for spoiled material and a monthly settlement ledger. Not everyone liked these arrangements, but those anxious to keep their goods immediately willing to pay a higher price.

Later, when he received new customers, he no longer first asked which kind of wheel they wanted. Instead he asked: when did you get stuck? What will you lose if it isn't solved? What workaround are you using now? Who decides that this is done? Good wheels finally served a clear purpose. The miller also discovered that the same person wanted different things during the day and at night. In the daytime they bought cheapness; at night they bought certainty. When grain was dry they bought speed; when it was about to sour they bought assurance. If you only listened to feature preferences, you would miss the moment when they were most willing to pay. Later the mill's ledger changed too: it was no longer divided by wheel model, but by night milling, freshness, market day, and compensation. Where the money comes from is where the task is. Later, during the review, everyone wrote this lesson into their daily rules: first look at where the site truly changed the outcome, then decide what to keep, what to repair, and what to restrict. If you only look at surface excitement, you will fall in the same place next time.

Reveal

This story is about: jobs to be done

Jobs to Be Done (JTBD) focuses on the progress users want to make in a specific situation, rather than understanding needs only by identity, industry, or feature preference. When implementing enterprise Agents, JTBD can break "the customer wants AI" into real tasks: reducing customer service backlogs, accelerating pre-sales material production, lowering compliance review costs, shortening ticket handling time. It requires teams to probe the triggering scenario, existing alternatives, cost of failure, acceptance criteria, and budget source, thereby avoiding building the Agent into a showy feature.

What matters is that it is not just a term, but a structural problem that repeatedly appears in real AI / Agent systems. Understanding it lets you judge when to add a model, when to change a process, and when to add governance.

Metaphor mapping

  • Three waterwheels: product features, model capabilities, and technology options
  • Long-table preference: requirements research that collects only feature wishes
  • Sour wet wheat: the customer's business pressure in a specific situation
  • Night milling dedicated line and compensation: restructure delivery around task outcomes
  • Can it be ground by tonight: the progress goal in JTBD
  • What is being used to make do now: existing alternatives and budget sources
  • Who judges completion: acceptance reviewers and success criteria in enterprise procurement

Andy Ren Xu's judgment

The practical value of this concept is that it helps you break "AI features that look smart" into deliverable, verifiable, and governable units of work.

User researchpersonavalue propositionworkflowpain pointoutcome

Related distinctions

This concept is easily confused with adjacent concepts; it is recommended to continue to the comparison page.

京ICP备2026014272号