The lens you lead with shifts as demand matures.

Because they were built to solve different problems, and the split is real rather than cosmetic. The Progress lineage — Bob Moesta's demand-side work out of Clayton Christensen's Jobs Theory — grew up answering why people switch, and its primitives are struggling moments, forces, and timelines. The Process lineage — Tony Ulwick's outcome-driven innovation, with Jim Kalbach's job mapping — grew up answering which needs are underserved, and its primitives are job steps, desired-outcome statements, and importance-versus-satisfaction scores. Practitioners have argued for years about which one is the real JTBD. The more useful question is which business problem you are facing: early demand needs discovery, mature demand needs prioritization, and those require different kinds of evidence.

They are two lenses on the same theory, built to answer different business questions. The Progress lens is qualitative and asks why people switch — the struggling moment, the forces at play, the progress someone is trying to make. The Process lens is quantitative and asks how much — how many customers hold a job, which outcomes are most underserved, and which segment is largest. Progress leads Latent and Emerging demand; Process leads Patterned and Scalable demand. Neither ever drops to zero — only the balance shifts.

No, and the question itself is the mistake. Asking whether qualitative beats quantitative is like asking whether interviews beat analytics — the answer depends entirely on what you do not yet know. When the job is unnamed and you cannot say whether the struggle is even real, stories are the only evidence that can answer the question. When the job repeats across buyers and you have to decide what to fund, counting is the only evidence that can. The lens should follow the uncertainty, not the preference.

Diagnose from symptoms rather than from how mature the company feels. In Latent demand, nobody on the team states the problem the same way twice and you cannot name a single person who actively switched to solve it. In Emerging demand, you can name the customer but not what they are hiring you for. In Patterned demand, you have more validated jobs than you can fund. In Scalable demand, every research request arrives as a dashboard question while the growth curve flattens.

Yes, and that is the normal case rather than the exception. Demand does not mature evenly across products, segments, geographies, or channels: the same product can sit in Patterned demand with enterprise buyers, Emerging demand with SMB, and Latent demand in a vertical nobody has sold into yet. Diagnose per segment rather than per company, and lead with whichever segment carries the biggest business constraint right now. A single company-wide research plan applied across segments at different stages is how teams end up surveying one audience too early and interviewing another too late.

Switch during the transition from Emerging to Patterned demand, when the same struggle repeats across buyers who share nothing else in common. That repetition is what makes a job stable enough to be measured: a repeating struggle becomes a fixed outcome set, and a fixed outcome set is something people can rate. Practical signals that you have arrived are that job definitions stop moving between studies, new interviews confirm rather than reshape the definition, and the customer's language has stopped changing. As a rule of thumb: if interviews are still changing your job definition, you are still too early to survey.

You can, but the results will not mean what you think they mean. An outcome survey asks people to rate importance and satisfaction against a list of outcomes, and that list has to come from somewhere — usually from qualitative work that discovered which outcomes matter. Survey first and you are asking customers to rate outcomes you invented, which produces a confident-looking chart built on noise. A team with twenty interviews and no repeating pattern is not ready to field an ODI-style survey. This is the most common and most expensive JTBD mistake.

No. Neither lens ever drops to zero — only the balance shifts. Even in Scalable demand, where measurement leads, a thin qualitative thread is what catches the next struggle forming at the edges of your data, because a survey can only ask about outcomes you already know exist. Teams that let qualitative atrophy entirely keep optimizing a job the market is quietly leaving.

Yes, and mature categories are where it happens most. A new technology, a competitor that educates the market, a regulatory shift, or a newly reachable segment can all produce a struggle your existing outcome set has no language for — at which point that part of your demand is Latent again, however well you measured the old job. This is why the Scalable stage on this page carries loop-back signals rather than exit criteria: opportunity scores flattening, growth arriving only through optimization, and customers leaving for something your segmentation cannot see.

Early stages call for switch and non-switch interviews, forces mapping, and struggle prevalence scanning. Later stages call for opportunity scoring, outcome segmentation, thematic analysis, and paired analytics. A small set of methods bridges the two: job mapping, desired-outcome statements, and story clustering are qualitative craft whose product is a quantitative instrument, which is why they sit at the handoff between the lenses.