The problem
A two-person coffee-roasting consultancy had already lived through a graveyard of failed bets. A wholesale model selling to cafes. A direct-to-consumer specialty store. An office-supply channel. A small production line. Each one had been launched on a hunch and quietly shut down when the market shifted or the demand never showed up.
Relaunching as a roasting-consultancy — hands-on training and equipment setup on a client's own production floor — the founders wanted to add a scalable digital product. But they refused to guess again. The question was: what does a beginning roaster actually need, and would they pay for it?
The segment: the "blind kitten" roaster
Jobs-based interviews surfaced one sharp, fundable segment: newcomer roasters with zero to eighteen months of experience, running small operations, working in isolation with no mentor nearby and no budget for an expert to fly in.
Their core jobs weren't about coffee chemistry in the abstract. They were:
- "When I hit a technical problem and there's no mentor around, I want a fast, specific expert answer so I don't waste beans on mistakes."
- "Working alone, I want access to expertise and calibration so I don't feel like I'm operating blind."
- "I want to ask a 'stupid' question safely, without being judged."
The big job underneath all of it: become a competent roaster quickly enough to feel confident, proud, and worth hiring.
What changed
The interviews — fifteen of them, starting with two domain experts to sanity-check the idea, then roasters and the managers who actually hold the budget — drove every downstream decision:
- The product scope: an assistant with two distinct modes, a quick reference lookup and a step-by-step instructor, because the interviews showed those were two different jobs.
- The buyer/user split: roasters use the tool, but managers approve the spend, so each got its own messaging.
- The landing page: written in the exact language customers used to describe their pain, not in the founders' insider jargon.
The results
This case is a go-to-market story, not a revenue story — the product is still pre-launch. But the methodology replaced guesswork with a validated plan: a clearly defined segment, a scoped two-mode product, a customer-language landing page, and a low-cost field test plan to confirm demand before building. After three ventures killed by untested assumptions, that shift — from guessing to knowing — is the whole point.