Field research beats focus groups for technical products.

Ask ten lab technicians in a conference room what they want from an instrument and you'll get sensible, articulate answers that are mostly wrong. Not because people lie, but because expertise is embodied: the workarounds, the muscle memory, the things they stopped noticing years ago don't surface in conversation. They surface at the bench, at 3pm, when the assay is late and someone is talking to them while they load samples.

The focus-group trap

Stated preference is a weak signal for products that live inside workflows. In a room, everyone wants more features and a bigger screen. In the field, you discover the screen is unreadable through a fume-hood sash, the gloves defeat the touchscreen, and the "rarely used" function runs forty times a shift. The gap between what people say and what their hands do is where technical products fail.

What the field shows you

Watching real operators in real environments yields a different class of finding: the gowning procedure that dictates how a device must be carried, the ambient noise that makes confirmation tones useless, the interruption pattern that demands resumable workflows, the bench depth that decides your enclosure footprint. None of these appear in a survey. All of them are design requirements — and each one we log with evidence attached: where we saw it, how often, what it costs.

From observation to requirement

Raw ethnography is not a deliverable. The craft is conversion: turning observations into requirements your engineers can argue with — specific, testable, ranked by consequence. "Operable with two fingers of a gloved hand while holding a rack" is a requirement. "Easy to use" is a wish. Our research practice exists to make that conversion, and to feed it straight into design decisions the same week.

Footage ends debates

One more reason to go to the field: internal arguments. Every product organisation has a loud opinion about what users do. Thirty seconds of footage of an actual operator fighting the current product ends debates that decks never could. For regulated categories like medical devices, the same observations feed the usability-engineering file — evidence doing double duty.