Story 01 · Washed lot
A washed lot that needed longer development
The first sample roast cupped sharp and grassy at day five — classic under-development despite an attractive colour reading on the tray. First crack had arrived early and we dropped on habit rather than curve. The second sample extended development by forty seconds with a lower charge temperature, holding rate-of-rise steady through mid-phase. At day eight the cup showed citrus and panela without the vegetal edge. We scaled to a six-kilo batch, marked the bags with the later roast date, and held filter service until day ten. Guests who bought early received a note on the bag suggesting waiting — honesty costs less than refunds.
Story 02 · Natural lot
A natural that smelled loud on the tray
Naturals deceive on the cooling tray. This lot smelled like strawberry jam at drop — staff were enthusiastic before the first cupping bowl cooled. At day three the same coffee tasted thin and perfumed, all aroma and no structure. We delayed retail release, cupped daily, and found structure arriving around day twelve in Singapore humidity. At the filter bar we opened grind one notch coarser and lowered water temperature two degrees to tame the ferment note without killing sweetness. The lot never became subtle — but it became drinkable, which was the brief.
Story 03 · Office subscription
A River Valley office subscription after humidity week
An office three streets from Kim Yam had been on a fortnightly subscription of our batch-brew lot. During a humid week their kitchen storage softened the coffee faster than our test kitchen predicted. Complaints were polite but clear: "tastes fine here, flat upstairs." We visited, moved their storage off a window ledge, adjusted the subscription to smaller bags delivered weekly, and swapped one origin for a more forgiving washed lot with longer development. No contract penalty — just a rebuild. Subscriptions are relationships, not software licences.
Story 04 · Decaf blend
A decaf that finally cleared the mute cup
Decaf often smells better on the tray than in the cup. Our sugar-cane decaf sample tasted woody at standard development. We slowed the roast after first crack, accepted a lighter colour, and extended rest to sixteen days before service. Espresso still lacked body — we never promised that — but filter showed chocolate and orange peel at a ratio guests could repeat at home. We label decaf clearly on the calendar so no one orders it expecting the same jolt as a caffeinated lot — we sell coffee flavour, not stimulant theatre.