Skip to content
We've been running Driftform's subscription program for seven years now, and I've changed my mind about several things in that time. Here's the short version of what I've come to believe matters and what doesn't.
What matters
Coffee shipped within 48 hours of roast. This is the single biggest input. A coffee at 14 days post-roast is a different drink than one at 2 days post-roast. The difference is more obvious in lighter roasts than dark, but it's there in both. We ship every Friday because we roast on Tuesday and Wednesday. If we shipped Monday, the math wouldn't work.
Cadence matched to consumption. A 340g bag is about 28 cups at 12g/cup. Two cups/day, that's two weeks. Three cups/day, that's nine days. The number-one reason subscribers cancel is "I have too much coffee piling up" — which is the cadence being wrong, not the coffee being wrong. We push hard on "choose biweekly if you average under 2 cups/day" because retention is dramatically better when the math fits.
Easy skip/pause. If skipping a shipment requires emailing customer service or calling, your churn rate is going to be brutal. We let subscribers skip up to four weeks ahead in their account dashboard. No retention specialist, no "are you sure" modal. Reciprocity matters: people skip occasionally, they don't cancel.
Consistent quality, not chasing perfection. A coffee that's an 87 every week is more valuable to a subscriber than a coffee that's 90 one week and 84 the next. We've turned down a few lots over the years that were objectively the most exciting cuppings we'd done because they wouldn't be repeatable — and a subscription on inconsistent coffee is worse than a subscription on consistent slightly-less-exciting coffee.
What doesn't matter as much as people think
Origin variety. A subscriber doesn't actually want 18 different origins. They want 4–6 origins on a rotation that they can learn and have favorites in. We over-rotated for the first three years and got steady feedback that subscribers wanted to know what was coming and have it sometimes be a favorite they recognized.
Fancy packaging. Resealable foil bag with a roast date is what people want. We tried compostable packaging for two years and the freshness loss was measurable in cup quality. We went back to foil with a clear conscience after running the math on shipping CO2 vs. landfill impact. Coffee bags are about 3% of the carbon footprint of a coffee shipment; the freight is the rest.
The grinder included with subscription. Several large coffee subscription brands ship a free grinder with annual plans. The grinder is bad. Bad grinders make worse coffee than pre-ground from a good roaster. We don't ship grinders. We do recommend that if you don't have a grinder, get pre-ground (medium grind for drip, coarse for French press) and don't worry about it.
The honest math
We've watched our retention curve for seven years. Subscribers who pick a cadence that matches their actual consumption, with grind matched to their brew method, stay for an average of 14 months. Subscribers where one or both of those is off stay 3.5 months. That's the entire ballgame.
— Caleb