A Coffee Cupping Protocol for a Small Roastery Without a Lab
By Caffé Premium
A coffee cupping protocol written for a certified lab assumes a calibrated sample roaster, a controlled room, a trained panel, and a scoring form nobody in a two-person roastery has time to complete properly. That is a fine standard and a poor fit for the actual decision in front of most buyers, which is simply whether to order a trial lot of this coffee or not.
What follows is a protocol built for that decision. It borrows the useful structure of formal cupping and drops the parts that only matter if you are certifying a score for somebody else.
What a working coffee cupping protocol needs
Cupping bowls of roughly equal size, ideally five to eight ounces, which can be glass tumblers as long as they match each other. Cupping spoons or any deep round soup spoons. A gram scale that reads to at least one gram. A kettle. A grinder you can set consistently. A timer. Hot water at just off boiling.
The equipment does not need to be specialized. It needs to be identical across every bowl on the table, because everything you are about to do is comparison.
Roast the sample for evaluation, not for sale
A cupping roast is deliberately different from a production roast. The aim is to expose the coffee's raw character, not to flatter it. Target a light to medium level, developed enough to avoid grassy underdevelopment but well short of any roast-derived darkness that would mask origin character.
Roast all samples you intend to compare on the same day, on the same machine, at the same batch size, aiming for the same color. If one sample is roasted darker than another, you will be tasting your roaster rather than the coffee.
Rest the roasted samples somewhere between eight and twenty-four hours before cupping. Cupping straight off the roaster gives you carbon dioxide and confusion.
Grind, ratio, and setup
Use a ratio in the region of 8.25 grams of coffee to 150 millilitres of water, which is close to the conventional cupping standard and easy enough to scale. What matters is that every bowl on the table uses the same ratio measured on the same scale.
Grind coarser than filter, roughly the texture of coarse sea salt. Grind each bowl immediately before pouring, and purge a small amount of coffee through the grinder between samples so you are not tasting the previous lot.
Set up at least two bowls per coffee. Green coffee is not perfectly uniform, and a single defective bean can define a single bowl and mislead the whole decision.
The sequence, step by step
First, dry fragrance. Smell the ground coffee before any water touches it. Note first impressions quickly and without deliberating.
Second, pour. Fill each bowl with water just off the boil, saturating all grounds, and start the timer. Let it stand four minutes undisturbed. A crust will form on the surface.
Third, break the crust. At four minutes, push the crust gently back with the spoon while leaning in to smell what is released. This is the single most informative moment in the session, and it lasts a few seconds per bowl.
Fourth, skim. Remove the foam and floating grounds from the surface with two spoons. Rinse spoons between bowls.
Fifth, taste as it cools. Begin somewhere around eight to ten minutes in, when the liquid is hot but not scalding, and keep returning as it drops toward room temperature. Slurp sharply to aerate across the whole palate. Acidity and sweetness reveal themselves as the cup cools, and a coffee that tastes flat when hot and sweet when cool is telling you something a single early sip would have missed.
Scoring without a Q license
Formal scoring across ten attributes is valuable and largely unnecessary for a buying decision. Four attributes carry most of the useful signal for a small roastery.
Rate acidity, sweetness, body, and finish, each out of ten, and write two or three concrete flavor descriptors. Then add one line stating whether you would put this coffee on your menu and in which position. That last line is the actual output of the session and the one most cupping forms leave out.
Note any defect distinctly. Fermented, phenolic, musty, papery, or bag-like notes are not preferences to be weighed against acidity. They are grounds to stop.
Always cup against something you already buy
Put a coffee currently in your lineup on the table alongside the samples, roasted the same way on the same day. Do not tell yourself which bowl is which if you can arrange for someone else to set the table.
A sample cupped alone almost always tastes promising, because your palate has no reference point in the room. The same sample cupped beside a coffee you know well often turns out to be a lateral move, and lateral moves are expensive once they arrive as a pallet.
Record it so the next decision is comparable
Keep the records in a consistent form. Date, origin, lot identifier, roast level, days rested, your four scores, your descriptors, and your buy or pass line.
Cupping notes only become valuable in aggregate. After a dozen sessions you can see which origins your palate consistently rewards, where your roast approach flatters certain profiles, and how well an arrival lot matched the sample that convinced you to buy it.
Samples of 350g to 1kg are available from India, Vietnam, and Ethiopia. One kilogram is the useful size if you plan to test-roast. Request one at caffepremium.com/request-a-free-sample.
Comments