Cupping & sensory

Coffee Cupping

Every green coffee transaction on this site ultimately rests on a cupping table. This page explains what that table controls for, why each control exists, and how to make sure a sample approval means what both parties think it means.

Last reviewed July 20263 min read

Quick answer

Cupping is standardised sensory evaluation: coffee is sample-roasted to a defined level, ground at a set particle size, infused with hot water at a fixed ratio, left to steep, and then the crust is broken and the liquor assessed. The standardisation exists so that differences observed on the table are differences in the coffee, not in the method.

What cupping is for

Cupping is not tasting coffee the way a customer would drink it. It is a deliberately austere method designed to strip out every variable except the coffee itself — no brewing skill, no equipment differences, no recipe. What remains is a comparison you can trust.

It serves three distinct commercial jobs, and it is worth being clear which one is happening at any given moment.

1Quality control. Is this parcel free of taints and faults? A pass/fail question, run on almost everything.
2Selection. Which of these samples do I want to buy? A comparative question, run across an offer list.
3Approval. Does this pre-shipment sample match the sample I approved? A conformity question, and the one that matters at contract stage.

The controls that make it work

What is held constant, and why
Control	Convention	What it prevents
Sample roast level	Light, even, to a defined colour, rested a set period	Roast differences masquerading as coffee differences
Roast-to-cup interval	A fixed rest window before cupping	Comparing a freshly roasted sample against a rested one
Grind size	Coarse, consistent, ground immediately before cupping	Extraction differences between cups
Dose and ratio	A fixed coffee-to-water ratio across every cup	Strength differences reading as body differences
Water	Consistent, defined mineral profile, off the boil	Water chemistry altering perceived acidity and sweetness
Cup and bowl	Identical vessels and volumes	Thermal differences between cups
Number of cups per sample	Multiple cups from the same sample	A single defective bean going undetected
Blind presentation	Samples coded, not labelled	Expectation bias about origin or price

The sequence at the table

1Fragrance. Assess the dry grounds before water is added.
2Aroma. Add water, wait the defined steep, and assess the wet crust.
3The break. Push the crust back with a spoon and smell the released aromatics — the single most information-dense moment on the table.
4Skim. Remove foam and floating grounds.
5Tasting. Begin as the liquor cools enough to sip, and continue as it cools further. Different attributes present at different temperatures.
6Evaluation. Record attributes and any faults, cup by cup, not as a single impression.

Temperature progression is not a formality. Sweetness and body read more clearly as the cup cools; some faults only become obvious at low temperature. A table that finishes while the cups are still hot has not finished.

Making sample approval mean something

The commercial value of a cupping is entirely dependent on the sample being representative. This is where most disputes originate — not in the tasting, but in what was in the bag.

Agree the sample type. Offer sample, type sample, pre-shipment sample and arrival sample are different things with different standing in a contract.
Agree the draw method. A sample drawn from the top of the accessible bags is not a composite of the parcel.
Agree the sample size. Too small a sample cannot represent a container.
Cup against a retained reference. Approval means "matches the approved sample", which requires that the approved sample still exists.
Agree what happens on a mismatch. Re-sampling, allowance or rejection — decided in the contract, not after the fact.

Lots with the identity stated

Lots are identified by station, factory or outturn where the origin supports it, with the moisture and preparation stated.

Request Current Crop Offer

Frequently asked questions

Why is cupping done at such a light roast?

Because roast development masks the coffee. A darker roast adds its own character and flattens differences between samples. A light, even sample roast keeps the origin, variety and processing visible, which is the point of the exercise.

Why break the crust?

Breaking releases the aromatics trapped under the floating grounds in one concentrated burst. It is the most information-dense moment of a cupping and the point at which many faults first announce themselves.

How many cups should be prepared per sample?

Multiple — commonly five in professional practice. Faults such as potato taste occur at individual bean level, so a single cup can easily miss a problem that affects a meaningful share of the parcel.

Does water quality really change the result?

Substantially. Mineral content affects extraction and directly changes perceived acidity, sweetness and body. Two tables using different water will disagree about the same coffee, which is why water is one of the controlled variables.

Tell us the coffee you need

Lots can be specified by origin, region, process, grade, screen, moisture, defect tolerance, crop year and packaging. Send what you know and we will confirm what each origin realistically supports.