Start with a sampling question
A sample represents something: a bag, a delivery, a production batch or a retained reference. Define that population before opening the container. A handful from the top of one bag may be convenient, but it cannot automatically support a claim about an entire shipment. Record who collected the material, which units were sampled, how portions were combined and what happened during transport. ISO 4072, Green coffee in bags — Sampling, is a relevant starting point for bagged green coffee, not a universal instruction for every format.
Keep the received sample and the tasting portion conceptually separate. Assign a traceable identifier on receipt, then use a separate blind code at the table. Preserve the connection in a restricted key. Check for torn packaging, moisture exposure and unusual odour before preparation. If the sample's history is incomplete, record that limitation rather than filling the gap with a confident assumption.
Choose a protocol, not a collection of habits
The Specialty Coffee Association's Coffee Value Assessment provides a current framework for describing and assessing coffee. ISO 6668 addresses preparation of samples for sensory analysis. They serve related purposes but should not be casually combined into an unnamed hybrid. Select the method appropriate to the decision, consult its current official documentation, and record the version you actually used. An internal screening procedure can also be useful, provided its conditions and limitations are explicit.
Translate the selected method into a bench checklist. Include coffee mass, water mass or volume as specified, grind setting, vessel specification, water conditions, roast age and the timing of observations. Use a balance and timer suitable for the work. Verify repeatability by checking repeated preparations, not just by trusting a display with many decimal places. Greater numerical precision on the screen does not guarantee more accurate preparation.
| Control | Record | Why it matters |
|---|---|---|
| Sample identity | Receipt code and blind-code key | Links a cup to the correct material |
| Water | Source, treatment and relevant checks | Changes extraction and flavour expression |
| Grinding | Equipment, setting and cleaning routine | Affects extraction and carryover |
| Timing | Pour order and observation stage | Separates coffee differences from cooling differences |
Make water and grinding visible variables
Water that tastes neutral is not necessarily equivalent to yesterday's water. Mineral composition, alkalinity, treatment and storage can affect the cup. Use a documented supply and check the properties relevant to your chosen method. A conductivity or total-dissolved-solids meter alone does not identify the mineral composition. For a repeat session, reproduce the water conditions where possible; if they change, flag the comparison.
Clean the grinder according to its manufacturer's instructions and manage retention consistently. A Mahlkönig EK43 and a Ditting laboratory grinder can both be used in professional settings, but the machine name alone does not establish an equivalent particle distribution. Record the actual equipment and setting. If changing coffees requires a purge, define it in advance. Do not improvise a different cleaning routine only for a sample you expect to dislike.
Build a table that does not give the answer away
Use neutral codes and avoid telling assessors the origin, price or producer story before independent observations. Ethiopia's Yirgacheffe, Kenya's Nyeri and Colombia's Huila are useful geographical references, but their names can also create expectations. Blind presentation reduces that particular cue; it does not eliminate all bias. Balance serving order where practical, use comparable vessels and keep the environment free of distracting odours.
Follow the chosen method's requirements for replicate cups. Multiple cups can reveal unevenness or an isolated problem that one cup would miss. They are not automatically independent samples of an entire lot. Track which observation belongs to which cup, and do not erase an unusual cup simply because the others were pleasant. Investigate whether the difference arose from the material, preparation or an identifiable handling error.
Observe first, interpret second
Write sensory observations before group discussion. Separate aroma character, perceived intensity and personal liking. “Citrus-like acidity” describes a perception; “too acidic for me” expresses preference. Both may be useful, but they belong in different fields. Taste through the relevant cooling stages and record when an impression occurred. A descriptor appearing late in the session should not be presented as though it dominated every stage.
Finish with a conclusion proportionate to the setup. You may have evidence that the prepared samples differed under these conditions. You may not yet know why. A fair next step could be a repeat preparation, a new representative sample or a roast check. Keep the blind-code key, preparation log and independent notes together so the conclusion remains inspectable after the table is cleared.