Building a Practical Sensory Vocabulary for Matcha

“Strong tea taste”, “too bitter” and “dry finish” may all be used for the same matcha sample, even when the speakers are noticing different sensations. If the vocabulary is not aligned, procurement, R&D, quality and supplier teams can appear to agree while assessing different evidence. That affects sample selection, formulation adjustment and, where sensory comparison forms part of an agreed control, lot-acceptance decisions.

A shared sensory vocabulary is not a universal matcha ranking. Its purpose is to identify the perceived attribute, describe its intensity and note when it appears. ISO publishes separate frameworks for sensory vocabulary, sensory profiling and the selection and training of sensory assessors. Those standards support disciplined terminology and procedure, while an internal project sheet still needs its own matcha definitions and validation.

Separate Bitterness from Astringency

Bitterness is a taste. In a project assessment, it can be described by its intensity, onset and persistence after the sample enters the mouth. Astringency is better treated as an oral sensation or mouthfeel: drying, tightening, roughness or friction may develop across the mouth and may continue or build after swallowing.

The two frequently occur together, which makes “bitter-astringent” a tempting shortcut. It is not a useful single measurement. A study of matcha prepared under different agitation conditions also evaluated bitterness and astringency as separate attributes. The study supports separating the fields in a working form; its reference solutions, agitation settings and sample results are not commercial thresholds.

R&D can replace “quite bitter and astringent” with a timed description: moderate bitterness at entry, followed by a more persistent drying sensation after swallowing. This gives the next formulation discussion a clear focus on taste, mouthfeel or both.

Record Umami, Sweetness and Aftertaste Independently

Umami and sweetness are different tastes and should not be used as substitutes for one another. Matcha studies may evaluate both, but a chemical result cannot be translated directly into the perceived strength of either attribute in a finished product. Raw-material composition, preparation, temperature, agitation and the formulation matrix can all influence perception.

The meaning of “sweet” also needs context. A slight sweetness perceived in a water-only sample is not the same assessment as the integrated sweet impression of a latte or confectionery product. Sugar, dairy or plant base, fat and aroma may all contribute to the finished experience. A project form can distinguish sweetness perceived in the water-only sample from the finished product’s integrated sweet impression.

Aftertaste is a time dimension rather than a single taste. The assessor should identify what remains after swallowing: umami, sweetness, bitterness, astringency, tea aroma or another defined sensation. Intensity and persistence can then be added. Commercial phrases such as “clean finish” or “long finish” become more useful when the lingering perception is stated.

The same principle applies to the word “balance”. Balance is a conclusion about the relationship among attributes, not an attribute with one fixed meaning. The underlying bitterness, astringency, umami, sweetness, aroma and mouthfeel should remain visible before an overall balance judgement is added.

Give Every Descriptor a Definition and a Reference

A sensory lexicon is more than a list of attractive words. Peer-reviewed work on food sensory lexicons describes trained assessors using defined descriptors, references and standardised procedures so that the language can support product development and quality work. That literature covers many food categories rather than matcha specifically, so it provides a method principle rather than a ready-made matcha vocabulary.

R&D and quality can begin with a compact set of attributes that distinguish the samples actually under review. Each entry can contain:

• a plain-language definition;
• the sensory domain, such as taste, aroma or mouthfeel;
• an internal reference or agreed comparison sample;
• an intensity scale with anchored points;
• the evaluation moment, such as entry, mid-palate or after swallowing;
• any known confusion with another descriptor.

Terms should be descriptive rather than hedonic. “Roasted aroma, medium intensity” records a perception; “pleasant aroma” records liking. Both may matter to a project, but they answer different questions and should not occupy the same field.

Internal references also need a controlled identity. An approved sample, preparation or formulation may anchor a team discussion without becoming a universal sensory standard. If the reference changes, quality should retain the comparison history and reason for change.

Standardise Sample Preparation Before Comparing Scores

Vocabulary alone cannot make results comparable. Temperature, powder and water quantities, water condition, agitation method, standing time, vessel, serving order and sample coding can all affect an evaluation. The matcha agitation study provides a clear caution: changes in preparation conditions were associated with changes in the measured sensory profile of the samples in that experiment.

R&D should write the preparation method before blind evaluation begins. Procurement, R&D and quality participants can score coded samples independently and then discuss the descriptors with the greatest disagreement. Supplier identity and price are better introduced after the first sensory comparison, reducing the chance that commercial information shapes the initial description.

A water-only comparison and a finished-product trial serve different purposes. The first can create a common baseline under a controlled preparation. The second establishes performance in the real formulation and process. Scores from the two should remain separate rather than being averaged into one apparent result.

Traditional or research-oriented matcha assessment may also observe appearance, particle size, foaming, foam colour and taste as distinct items. That separation is useful because taste vocabulary cannot carry information about colour, powdery mouthfeel or foam. The published Japanese method has its own preparation and quality context, however, and should not be presented as a universal acceptance procedure for every food or beverage application.

Use Instrumental and Composition Data as Supporting Evidence

Researchers have used taste-sensor systems to evaluate astringent and umami intensities in sets of commercial matcha, while other studies have explored relationships among chemical, instrumental and sensory results. These approaches can help compare samples, examine trends or formulate a hypothesis about an observed difference.

Their limits matter commercially. A sensor output depends on its method, references and calibration range. A model reflects the samples and measurements used to build it. A higher concentration of one compound does not, by itself, prove that a finished product will taste more bitter, more umami or more suitable for its application.

Sensory observations, instrumental or analytical results and finished-product performance should remain related but separate evidence. A relationship seen across multiple lots may support investigation or control design, but neither a single human score nor a single instrument result should replace application testing and agreed lot criteria.

Build a Repeatable Project Record

A practical sheet can pair each attribute with its definition, reference, intensity scale and evaluation time. It should also retain sample code, lot identity, preparation details, assessor, date and any unusual observations. Where the method is used repeatedly, variation among assessors can be as informative as the average.

Repeated disagreement about one term is a reason to revise the definition, check the reference or recalibrate the group before rejecting a material lot. The sensory target depends on the intended application. Lower bitterness, higher umami or a longer aftertaste does not automatically indicate a higher grade, higher price or better result in every formula.

The assessment becomes most useful when it connects three layers without collapsing them:

1. controlled baseline preparation;
2. actual product formulation and process;
3. agreed incoming or lot-comparison procedure, where applicable.

Each layer should retain its own purpose and acceptance criteria. The result is a vocabulary that supports discussion and traceability without claiming that one internal method is a global matcha standard.

Discussing Sensory Targets with CHA STELLA

CHA STELLA can provide candidate matcha samples for the intended application and work with procurement and R&D teams to align the target sensory profile, including bitterness, astringency, umami, sweetness, aroma, mouthfeel and aftertaste. Evaluate the results in the context of the actual formulation and project specification. No single descriptor, chemical result or instrument score replaces finished-product testing, agreed acceptance criteria or lot confirmation.