How We Test Matcha

A ranking is only worth reading if you can see how it was produced. This page documents our testing protocol and scoring model in enough detail that someone else could repeat it and check our work.

How we obtain what we test

• Wherever possible we buy products ourselves at retail price, through the same channels a reader would use.
• Where a brand supplies a sample, the review says so explicitly. A supplied sample does not guarantee coverage, and it does not change the score.
• We record the purchase date, channel, price paid, pack size and batch or lot code where visible on the tin. Matcha is perishable, so a review without a date is close to meaningless.

Fixed brewing protocol

Every product goes through the same two preparations so results are comparable.

Variable	Straight (usucha) test	Latte test
Matcha	2.0 g, sifted	2.0 g, sifted
Water	70 ml at 80°C	30 ml at 80°C to make a paste
Milk	,	150 ml full-fat dairy at 60°C; repeated with unsweetened oat milk
Sweetener	None	None, sweetening hides faults
Whisking	Bamboo chasen, 30 seconds, W motion	Chasen to a smooth paste, then combined
Assessment	Within 2 minutes, before the foam falls	Within 2 minutes

Water is filtered and consistent across tests. Tasting is done without sweetener, in daylight where colour is being judged, and alongside at least one previously tested reference product so scores do not drift over time.

What we record before tasting

• Powder colour in daylight, and whether it has dulled toward olive or yellow-green.
• Aroma dry, straight from the tin.
• Texture, grind fineness, clumping, how much sifting it needs.
• Label claims recorded verbatim and kept separate from what we observe. A tin saying “Uji, first harvest” is a claim, not a finding.

Taste dimensions

Each preparation is rated on five dimensions, which are reported alongside the score rather than hidden inside it:

• Umami, savoury depth and body
• Sweetness, natural sweetness without added sugar
• Bitterness, recorded as intensity, not automatically a fault
• Astringency, the drying, grippy sensation
• Finish, how long and how cleanly the flavour lasts

The KYM Score (100 points)

Dimension	Points	What earns them
Taste	25	Balance and quality across the five dimensions, judged against intended use.
Origin transparency	15	Named region, and cultivar or producer where disclosed. Vague “Japanese origin” scores low.
Freshness	15	Harvest or packing date on the tin, colour and aroma condition on arrival.
Texture & whiskability	10	Grind fineness, ease of suspension, clumping.
Colour	10	Vibrancy of the dry powder and the prepared bowl.
Value	10	Quality delivered per rupee, using ₹/g. Expensive is not penalised if it earns it.
Packaging & storage	5	Airtight, opaque, resealable. Matcha in a clear jar loses points.
Producer transparency	5	Whether the brand answers questions about sourcing.
Availability	5	Whether an Indian reader can reliably buy it.

What we deliberately exclude

• Marketplace star ratings and review counts. These measure popularity and shipping speed, not matcha quality. We may mention them as context; they do not feed the score.
• Grade wording on the packaging. “Ceremonial”, “AAA” and “imperial” are marketing terms, not regulated Japanese grades, and earn no points by themselves.
• Brand size, funding or social following.
• Any commercial relationship with us, including a free sample, stated on the entry itself. See our editorial policy.

Café assessment

Cafés are handled separately from powders. A café entry is only marked visit-verified when we have been there in person, and it then records the date, what we ordered, the price paid, which matcha they use if disclosed, whether straight matcha is available, milk options, and whether drinks are sweetened by default. Entries compiled from menus or public information are labelled as such and carry no KYM Score.

Re-testing and versioning

• Matcha degrades, and brands change suppliers. We aim to re-test ranked products annually and after any visible reformulation.
• Prices and availability are checked separately from taste and carry their own last-verified date.
• Changes to this methodology are versioned, dated and noted here, so an older score can be read in the context of the rules it was given under.

Where we are honest about limits

We are one palate, not a trained tasting panel, and our sample sizes are small. We cannot verify farm-level sourcing claims independently, and we do not run lab analysis for catechin content or heavy metals. Where a conclusion rests on a brand's own claim rather than our observation, the review says so.

Last reviewed 15 August 2026 · Report an error