- Buying Tea Samples: How to Judge Before Committing
- The practice is older than the sample bag
- Why most samples teach you nothing
- Order the right size
- Brew it your way, twice
- What ISO 3103 is, and what it is not
- Three brews, three questions
- Judge the leaf before the liquor
- Keep a record or the samples are wasted
- What a sample can and cannot tell you
- What a sample costs you, in time
- Reading the seller, not just the tea
- When to stop sampling
- How big should a tea sample be?
- Should I brew a sample the way the seller recommends?
- What is ISO 3103?
- Why brew a sample harshly?
- How many samples should I order?
Buying Tea Samples: How to Judge Before Committing
The practice is older than the sample bag
Buying a small quantity before committing is not a modern invention of online tea shops. In the Chinese tea trade a yang, or sample, is formally defined as a portion drawn from a batch for assessment and comparison, and it is the basis on which standardised assessment brewing, buying and pricing all rest. Older buyers would ask for a handful before discussing a whole lot: look at the dry leaf, smell it, brew it, and only then talk about the consignment.
The trade phrase for this is ‘setting a price by looking at the tea’, and by the modern period tea houses were keeping standard samples to make transactions comparable. Two things follow from that history and both apply to a buyer today. A sample is only useful if it represents the batch, and it is only useful if it is brewed under fixed conditions.
Why most samples teach you nothing
A sample exists to answer one question: would I buy this again? Most people answer a different question by accident. They brew the sample once, at whatever strength they use for everything else, in whatever vessel is nearest, and decide on the first impression. That measures whether the tea survives a careless brew, not whether it is good.
The mistake is structural. A sample is a small quantity of a specific lot, and it is only informative if the test matches how you will actually drink the tea. A 5 g sample brewed at 3 g per 150 ml gives you one cup and no second infusion. A sample tasted in a crowded shop tells you about the shop. What follows is a way to test that produces a decision rather than an impression.
Order the right size
The useful sample size is the one that gives you at least three sessions in your own vessel. For a 100 ml gaiwan at 5 g per session that means 15 g, and most sellers who offer samples sell in 10 g or 25 g units. Below 10 g you are buying a smell, not a tea.
Two samples of the same tea from different lots are worth more than one larger sample, because the trade principle
is that a sample must represent its batch evenly rather than being a lucky handful from the top of the tin, and because the question a buyer actually has is not “is this tea good” but “is this tea consistent”. A single sample cannot answer that. Two can, cheaply. Order the sample in the packaging you would receive at full size if you can. A tea that arrives in a foil pouch and a tea that arrives in a paper bag are different products by the time they reach you, and the sample should not hide that.
Brew it your way, twice
First, brew the sample exactly as you brew tea at home. Same vessel, same leaf-to-water ratio, same water. If it is disappointing at your normal parameters, that is the answer, and no amount of careful brewing later rescues it.
Second, brew it at the parameters the seller recommends, if they give any. The gap between the two results is diagnostic. A tea that is dull at your ratio and excellent at theirs is telling you that it is a narrow tea, and you need to decide whether you are willing to brew narrowly. A tea that is good at both is forgiving, which for a daily drink matters more than peak quality.
Third, if the sample allows, brew it a third time at deliberately harsh parameters: hotter water, longer steep. This is the test professionals use, because it exposes the faults that a careful brew hides. Bitterness that appears immediately and does not clear is a processing fault. Bitterness that arrives late and fades into sweetness is a characteristic, not a defect.
What ISO 3103 is, and what it is not
There is a standard method for tasting tea, and it is worth knowing what it says before anyone cites it at you. ISO 3103 specifies a standardised brew for sensory comparison: a fixed leaf-to-water ratio, boiling water, and a six-minute infusion, served in a white porcelain set so the liquor and the leaf can both be examined.
That method exists so that two tasters in different buildings can compare notes, and the Chinese assessment tradition reaches the same conclusion by a different route: a fixed leaf weight, fixed water temperature and fixed steep time, so that the tea rather than the brewing is what varies. It is a comparison tool, not a recommendation. Its own scope is limited, and it is a poor guide for green tea in particular, which is normally brewed well below boiling for under three minutes. Applying ISO 3103 to a green tea produces a cup that no producer intended you to drink.
The practical consequence for a buyer is that a seller who cites the standard has told you they can be compared fairly with another seller, which is useful. A seller who implies the standard defines the correct way to brew the tea has told you something else.
Three brews, three questions
| Brew | Parameters | What it answers |
|---|---|---|
| Your normal | Your vessel, your ratio | Would I drink this daily? |
| Seller’s | As recommended on the pack | Is it narrow or forgiving? |
| Harsh | Hotter water, longer steep | Does it have faults worth avoiding? |
Judge the leaf before the liquor
Before any water, do what the old buyers did: look at the dry leaf and smell it. Before any water, look at the dry leaf. It should be intact rather than dusty, uniform rather than a mixture of sizes, and it should smell of something specific rather than of “tea”. A sample that has broken into powder in transit has already told you how it was handled.
Then look at the leaf after brewing, which most buyers never do. Fully opened leaf tells you the picking standard: whole leaves and buds mean a higher grade than chopped fragments. It also tells you whether the leaf is even, which is the difference between a sorted lot and a sweepings pile. This is the one part of judging that a sample can do better than a full purchase, because you are willing to destroy a small quantity to look at it.
Keep a record or the samples are wasted
The value of sampling compounds only if you write something down. A note that says “good” is worthless three months later. A note that records the seller, the lot, the price, the parameters you used and one sentence about what the cup did will let you decide a repeat order in seconds. The minimum useful record is four fields: what it was, what it cost, how you brewed it, and whether you would buy it again. Anything beyond that is a bonus. Without the fourth field you have collected data rather than made a decision.
What a sample can and cannot tell you
- It can tell you whether you like this lot at your own parameters.
- It can tell you whether the leaf is intact and evenly sorted.
- It cannot tell you how the next lot from the same seller will taste.
- It cannot tell you how the tea ages, because you will have drunk it long before then.
What a sample costs you, in time
The arithmetic that decides whether sampling is worth doing is not the price of the sample but the time it takes to test properly. Three brews at 5 g in a 100 ml vessel is roughly fifteen minutes of attention, plus the notes. At that rate, ten samples is two and a half hours of work, which is more than most buyers budget for a purchase decision.
That is why sampling works best when it is narrow. Two lots of one tea, tested properly, will settle a question that twenty random samples will not. The constraint is not money but attention, and attention is what the sellers who offer free samples are counting on you not to spend. A practical compromise is to test properly once and then taste quickly. The first sample of a new seller gets the full three brews; subsequent ones from the same seller can be one brew each, because you have already established how that seller’s tea behaves.
Reading the seller, not just the tea
The sample is also a test of the seller, and some of the information arrives before the water does. A pack that states the lot, the harvest date and the picking standard has told you the seller knows what those things are. A pack that states a poetic name and nothing else has told you what they think sells.
Delivery matters more than buyers expect. A sample that arrives crushed, damp or in a paper envelope has been handled carelessly, and the tea inside has already lost the argument about how it will be treated at full size. A seller who packs a 15 g sample in foil with the air pressed out will pack a 250 g tin the same way.
The third signal is whether the sample matches the description in the cup. If the listing promises a smoky tea and the sample is not smoky, that is not a matter of taste, it is a mismatch, and it is the single most useful thing a sample can reveal about a seller.
When to stop sampling
Sampling has a failure mode: buying samples becomes the hobby, and the full-size purchase never happens. The signal is a shelf of 10 g bags and no tea you drink regularly. The fix is to set the decision rule before you order. If the sample is good at your normal parameters and survives the harsh brew, buy the full size in the next size up rather than collecting more samples of the same thing. If it fails, note why and move on. Two outcomes, one decision, no accumulation.
There is a second reason to stop: a sample is a small quantity of a lot that may not exist in six months. The tea you loved as a sample and did not buy is not a tea you can return to. That is not a reason to buy everything; it is a reason to decide rather than to keep evaluating. How the same logic applies to a whole category rather than a single lot is set out in the buying steps for green tea, and the failure modes that a sample is designed to catch are covered in green tea traps.
How big should a tea sample be?
At least 15 g, so you get three sessions in your own vessel rather than one cup.
Should I brew a sample the way the seller recommends?
Yes, but only as a second brew. The first should be your own normal parameters.
What is ISO 3103?
A standardised tasting method for fair comparison between tasters, not a brewing recommendation.
Why brew a sample harshly?
It exposes faults that a careful brew hides, which is what professional tasters test for.
How many samples should I order?
Two lots of the same tea tell you about consistency; ten different teas tell you nothing.
