What each channel actually offers
In person you can see the dry leaf, judge uniformity, smell it, ask questions of someone who may have tasted it, and in the better shops taste it before buying — which is worth more than any amount of copy. What you usually cannot get is harvest data, provenance detail below country level, or a considered answer about processing, because most shops selling tea are not specialists. Online the position reverses: a good speciality site publishes harvest year and season, district or garden, cultivar, processing detail, the maker’s name, and its own tasting notes, and none of it can be verified before purchase. The rational response is to use each channel for what it is good at — in-person for judging leaf and for tasting, online for information and for range — and to regard the first purchase from any new seller as a test of whether their descriptions match their tea. There is a third channel worth naming because it behaves differently from both: buying at a tea house or a shop that brews before it sells. Tasting first eliminates almost every uncertainty this page is about, and the price reflects the service. For anyone new to a category it is the fastest and cheapest route to knowing what they actually like, because a single afternoon of tasting four teas replaces months of ordering samples and guessing at descriptions.
What a good description contains
Six things, in rough order of how rarely they appear. Harvest year and, where relevant, season or picking. Origin at the tightest level the seller actually knows, stated at that level and not inflated. Cultivar, where the tea is single-cultivar and the information exists. Processing detail specific to the category — steaming depth for a Japanese green, roast level and roasting method for an oolong, whether a sheng was sun-dried, shading period for a matcha, oxidation described as a descriptor rather than a percentage. Brewing parameters given for a stated method and vessel size rather than as a universal instruction. And a statement of what the seller does not know, which is the rarest and the most informative of the six. A description containing four of these is written by someone who handled the tea. What a good description does not need is length. A short entry giving garden, year, cultivar and roast level is worth more than four paragraphs of atmosphere, and the two are inversely correlated often enough to be a rough signal in themselves. The test to apply is the one used on the empty-description page: count the statements that could be false, and ignore the rest of the word count.
The tells that a description was written from a spreadsheet
Recurring phrases across unrelated teas in the same shop, which indicates copy generated from a template rather than from tasting. Tasting notes that could apply to any tea in the category — smooth, refreshing, well-balanced. Brewing instructions identical across every product regardless of category, which is impossible to mean. Oxidation percentages, which are quoted everywhere and measured almost nowhere and are a reliable signal that the writer is repeating rather than knowing. Origin stories that are historical romance with no dates. And a range in which everything is described as rare, ancient or the finest available, which is a statement about the shop’s marketing rather than about its tea. None of these proves the tea is bad. They establish that the description is not evidence about it. A further tell is a photograph that does not match the product. Stock imagery, a picture of a different tea in the same category, or an image reused across several listings all indicate a catalogue assembled rather than tasted. Conversely a seller who photographs the dry leaf, the liquor and the wet leaf of the actual lot has handled it, and those three images together are more informative than most of the copy that accompanies them.
Judging leaf in a shop
Four things are visible without any expertise. Uniformity — whether the lot is one thing or a mixture of sizes, which affects extraction and often indicates careless sorting. Condition — whether the leaf is intact and crisp or has been broken down by handling and time. Colour — vivid for a fresh green tea, dull olive or brown for a tired one, and the same logic in reverse for aged and roasted teas. And smell, which is the most informative and the least used: ask to smell the leaf, and if the answer is no, that is also information. What you cannot judge by eye is nearly everything else, so the shop’s advantage is real but narrow. A shop that will let you smell, that stores its tea in opaque sealed containers rather than open display jars, and that can tell you when a tea arrived is worth returning to. How the shop stores its stock is the fifth thing to look at and it is visible from the doorway. Open display jars in daylight, tea in clear containers on a sunny counter, and large decanted quantities of green tea all indicate a shop whose stock is being degraded on the premises, however good the leaf was on arrival. A shop that keeps its working stock in sealed opaque containers and decants only what it sells has understood the same argument this catalogue makes about storage.
Buying online without wasting money
Buy samples on the first order, always, and buy several rather than one so that the order teaches you something by comparison. Order across a range that holds one variable — three teas of the same category from different origins, or the same tea at two roast levels — rather than one of everything. Note the shipping time and the packaging on arrival: a tea shipped in a sealed foil pouch by a seller who states harvest year is behaving consistently, and a tea that arrives in a paper bag is telling you about the seller’s storage assumptions. And check the description against the cup once, deliberately. A seller whose notes match what you find is a seller whose notes you can use for years, which is the actual thing being purchased on a first order. Consolidating orders is worth a thought too, since shipping is a real cost and a real delay. Fewer, larger sample orders from one seller generally cost less per tea and arrive faster than a scatter of single purchases, and they make the comparison within the order easier to set up. The exception is the fragile seasonal teas, where ordering promptly at the right time of year matters more than the postage.
The one advantage of buying at origin, and its limits
Buying tea in a producing country is often cheaper, sometimes much cheaper, and is occasionally the only way to get certain products. The limits are worth knowing before travelling with an empty suitcase. Tourist-facing tea shops in producing regions are frequently no better informed and no better priced than a good importer, and are sometimes worse on both counts. Storage between purchase and home is a real risk for the fragile categories. And the ability to judge what you are being offered matters more, not less, in a market where the range is enormous and the language barrier is real. The genuinely reliable version is buying at origin with a knowledgeable guide or from a producer you were introduced to, which is a different activity from buying tea while on holiday. There is a customs and quantity dimension as well, which travellers routinely overlook: countries have rules about importing agricultural products, and a suitcase of loose tea can attract attention or charges depending on the destination and the quantity. That is a practical matter to check before buying rather than after, and it is one more reason that the romantic version of buying at origin works better as a small, deliberate purchase than as a bulk one.