What actually tells you something
HARVEST YEAR AND SEASON, for anything where freshness matters — a Japanese green tea without a year on it is a warning. ORIGIN AT DISTRICT LEVEL rather than country: 'Ceylon' spans teas that share almost nothing, and 'Uva' or 'Nuwara Eliya' is a real claim. CULTIVAR, where the seller states it, because it explains character. And PROCESSING DETAIL — steaming depth, roast level, whether pu-erh was sun-dried — which is the information most sellers omit and the most useful when present.
What tells you less than it appears to
GRADE CODES describe leaf size and appearance, not quality: FTGFOP tells you there are golden tips, not that the tea is good. VISIBLE FLOWERS in scented tea usually indicate decoration compensating for weak scenting rather than quality, with the deliberate Sichuan style as the exception. AGE CLAIMS on pu-erh and white tea are unverifiable by a buyer at any price, and colour deepens with warm storage in months as readily as with genuine years.
Where substitution is routine
Some names are sold far more than they are produced. Genuine Da Hong Pao mother-bush tea does not exist commercially. Lao Ban Zhang production is a fraction of what carries the name. Wu Niu Zao is sold as early Longjing every spring, Ben Shan as inexpensive Tie Guan Yin, and Nepali tea as Darjeeling. None of these substitutes is a bad tea — the problem is the label and the price, not the leaf.
Buy small first
The single most useful habit. A 25 g sample of four teas teaches more than 100 g of one, and it caps the cost of discovering you dislike something. Sellers who offer samples are generally the ones confident their tea survives comparison.
Signals a seller is worth trusting
They state harvest year and district without being asked. They say when they do not know something. They explain processing rather than only romance. They sell samples. And they price consistently — a shop where everything is described as rare and ancient is telling you about its marketing rather than its tea.