Two kinds of date, doing two different jobs
Food-labelling regimes generally distinguish a date about quality from a date about a much harder limit. The quality date, printed in English-speaking markets as best before, indicates the period over which a product is expected to retain the properties the buyer would expect of it — its aroma, its texture, its appearance. The other kind of date is applied to a narrow class of highly perishable products and carries a different meaning entirely. Dry tea belongs unambiguously to the first group, which is why an old tea is a disappointing tea rather than anything else, and why this catalogue frames age as a question about character. TeaHQ describes the architecture rather than the law: requirements, wording and exemptions differ between the United Kingdom, the European Union, the United States, Japan and other markets, are revised, and have not been verified here. What does not depend on jurisdiction is the structural point that a durability date is a manufacturer’s statement of expected quality, made by the business responsible for the food, on the basis of its own judgement — not an assessment made by any authority and not a measurement made on the leaf in that particular packet.
Why the interval on tea is a policy rather than a finding
A packer setting a durability date for a dry, shelf-stable product is not usually running a study on that product. It is applying a period that its category, its packaging and its trade practice support, often the same period across a whole range, and building in margin so that stock has time to move through distribution and sit on a shelf. Two and three years are common. That is a defensible commercial decision and it produces a number that carries almost no information about a particular tea, because the same interval lands on a robust CTC blend whose character will barely move and on a spring sencha whose selling point is measured in months. The date is accurate about the packer and inaccurate about the leaf. It also encodes an assumption the packer cannot control, which is that the product will be stored reasonably. A tea kept above a cooker in a half-open carton has been in worse conditions than the estimate assumed, and one sealed and cool has been in better, and the printed date does not move for either. This is the deeper reason a date is a poor instrument here: it describes a gradient with an endpoint, and the rate along the gradient belongs to whoever has the tea.
What the date does tell you, read sideways
One thing reliably, and it is worth having. The interval a packer chooses reveals how that business classifies its own product. A three-year date on a green tea says the packer regards it as shelf-stable stock moving through a long supply chain, which is a fair description and tells a buyer what kind of tea they are looking at and how it has probably been handled. A short dated window on a Japanese green tea, or a harvest statement in place of a durability estimate, says the opposite. Reading the date as evidence about the business rather than about the leaf converts an uninformative field into a modestly informative one, and it costs nothing. The second sideways reading is the gap between the date and today. A best-before two months away on a category with a two-year default implies a tea packed some considerable time ago, which for a fragile category is the question worth asking. That inference is rough, since the default period is a guess, and it is better than nothing on a packet that carries no other date. The reliable move remains the one Sprint 31 gives: buy from sellers who state a harvest, and read the durability date as a packaging requirement.
What happens after the date
Nothing in particular, because nothing in particular was happening before it. The drift that a durability date is estimating is continuous, the endpoint is a judgement about acceptability rather than a transition, and a tea a week past its date is indistinguishable from the same tea a week before it. This is not an invitation to disregard the field; it is a description of what the field is. What actually changes a tea is the storage it has had, which is why a well-kept black tea can be entirely serviceable long past its printed date and a badly kept green tea can be tired well before it. The relevant question is always the cup and the leaf: does the dry leaf smell of anything, does it snap, does the wet leaf come alive after the first infusion. There is one respect in which the date remains useful past its expiry, which is as a record. A dated carton is often the only evidence a household has about when a tea entered the house, since almost nobody writes the purchase date on a packet. Keeping the printed packaging with the tea rather than decanting into an anonymous jar preserves that, and it is the same argument the storage pages make about keeping the laminate pouch.
The categories where the field is actively misleading
Three. Matcha, where a durability date printed for an unopened tin says nothing at all about the horizon after opening, which is the number that matters and which is measured in weeks. Deliberately aged teas, where a durability date is a category error — a pu-erh cake with a best-before date has been given a field designed for a product that declines, and some packers print one anyway because a label requires a date. And flavoured teas, whose added aromatics run on a shorter clock than the leaf underneath and whose durability estimate is usually inherited from the base. In each case the printed field is being asked to describe something it was not built for. The general position this catalogue takes is to record durability dates as a fact about the packet rather than as evidence about the tea, and to prefer a harvest or production date wherever one exists. That is not scepticism about labelling, which is doing its job perfectly well for the products it was designed around. It is an observation that tea spans a range from a two-week product to a thirty-year one, and that no single date field describes both ends.
What TeaHQ does not claim here
No jurisdiction’s rules are stated. The definitions of the date categories, which products must carry which, the permitted wording, the exemptions and the responsibilities they place on a business differ between markets and change; TeaHQ has not verified the current requirements anywhere and nothing above should be relied on as a description of the law. Nobody designing a pack, disputing a claim or advising a business should use this page for that purpose, and the applicable regulation is the document to read. What is asserted is only the reading habit and the reasoning behind it: durability dating is a quality estimate made by the packer, the interval reflects a policy applied to a category, and the field therefore describes the business more reliably than it describes the leaf. Nor does this page say anything about whether any tea is fit to drink at any age, which is a question this catalogue does not answer and which `guide-does-tea-expire` already declines. The subject here is what a printed date is as a piece of information, and what a buyer can honestly infer from its presence, its absence and its length.