Your first twelve bottles.

Ask most people to build a first cellar and they produce a list of famous names. It is an understandable instinct and it produces a cupboard that is impressive to look at and useless to live with: twelve wines that all need ten more years, nothing to open on a Thursday and a nagging sense that opening any of them would be a waste.

A working cellar is not a trophy shelf. It is a supply chain with a schedule. Build it around when you will drink and the rest follows.

The three-quarters rule

Split the first twelve into three groups of four.

• Four for now. Ready today, good enough to be worth opening, cheap enough that you will. This is the group that makes the cellar a habit rather than a museum.
• Four for three to five years. Wines with enough structure to improve but which will not punish you for early curiosity. This is where you learn what ageing actually does, because you can open one now and one later and taste the difference.
• Four for ten years and beyond. The wines you are genuinely laying down, drawn from the minority of wines that improve with age. Ideally two bottles each of two wines rather than one each of four, for reasons below.

What goes in each group

These are categories rather than shopping lists, because the right bottle depends on your budget and what is available where you are.

For now: grower Champagne, village white Burgundy or a good Chablis, a Northern Rhône syrah from a lighter appellation, a mature-leaning Rioja reserva. All food-friendly, all rewarding immediately, none of them requiring an occasion.

For three to five: a premier cru Chablis, a serious Barbaresco or a mid-tier Barolo, a Loire cabernet franc from a good year, a right-bank Bordeaux from a lesser vintage. These are the teaching wines.

For ten and beyond: classed growth Bordeaux from a strong year, premier cru red Burgundy, top Northern Rhône or Sauternes, which is both the most undervalued category in fine wine and the one that survives longest.

Buy two of a wine you intend to age, never one. A single bottle can only be opened once and always too early or too late.

Why two bottles, not one

This is the single most useful habit in cellaring. With one bottle you get one guess. With two you open the first at what you think is the early end of the window, learn something concrete and hold the second with real information rather than a critic's estimate.

With six, which is what most serious collectors do with wines they love, you can open one every eighteen months and watch the entire arc. That is the education. No amount of reading substitutes for it. It is also the only reliable way to locate a drinking window.

What to skip at the start

• Very old vintages. Tempting and the category where provenance risk is highest and your ability to judge it is lowest. Come back to these in year three.
• Large formats. Magnums age beautifully and are a nightmare to open casually. One is a pleasure, four is a commitment.
• Anything bought because it appreciated. If the plan is to sell it, that is a different activity with different rules. A cellar you drink and a portfolio you hold are not the same object.
• Filling the last slot. Leave room. The best bottles you buy in year two will be ones you did not know existed in year one.

Do not spend the budget evenly

The instinct is to divide the money twelve ways. Resist it. The three groups do different jobs and deserve different money.

• Drink-now: roughly a quarter of the budget. These are working bottles. If they cost as much as the cellared ones you will hesitate to open them, which defeats the purpose entirely.
• Three to five years: roughly a third. This is where the learning happens, so buy enough quality to be worth studying but not so much that opening one early feels like vandalism.
• Ten years and beyond: the balance. Concentrate it. Two wines at a serious price beat four at a moderate one, because with two bottles each you get to see the arc.

Put differently: if the budget is fixed, the mistake is spreading it thin. Twelve mid-priced bottles from twelve producers teach you almost nothing. Six bottles across three producers, two each, teach you a great deal.

The second twelve

A year in, you will know things you cannot know now: which styles you reach for, which you bought because you thought you should and whether you actually enjoy mature wine or prefer it young. Most people discover at least one uncomfortable truth here, commonly that they like fresher wines than their reading suggested.

Let the second twelve follow the evidence rather than the theory.

• Double down where you were right. If the Northern Rhône bottle was the one you thought about afterwards, buy three more, not one of something else.
• Add depth, not breadth. A vertical of one producer across three vintages is more instructive than three producers in one vintage and more satisfying to own.
• Fill the gap in the calendar. Look at your earliest-drink dates. If nothing comes ready for four years, you have built a cellar you cannot use.
• Buy one thing that scares you. Something older or from a region you do not understand. This is how the palate stops narrowing. A blind tasting is the fastest way to find out which.

The cellar that works is not the one with the best bottles. It is the one where something appropriate is always ready and where opening it does not feel like a decision. The same rule holds as you grow a cellar from fifty bottles to five hundred.

Write it down

Keep a cellar inventory you will actually use: a spreadsheet with wine, vintage, quantity, what you paid, where it is stored and the earliest date you intend to open it. Six columns. It takes ten minutes and it is the difference between a cellar and a pile.

Add a seventh column for tasting notes when you open one. In three years it will be the most valuable document you own about your own palate.