Keeping a cellar inventory that stays accurate
Published September 13, 2026 · 3 min read
The fields that earn their place, the ones that clutter, and the three moments where an inventory drifts away from reality.
A cellar inventory has one purpose: answering two questions without going down to the cellar. What do I have? What should I be drinking? Anything that does not serve those two questions is padding.
The fields that earn their place
Producer and cuvée. This is the wine's identity. The estate name alone is not enough: one estate often makes five cuvées with different prices and different ageing.
Vintage. Essential, and the field most often entered wrongly, because it reads poorly on older labels.
Quantity. The only value that changes constantly, and therefore the only one that drifts.
Location. A rack or row identifier. Without it, a two-hundred-bottle inventory tells you that you own the wine, not where it is.
Drinking window. Two years, start and end. This is what turns a list into a decision tool.
The fields that clutter
Detailed tasting notes, label photographs, price history, provenance, lot numbers: all legitimate for a collector, useless for almost everyone else. Every optional field adds friction to data entry, and entry friction is the number one cause of abandoned inventories.
Start with five fields. Add the sixth the day its absence genuinely got in your way.
The three moments where an inventory drifts
1. At purchase, when entry is put off. Six bottles brought back from a weekend, put in the cellar "for now", and never recorded. This is the most common case. The fix is to enter before shelving, not after.
2. At opening, when the decrement is forgotten. You pull the cork at the table, not in front of a screen. This is structural: the moment of drinking is the worst possible moment for data entry. The only workable fix is for "I drank a bottle" to be a single tap on a phone, with no form.
3. Gifts and loans. Bottles that leave without being drunk leave no trace. They need their own exit category, otherwise the discrepancy surfaces a year later with no explanation.
The test that tells you whether your inventory is alive
Go down to the cellar, pick three bottles at random, and check they are in the inventory at the right location and quantity. Then take three rows at random from the inventory and check the bottles exist.
An inventory that passes in both directions is usable. If it fails in only one direction, you already know which of the three moments above is to blame.
What to remember
- Five fields are enough: producer, cuvée, vintage, quantity, location.
- The drinking window is the sixth, and it is what makes the inventory useful.
- Entry friction kills inventories more reliably than forgetfulness.
- Drift comes from opening bottles and giving them away, not from counting.
Frequently asked questions
- Which fields are genuinely required?
- Five are enough for an inventory to be useful: producer, cuvée name, vintage, quantity and location. Everything else — appellation, grape, price, score — is nice to have but does not drive daily use, which is finding a bottle and knowing how many are left.
- Should you record the purchase price?
- Yes, but for a reason rarely given: it tells you what you actually buy, by region and price band. Valuing the cellar is secondary and often misleading, since a wine is worth what it is worth only if you sell it, which almost nobody does.
- How often should you physically recount?
- Once a year is enough if bottles are logged out at the moment they leave. If they are not, no frequency closes the gap: the problem is not the count, it is the entry made when the bottle is opened.