The Hidden Revenue Leak in Most Restaurant Wine Programs

Ask a fine-dining operator about their wine program and they will tell you about the list — the depth, the verticals, the by-the-glass curation. They are proud of it, and often they should be. But the list is the inventory. The program is what happens between a guest sitting down and a bottle being poured, and that is where the money quietly disappears.

The distinction matters because the two are funded, staffed, and judged completely differently. Inventory is a capital decision made once a season by a buyer. The program is an operational decision made a hundred times a night by whoever happens to be standing at the table. Most rooms invest heavily in the first and leave the second to chance — and then wonder why a cellar they are proud of produces a beverage number they are not.

Where the money actually leaves the room

Here is the pattern we see most, across markets and price points:

A guest orders a serious bottle. No sommelier visits the table. The single moment most likely to produce a trade-up, a pairing, or a second bottle simply does not occur.
The tasting menu is offered. By-the-glass pairings are not. A guest who would happily have spent another $60 is never given the chance.
A wine that plainly wants decanting is poured straight. The ritual that justifies the price — and signals the room's seriousness — is skipped.

None of these is a failure of the list. Each is a failure of service design. And each is measurable in dollars per cover.

The cellar is the asset. The pour is the business.

Consider the specimen finding from a recent evaluation of a landmark, wine-forward dining room. The guest ordered a four-figure bottle. It was opened tableside, competently, by the server — and poured without a decant, without a sommelier ever appearing, over a three-hour dinner built entirely around wine. The list scored well. The program scored 63. That gap is not a matter of taste; it is a matter of money, and it repeats every night the same choreography is left undesigned.

The four leaks, priced

When we run a wine-program audit, we do not editorialize about the leak — we quantify it, line by line, per cover. Four categories account for most of it.

The unoffered by-the-glass pairing. A tasting menu without a pairing offer is a menu selling itself at half price. The guest who ordered the nine courses has already signaled they are here for the experience; not offering the wine that completes it is the clearest revenue miss in the room. In a strong house we typically price this at $38 to $62 a cover — before you count the guests who would have said yes and simply were never asked.

The missed sommelier-led trade-up. A guest reaching for a bottle in the middle of the list is often one informed sentence away from the bottle just above it. That sentence is worth $75 to $140 on a single order, and it is only ever spoken by someone who came to the table. No touch, no trade-up.

The skipped ritual. Decanting, correct glassware, service temperature — these are not affectations. They are the visible proof that the price on the list is real. A wine poured warm and straight from a cold cellar tells the table the room does not take its own list seriously, and a table that senses that does not order the second bottle.

The unconverted occasion. The anniversary, the deal closing, the birthday noted in the reservation and never acknowledged — each is a wine moment that a program designed to notice would have converted into a memorable bottle and a returning guest. Its value is not a number on tonight's check; it is lifetime value, walking out the door unrecognized.

Add these across a full book of covers and the figure is rarely small — often $35 to $60 a cover, every night, compounding across a year into a number that would comfortably fund the very sommelier position whose absence caused the leak in the first place.

Why it breaks — and it is never the list

Operators reach instinctively for the list when the beverage number disappoints: buy deeper, add verticals, chase allocations. It almost never helps, because the list was not the problem. The program breaks for three structural reasons, and each has a structural fix.

It depends on a person, not a standard. The room pours beautifully on the nights the sommelier is on the floor and poorly on the nights they are not. That is not a wine problem; it is the absence of a rule that survives any given shift.

The reservation's intelligence never reaches the table. The system knows a guest ordered top-of-list last visit, or flagged a celebration this time. The server does not, because nothing carries that fact from the screen to the floor.

Wine service is treated as a personality, not a discipline. "Our somm is great with guests" is a description of one employee, not a description of a program. When that employee leaves, the margin leaves with them.

The fix is a standard, not a bigger cellar

The remedy is not more inventory. It is a small number of non-negotiable rules, written into the service sequence and drilled until they hold under a full room:

A sommelier-touch threshold. Above a set bottle price, a sommelier must touch the table — not as a courtesy, but as a standard with a name and an owner. This single rule recovers most of the trade-up leak on its own.
A scripted by-the-glass offer with every tasting menu. Not improvised, not left to the server's mood — a specific offer, made the same way every time, so the guest is always asked.
A decant decision built into the sequence. For every bottle above a threshold, the standard requires a yes-or-no decant decision, made out loud, at the table.
A pre-shift recognition brief drawn from the reservation. Two minutes before service, the floor learns which tables carry a flagged occasion, a returning guest, or a notable prior order — so the wine moment is anticipated rather than missed.

None of this requires a new hire or a deeper list. It requires deciding that the program is a system to be run rather than a talent to be hoped for.

What to measure this quarter

An operator who wants to find the leak before we do can start with four questions, asked honestly of last week's covers:

On what percentage of bottles above your threshold did a sommelier actually touch the table? Most rooms overestimate this by half.
What is your by-the-glass attachment rate on tasting menus — and is anyone accountable for the offer being made?
How many decants happened last week, and were any of them decisions rather than requests?
Of the reservations that flagged an occasion, how many produced a wine that marked it?

Inventory is bought once. The program is run every service — and most, quietly, are not run at all. The list will not tell you where the money is going. The pour will.