Chapter 16: The Wine Program: Buying, Listing, Pricing, Storing, and Actually Selling Wine

"Every other thing in this building gets sold, eaten, and replaced inside a week. The wine just sits there being worth money." — constructed; a beverage director explaining inventory to a chef

Overview

Walk into the storage room of almost any independent restaurant and you will find, on a rack in the back corner, somewhere between two and eight thousand dollars of somebody's money doing absolutely nothing. It has been there for months. Some of it has been there since the week before opening. The owner bought it because a distributor rep was persuasive on a Tuesday in March, or because it seemed like a list ought to have a Barolo on it, or because the case price was better than the six-bottle price and the arithmetic on that seemed obvious at the time.

Wine is the only inventory in a restaurant that behaves like an investment portfolio. Your produce turns every two or three days. Your proteins turn weekly. Your dry goods turn monthly. Your wine, if you buy it the way most people buy it, turns eleven or twelve times a year — and within that average there is a shelf that turns twenty-six times and a shelf that turns fewer than three. Those two shelves sit four feet apart, look identical from across the room, and have nothing whatsoever in common as financial objects.

That is the first half of this chapter. The second half is harder and matters more.

A wine program's margin does not exist until somebody sells it. A cook produces the food whether or not anybody is charming; wine sits on a list and is chosen, or isn't, by a guest who is frequently nervous. Most people ordering wine in a restaurant are quietly worried about three things at once: mispronouncing something, spending more than they meant to, and being silently judged by a stranger in an apron. Every one of those fears suppresses spending, and every one of them is solvable by design — by how the list is organized, by what the server is trained to say, and by whether there is genuinely good wine at the bottom of the ladder.

Bellwether has no sommelier. A sixty-eight-seat neighborhood restaurant cannot carry one and should not pretend otherwise. That fact is not a limitation this chapter apologizes for. It is the design constraint that makes everything else in the chapter necessary: if there is no expert on the floor, then the list itself and the training program have to do the expert's job, and they have to do it with people who are also carrying plates.

In this chapter, you will learn to:

• State what a wine list is for in a specific restaurant, and size it to the menu, the room, and the check average rather than to the buyer's enthusiasm.
• Describe the three-tier system, its post-Prohibition origins, and the four concrete ways it constrains a small independent buyer.
• Build a bottle-price ladder with declining multiples, and explain — with worked arithmetic — why the target-cost pricing you learned in Chapter 11 fails at both ends of a wine list.
• Compute by-the-glass yield and per-glass cost, and quantify what the unsold fifth glass costs.
• Calculate the capital tied up in a list, its turns by price band, and what a slow bottle costs in money over time.
• Train service staff to sell wine in a restaurant with no wine expert on the floor, and put a dollar value on doing it well.
• Set a corkage fee, a pruning rule, and an exit plan for a bottle that will not move.

Learning Paths

🏗️ Opening — §16.6 is your chapter. The capital and storage decisions get made before you open and are expensive to reverse. Do the par-level arithmetic before you write a single purchase order. 📋 Managing — weight §16.5, §16.7, and §16.8. The open-bottle log, the staff tasting, and the ninety-day sales report are three Monday-morning habits that carry the entire program. 🍸 Beverage — all of it, and pair it with Chapter 15. The wine half of the beverage line has a different cost structure, a different capital profile, and a different sales problem from the bar. 🚚 Small Format — §16.2 and §16.4 scale down cleanly. A twelve-selection list on a truck or in a forty-seat room obeys the same ladder; you simply have fewer rungs and less room to be wrong.

16.1 What a wine list is for: this restaurant, this menu, this guest

Before you buy a bottle, answer a question that almost nobody answers: what is this list obligated to earn?

Most wine lists get built the other way around. Somebody who likes wine writes down wines they like, prices them at a multiple somebody mentioned once, and prints it. Two years later the list has seventy selections, the storage room is full, and nobody can say whether the program makes money, because "the program" was never a financial object. It was a hobby with a menu attached.

A list is a financial object. It occupies capital, consumes square footage you paid rent on, requires glassware and training and a printing budget, exposes you to spoilage, and produces a specific number of gross-profit dollars per year. Start with that number.

The revenue the plan hands you

Chapter 15 established the beverage line and its blended cost. Beverage is 28% of Bellwether's plan revenue — \$434,000 on \$1,550,000 — at a blended pour cost of 22%. That single line splits three ways, and the split is the first real decision in this chapter.

🧮 Run the Numbers

Splitting \$434,000 three ways, and why wine costs more.

Category	Revenue	Share of beverage	Cost %	COGS
Wine	\$164,920** | **38.0%** | **28.0%** | **\$46,180
Spirits and cocktails (Ch. 15)	\$173,600 | 40.0% | 17.0% | \$29,500
Beer — draft and package (Ch. 15)	\$69,440 | 16.0% | 24.2% | \$16,800
Non-alcoholic (Ch. 15)	\$26,040 | 6.0% | 11.5% | \$3,000
Total beverage	\$434,000** | **100%** | **22.0%** | **\$95,480

Wine carries thirty-eight cents of every beverage dollar at Bellwether — a little over a third, which is right for a hearth-driven American menu in a room that also has a real cocktail list. It does not carry the bar's cost percentage. Wine runs 28% against the bar's 18%, and the blend of the three lands at the plan's 22%.

That ten-point gap is not a failure of purchasing. It is structural, and you should understand it before you try to fix it.

A cocktail sells for \$14 with about \$2.52 of liquor in it — an 18% pour cost and \$11.48 of gross profit. A five-ounce glass of wine sells for \$14 with \$2.80 of wine in it — a 20% pour cost and \$11.20 of gross profit. Twenty-eight cents apart. But the cocktail required a bartender to juice citrus, make syrup, batch, shake, strain, garnish, and wash the tin. The glass of wine required somebody to pull a cork and pour.

Wine's percentage is worse and its dollars per minute of labor are far better. That is the whole trade, and it is why a category running 28% belongs in a business chasing a 60% prime cost.

Against total sales, wine COGS of \$46,180 on \$1,550,000 is 3.0 points of the plan's 27.8% blended cost of goods — about a ninth of everything Bellwether spends on product.

Now put it on the check. Chapter 11's check build is \$33.12 food plus \$12.88 beverage = \$46.00 per dinner guest. Split the beverage side the same three ways:

Four dollars and ninety cents. That is what a wine program is at Bellwether — a hair under five dollars a head, or about one glass of wine for every third guest through the door. Every argument in the rest of this chapter is about moving that number, protecting it, or refusing to spend capital that does not move it.

Hold onto the leverage, because it is enormous. Fifty cents a guest — one additional glass sold for every ten covers — is a 10.2% lift on \$164,920, or \$16,830 of incremental wine revenue. At a 72% gross margin that is \$12,118 of gross profit on a plan whose entire operating profit before debt service is around \$260,000. One glass per ten guests is roughly five percent of the year's profit.

The three jobs a list actually does

With the number in hand, you can say what the list has to do. Three jobs, in order of the money they produce:

One: raise the check. This is the big one and the reason wine exists on the plan. A guest who adds a glass of wine to a \$29 Hearth Chicken has moved that table's contribution by more than half — we work that arithmetic in §16.7 — and has done it without adding a single second of kitchen labor or a square inch of hearth space. Wine is the cheapest incremental revenue in the building.

Two: complete the menu. A hearth menu — smoke, char, fat, acid, roasted vegetables, whole fish, half a chicken with salsa verde — has specific needs. It needs bright, high-acid whites for the brined and the pickled. It needs light, savory reds that do not fight char. It needs one structured red for the steak and the lamb, and it needs something with bubbles for brunch and for the table celebrating something. A list that cannot do those five things is not serving this restaurant, no matter how interesting it is.

Three: say what kind of place this is. The list is read before anything is tasted. A list with nothing under \$40 says one thing. A list with nothing over \$60 says another. A list organized by appellation says we assume you know what an appellation is. That signal is free, it reaches every guest who opens the folder, and most restaurants send it by accident.

🤝 Hospitality

The list is a document about the guest, not about you.

Here is a test I have used on every list I have ever built, and it has never once failed to find a problem.

Hand the list to somebody who does not care about wine. A line cook, a friend from another industry, your accountant. Say nothing. Watch what they do.

If they scan it top to bottom, stop, and hand it back with a slightly apologetic face — your list has failed. If their eye lands somewhere, stays, and they say "what's this one like?" — your list is working.

The failure mode is almost always the same. The list is organized by a category the guest does not possess. Grape varieties are a category. Regions are a category. Producer alphabetization is a category. All three assume the reader already has a map. Most readers don't, and a person handed a document they cannot navigate does the safe thing, which is order a beer.

What a guest reliably knows is what they want the wine to do. Light or heavy. Crisp or round. Something for the fish or something for the steak. Those are categories every adult already owns, and a list built on them is legible to everyone — including the guest who does know grapes, who will find the grape printed right there in the second line.

This is not dumbing anything down. It is putting the information the guest has at the top of the hierarchy and the information you have underneath it. §16.7 builds the whole program on this idea.

Wine list architecture is the deliberate structure of a list: how many selections, how they are grouped, how the prices ladder from bottom to top, which of them are also available by the glass, and what each section is for. It is a design decision made once and revised quarterly — not the residue of two years of purchasing.

16.2 List architecture: size, coverage, price ladder, and the by-the-glass core

Why forty, and not ninety, and not fifteen

Bellwether's list is forty selections. That number is in the business plan, and it deserves a defense, because a buyer's instinct is always to add.

Forty is what you get when you solve three constraints simultaneously.

Coverage. The menu needs five or six distinct things from wine (§16.1). Each of those needs at least two options at different prices, because a guest who wants a light red and finds exactly one light red has been given a choice between yes and no. Two options per need across six needs is twelve; you want depth in the categories that actually sell, which for an American hearth menu means the reds. Call it thirty-five as a floor for genuine coverage.

Memory. Nobody on this floor is a sommelier. A server working four shifts a week has to be able to hold the list in their head well enough to answer "what's the Syrah like?" without leaving the table. Twelve by-the-glass wines can be memorized in a shift. Forty selections can be navigated with a list in hand and one round of tasting. Ninety cannot, by anyone who is also running food.

Capital and space. §16.6 does this arithmetic properly, but the headline is that forty selections at Bellwether's depth is 207 bottles and \$4,068, and it fits — barely — into a 46-bottle wine refrigerator, two under-bar drawers, and a 180-bottle wall rack in a storage room that is also holding paper goods. Ninety selections is roughly 380 bottles and about \$8,000, and there is physically nowhere to put it.

Forty is not a magic number. It is the number where coverage, memory, capital, and square footage intersect for this restaurant. Do the same three-constraint solve for yours and you may land at eighteen or at sixty-five.

The list at a glance

Read the section names again. Not "Chardonnay," not "Burgundy," not "France / Italy / Domestic" — Crisp & Bright, Round & Textured, Light & Savory, Dark & Structured. Each entry then carries its grape and its origin on the second line, in smaller type, for the guest who wants it. The guest who does not want it never has to engage with it.

That is the single highest-leverage decision in this entire chapter, it costs nothing, and it is available to any restaurant that is willing to reprint a page.

The by-the-glass core

By-the-glass (BTG) wine is wine sold by the individual pour rather than by the bottle. At Bellwether, twelve of the forty selections are poured by the glass; all twelve are also available as full bottles.

Those twelve selections carry 60% of wine revenue — \$99,000 of the \$164,920 — from a fifth of the list. That concentration is not an accident of demand. It is the design.

Here is the logic. In a restaurant with no wine expert on the floor, the program has to be operable by a server who has worked here six weeks. Twelve wines is a memorizable set. Twelve wines can all be tasted in a single forty-five-minute pre-shift. Twelve wines fit on one index card in the server station. Twelve wines means every server can answer every by-the-glass question at the table, immediately, without hedging — and the confidence that produces is worth more in sales than any amount of curatorial cleverness in the other twenty-eight.

The remaining twenty-eight bottle-only selections carry \$66,000, and they carry it at a much worse cost percentage and a far worse turn rate. They are still worth having, for reasons §16.8 gets into, but understand the structure clearly: the by-the-glass core is the business. The bottle list is the range.

👨‍🍳 On the Line

The list is a printed document in a building full of liquid.

Practical things nobody tells you until you have already spent money.

Print it yourself, in house, on decent stock. The moment you send a wine list to a commercial printer in quantity, you have committed to that list for as long as the box lasts, which means you will keep selling a wine you no longer stock because reprinting feels wasteful. Print twenty at a time on a laser printer. Reprint when something changes. The marginal cost of being current is about forty cents.

Assume every list will be destroyed. Wine lists get set down in water rings, handled by people who have been eating with their hands, and dropped. Budget replacement at roughly a third of the stack per month and stop being annoyed about it.

Date the list in the footer, in small type. "Rev. 10/14." When a guest says "you had a Muscadet last time," the manager can tell in two seconds whether the guest is right and which list they saw.

The 86 problem is worse for wine than for food. When the kitchen 86s the halibut, one server tells the floor at pre-shift and everyone adapts. When you sell the last bottle of something at 8:40 on a Friday, the server who sold it knows and nobody else does — and the next table orders it forty minutes later, is told no, and has now had a small bad experience over a document you printed. Fix: a whiteboard in the service station, one line per 86'd wine, updated by whoever pulled the last bottle, checked by everyone on the way out of the kitchen. It is a fifteen-second habit that prevents about two hundred small disappointments a year.

16.3 The three-tier system and how it shapes what you can buy and what you pay

You cannot buy wine from a winery. In most of the United States, most of the time, that transaction is illegal, and the structure that makes it illegal shapes your list more than your taste does.

The three-tier system is the post-Prohibition regulatory architecture that separates alcohol producers, wholesalers, and retailers into three legally distinct tiers and, as a general rule, requires product to pass through all three in order. A producer sells to a licensed wholesaler; the wholesaler sells to a licensed retailer or on-premise licensee; the retailer sells to the public. Crossing a tier — a winery selling straight to a restaurant, a wholesaler selling to a consumer — is prohibited except where a state has carved out a specific exception.

FIGURE 16.1 — THE THREE-TIER SYSTEM                     [structural; details vary by state]

    TIER ONE                  TIER TWO                      TIER THREE
    producers &        →      wholesalers /          →      retailers &
    importers                 distributors                  on-premise licensees
  ┌────────────────┐       ┌────────────────────┐        ┌──────────────────────┐
  │ wineries       │       │ licensed in YOUR   │        │ liquor stores        │
  │ importers      │       │ state; frequently  │        │ bars, RESTAURANTS    │
  │ (domestic and  │       │ the ONLY legal     │        │ ─── Bellwether ───   │
  │  foreign)      │       │ route to you       │        │                      │
  └────────────────┘       └────────────────────┘        └──────────────────────┘
          │                          │                             │
          │  generally may NOT       │  generally may NOT          │  generally may
          │  sell direct to tier 3   │  sell to the public         │  buy ONLY from
          │                          │                             │  tier 2
          └──────────── X ───────────┴─────────── X ────────────────┘

  WHY IT EXISTS
  Prohibition ended in 1933 with the 21st Amendment, which handed the states broad
  authority over alcohol inside their own borders. Most states used it to dismantle
  the pre-Prohibition "tied house" — the arrangement in which a brewer or distiller
  owned, financed, or controlled the saloon that sold its product, and pushed volume
  through it. The remedy was to force an independent, separately licensed party to
  stand between the maker and the seller.

  WHAT THAT MEANS FOR YOU
  Fifty states built fifty versions. Some permit limited winery self-distribution.
  Some are "control" jurisdictions where the state itself occupies part of tier two,
  tier three, or both — most often for spirits, occasionally for wine. Assume nothing.
  Ask your state's alcoholic beverage control authority, in writing, before you plan
  a program around an assumption.

Chapter 8 established that Bellwether sits in an open-issuance liquor jurisdiction — its license is obtainable for a fee rather than bought at auction from a quota holder — and Chapter 8 owns the licensing process. What Chapter 8 does not tell you is how the tier structure feels from behind a sixty-eight-seat buyer's desk. Four things, and every one of them is a constraint on the list you just designed in §16.2.

One: exclusivity decides your list more than you do. In most states, a given wine is represented by exactly one wholesaler. If that wholesaler does not call on accounts your size, that wine is not available to you at any price. Not expensive — unavailable. A buyer who designs a list first and sources it second will spend three weeks discovering this. Design the list as a set of roles (§16.2's seven sections) and fill each role from what your actual distributors carry.

Two: you may not be able to negotiate your price. Many states require wholesalers to post their prices with the state, often monthly, and to sell to every licensee at the posted price. Where that is the law, there is no such thing as your special deal, and a rep who implies otherwise is either confused or describing something you should not participate in. Your leverage is which posted item you choose and when — posted monthly specials are real, legal, and where the actual savings live.

Three: quantity discounts exist in some states and not others. Where they are permitted, they are usually structured per case or per invoice. Where they are restricted, a five-case order costs exactly five times a one-case order. Know which regime you are in before you build a purchasing strategy around volume, and read §16.6 before you build one anyway.

Four: alcohol may be the one category you cannot float. Several states impose credit restrictions on alcohol sales to licensees — cash on delivery, or a short mandated term, with a published delinquency list that can cut a late-paying licensee off from every wholesaler in the state at once. This is a cash-flow fact before it is a purchasing fact. If your produce vendor gives you thirty days and your wine distributor gives you seven, then wine is disproportionately expensive in the currency that actually kills restaurants, which is timing.

⚖️ Code and Compliance

Questions to put to your state's alcoholic beverage authority before you buy anything.

All of the below vary by state, and several vary by county or city. Verify locally, in writing, and use an attorney for anything you are betting money on.

1. May a licensee purchase wine from anyone other than a licensed in-state wholesaler? Some states permit limited direct purchase from small wineries or limited self-distribution.
2. Is wholesale pricing posted, and if so, on what cycle? This determines whether "negotiating" is a real activity in your market.
3. Are quantity or cumulative-volume discounts permitted? And if so, are they per invoice, per case, or per period?
4. What are the mandated credit terms for alcohol purchases, and is there a delinquency list? Ask what triggers listing and how removal works.
5. May a wholesaler provide samples, glassware, staff training, or list-printing support? "Inducement" rules are strict, vary widely, and the licensee — you — is usually the one who gets penalized. A rep offering you something free is not the person to ask whether it is legal.
6. Is BYO / corkage permitted at a licensed premises? See §16.8. In some states it is flatly prohibited; in others it requires a specific endorsement.
7. What are the recordkeeping and invoice-retention requirements? Alcohol invoices are frequently subject to a longer retention period than your other records, and an ABC audit that finds a gap is a bad day.

None of this is exotic. It is the ordinary homework of holding a license, and Chapter 8's compliance calendar is where it goes.

Working the distributor relationship

Practical, and it matters more than any of the above at a sixty-eight-seat account.

Consolidate to two or three houses. A buyer spread across seven distributors gets seven small, low-priority accounts, seven delivery days, seven invoices to reconcile, and no leverage anywhere. A buyer who does real volume with two houses gets return calls, gets shown the allocations, and gets told when a posted special is coming.

Take the appointment, on your schedule. Set one morning a month, ninety minutes, tasting at the bar with the FOH partner present. A rep who has a fixed slot will bring the right things. A rep who has to ambush you at four o'clock on a Friday will bring whatever they need to move.

Taste everything before you buy it, and taste it against the food. A wine that is delicious alone and dies next to salsa verde has no job here. Bring a plate to the tasting. It takes six minutes and prevents a case of dead inventory.

Never buy on the first call. The pressure to commit at the table is the whole method. "Send me the posted price and I will put it on the order Thursday" costs nothing, and about a third of the time you will not put it on the order Thursday, which is the point.

16.4 Pricing wine: multiples, laddering, and why straight multiples fail at both ends

Chapter 11 taught you target-cost pricing: take the plate cost, divide by the target food cost percentage, and you have a price.

$$\text{Menu price} = \frac{\text{plate cost}}{\text{target cost \%}}$$

It also taught you when to ignore it. Wine is the purest case in this book of a category where cost-percentage pricing fails at both ends of the range simultaneously. Not "needs adjustment" — fails, in opposite directions, on the same list, on the same page.

Watch it happen.

The two vocabularies

First, two terms that get used interchangeably and are not the same thing.

Wine markup is the multiple applied to the wholesale bottle cost to reach the menu price: a \$14 bottle at a 3× markup is priced at \$42.

Wine cost percentage is the category's product cost divided by its sales — wine COGS ÷ wine sales. Bellwether's plan target is 28%.

They are reciprocals of one another on any single bottle, and the conversion is worth memorizing because distributors and operators speak in multiples while accountants and P&Ls speak in percentages.

Where the straight multiple breaks

At the bottom. Take a wine that costs you \$8 wholesale. A flat 3× puts it on the list at \$24 and produces \$16 of gross profit. Sixteen dollars has to cover the glassware it is poured into, the labor to receive it, store it, carry it, open it, pour it, clear it and wash the glass, the roughly three percent the card processor takes, the breakage, and the capital that sat in the bottle. It is not nothing — but it is the least profitable transaction on your list, and because it is the cheapest bottle, it is also one of the highest-volume ones. Your highest-volume price point earns your fewest dollars, which is exactly backwards.

Push that bottle to 3.5× and it lists at \$28 with **\$20 of gross profit — twenty-five percent more margin — and not one guest in a hundred experiences \$28 as meaningfully different from \$24. At the bottom of a list, price sensitivity is far lower than buyers assume, because the guest choosing the cheapest bottle is choosing a position, not a number.

At the top. Now take a wine that costs \$80 wholesale. A flat 3× puts it on the list at **\$240.**

In a neighborhood restaurant with a \$46 check average, that bottle will not sell. Not rarely — never. A guest who knows enough to want that wine knows roughly what it costs in a shop, and at three times wholesale they will order the \$65 bottle instead, or a round of cocktails, or nothing. The bottle sits.

And here is the part that breaks the percentage framework entirely: a bottle that never sells does not have a cost percentage. It has zero sales and zero cost of goods. It contributes nothing to your 28%. What it actually has is eighty dollars of your money, immobilized, aging in a room that runs seventy-six degrees in August, for as long as you own it.

Price the same bottle at \$135 — cost plus fifty-five dollars, a 59.3% cost percentage that would horrify anyone trained on food cost — and it sells a few times a year and earns \$55 of gross profit each time. Fifty-five dollars is nearly triple what your \$28 bottle earns.

You bank contribution margin, not cost percentage. This book has been saying that since Chapter 11. Wine is where the sentence becomes undeniable.

The ladder

Bottle-price laddering is the practice of applying a markup multiple that declines as wholesale cost rises, with a floor on gross-profit dollars at the bottom and a dollar-margin rule at the top. The multiple falls the whole way up the list; the gross-profit dollars rise the whole way up. Both facts are the point.

FIGURE 16.2 — THE PRICE LADDER: the multiple falls, the dollars rise    [the Bellwether plan]

  wholesale   markup      menu     gross profit    cost %    GP dollars
   cost       applied     price    dollars                   (bar length)
  ──────────────────────────────────────────────────────────────────────────────
  $  8.00     3.5x        $  28    $20             28.6%     █████
  $ 12.00     3.2x        $  38    $26             31.6%     ███████
  $ 18.00     2.8x        $  50    $32             36.0%     ████████
  $ 26.00     2.5x        $  65    $39             40.0%     ██████████
  $ 38.00     2.2x        $  84    $46             45.2%     ████████████
  $ 55.00     1.9x        $ 105    $50             52.4%     █████████████
  $ 80.00     cost+$55    $ 135    $55             59.3%     ██████████████
  ──────────────────────────────────────────────────────────────────────────────

  THE RULE, in four lines:
    under $10 wholesale ......... 3.5x, with a $20 gross-profit floor
    $10-$22 wholesale ........... 3.2x falling to 2.8x
    $22-$45 wholesale ........... 2.5x falling to 2.2x
    over $45 wholesale .......... 1.9x, or cost + $55, whichever is LOWER

  A flat 3x would price the $8 bottle at $24 (giving up $4 of margin nobody
  would have noticed) and the $80 bottle at $240 (giving up the entire sale).
  The ladder gives up neither.

  ROUNDING: to the nearest whole dollar, and never to .99. A wine list with
  $37.99 on it reads like a supermarket shelf tag and undoes the work.

Every figure in that ladder resolves. \$8 × 3.5 = \$28.00, gross profit \$20, cost 8 ÷ 28 = 28.6%. \$18 × 2.8 = \$50.40, rounded to \$50, cost 18 ÷ 50 = 36.0%. \$38 × 2.2 = \$83.60, rounded to \$84, cost 38 ÷ 84 = 45.2%. \$80 + \$55 = \$135, cost 80 ÷ 135 = 59.3%. Check them.

🧮 Run the Numbers

Three pricing methods on the same seven bottles.

Wholesale	Flat 3×	Cost + \$32 flat	The ladder
\$8 | \$24 (GP \$16)* | \$40 (GP \$32)* | **\$28 (GP \$20)
\$12 | \$36 (GP \$24)* | \$44 (GP \$32)* | **\$38 (GP \$26)
\$18 | \$54 (GP \$36)* | \$50 (GP \$32)* | **\$50 (GP \$32)
\$26 | \$78 (GP \$52)* | \$58 (GP \$32)* | **\$65 (GP \$39)
\$38 | \$114 (GP \$76)* | \$70 (GP \$32)* | **\$84 (GP \$46)
\$55 | \$165 (GP \$110)* | \$87 (GP \$32)* | **\$105 (GP \$50)
\$80 | \$240 (GP \$160)* | \$112 (GP \$32)* | **\$135 (GP \$55)

The flat multiple looks best on paper and produces the highest gross profit per bottle at every tier. It is also a fantasy above about \$60, because the top four rows will not sell at those prices in this room. A \$160 gross profit realized zero times is \$0.

Cost-plus-a-fixed-dollar-margin fixes the top beautifully — \$112 for that \$80 bottle is a price a guest will actually pay — and destroys the bottom. Forty dollars for an \$8 bottle prices your entry point above your mid-list, inverts the ladder, and tells every price-conscious guest that this restaurant is not for them. It also produces the same \$32 whether you sold a \$28 bottle or a \$135 bottle, which means your buyer has no incentive ever to trade a guest up.

The ladder takes the top from cost-plus and the bottom from the multiple. Gross profit rises monotonically — \$20, \$26, \$32, \$39, \$46, \$50, \$55 — so every trade-up is worth more to the house, which is exactly the incentive you want a server to have. And no price on the list is one a guest in this room would refuse on principle.

What the ladder produces at the category level

Run the ladder across the whole list and the by-the-glass program (§16.5) and Bellwether's wine cost lands at 28%, split like this:

And the bottle side breaks down by shelf:

Look at the shape of that. The cost percentage gets worse the further up the list you go — 30%, 35%, 44%, 52.8% — and that is the ladder working exactly as designed. An operator who tried to hold 30% across the whole list would have priced the top shelf at \$260 and sold nothing off it.

⚠️ Where the Money Leaks

The second-cheapest bottle.

Restaurant people have observed for as long as there have been wine lists that a large share of bottle orders land not on the cheapest wine but on the second-cheapest — the guest who does not want to overspend but also does not want to be seen ordering the bottom of the list. This is trade lore rather than a peer-reviewed finding, and you should treat it as a pattern to watch for in your own POS data rather than a law. But watch for it, because in most rooms it is there.

Now consider what most lists put in that position. It is almost always the bottle the buyer cared about least: an anonymous filler wine bought on price to make the list "have a range."

You have placed your worst wine in one of your highest-volume slots. The guest who orders it has a mediocre bottle, concludes the wine here is not very good, and orders a cocktail next visit — a guest lost through a purchasing decision the buyer made in four seconds.

The fix costs nothing. Put something you are genuinely proud of second from the bottom, and be willing to take a slightly worse cost percentage there. A \$15 wholesale bottle at \$48 runs a 31.3% cost — three points above target — and it is the best thirty-one percent you will spend all year, because that slot is a recruiting tool for everything above it.

Then sort the list so the eye starts at the bottom of each section. Guests read wine lists from the price column inward. Design for how they actually read, not how you wish they did.

16.5 By-the-glass economics: yield per bottle, oxidation, preservation, and waste

By-the-glass is where sixty percent of Bellwether's wine revenue lives and where nearly all of its waste happens. It also has the most unforgiving arithmetic in the chapter, because a bottle you open is a bottle you have already paid for whether or not anyone drinks the rest of it.

The pour that divides the bottle

Chapter 15 set the house standard pour for spirits. Wine needs its own, and the number is chosen for a different reason: pick a pour size that divides the bottle.

A 750 milliliter bottle holds about 25.4 fluid ounces.

$$\text{BTG yield} = \frac{\text{bottle volume in ounces}}{\text{pour size in ounces}}$$

BTG yield is the number of saleable glasses a bottle produces at your house pour. At Bellwether the pour is five ounces, which yields five glasses and leaves about four-tenths of an ounce of slack for spill, the drip on the neck, and the fact that no human pours identically twice.

Now watch what a six-ounce pour does. Six into 25.4 is 4.23 — four glasses, with 1.4 ounces stranded in the bottom of the bottle. You either pour a visibly short fifth glass, which the guest notices and resents, or you pour it out. Across the roughly 1,650 by-the-glass bottles Bellwether opens in a year, 1.4 ounces each is 2,310 ounces — ninety-one bottles of wine, about \$1,200 at Bellwether's average by-the-glass cost — poured down a drain purely because of a pour-size decision.

Six ounces is a perfectly defensible pour in a room that wants it. Just price it knowing it costs four glasses to the bottle, not five, and never let the two arithmetics get mixed up.

The twelve, costed

One bottle of each of the twelve costs \$160** and produces **\$841 of revenue at perfect yield — 19.0%. Weighted by what actually sells, Bellwether's plan carries a theoretical by-the-glass cost of 19.6%, because the wines that move fastest are not quite the cheapest ones.

Notice the pricing convention embedded in that table: for most of these, the glass price is approximately the wholesale bottle cost. A \$14 bottle pours at \$14 a glass. That old rule of thumb works because at a five-glass yield it lands you at a 20% pour cost automatically — the first glass pays for the bottle and the other four are gross profit.

Notice also where the rule is broken, deliberately. The \$9 and \$10 and \$11 bottles are priced at \$11 and \$12, not \$9 and \$10, because an \$9 glass of wine at a restaurant with a \$46 check reads as cheap in a way that costs you more than the two dollars gains you. And nothing is poured above \$17, because a \$25 glass of wine in this room sells twice a week and then oxidizes — which brings us to the central problem.

FIGURE 16.3 — THE LAST GLASS: what one bottle earns, by glasses sold   [the Bellwether plan]
  One bottle. $14.00 wholesale. Five 5-ounce pours at $14.00 a glass.

  glasses    revenue    wine cost      POUR      gross     gross profit
   sold      earned     (always $14)   COST      profit
  ────────────────────────────────────────────────────────────────────────
    1        $ 14       $ 14.00        100.0%    $  0
    2        $ 28       $ 14.00         50.0%    $ 14      █████
    3        $ 42       $ 14.00         33.3%    $ 28      ██████████
    4        $ 56       $ 14.00         25.0%    $ 42      ███████████████
    5        $ 70       $ 14.00         20.0%    $ 56      ████████████████████
  ────────────────────────────────────────────────────────────────────────

  The bottle is paid for by glass ONE. Glasses two through five are gross
  profit at a rate nothing else in the building matches.

  The fifth glass is worth $14.00 of pure margin if it sells and costs you
  FIVE POINTS of pour cost on that bottle if it doesn't. There is no middle.

That is the whole by-the-glass discipline in one figure. Every open bottle is a countdown. The question at 9:40 p.m. is never "should I open another one" in the abstract; it is "will the glass I have left sell tonight, and if not, will it still be good tomorrow?"

What oxygen does, and how long you actually have

The moment the cork comes out, oxygen goes to work. Fruit flattens first, then the wine goes nutty, then it turns — and somewhere in the middle it stops being what you are charging fourteen dollars for, well before it becomes anything a guest would describe as spoiled.

Practical windows, refrigerated, with a stopper in the neck. These are working estimates from the floor, not laboratory findings:

Two operating rules follow, and they are non-negotiable.

Date every bottle when you open it. Grease pencil on the label or a dot sticker, day and hour. This is the same FIFO and date-labeling discipline Chapter 13 built for the walk-in, applied to the back bar, and it fails for the same reason it fails in the walk-in — because it depends on somebody doing a two-second thing under pressure, every time.

The test for whether to pour it is not "is it bad." It is: would I set this in front of table 12 without saying anything? If you would have to explain it, or hover, or watch their face — dump it. A guest who gets a tired glass of wine does not complain. They quietly downgrade their opinion of the restaurant and order a beer next time, and you never learn why.

🧾 Read the Numbers

```text FIGURE 16.4 — "The open-bottle log, week 31" [the Bellwether plan] THE ARTIFACT The bar's by-the-glass open-bottle log for one week — every BTG bottle opened Tuesday through Sunday, glasses poured off it, and its disposition at close on Sunday. THE CONTEXT Bellwether, last week of July. Patio open, six services (five dinners plus two brunches counted as one line). Twelve BTG selections, five-ounce house pour, five glasses to a bottle. Whites and rosé live in the under-bar drawer at 42-46F; open reds sit on the back bar under an argon blanket with a vacuum stopper.

       BTG selection              btls    glasses   glasses   glasses
                                 opened   possible    sold     dumped
       ────────────────────────────────────────────────────────────────
       Bubbles - bright & dry        3       15        12         3
       Crisp - high-acid white       5       25        24         1
       Crisp - aromatic white        3       15        13         2
       Round - oaked Chardonnay      4       20        18         2
       Round - skin-contact          2       10         7         3
       Rose - dry                    5       25        24         1
       Light - Gamay                 3       15        14         1
       Light - Sangiovese            3       15        13         2
       Dark - Cabernet blend         4       20        18         2
       Dark - Syrah                  3       15        14         1
       Dark - Malbec                 2       10         9         1
       Fortified - tawny (3 oz)      1        8         6         2 *
       ────────────────────────────────────────────────────────────────
       TOTAL                        38      193       172        21

       * the fortified bottle carries over to next week and is still
         perfectly sellable. Genuinely dumped glasses: 19.

       Glasses sold          172  x  $13.47 avg  =  $2,317 of BTG revenue
       Wine opened            38  x  $13.20 avg  =  $  502 of wine cost
       ACHIEVED POUR COST                 502 / 2,317  =  21.6%
       Pour cost if all 193 possible glasses had sold  =  19.3%

WHAT IT SHOWS A normal week. 89% of the wine opened was sold; 11% was poured out. Nineteen dumped glasses at $2.64 of wine each is $50 of product into a drain over six services, and it lands the week at 21.6% against a plan of 22.0% — slightly better than budget, on a heavy summer week. Two selections did the damage. The skin-contact white sold seven of a possible ten (30% waste) and the bubbles sold twelve of fifteen (20%). Everything else ran between 4% and 13%. WHAT IT DOESN'T It does not say WHEN in the week the dumps happened, which is the diagnostic question. Three bubbles dumped on a Tuesday is an opening discipline problem — somebody opened a bottle for one glass at 6:15 on the slowest night of the week. Three dumped on Sunday is a forecasting problem. The log needs a date column and does not have one. It also does not capture tastes poured for guests, comps, or the pours that went to the staff line-up. Those are real wine and they are inside the 38 bottles, uncounted. THE DECISION Add a DATE column to the log this week. Take the skin-contact white off by-the-glass and leave it as a bottle-only selection; a wine that wastes three of every ten glasses is not paying for its slot. Replace it with a second round white in the $12-13 range. Institute a Tuesday rule: no bubbles bottle opens before 7:00 p.m. unless a guest has ordered a glass. THE LESSON By-the-glass waste is not a wine problem, it is an OPENING problem. Every dumped glass traces back to a moment when somebody decided to pull a cork. The log is only useful if it records that moment. ```

The annual number, and where the two points went

Roll the log up over a year and the gap between the theoretical and the achieved becomes a line item.

\$2,376 a year.** At Bellwether's \$13.20 average by-the-glass bottle cost, that is 180 bottles of wine** poured down a drain — about half a bottle every service, 364 services a year.

And that is before the bottle-list losses, which §16.6 adds. Together they are the difference between the 26% wine cost you would run if nothing ever went wrong and the 28% in the plan.

⚠️ Where the Money Leaks

The preservation-system question, answered with arithmetic instead of enthusiasm.

Every wine rep and every equipment catalog will eventually put a nitrogen or argon dispensing cabinet in front of you. It keeps a bottle sellable for weeks because you never pull the cork. Should you buy one?

The \$4,200 answer.** An eight-bottle dispensing cabinet, installed, at an illustrative \$4,200, plus roughly \$40 a month in gas — \$480 a year. Best case it eliminates every dumped glass. It won't: it covers eight of your twelve selections and does nothing about the bottle somebody opened at 6:15 on a Tuesday for one glass. Call it 60% of the \$2,376 — **\$1,426 saved, less \$480 of gas = \$946 of net annual benefit.** Payback on \$4,200 is 4.4 years, on equipment that will need servicing inside that window. (Illustrative pricing — get quotes.)

The \$180 answer. Six vacuum stoppers, two proper sparkling stoppers, and a case of inert-gas spray: about \$180** at open, plus roughly **\$310 a year in gas cans. If it recovers 40% of the dump — \$950 — the net benefit is **\$640 a year on a \$180 investment. Payback: under four months.**

At Bellwether's volume the \$180 answer wins by a mile, and it is not close.

Now the honest other side, because the cabinet is not a bad machine — it is a machine for a different restaurant. Its real value is not saving waste; it is unlocking revenue that otherwise cannot exist. A room whose guests will pay \$32 a glass can pour a \$60-wholesale bottle five ways for \$160 — a 37.5% pour cost and \$100 of gross profit per bottle — with no risk of losing the remainder. Three such glasses a night is \$60 of gross profit, or roughly **\$21,840 a year**, and the machine pays for itself in about three months.

The machine does not create those guests. It monetizes them if they already walk in. Bellwether's guests, at a \$46 check, do not. Buy the equipment your demand justifies, not the equipment your aspirations do.

🔍 Check Your Understanding

1. A restaurant pours 6 oz from a 750 ml bottle and prices the glass at the wholesale bottle cost. What pour cost does it actually run at full yield, and how does that compare to the 20% a 5 oz pour would produce?
2. A by-the-glass selection sells three of five possible glasses on average. What is its pour cost if the bottle costs \$15 and the glass sells for \$15?
3. Why is by-the-glass waste described in this section as an opening problem rather than a storage problem?

(1: 6 oz gives 4 saleable glasses. At a \$14 bottle priced \$14/glass, revenue is \$56 and cost is \$14 — 25.0%, five points worse than the 20% a five-glass yield produces, before you count the 1.4 oz stranded each time. 2: Revenue \$45, cost \$15 → 33.3%. 3: Because the cost is incurred the instant the cork comes out; no amount of preservation recovers a bottle that should never have been opened that night.)

16.6 Storage, cellar depth, and capital tied up in bottles

This is the section no other chapter in this book can write, because no other category behaves like this. Wine is the only inventory in the building that is simultaneously perishable, appreciating in nobody's estimation but the buyer's, and slow.

Where the wine actually sits

Start with the physical truth, because it constrains everything financial that follows.

Bellwether does not have a cellar. Chapter 7's build-out ran into an undersized Type I hood and an inadequate grease interceptor, and absorbing that \$75,000 scope meant cutting the reclaimed-plank floor, the stone bar top, and the custom millwork. What did not survive that meeting was any notion of a temperature-controlled wine room. Nobody argued for it. There was no money and there was a hood.

So here is where \$4,068 of wine lives:

FIGURE 16.5 — WHERE THE WINE ACTUALLY SITS                        [the Bellwether plan]

  BAR (12 seats)                 SERVICE CORRIDOR            STORAGE / OFFICE (200 sq ft)
  ┌────────────────────┐        ┌──────────────────┐        ┌──────────────────────────┐
  │ under-bar refrig.  │        │ glass-door wine  │        │ wall rack, bolted        │
  │ drawers, 2 bays    │        │ refrigerator     │        │ 180-bottle capacity      │
  │                    │        │ 46 bottles       │        │                          │
  │ 42-46 F            │        │ dual zone 45/55F │        │ NO climate control       │
  │                    │        │ $1,600 used      │        │   66-68 F in winter      │
  │ ~24 bottles        │        │                  │        │   74-78 F in summer      │
  │                    │        │ ~46 bottles      │        │   shares a wall with the │
  │ OPEN BTG whites,   │        │                  │        │   dish area              │
  │ rose, sparkling;   │        │ white/rose/      │        │                          │
  │ tonight's reds     │        │ sparkling        │        │ ~137 bottles: ALL REDS   │
  │                    │        │ backstock        │        │ plus all bottle backstock│
  └────────────────────┘        └──────────────────┘        └──────────────────────────┘
        24 bottles                   46 bottles                    137 bottles
        12% of inventory             22% of inventory              66% of inventory

  IDEAL long-term wine storage: about 55 F, dark, still, humid, bottles on their side.
  Bellwether provides none of that for two-thirds of its inventory. Chapter 7 spent the
  millwork budget on a hood. THIS is what that decision looks like four chapters later.

Wine storage, as a discipline, means holding bottles at a stable cool temperature, in the dark, free of vibration, and — for cork-finished bottles — lying down so the cork stays wet. Bellwether achieves the dark and mostly achieves the lying-down. It does not achieve stable or cool.

Be honest about what that costs, without inventing precision. Warm storage accelerates every chemical reaction in a bottle; wine held in the mid-seventies ages materially faster than wine held at fifty-five, and the difference compounds over months rather than weeks. Repeated heating and cooling — which is what an interior room next to a dish pit does daily — pushes the cork and can produce seepage past the capsule, and a bottle weeping wine down its neck is a bottle you cannot put on a table. Separately and unrelated to your storage, a small but real percentage of cork-finished bottles arrive affected by cork taint; industry estimates have generally run in the low single digits, and the spread of screwcaps and technical corks has reduced it.

Put it in the plan as a number:

About one bottle a week that you paid for and cannot sell. Add the \$2,376 of by-the-glass dump from §16.5 and Bellwether's total wine shrink is \$3,353 a year — 2.0% of wine revenue.

That is the two-point gap between a theoretical 26.0% wine cost and the plan's 28.0%. The 28% is not a purchasing number. It is a purchasing number plus a drain number, and the drain is a fifth of it.

To feel that: at the plan's 16.8% operating margin, replacing \$3,353 of lost gross profit requires about \$19,958 of additional sales — 434 dinner covers, roughly four and a half full nights of service. Four and a half nights of a sixty-eight-seat dining room, to replace wine nobody drank.

👨‍🍳 On the Line

The cheap fixes, and the expensive one you should not buy.

Bellwether cannot build a cellar. It can do all of this, and most of it is free.

• Move the rack off the dish-pit wall. The far wall of that room is three or four degrees cooler on a summer afternoon, and moving a rack costs an hour and some anchors. Nobody does this because nobody measures.
• Put a \$25 min/max recording thermometer in the rack. Read it every Monday. You cannot manage a temperature you have never observed, and "it feels warm back there" is not data. This is the single best twenty-five dollars in the chapter.
• Keep the \$96–\$135 shelf in the wine refrigerator, not the storage room. Fifteen bottles fit, they are the most expensive wine you own, and they are — by §16.6's turn table — the inventory that sits longest and is therefore most exposed to time. Free.
• Store cork-finished bottles on their side. Free. Screwcaps do not care.
• Buy the slow shelf in threes, not sixes. Free, and it is the entire pruning discipline in §16.8.
• Receive properly. A summer delivery that sat on a truck in the sun is already damaged; check the cases at the door, in front of the driver, and refuse what is hot. You have about four minutes of leverage and then it is your wine.

What you do not do is buy a \$9,000 climate-controlled cellar unit to protect \$4,068 of wine. The machine costs more than twice the asset. If a category's protective equipment exceeds the value of the category, you have either bought the wrong equipment or built the wrong list — and it is nearly always the second one.

The capital

Now the money. Every bottle on that rack is cash you converted into an object.

Steady-state par — what Bellwether holds on an ordinary Wednesday:

Opening inventory — what you buy before you know anything:

You open deeper than par on the fast shelves and no deeper on the slow ones, because on opening day you do not know what sells and running out of a by-the-glass wine in week two is worse than carrying two extra cases of it.

And the hardware, which is capital too:

Eight thousand one hundred fifteen dollars, inside a \$620,000 project. It is 1.3% of the build and it generates 10.6% of revenue, which is a very good trade — and it is also \$8,115 that could have been \$5,000 if somebody had been undisciplined in the opposite direction and simply carried a shorter list.

Two ongoing capital notes. Glassware breakage runs at roughly 60% of the opening set per year — about 58 glasses at \$6.50, or **\$377 a year, which belongs in other operating, not in wine cost. And the reason Bellwether stocks two shapes only** is precisely this: every additional glass shape is another SKU to store, wash, break, and reorder, and no guest in this room has ever ordered a second glass of wine because it arrived in the correct bowl.

Turns, by shelf — the argument of the chapter

Here is where wine stops looking like inventory and starts looking like a portfolio.

FIGURE 16.6 — WHAT THE LIST COSTS TO OWN                            [the Bellwether plan]
  207 bottles at par. $4,068 of capital. 11.4 turns a year, 32 days on the shelf.

  shelf                sel.  btls   capital   annual COGS   turns   days on shelf
  ────────────────────────────────────────────────────────────────────────────────────
  BTG core              12    72    $  972     $25,680      26.4    ██ 14
  $38-$65 shelf         14    84    $1,344     $12,250       9.1    ██████ 40
  $66-$95 shelf          9    36    $  972     $ 6,160       6.3    █████████ 58
  $96-$135 shelf         5    15    $  780     $ 2,090       2.7    ████████████████████ 135
  ────────────────────────────────────────────────────────────────────────────────────
  TOTAL                 40   207    $4,068     $46,180      11.4    █████ 32

  For comparison, the same restaurant's FOOD inventory turns about 33 times a year
  (roughly 11 days). Per dollar of annual revenue, wine ties up about 2.7 TIMES as
  much inventory as food does.

  Nineteen percent of the wine money sits on a shelf that turns fewer than three
  times a year and averages four and a half MONTHS in the rack.

  None of that is automatically wrong. All of it should have been decided on purpose.

Every figure resolves. \$25,680 of by-the-glass and bottle COGS against \$972 of inventory is 26.4 turns, or 365 ÷ 26.4 = 14 days. \$2,090 against \$780 is 2.7 turns, or 135 days. The four shelves' COGS — \$25,680 + \$12,250 + \$6,160 + \$2,090 — sum to \$46,180, the plan's wine COGS exactly.

Fourteen days versus one hundred thirty-five days. Those two shelves are four feet apart and they are not the same kind of asset. One is working capital. The other is a fixed asset that happens to be liquid.

What a slow bottle costs in money over time

$$\text{Carrying cost} = \text{bottle cost} \times \text{annual cost of capital} \times \frac{\text{days held}}{365}$$

The capital in Bellwether's stack costs about 10.5% a year. Apply it.

A \$52 bottle on the top shelf, held 135 days: \$52 × 0.105 × (135 ÷ 365) = \$2.02 of financing. Add a loss allowance of 4% annualized over the same 135 days — \$52 × 0.04 × (135 ÷ 365) = **\$0.77 — and the bottle costs \$2.79 to own** before anybody drinks it. Against \$55 of gross profit when it finally sells, that is about 5% of the margin, quietly gone.

Five percent is not a crisis. The crisis is what happens when a buyer scales that behavior, which is the single most common way a wine program destroys cash.

🧮 Run the Numbers

The case-stack deal, which is how it actually happens.

A distributor rep offers ten percent off on a ten-case buy — 120 bottles — of a wine that costs \$22 wholesale. Bellwether currently sells about four bottles a month of it. The rep is friendly, the math sounds obvious, and the decision gets made standing up.

What the discount is worth: 120 × \$22 = **\$2,640 at list. Ten percent off = \$2,376.** Savings: **\$264.**

What the deal actually costs:

At four bottles a month, 120 bottles is thirty months of supply. Two and a half years.

Financing. You will hold, on average over that depletion, about half the lot: 60 bottles at the discounted \$19.80 = **\$1,188 of average inventory, for 2.5 years, at 10.5%. \$1,188 × 0.105 × 2.5 = **\$311.85.

The interest alone exceeds the discount. \$312 against \$264. The deal is already underwater and we have not touched the rest.

Spoilage. Thirty months on a rack that runs mid-seventies each summer. Call 8% of the lot degraded or unsaleable: about 10 bottles at \$19.80 = **\$198.**

Total cost of the deal: \$510 against a \$264 discount. Net: minus \$246.**

And that is before the things you cannot put a number on: 120 bottles of a 180-bottle rack occupied by one SKU for two and a half years; a list that cannot change because you are still working through March of two years ago; a vintage that has rolled over twice while you sold the old one; and \$2,376 of cash that left the account in a week when it could have been sitting there in February.

The disciplined answer: buy two cases at full price. \$528, six months of supply, **\$1,848 of cash still in the bank**, and the freedom to change your mind in June.

A rule that has never once cost me money: never buy more than ninety days of supply of anything on the bottle list, and never more than thirty days of anything by the glass. If the discount only works at a quantity that breaks that rule, the discount is not a discount. It is a loan you are making to a distributor at a negative interest rate.

16.7 Selling it: training, tasting notes, pairing, and the guest who doesn't want to look foolish

Everything above this section is an input. None of it produces a dollar until a guest says yes.

And here is the uncomfortable fact about wine: it is the only thing on your menu that a large share of your guests are actively nervous about ordering. Nobody is embarrassed to order the chicken.

The three fears, and the three fixes

Talk to enough guests and the anxiety resolves into three specific things.

Fear one: mispronouncing something. A guest who cannot say a word out loud will not order it. This is not a small effect and it is invisible in your data, because the guest simply orders something else and you never know what happened.

The fix, in three parts. Put a short phonetic in small type on the list for anything genuinely hard. Train servers to offer the name first — "the Gamay, the light red about a third of the way down" — so the guest can point rather than pronounce. And train them, absolutely without exception, never to correct a guest's pronunciation. Not gently, not helpfully, not by saying it correctly back. The one time a server does that, that table's wine spending stops for the evening and possibly forever.

Fear two: spending more than they meant to. The guest has a number in their head and no comfortable way to say it out loud in front of the table.

The fix: the second question below, and a genuinely good bottle at the bottom of the ladder (§16.4).

Fear three: being judged by somebody who knows more. This one is about the server, not the wine, and it is entirely under your control.

The fix: a service standard that treats every wine question as an ordinary question. No raised eyebrows. No "oh, interesting choice." No unrequested education. The correct posture toward a guest who orders the second-cheapest bottle is exactly the posture toward a guest who orders the \$135 one.

🤝 Hospitality

Two questions, and the most valuable one costs nothing.

There is no script here. There are two questions, asked in order, and they carry more wine revenue than any tasting note ever written.

Question one: "Red or white tonight — or shall I bring you both to look at?"

It is a binary. Every guest can answer it. It converts an intimidating open field into a choice with two options, and the third clause gives the indecisive table an exit that is not "no."

Question two — and this is the one — the server puts a finger on the list, at a price, and says: "Something around here?"

That is it. That is the whole move.

The guest can now say "a little less" or "that's fine" or "maybe a bit more" without ever saying a number out loud in front of their table. The single largest suppressor of wine spending in an American restaurant is the social cost of naming a price, and this gesture removes it entirely.

Where the server puts the finger matters. Train it to land just above the middle of the section — not at the bottom, which caps the table, and not at the top, which reads as a push. In a section running \$38 to \$72, the finger goes on the \$55.

I have watched this one gesture move a room's bottle average by six or seven dollars in a month. It requires no product knowledge whatsoever. It is available to a server on their fourth shift. And it is the reason a restaurant with no sommelier can still run a serious wine program.

Training, and what it is worth

Bellwether has no sommelier and will not have one. The program is therefore the training program.

The quarterly full-list tasting. Every server, every bartender, the FOH partner, and the chef, in the dining room before service, four times a year when the list changes with the season. Twelve by-the-glass wines plus the six rotating bottle-list selections — eighteen bottles at about \$20 average = \$360 per session.** Add ten people for forty-five minutes at an \$18 blended rate = \$135.** Four sessions a year:

Now the return. Recall §16.1: fifty cents a guest of additional wine spend — one more glass sold per ten covers — is \$16,830 of incremental revenue and \$12,118 of gross profit.

\$1,980 against \$12,118. It pays six times over on a movement so small you would struggle to see it in a single night's numbers. And it does not require anyone to become an expert. It requires that when a guest asks "what's the Syrah like?", every single person on the floor has actually tasted the Syrah and can say one true sentence about it without their voice going up at the end.

The tasting-note card. One index card per by-the-glass wine, laminated, in the server station. Not a paragraph. Four lines:

  THE BTG CARD — one per wine, four lines, no more            [the Bellwether plan]

  ┌──────────────────────────────────────────────────────────────────┐
  │  LIGHT & SAVORY  -  Gamay                          $14 / $52     │
  │                                                                  │
  │  TASTES LIKE ...  tart red fruit, a little pepper, low tannin    │
  │  DRINKS ...       cold-ish. Ask the bar to pull it 10 min early. │
  │  GOES WITH ...    the Hearth Chicken. The pork. Anything charred.│
  │  SAY THIS ...     "It's light enough for the fish and still      │
  │                    stands up to the hearth. It's what I drink."  │
  └──────────────────────────────────────────────────────────────────┘

  The SAY THIS line is the one that matters and the one every list omits.
  A server who has been handed language uses it. A server who has been
  handed a flavor profile has to invent language at the table, in front
  of four people, while carrying two plates. They will not.

The pre-shift wine of the day. Ninety seconds at line-up, one wine, one sentence, and a pour the size of a thimble for anyone who has not tasted it. Chapter 18 builds the pre-shift properly; this is its cheapest and highest-return use.

🧮 Run the Numbers

What one glass does to the Hearth Chicken.

The Hearth Chicken is Bellwether's costing lab, and its card is frozen: \$8.52 plate cost, \$29.00 menu price, 29.4% food cost, \$20.48 contribution margin. The single most profitable dollar contributor on the menu.

Now the server sells that guest a glass of the Gamay — \$14, with \$2.80 of wine in it and \$11.20 of contribution.

	Cost	Price	Contribution
Hearth Chicken	\$8.52 | \$29.00	\$20.48
Glass of Gamay, 5 oz	\$2.80 | \$14.00	\$11.20
Together	\$11.32** | **\$43.00	\$31.68

One glass of wine raises that guest's contribution by 54.7%.

Read that again, because it reframes the entire chapter. The kitchen cannot do this. To add \$11.20 of contribution through food, the kitchen has to cook, plate, and fire another course — more labor, more hearth space, more ticket time, more risk on a Friday. The server adds it by pulling a cork.

This is why a wine program at a 28% cost percentage belongs in a restaurant chasing a 60% prime cost. The percentage is worse and the marginal contribution per minute of labor is not close.

Responsible service does not soften because the product is nicer

⚖️ Code and Compliance

Wine is alcohol. All of it applies.

There is a comfortable fiction in nice restaurants that wine service is a different activity from bar service. It is not, legally or physiologically, and the fiction is precisely how over-service happens in dining rooms.

• A 750 ml bottle is roughly five standard drinks. A couple that shares two bottles over dinner has each had about five. That is not a moderate evening because it arrived in stemware.
• Dram shop liability applies to the wine list. Chapter 8 covers the exposure; the specifics vary enormously by state, and several states impose liability for service to a visibly intoxicated guest regardless of what was poured. Verify locally and carry the coverage.
• ID obligations are identical. A twenty-year-old ordering a glass of rosé on a patio in July gets carded exactly like a twenty-year-old ordering a beer. "They ordered wine" has never been a defense anywhere.
• Alcohol-server training — where your state requires or credits it — covers wine service. Do not let a server certify on the bar program and assume the dining room is different.
• The server pours; the table does not self-pour. This is usually taught as a courtesy standard. It is at least as much a control. A table that pours its own bottle is a table nobody is counting. The server who refills is the only person in the building tracking how much that four-top has actually had, and taking that job away from them removes your last line of sight.
• Have a written refusal procedure and back it. A server who declines to sell the third bottle must know, in advance and in writing, that the manager will support them and that no one will discuss the check average afterward.

Jurisdictions differ on every one of these. Verify locally, put it in the training manual (Chapter 18), and document that every server completed it.

Certification: what it is, and what Bellwether should actually do

Sommelier certification levels are the tiered credentials offered by the trade's education bodies. The main ones are real and worth knowing by name:

Costs vary by provider and region and change; get current pricing locally rather than budgeting from anything printed in a book.

Here is the operator's judgment, plainly. Nobody at Bellwether needs an advanced credential, and buying one would be a mistake in year one. A \$164,920 wine program does not require a Master Sommelier; it requires twelve servers who have tasted twelve wines and are not frightened of a list.

What is worth doing, and belongs in the plan: the FOH partner takes an introductory or Level 2 course in year two, after the restaurant is stable, for three reasons that have nothing to do with the certificate. It builds a vocabulary for buying. It builds credibility with distributor reps who will otherwise treat a sixty-eight-seat account as a route stop. And it gives one person in the building the confidence to say no to a wine, which is worth more than the ability to say yes to forty.

Year one's wine education budget goes to staff tastings, not credentials. \$1,980 spread across ten people beats one certificate held by one person who might leave.

16.8 Corkage, list maintenance, and pruning what doesn't move

Two housekeeping disciplines that quietly decide whether the program stays healthy or slowly silts up.

Corkage

Corkage is the fee a restaurant charges to open and serve a bottle of wine a guest has brought in. It exists because the restaurant is providing everything except the wine: the glassware, the service, the opening and pouring and decanting, the washing, and the seat.

First, and this is not optional: find out whether it is legal where you are. Some states prohibit bringing alcohol onto a licensed premises entirely. Some permit it only with a specific endorsement. Some leave it to the licensee. Some counties differ from their states. Ask your alcoholic beverage authority in writing, and do not take a competitor's practice as evidence — they may simply not have been caught.

Assume it is permitted where Bellwether sits. What should it cost?

🧮 Run the Numbers

Setting a corkage fee against what you would otherwise have earned.

Bellwether's bottle sales are \$66,000 at 37% cost, so the average bottle sold produces:

• Average bottle price: \$53.10
• Average wholesale cost: \$19.65
• Average gross profit per bottle: \$33.45

Now set corkage at \$25.

Twenty-five dollars recovers 74.7% of the gross profit of an average bottle sale — and it does it with zero inventory cost, zero capital tied up, zero spoilage risk, zero storage, and no distributor relationship required. On a pure margin basis, corkage is the best-performing product on the wine list.

The catch, and it is the whole catch: a corkage table is frequently a table that would have bought a bottle. If corkage cannibalizes a \$53 sale, you traded \$33.45 for \$25 and lost \$8.45. If it brings in a table that would have gone somewhere else, you gained \$25 of nearly pure margin plus whatever they ate.

Which effect dominates depends entirely on the fee. Set it too low — \$10 — and you have invited your own guests to stop buying wine. Set it too high — \$45 — and nobody uses it and you have simply declined a category of guest. Twenty to thirty dollars, in a room with a \$46 check, is where the two effects roughly balance.

The policy that resolves it: \$25 per 750 ml, limit two bottles per table, and we will waive one corkage fee for every bottle purchased from our list.

That last clause is the clever part. It converts a corkage table into a list-buying table — the guest brings the special bottle they wanted to drink and buys one from you to get the fee waived, and you have made the \$33.45 anyway. It costs you nothing when it works and it costs you \$25 when it doesn't.

Two more corkage rules worth writing down:

Do not accept a bottle that is already on your list. Otherwise you are hosting a demonstration that your markup exists, at your own table, using your own glassware.

Pour a taste for the host, and pour it as well as you would pour a \$100 bottle of your own. A guest who brings a bottle has an occasion — an anniversary, a wine they have been saving, a bottle from a trip. That is a guest with a reason to be here, which is the most valuable kind, and the twenty-five dollars is not the transaction. The transaction is whether they come back for the next occasion.

Reading the list like a menu-mix report

Chapter 12 taught menu engineering on food. The wine list is a menu and it deserves the same report, run monthly and read seriously once a quarter.

🧾 Read the Numbers

```text FIGURE 16.7 — "Ninety days of bottle sales" [the Bellwether plan] THE ARTIFACT POS wine sales report by SKU, bottle sales only, 1 August - 30 October. The 28 bottle-only selections; the 12 BTG wines report separately. Sorted ascending by units sold. Ten of the 28 shown. THE CONTEXT Bellwether's first full autumn. The list has not been pruned since opening eleven months ago. Storage is a 180-bottle wall rack plus a 46-bottle refrigerator, and it is full. Every slow SKU on this report is occupying a slot a faster one wants.

on the list as              shelf      cost   price   sold  revenue    GP$   days
                                                       90d              90d  cover
─────────────────────────────────────────────────────────────────────────────────────
Dark - Nebbiolo             $96-135  $78.00  $133      0   $     0  $    0    ---
Dark - Napa Cabernet        $96-135  $62.00  $117      1   $   117  $   55    270
Round - white Burgundy      $66-95   $34.00  $ 75      2   $   150  $   82    180
Dark - Rioja Gran Reserva   $96-135  $52.00  $ 99      3   $   297  $  141     90
Light - cru Beaujolais      $66-95   $28.00  $ 70      6   $   420  $  252     60
Dark - Barbera              $38-65   $19.00  $ 53     14   $   742  $  476     39
Crisp - Muscadet            $38-65   $13.00  $ 42     19   $   798  $  551     28
Round - Alsace Riesling     $38-65   $15.00  $ 48     22   $ 1,056  $  726     25
Dark - Cotes du Rhone       $38-65   $14.00  $ 45     31   $ 1,395  $  961     17
Light - Oregon Pinot Noir   $38-65   $24.00  $ 60     34   $ 2,040  $1,224     16
─────────────────────────────────────────────────────────────────────────────────────
these 10 of 28                                       132   $ 7,015  $4,468

"days cover" = bottles held at par, divided by 90-day units, times 90.

WHAT IT SHOWS A textbook long tail. The bottom four selections sold SIX bottles in ninety days between them and hold $520 of inventory; the top four sold 106 bottles and hold $312. The Nebbiolo has sold nothing in three months and represents $234 of capital and three of the rack's most valuable slots. The $38-65 shelf is doing what it was designed to do -- 17 to 39 days of cover and $3,914 of the $4,468 of gross profit on this page. WHAT IT DOESN'T It does not say whether the slow bottles were ever OFFERED. A wine with zero sales might be badly priced, badly placed on the page, or simply something no server has ever mentioned at a table. Before you cut a SKU, find out which. It also cannot see the anchoring effect: the $133 Nebbiolo may be earning its place by making the $75 white Burgundy read as reasonable, and a report of units sold will never show that. And it is one autumn. A big red in October is not a big red in June. THE DECISION Do not cut the top shelf. REDUCE it. Take the Nebbiolo from three bottles to one and the Napa Cabernet from three to one; take the white Burgundy from four to two. That frees $348 of capital and six rack slots while keeping every price point on the page. Before the next order, put the Nebbiolo on the pre-shift board for two weeks and see whether zero sales was a demand fact or a mentioning fact. THE LESSON The top of a wine list is marketing inventory, not selling inventory. Its job is to make the middle of the list look reasonable. ONE bottle anchors exactly as well as three, and costs a third as much to own. ```

The pruning rule, and its exception

The rule Bellwether runs:

• Any bottle-only selection that has not sold in 90 days goes on notice. It gets two weeks on the pre-shift board and a line on the specials card.
• If it has not sold in 120 days, it comes off the list — or comes down to one bottle if it is serving as a price anchor.
• No more than 10% of selections and no more than 12% of inventory dollars may sit above the "sells regularly" ceiling, which at Bellwether is about \$95.

And the honest self-assessment, because the plan should be able to criticize itself: Bellwether is currently at 12.5% of selections and 19.2% of dollars above that ceiling. Five selections at three bottles deep is \$780 of a \$4,068 inventory sitting on a shelf that turns 2.7 times a year. That is heavy. The year-one discipline is to go two deep instead of three on the top shelf and revisit in the fall.

Here is the arithmetic that makes it obvious, and it is the sharpest number in the chapter:

\$348 of capital held on the \$96–\$135 shelf, at 2.7 turns and a 52.8% cost, produces about \$840 of gross profit a year.** The same \$348 held on the \$38–\$65 shelf, at 9.1 turns and a 35% cost, would produce about \$5,881. Seven times the gross profit from the same dollar.**

But — and this is why the rule has an exception — you cannot actually redeploy it that way. The fast shelf is demand-limited. Owning more Côtes du Rhône does not sell more Côtes du Rhône; it just means you have more Côtes du Rhône. So the freed \$348 does not become wine at all.

It becomes cash, and cash in February is the point of this entire book.

The exit you decide before you buy

The last discipline, and the one that would have prevented most of the dead inventory I have ever seen: decide the exit before you buy the wine.

Ask, at the tasting, in front of the rep: if this does not sell, what do I do with it? There are exactly four answers, and you should know which one applies before you sign the order.

1. Pour it by the glass at a price that moves it. Three bottles of a \$52 wine sold as fifteen glasses at \$18 is \$270 against \$156 of cost — a 57.8% pour cost that would look terrible on a report and returns \$114 of gross profit you were otherwise never going to see. A bad percentage beats a zero.
2. Feature it. A bin-end special on the pre-shift board, one line on the specials card, two weeks. Cheap, and it distinguishes "nobody wants this" from "nobody has mentioned it."
3. Make it staff education. A dead bottle poured at a Tuesday line-up is not a loss; it is the \$1,980 training budget doing its job with wine you already own.
4. Return it — if you asked first. Some distributors will take back recent purchases in original cases. Almost none will take back something you have held a year. Ask about the return policy before the first order, not after the twelfth month.

There is a fifth answer people reach for and you should mostly refuse: cooking with it. A \$12 bottle into a braise is fine and Chapter 13's costing handles it. A \$52 bottle into a braise is a fifty-two-dollar bottle in a braise, and no guest has ever detected the difference.

🔍 Check Your Understanding

1. Bellwether's average bottle sale produces \$33.45 of gross profit. Why might a \$25 corkage fee be a better transaction than that bottle sale, and why might it be worse?
2. A selection has sold zero bottles in 120 days. Give two reasons you might keep it anyway, and state what you would do to test each one.
3. Why does the chapter argue that capital freed from the top shelf should go into the bank rather than into more of the fast-selling wine?

(1: Better — \$25 with no inventory, capital, spoilage, or storage attached, and it may bring in a table that would otherwise have gone elsewhere. Worse — if it cannibalizes a bottle sale you would have made anyway, you traded \$33.45 for \$25. 2: It may be functioning as a price anchor that makes the middle of the list read as reasonable — test by reducing depth to one bottle rather than cutting it; or it may never have been offered — test with two weeks on the pre-shift board before deciding. 3: Because the fast shelf is demand-limited; more inventory does not create more demand, so the capital would simply sit in a different place. Cash is the only redeployment that actually earns anything.)

🍽️ The Business Plan

Checkpoint 16 of 40 — the Wine section, and Part III closes.

Chapter 15 built the bar. This chapter builds the other thirty-eight cents of the beverage line, and with it the last piece of Part III: Bellwether now has a costed menu, a priced menu, a purchasing and inventory system, a kitchen that can execute it, a bar program, and a wine program.

What this chapter contributes to the plan:

BELLWETHER — THE WINE PROGRAM (constructed teaching example)

List size	40 selections; 12 poured by the glass
Sections	Bubbles (3) · Crisp & Bright (5) · Round & Textured (7) · Rosé (2) · Light & Savory (8) · Dark & Structured (12) · Sweet & Fortified (3)
Organized by	style, not grape or region; grape and origin on the second line
Bottle price range	\$35 to \$135
BTG price range	\$11 to \$17 for 5 oz; \$12 for a 3 oz fortified pour
House pour	5 oz, yielding 5 glasses per 750 ml
Pricing method	laddered multiples: 3.5× under \$10 wholesale, falling to 1.9× or cost + \$55 above \$45
Wine revenue, Year 1	**\$164,920** — 38% of the \$434,000 beverage line, 10.6% of total sales
Wine cost target	28.0% — \$46,180 (BTG 22.0% / bottle 37.0%)
Per-guest wine spend	**\$4.90** of the \$12.88 beverage component of the \$46 check
Inventory at par	207 bottles, \$4,068 — 11.4 turns, 32 days
Opening inventory	*259 bottles, \$4,840** *(inside the \$35,000 pre-opening line)
Program hardware and glassware	*\$3,275** *(inside the \$45,000 smallwares & FF&E line)
Total capital committed at opening	\$8,115
Storage	46-bottle dual-zone refrigerator, two under-bar drawers, 180-bottle wall rack. No temperature-controlled cellar. 66% of inventory is held at ambient.
Shrink allowance	**\$3,353/yr** — \$2,376 BTG dump + \$977 damage, taint, and breakage
Sommelier	none, and none planned. The program is the training program.
Training	quarterly full-list tasting for all FOH — \$1,980/yr. FOH partner takes an introductory certification in year two, not year one.
Corkage	\$25 per 750 ml, two-bottle limit, one fee waived per bottle purchased from the list. Not offered for wines on our own list. (subject to state law — verify)
Pruning rule	90 days without a sale = on notice; 120 days = off the list or down to one bottle. Ceiling shelf capped at 10% of selections, 12% of inventory dollars.

Every figure above is constructed and internally consistent with Chapters 11, 13, and 15.

What this section settles. The wine line of the pro forma is now real rather than assumed. \$164,920 of revenue at a 28% cost is a defensible number built from a specific list, a specific pour size, a specific ladder, and a specific waste allowance — not from a benchmark someone quoted. The inventory investment is sized and located. The storage constraint is named honestly rather than wished away.

What it does not settle, and you should say so in the plan:

1. The 38% wine share of beverage is an estimate, not a measurement. It is built on an American hearth menu in a gentrifying neighborhood, and it could plausibly land at 32% or at 45%. The whole \$434,000 does not move; the mix does, and wine and spirits have a ten-point cost gap between them. A shift of five points of mix toward wine moves blended beverage cost by half a point.
2. The storage room is a known, accepted, unfixed risk. The plan should say plainly: two-thirds of the wine is held at ambient temperature in a room that runs mid-seventies in August, we have budgeted \$977 a year for the consequence, and we have chosen a \$25 thermometer and a discipline over a \$9,000 cellar. That is a defensible answer. Pretending the risk is not there is not.
3. The 22% by-the-glass cost assumes the log gets kept. It is 2.4 points worse than theoretical by design, because we expect to dump about 180 bottles a year. If the log lapses, that number does not stay at 2.4 points. It goes to four or five and nobody notices for a quarter.
4. A neighborhood restaurant with no sommelier is betting the program on training that has to be repeated forever, in an industry with roughly 75% annual turnover. Every departure takes twelve tasted wines out the door with it. Chapters 17, 18, and 21 have to answer this, not Chapter 16.

Open questions carried forward:

1. Does a \$46-check room actually buy \$4.90 of wine a head, or is that number aspirational? (Chapters 22, 24, 31 — the first ninety days of POS data settles it, nothing before.)
2. Can training survive turnover well enough to hold the by-the-glass attachment? (Chapters 17, 18, 21)
3. How does the wine line behave in February, when covers drop and the top shelf still sits there? (Chapters 32, 33)
4. What happens to the corkage policy if the neighborhood's demographics keep moving? (Chapter 26)
5. Does the \$8,115 of wine capital belong in the plan at all, or should year one open with a twenty-eight-selection list and add the top shelf in year two? (Chapter 33, and honestly a question the partners should argue about before opening.)

Conclusion

Wine is the strangest line on a restaurant's P&L, because it is three different businesses wearing one name.

The by-the-glass program is a fast-moving, high-margin, waste-prone operating business. It turns twenty-six times a year, produces sixty percent of the revenue from twelve selections, runs a 22% cost, and loses about a hundred and eighty bottles a year to the simple fact that somebody has to decide when to pull a cork. It is managed with a log, a pour size that divides the bottle, and a discipline about Tuesdays.

The bottle list is a slow-moving capital business. The top shelf turns fewer than three times a year and sits four and a half months in a rack in a room that runs seventy-six degrees in August. It is managed with a ninety-day report, a pruning rule, an exit decided before purchase, and a willingness to say out loud that nineteen percent of the money is in the slowest nineteen percent of the inventory.

And the selling of it is a hospitality business — the only one of the three that produces any margin at all, because none of the arithmetic above matters until a nervous guest says yes. A server who puts a finger on the list at fifty-five dollars and says "something around here?" is worth more to the wine program than any curatorial decision the buyer will ever make. Fifty cents a guest is twelve thousand dollars a year, and fifty cents a guest is one training session and one gesture.

Along the way, wine has done something no other category in this book does: it has broken cost-percentage pricing at both ends of the same page, and made the case for contribution margin undeniable. A 28.6% bottle earning \$20 and a 59.3% bottle earning \$55 sit six lines apart on Bellwether's list, and the one with the terrible percentage is worth almost three times as much every time it sells. You bank dollars. You have never once deposited a percentage.

Part III ends here. Bellwether now has a menu that is costed, priced, engineered, purchased, inventoried, produced, and — with Chapters 15 and 16 — drunk alongside. What it does not have is a single person to cook or serve any of it.

That is Part IV, and it is the hardest part of the business. Half of prime cost, all of the guest experience, and roughly seventy-five percent annual turnover. Chapter 17 starts where every restaurant starts and most restaurants stumble: finding people.

Key Terms

Wine list architecture — the deliberate structure of a wine list: how many selections, how they are grouped, how prices ladder from bottom to top, which are poured by the glass, and what each section is for. A design decision made once and revised quarterly, not the residue of two years of purchasing. (Ch. 16)

By-the-glass (BTG) — wine sold by the individual pour rather than by the bottle. At Bellwether, 12 of 40 selections, carrying 60% of wine revenue. (Ch. 16)

BTG yield — the number of saleable glasses a bottle produces at the house pour size: bottle volume in ounces divided by pour size. A 750 ml bottle holds about 25.4 oz, so a 5 oz pour yields 5 glasses and a 6 oz pour yields 4. (Ch. 16)

Bottle-price laddering — pricing wine with a markup multiple that declines as wholesale cost rises, floored on gross-profit dollars at the bottom and capped by a dollar-margin rule at the top. The multiple falls up the list; the gross-profit dollars rise. (Ch. 16)

Wine markup — the multiple applied to a wholesale bottle cost to reach the menu price. A \$14 bottle at 3× lists at \$42. The reciprocal of the bottle's cost percentage. (Ch. 16)

Wine cost percentage — wine cost of goods sold divided by wine sales, for the category as a whole. Bellwether's target is 28%, against a bar running 18% and a blended beverage cost of 22%. (Ch. 16)

The three-tier system — the post-Prohibition structure separating alcohol producers, wholesalers, and retailers into three legally distinct tiers, with product generally required to pass through all three in order. Created after the 21st Amendment (1933) to dismantle the pre-Prohibition "tied house"; every state's version differs. (Ch. 16)

Corkage — the fee a restaurant charges to open and serve a bottle of wine a guest has brought in, covering the glassware, service, and seat. Legality varies by state and locality. Bellwether's is \$25 per 750 ml, two-bottle limit, one waived per bottle bought from the list. (Ch. 16)

Wine storage — holding bottles at a stable cool temperature (about 55°F is the long-term ideal), in the dark, free of vibration, and lying down for cork-finished bottles. What a restaurant without a cellar actually achieves, and what the gap costs in spoilage, is a budgetable number. (Ch. 16)

Sommelier certification levels — the tiered credentials of the trade's education bodies: the Court of Master Sommeliers (Introductory / Certified / Advanced / Master Sommelier), WSET Levels 1–4, the Society of Wine Educators' CSW, and the Institute of Masters of Wine's MW. Useful for a buyer's vocabulary and credibility; not a prerequisite for running a profitable list. (Ch. 16)

Spaced Review

1. Without looking back: a 750 ml bottle is poured at 5 oz for \$16 a glass and costs \$16 wholesale. What is the pour cost if all five glasses sell? What if only four sell, and how many points is the difference?
2. From Chapter 11: target-cost pricing says price equals cost divided by target cost percentage. Apply it at a 28% target to an \$8 wholesale bottle and to an \$80 wholesale bottle. State what goes wrong in each case and what the ladder does instead.
3. From Chapter 1: prime cost is COGS plus total labor as a percentage of sales. Bellwether's wine COGS is \$46,180 on \$1,550,000 of sales. How many points of prime cost is that, and would a buyer who cut it by three points be doing the business a favor? Defend your answer with §16.4's argument.
4. From Chapter 13: the inventory usage formula is beginning plus purchases minus ending. Explain why an operator who computes wine cost from invoices alone will be badly misled in a month when the buyer took a ten-case deal — and quantify it using the \$2,376 case-stack example from §16.6.
5. The recurring question: a distributor offers a wine that would sell about two bottles a month at a price that requires a five-case minimum. Walk through the four things you would compute before answering, and say which one usually settles it.