The menu design read
Where items sit, how they read, where the eye lands first — placement and description calibrated to the room. Abstract layout shown; not La Brisa’s real menu text.
Sample Verdict
One real Verdict, anonymized — read end to end.
First, a real Couverté Verdict with every identifying detail removed: a luxury hotel restaurant in a major US metro, figures exactly as the audit measured them, in dollars. Then the same format as two worked examples — a fictional boutique hotel with three outlets, and a fictional independent restaurant. The methodology, the deliverables, and the critic pass are exactly what your property receives.
See a Verdict like this — on your own menu.
Your first property-specific Verdict is free, and we reply within one business day.
The Verdict, at a glance
The score, the matrix, the plan — on one screen.
What an operator opens to — shown here on the restaurant example. Hover a class to isolate it across the matrix and the plan. Every figure is illustrative.
72Couverté Score
Couverté Score
Strong menu, with margin leaking on three plowhorses.
Items analysed
48Stars
9Plowhorses
6Projected upside
+$13.7k/moClassification matrix
Hover or focus a class to isolate itTop moves
- 01Reprice Plowhorses+$8,400/moest. impact
- 02Promote Stars+$3,200/moest. impact
- 03Retire Dogs+$2,100/moest. impact
Five deliverables
- 01Score
- 02Action plan
- 03Menu-design read
- 04Market position
- 05Review cycle
The one real example — anonymized
A real Verdict, anonymized — $63,880 a month, recoverable.
Our first published Verdict, with every identifying detail removed: the signature restaurant of a luxury hotel in a major US metro. The figures are exactly as the audit measured them, in US dollars — not fabricated, not converted. The headline finding was a signature dessert ringing zero sales for a full year — a point-of-sale blind spot masking an estimated $52,360 a month.
Real audit · anonymized
- Couverté Score 44.5
- Confident recoverable / mo $63,880
- Identified potential / mo $150,677
- Menu items audited 137
Figures exactly as audited, in US dollars. Identity removed for publication.
Read the full anonymized Verdict
Real audit, anonymized. This is an actual Couverté Verdict on a real property, not a worked example. The property name, brand, city, outlet name, suppliers, and signature-dish names have been removed so it can be shown publicly. Every figure below is exactly as the audit measured it, in US dollars — nothing here is fabricated or converted. Following our conservative method, the opportunity is shown as two numbers: what we are confident is recoverable, and the larger total we identified. Neither is a guarantee.
The property
A luxury hotel’s signature restaurant in a major US metropolitan market — a 110-seat room running roughly 14,859 covers a month, attached to a hotel averaging about 140 occupied rooms a night. International and modern menu, analysed in the Luxury — Hotel F&B voice: recommendations written for an F&B director answerable for guest experience and margin at once.
Everything that follows is the Verdict the property’s leadership received — the same five-layer methodology, the same five deliverables, the same critic and QA pass as every engagement. Only the identifying details are removed.
The two numbers the Verdict opens with
The menu portfolio scores 44.5 out of 100 against a tier floor of 70, with a financial-health read of 42 — a property leaving real money on the table, not a failing operation. The opportunity splits into two figures, by design:
- Confident recoverable: $63,880 / month — roughly 9% of F&B revenue. The high- and medium-confidence moves, each already discounted by a realization factor so the number under-promises.
- Identified potential: $150,677 / month — the full opportunity surfaced across every action before that discount, including the lower-confidence items that need clean data to confirm.
We lead with the smaller, defensible number on purpose. The larger one is the ceiling, not the promise.
What the Verdict found that the operator did not know
The single most valuable finding was a blind spot: the restaurant’s signature tableside dessert — a chocolate sphere cracked open at the table, rated high-popularity and priced correctly at the comp-set median of $22 — showed zero point-of-sale volume across twelve months.
That is not a demand failure. It is a POS configuration failure — the item was almost certainly being rung as a modifier rather than a tracked line, so its revenue was invisible. The audit estimates it is masking $52,360 a month in dessert-stage contribution. The fix is a 72-hour data-integrity task, not a menu redesign.
Where the check is leaking
Four structural gaps account for an estimated $107,960 a month in check-path leakage — every dollar recoverable from guests already seated, without adding a single new cover.
| Leakage point | Estimated monthly | The gap |
|---|---|---|
| Dessert POS capture | $52,360 | Desserts rung as modifiers, not tracked items — the signature dessert reads as zero volume. |
| By-the-glass wine gap | $28,500 | Two house pours at $6, then a bottle ladder that jumps $70 → $180. No mid-tier by-the-glass program at all. |
| Starter check-anchor drag | $18,200 | A 28-item starters category spanning $5 to $28 — a $5 item dominates the mix and pulls table-level check expectations down. |
| Opening-beverage non-attach | $8,900 | No pre-meal beverage attach at a property whose guests expect one. |
The financial read
Food cost runs 32.7% against a tier target of 25–30% — 2.7 to 7.7 points high, consuming an estimated $37,095 a month in excess contribution at the midpoint. Labor at 32.0% sits at the lean end of its band, so the combined prime cost of 64.7% is carried almost entirely by food cost, not staffing.
The likely cause is structural: a 137-item menu — nearly three times the tier ideal of 30–50 — inflates kitchen complexity, spoilage, and low-velocity ingredients. Rationalizing toward 50–65 items projects an estimated $33,528 a month recovered in food cost alone.
The anchors that underperform
The two highest-volume food items are also the menu’s margin laggards. A 52oz tomahawk ($165, ~396 covers a month) runs a 62.5% contribution margin; a dry-aged rib eye ($98, ~376 covers a month) runs 60.3% — both below the 70% tier target. Several proteins at $54–$98 carry null or near-null descriptions, undermining the price anchor at the exact moment of decision.
The plan, in order
The sequencing is deliberate — data integrity before any pricing move, because raising prices on items with unconfirmed volume optimizes against phantom demand.
Days 1–30 — fix the data, open the easy gaps. Audit POS mapping for every dessert and confirm the signature dessert has a dedicated tracked item; launch a six-label by-the-glass still-wine program priced $14–$22 to fill the mid-tier; restructure the 28-item starters category into two visible tiers.
Days 31–90 — recover the margin. Bring food cost from 32.7% toward 28% through menu rationalization and protein-portion audits; rewrite the underperforming anchor descriptions and add a paid accompaniment upsell; introduce a tableside dessert prompt pairing the signature dessert with a digestif as a closing ritual.
What this is, and is not
This is a real Verdict on a real property, anonymized for publication. The figures are as measured, in US dollars, and the two-number split is deliberate — we would rather under-promise and over-deliver than quote a single optimistic headline. It is not a guarantee, and it is not your property. A Verdict on your menu produces your own numbers, from your actual menu mix, costs, and comp set. If you want the real version, start the conversation.
And the rest of the package
Three more deliverables, beyond the dashboard.
Beyond the score, matrix, and plan: the menu-design read, the market position, and the review cycle. Each shown as the shape of the thing — illustrative, not measured data.
-
01 -
02 The market position
Your pricing read against a named comp set, with convergence risks and quiet headroom flagged. The bars show relative position only — illustrative shape, not measured prices.
-
03 The review cycle
What to re-examine, and when — a recommendation without a date to check it is a guess. A conceptual cadence from delivery through the first re-read.
Featured worked example — hotel
Hotel Llevant: 120 rooms, three outlets, one Verdict.
A fictional urban boutique hotel — an all-day restaurant, a rooftop bar, and a breakfast room that doubles as the banquet space — read as one engagement, written for an F&B director: capture rate, daypart yield, banquet contribution, and cross-outlet pricing consistency.
Illustrative
- Occupied room-nights / month ≈2,800
- Breakfast capture 73%
- In-house dinner capture 12%
- Banquet share of F&B revenue 23%
Read the full hotel example
Illustrative — Hotel Llevant is a fictional property; every figure here is fabricated to demonstrate the format, not a real result. Hotel Llevant does not exist. The room counts, capture rates, outlet revenues, and interventions below are a worked example so you can read a multi-outlet Verdict end to end before you book your own. Nothing here is a measured outcome, a benchmark, or a guarantee.
The property
Hotel Llevant is an urban boutique hotel — 120 rooms near the seafront in Barcelona, running three food-and-beverage outlets: Terra, the all-day restaurant on the ground floor; Cel, the rooftop bar; and Sala Llum, the breakfast room that doubles as the banquet and events space. For this worked example it is analysed in the Boutique — Hotel F&B voice: recommendations written for an F&B director who answers for guest experience and margin at the same time, across three menus that have to make sense individually and together.
The month used as the basis (all figures fabricated for the example): 78% occupancy on 120 rooms — roughly 2,800 occupied room-nights, about 4,770 in-house guest-nights at 1.7 guests per room. Total F&B revenue of €305,000 across the three outlets: Terra €118,000, Cel €64,000, Sala Llum €123,000 (€52,000 breakfast, €71,000 banquets and events).
Everything that follows is the Verdict Llevant’s F&B director would receive — the same five-layer methodology run across all three menus, the same five deliverables, the same critic and QA pass before any of it ships. Only the property is invented.
The four numbers the Verdict opens with
A hotel Verdict leads with the reads a single-room restaurant never needs. All four are illustrative here:
- Capture rate. Breakfast captures 73 of every 100 in-house guests (3,480 covers against ~4,770 guest-nights). Dinner captures 12 — 590 of Terra’s 2,140 dinner covers come from guests already sleeping upstairs. The rooftop reaches 31. The dinner gap is the single largest commercial finding in this example: the guests are already in the building.
- Daypart yield. Cel takes 61% of its month between 19:00 and 22:00, and under 8% between 15:00 and 18:00 — staffed identically across both. Terra’s lunch holds on local trade; its Monday-to-Wednesday dinner runs soft.
- Banquet contribution. Banquets and events are €71,000 of €305,000 — 23% of F&B revenue from fourteen events, at €67.60 per cover, the highest-margin revenue in the building and the line most hotels price by habit.
- Cross-outlet pricing consistency. Shared items carry rooftop premiums ranging from 24% (patatas bravas: €8.50 at Terra, €10.50 at Cel — wait staff can defend the view and the serve) to over 100% (espresso: €2.40 downstairs, €4.90 on the roof — the one guests mention in reviews). The premium is not the problem; the absence of a stated logic is.
Five deliverables in five days
The same five deliverables as every Verdict — here, read across three outlets at once.
1. The item analysis
Every item on all three menus is scored through the five analytical layers and classified: Star, Plowhorse, Puzzle, Dog, plus the two Couverté additions, Vulnerable Star and Conversion Candidate. Below is an illustrative slice — six representative items across the outlets, not the full count.
| Item | Outlet | Classification | Diagnosis (illustrative) |
|---|---|---|---|
| Dorada a la sal | Terra · dinner | Star | High draw, €14.10 contribution on €26.50. Protect it — and put it in front of in-house guests. |
| Menú del día | Terra · lunch | Plowhorse | 41% of lunch covers at €4.20 contribution per cover, under the house average. It carries the local trade; it must not carry the P&L. |
| Gin del Llevant | Cel | Star | The signature serve: €9.80 contribution on €14.00. The rooftop’s anchor. |
| Vermut i conserves | Cel | Puzzle | Strong margin, low order count — buried under the cocktail list, invisible in the dead afternoon daypart it was built for. |
| Breakfast buffet | Sala Llum | Vulnerable Star | Capture is excellent, but cost per cover has crept three months running. One more supplier increase compresses it. |
| Menú Empresa (banquet) | Sala Llum | Conversion Candidate | €62.00 per cover with food cost below the banquet category average — headroom on the highest-margin line in the building. |
2. The action plan
Ranked by expected impact. Each action carries the full backbone: a diagnosis, a primary action, an implementation detail, an expected impact labelled as an estimate, and a measurement window. Four illustrative actions, in rank order:
Action 1 — Reprice the Menú Empresa (Conversion Candidate). Diagnosis: banquets are 23% of F&B revenue and the corporate set menu’s food cost sits below the banquet category average — the cost base can absorb a price move on the dates that sell out anyway. Primary action: Reprice €62.00 → €68.00 on peak corporate dates. New inquiries only; every signed contract is honoured at its agreed price. Implementation: update the events sales sheet and the proposal template; brief the events coordinator on which dates carry the new rate and why. Expected impact (illustrative estimate): at fourteen events a month, a six-euro move on peak-date covers compounds quietly without touching the product. Measurement window: re-read after the next eight contracted events, or eight weeks — whichever comes first.
Action 2 — Close the dinner capture gap. Diagnosis: 12 of every 100 in-house guests eat dinner in the building. Each point of dinner capture is roughly 48 covers a month at Terra’s €27.40 average check — about €1,300 of monthly revenue per point, from guests already upstairs. Primary action: a one-line dinner offer at check-in and on the in-room card — a guaranteed table and a starter of the house for same-day reservations. Not a discount. Implementation: front-desk script plus a QR card in the room; the offer routes to Terra’s soft Monday–Wednesday dinner first. Expected impact (illustrative estimate): moves capture without buying it — the starter costs the kitchen far less than the cover earns. Measurement window: re-read capture at four weeks.
Action 3 — Stop staffing the rooftop’s dead daypart like its peak. Diagnosis: 15:00–18:00 produces under 8% of Cel’s revenue with peak staffing. The Vermut i conserves Puzzle was built for exactly this window and nobody can see it. Primary action: a low-prep afternoon serve — the vermut hour as the only afternoon card — with staffing trimmed to match, instead of the full menu and full floor. Implementation: a single-card afternoon menu (vermut, conserves, two cold items already in the prep chain); shift one floor position to the 19:00 service. Expected impact (illustrative estimate): the daypart either earns its staffing or stops costing it — both outcomes improve the read. Measurement window: re-read after six weeks.
Action 4 — State the cross-outlet pricing logic. Diagnosis: rooftop premiums on shared items range from 24% to over 100% with no stated rule. Guests compare menus inside the same building; the espresso outlier shows up in reviews. Primary action: set a declared rooftop premium band — 20–30% over ground floor for shared items — and fix the outliers in both directions. Implementation: reprice the espresso into the band on the next menu print; hold the bravas where they are (already inside it); brief both floor teams on the one-sentence answer to “why is it more upstairs?”. Expected impact (illustrative estimate): margin holds; the review-bait disappears; pricing reads as intentional, which is what a premium needs to survive. Measurement window: re-read margins and review mentions at eight weeks.
3. The menu design read
Three menus, read on their own and against each other. The illustrative read for Llevant: Terra’s dinner card buries the Dorada a la sal mid-page below the rices, where the eye lands on the lowest-margin section first; Cel’s card lists the signature gin serve last under a long cocktail run, and the vermut card does not exist yet as a separate afternoon piece; and across the three outlets the description voice shifts from provenance-led (Terra) to ingredient-list (Cel) to none at all (the banquet proposals) — one house voice, applied per outlet, reads as one property instead of three vendors.
4. The market position
Where Llevant’s pricing sits against its comp set — nearby hotels with rooftop bars competing for the same evening guest, and the standalone restaurants within walking distance competing for the same dinner cover. The illustrative read: the rooftop’s cocktail pricing sits comfortably mid-set with headroom on signature serves; Terra’s dinner mains are priced against the neighbourhood rather than against the in-house alternative of room service and delivery apps, which is the comparison its own guests actually make; and the banquet rate card sits below what the events comp set charges for an equivalent package — consistent with Action 1.
5. The review cycle
What to re-examine, and when. Llevant’s illustrative cycle: dinner capture at four weeks, the rooftop afternoon serve at six weeks, banquet pricing after eight contracted events or eight weeks, and the cross-outlet premium band at eight weeks — the four-to-eight-week band, long enough for a change to show in the numbers, short enough that the season has not moved on.
The critic and QA pass
Before any of the above reaches the F&B director, the recommendations go through a critic and QA pass. For this illustrative Verdict the critic would, for example, check that Action 3 survives contact with the rota — trimming the afternoon floor must not leave the 19:00 turn short — and that Action 1 never reprices a signed contract. QA confirms each action carries its full backbone: diagnosis, primary action, implementation detail, expected impact labelled as an estimate, and a measurement window.
What this is, and is not
This is the shape of a multi-outlet Verdict — capture, daypart yield, banquet contribution, and pricing consistency read across three menus, with the same five deliverables and the same critic pass as every engagement. It is not a real result for a real property. Hotel Llevant is fictional and every figure on this page is fabricated to demonstrate the format.
A real Verdict produces property-specific numbers from your actual outlet mix, menu data, and comp set, and every expected impact in it is labelled as an estimate — never a guaranteed figure. If you run a hotel and want the real version, start the conversation.
The independent restaurant example
Five deliverables in five days — what La Brisa would receive.
Below is the Verdict La Brisa would receive — the five-layer methodology run on its menu, the five deliverables, the critic and QA pass before delivery. Only the property is invented; the shape is exactly what your restaurant gets.
Read the full worked example
Illustrative — La Brisa is a fictional property; every figure here is fabricated to demonstrate the format, not a real result. Restaurante La Brisa does not exist. The numbers, classifications, and interventions below are a worked example so you can read a Verdict end to end before you book your own. Nothing here is a measured outcome, a benchmark, or a guarantee.
The property
Restaurante La Brisa is an independent, owner-run restaurant in Barcelona — a single dining room, lunch and dinner service, a menu of around thirty items. It is the kind of room that lives and dies on neighbourhood regulars and table turns. For this worked example it is analysed as a Tier 1 — Local / Casual property, which sets the scoring weights, the metrics that lead, and the voice of every recommendation: plain, practical, and written for an owner who has five other jobs today.
Everything that follows is the Verdict La Brisa would receive — the same five-layer methodology run on a real menu, the same five deliverables, the same critic and QA pass before any of it ships. Only the property is invented.
Five deliverables in five days
A Verdict is the consulting engagement — the deep first read — delivered as five things over five days. Here is each one, as La Brisa would receive it.
1. The item analysis
Every dish on La Brisa’s menu is scored through the five analytical layers and sorted into one of six classifications: Star, Plowhorse, Puzzle, Dog, and the two Couverté additions, Vulnerable Star and Conversion Candidate. Below is an illustrative slice of that analysis — six representative items, not the full menu.
| Item | Classification | Diagnosis (illustrative) |
|---|---|---|
| Esqueixada de bacalao | Star | High draw, healthy contribution margin. Protect it. |
| Pollo a la brasa | Plowhorse | A top-three seller, but contribution margin sits well under the menu average. |
| Arroz negro (per-person) | Puzzle | Strong margin, low order count — buried mid-menu, thin description. |
| Ensalada de la casa | Dog | Low draw, low margin. Earns little, occupies a prime menu slot. |
| Fideuà de marisco | Vulnerable Star | Looks like a Star, but food cost runs well above its category average — exposed to the next seafood price spike. |
| Croquetas de jamón | Conversion Candidate | A Plowhorse with food cost below its category average — headroom for a low-risk move. |
The classification is the start of the action, not the end of it.
2. The action plan
A ranked set of specific changes. The order is part of the recommendation — the methodology ranks by expected impact, and an owner with limited Thursdays works the list from the top. Each action carries the full backbone QA checks for: a diagnosis, a primary action, an implementation detail, an expected impact labelled as an estimate, and a measurement window.
Three illustrative actions, in rank order:
Action 1 — Croquetas de jamón (Conversion Candidate). Diagnosis: high popularity, contribution margin under the menu average, but food cost sits below the category average — the cost base has room to absorb a price move. Primary action: Reprice +€2.10. Implementation: update the price on the next menu print and the daily board; brief the floor that the croquetas portion is unchanged so the team can answer guests directly. Expected impact (illustrative estimate): the below-average food cost absorbs the increase and lifts the dish toward Star; no material change to order count expected at this delta. Measurement window: re-read order count and margin after four weeks.
Action 2 — Arroz negro (Puzzle). Diagnosis: margin is already strong; the problem is visibility — the dish is mid-menu with a six-word description. Primary action: move it to the top-right panel and rewrite the description to lead with provenance. Implementation: on the next print, lift the arroz negro into the top-right panel and extend the description to roughly twelve words leading with the squid-ink and stock detail; no price change. Expected impact (illustrative estimate): a modest rise in order count as the dish becomes visible; margin per cover holds. Measurement window: re-read after six weeks.
Action 3 — Fideuà de marisco (Vulnerable Star). Diagnosis: strong draw and acceptable margin today, but food cost is running well above the seafood category average — one ingredient spike compresses it. Primary action: defend the margin before the market forces the issue — re-cost the recipe and trim the portion slightly toward the category average. Implementation: re-cost with the current supplier sheet, set a target plate cost, and adjust the marisco portion within what the kitchen can hold without changing the dish’s character. Expected impact (illustrative estimate): food cost pulled back toward the category average; the dish stays a Star instead of drifting. Measurement window: re-cost again after eight weeks, or sooner if a supplier price changes.
3. The menu design read
Where items sit, how they are described, and how prices are displayed — calibrated to Tier 1 display rules. For La Brisa the illustrative read flags three things: the Ensalada de la casa (a Dog) holds a prime top-left slot that a higher-earning dish should own; descriptions across the menu run short and value-light, where Tier 1 supports roughly eight-to-fifteen-word descriptions that lead with house-made and locally-sourced detail; and the menu currently highlights four dishes at once, which dilutes the eye — one highlighted item per section does more work.
4. The market position
Where La Brisa’s pricing sits against comparable Barcelona neighbourhood restaurants, and where the convergence risk and the headroom are. The illustrative read: La Brisa’s mains are priced toward the lower end of its comp set, which leaves quiet headroom on the dishes whose cost base can carry it (see Action 1) — and a convergence risk on the rice dishes, where several nearby rooms are clustered at a near-identical price and the menu reads as interchangeable. The recommendation is to use description and provenance, not price, to separate the arroz negro from that cluster.
5. The review cycle
What to re-examine, and when. A recommendation without a date to check it is a guess. La Brisa’s illustrative review cycle: re-read the repriced croquetas at four weeks, the repositioned arroz negro at six weeks, and re-cost the fideuà at eight weeks — typically the four-to-eight-week band, because that is long enough for a change to show in the numbers and short enough that the menu has not moved on.
The critic and QA pass
Before any of the above reaches the owner, the recommendations go through a critic and QA pass. For this illustrative Verdict the critic would, for example, check that Action 3 survives contact with the kitchen — trimming the fideuà portion must not demand an extra prep station during peak service — and that every action in the plan still fits the Tier 1 voice: a clear instruction to a working operator, not a paragraph of fine-dining restraint. QA confirms each action carries its full backbone: diagnosis, primary action, implementation detail, expected impact labelled as an estimate, and a measurement window.
What this is, and is not
This is the shape of a Verdict — five deliverables, five days, the full five-layer methodology behind it, and a critic pass before delivery. It is not a real result for a real property. La Brisa is fictional and every figure on this page is fabricated to demonstrate the format.
A real Verdict produces property-specific numbers from your actual menu mix and comp set, and every expected impact in it is labelled as an estimate — never a guaranteed figure. If you want the real version for your own restaurant, that is the Verdict.
The framework behind it
Read the methodology Every Verdict is run through the public, versioned Couverté methodology — five analytical layers, six classifications, four tier voice rules, a critic and QA pass. Read the methodology end to end before you book.
Get your free Verdict
Five days from now, you have a Verdict like this — on your property.
We reply within 24h · No card required — your first Verdict is free.