The methodology behind every Verdict.

“How hard is this dish working for us relative to the seat and hour it occupies?”

A dish does not exist on a spreadsheet. It exists in a service period, at a table, in a dining room with a fixed number of seats. Layer 2 measures the dish against that reality — the seat it holds and the hour it holds it for. A tasting plate can score beautifully on margin and still be the wrong dish if it pins a four-top for ninety minutes during your busiest service.

A one-sentence summary of the math

Divide the profit a dish generates by the seats and hours it consumes — so a dish is judged not just on what it earns, but on what it earns per seat, per hour.

What it measures

Layer 2 needs the outlet’s seat count, the operating hours per service period, the average dining duration by course type, and seat occupancy data where the reservation system can provide it.

From those it calculates the time-and-space metrics:

• RevPASH — revenue per available seat hour. How much top-line each seat generates per hour of service.
• ProPASH — profit per available seat hour. The sharper figure: a busy seat turning low-margin covers can post a strong RevPASH and a weak ProPASH. Revenue and profit do not always move together, and ProPASH is where that gap shows.
• Seat utilisation ratio — covers served against the turns the room could physically deliver. A casual room running well below its possible turns is leaving capacity unsold every service.
• Item ProPASH contribution — each dish’s weighted margin expressed against the seats and hours it occupies.

What Layer 2 means depends on the restaurant. For a casual room, turns are the whole game. For hotel F&B, ProPASH by daypart shows where the real money is — breakfast can be high-volume and low-margin while dinner is the opposite. For fine dining, low turnover is the point, so the figure that matters is profit per cover. For a Michelin kitchen, the tasting menu fixes both duration and price, and time-space efficiency nearly drops out as a primary measure.

One concrete intervention

A casual lunch outlet is turning its tables noticeably below the number the room could deliver. Layer 2 reads that gap as unsold capacity and points the recommendation at speed, not price: move two slow-firing dishes off the express-lunch section and replace them with items the kitchen can plate without bottlenecking the pass. The covers were always available — the menu was holding the seats hostage.

“One number that tells you how this item is really performing.”

Layer 3 is the proprietary part. It takes everything Layer 1 and Layer 2 produce and meshes it into a single score per dish — weighted by the kind of restaurant you run. It is the layer that turns a wall of metrics into one figure an operator can rank a menu by.

A one-sentence summary of the math

Normalise five components to a common 0–100 scale, multiply each by a weight that shifts with your restaurant tier, and add them up — so the same dish scores differently at a casual room than it does at a Michelin kitchen.

What it measures

The Couverte Score combines five components: contribution margin, popularity, food cost efficiency, ProPASH contribution, and brand alignment. Each is normalised to 0–100 within its category, then weighted.

The weights are not fixed. They shift across the four tiers:

• At a casual restaurant, the economics dominate — margin, velocity, and food cost efficiency together carry most of the score.
• At hotel F&B, ProPASH rises in weight, because profit per seat-hour across dayparts is where the real money is.
• At a fine-dining room, brand alignment climbs sharply.
• At a Michelin kitchen, brand alignment is the dominant weight — creative coherence and guest experience come first, profit second.

Brand alignment is the component no published model includes. It answers one question: does this dish belong on this restaurant’s menu? A truffle pizza can score high on margin and popularity at a casual Italian spot and still be wrong if the place positions itself as a simple, honest trattoria. A low-margin amuse at a Michelin restaurant can score poorly on economics and still be essential to the tasting narrative. It is assessed against the restaurant’s stated concept, its ingredient-sourcing identity, the kitchen’s technique level, and whether the dish matches guest expectations for its price tier.

One concrete intervention

A signature pasta at a fine-dining room scores in the mid-range — solid margin, modest velocity. At a casual tier, that score would push a recommendation to reprice or reposition for volume. But at the fine-dining tier the weight on brand alignment is high, and this dish scores at the top of that component: it is exactly what the restaurant is known for. The Score reads the dish as a brand anchor, not a margin problem, and the recommendation protects it — hold the price, keep its menu position — rather than chasing turns it was never meant to deliver.

“What does the world outside this kitchen demand?”

Layer 4 wraps the item-level analysis in market reality. The same dish, in the same kitchen, can earn a different recommendation depending on what is happening in the market around it. The first three layers can be done by anyone with a spreadsheet and patience. This one cannot — it needs data from beyond your four walls.

A one-sentence summary of the math

Compare each item’s price and each menu’s overall position against aggregated data from comparable restaurants in the same market — so a recommendation reflects not just your numbers, but where your numbers sit against everyone else’s.

What it measures

Layer 4 reads three things:

• Competitive positioning — where each item’s price sits relative to comparable dishes at comparable restaurants in the same market, whether the menu’s overall price point is aligned to its tier and location, and whether there are price-convergence risks, such as casual-dining items drifting close to fast-casual prices and triggering a value-perception problem.
• Consumer-segment behaviour — how spending is moving by income band. High-income guests are spending freely at fine dining and upscale casual; middle-income guests are pulling back from fast casual, trading down to value or up to “worth it” experiences. The casual price battleground is where that pressure concentrates.
• Seasonal and demand patterns — which items show real seasonal demand curves, where there are daypart opportunities, and how the local event calendar moves demand.

This is the layer where the benchmark database becomes critical. Saying “your grilled sea bass is priced below the market median for comparable hotel restaurants in this city, and items at the next price point up show stronger wine-pairing attach rates” requires aggregated data across many properties. It is the part of the methodology that grows sharper with every restaurant on the network — a dataset no single operator can assemble alone.

One concrete intervention

A 5-star hotel restaurant prices its grilled sea bass conservatively. Layers 1 through 3 see a healthy, well-behaved dish and leave it alone. Layer 4 sees something the in-house view cannot: the price sits below the market median for comparable rooms in that city, and the gap is pure margin left on the table. The recommendation closes part of that gap — a measured increase toward the market band — with the confidence that comparable guests are already paying it elsewhere.

“Here’s what to do, in language that fits your operation.”

Layer 5 is where the four analytical layers converge into something you can act on. This is the layer that separates the methodology from a report. The key insight: the same analytical finding produces a different recommendation depending on the tier.

A one-sentence summary of the math

There is no formula here — Layer 5 is judgment: it takes the score, the classification, and the market read, and writes the single highest-impact change, calibrated to what your kind of restaurant can realistically execute.

What it measures

Layer 5 does not measure — it prescribes. Take a Puzzle: a dish with a high margin and low popularity. The finding is identical across every tier. The recommendation is not:

• At a casual restaurant, the move is mechanical and volume-focused — reposition the dish to a prime spot on the menu, sharpen its description, test its price elasticity.
• At hotel F&B, it becomes a service play — feature it as a chef’s recommendation, brief the team on a two-sentence spoken description, test it through room service.
• At a fine-dining room, it becomes a menu-architecture play — work it into the tasting menu, or position it next to the highest-priced item as an anchor.
• At a Michelin kitchen, it becomes a narrative question — does the dish serve the menu’s arc? If it does, treat it as a signature course; if its popularity is low because it sits at the wrong moment in the sequence, reposition it rather than reprice it.

Every action plan, whatever the tier, has the same backbone: a diagnosis (which classification, which score components are weak, and why), a primary action (the single most impactful change — price, position, description, portion, or removal), an implementation detail specific enough to hand to the F&B manager, an expected impact, and a measurement window for when to re-evaluate, typically four to eight weeks out. Premium tiers add an experience-enhancement step — a hospitality moment, not just a profit move.

Menu psychology is woven in here, not bolted on. How a price is displayed, how long a description runs, how a section is laid out — all of it is calibrated by tier inside the recommendation itself.

One concrete intervention

A high-margin starter is barely selling at a casual bistro — a textbook Puzzle. Layer 5 does not say “promote it.” It writes the action: move the dish to the top-right of a two-panel menu, give it a short description that leads with provenance, and trim the price by a small, testable amount to probe demand. Then it sets a four-week window to look again — because a recommendation without a review date is a guess, not a plan.