There is a number every F&B operator has been quoted at some point. A consultant, a tool, a pitch deck — and a single, confident figure for how much margin they will find you. It is always precise. It is always before anyone has looked at your menu.
That number is invented. Not maliciously, usually — it is an average dressed up as a forecast. But it is invented, and once you have run a few operations you can feel it. So Couverte does not quote one. Where a competitor puts a single figure, we put a band — a range, with the research it came from named next to it. This post is about why, and exactly what the band is made of.
Why a single number cannot be honest yet
The output of the methodology is property-specific by design. The Couverte Score uses different weights for a casual room than for a Michelin kitchen. The action plan depends on your actual menu mix, your actual food costs, your actual comp set. The whole point of the framework is that the same finding produces a different recommendation depending on the restaurant — which means the result is different depending on the restaurant too.
So a precise uplift figure quoted before your audit has run is a contradiction. If it were genuinely specific to your operation, it could not exist yet, because the inputs that make it specific have not been gathered. If it exists anyway, it is not specific to your operation — it is a market average wearing your name. There is no third option.
The honest move is to say what we actually know, and to be precise about how much that is. Before your data, what we know is the published research: decades of menu-engineering and revenue-management studies that have measured what disciplined menu work tends to recover, across many properties. That is real knowledge. It is just knowledge about a population, not about you. And knowledge about a population is correctly expressed as a range.
What the band is built from
When Couverte shows an uplift band, it is modelled from published hospitality research — principally the foundational menu-engineering work from the Cornell School of Hotel Administration (Kasavana and Smith 1982) and revenue-management aggregates from the wider industry literature. These are sources about the hospitality sector at large. None of them are Couverte’s own measured results, and the band is labelled that way wherever it appears.
The band is also tier-aware, because the research does not say the same thing about every kind of restaurant. A casual room with room to move on table turns and pricing has a different recovery profile than a hotel F&B operation, which has a different profile again from a group with multiple outlets. A single sector-wide number would flatten exactly the distinction the methodology exists to make. So the band shifts by tier, and it is shown as a low, a midpoint, and a high — because that honestly represents what the research supports.
Two ranges in this paragraph are illustrative, included only to make the shape concrete: a casual-tier band might run from roughly the low single digits to the mid single digits of revenue, while a multi-outlet group band would typically sit tighter and lower, because scale narrows the variance. Illustrative — your numbers will vary, and the figures shown on the pricing page are the ones modelled from the research. The point is not the specific percentages. The point is that it is a range, it moves by tier, and it never pretends to be your result.
What replaces the band
A band is the right answer before your audit. It is the wrong answer after.
This is the actual function of the Verdict. When Couverte runs your audit, the research-cited band is replaced by a property-specific estimate built from your real menu mix, your real costs, and your real comp-set position. That estimate is still labelled an estimate — a modelled projection, not a guarantee, because the only honest version of a forward number is one that admits it is forward. But it is yours. It is built from your inputs, not borrowed from a sector average.
And there is a longer arc here worth being straight about. As Couverte accumulates measured outcomes — real audits with real before-and-after results — the bands themselves get rebuilt on Couverte’s own data instead of the published research. When that happens, the change will be stated plainly: the source will say so, and it will carry a sample size, because a coefficient drawn from a known number of measured audits is a different kind of claim than one drawn from a sector aggregate, and you deserve to know which one you are looking at. We are not there yet. Until we are, the research-cited band is what we have, and labelling it honestly is the whole discipline.
Honesty as the position, not the disclaimer
It would be easier to quote the confident single number. It converts better in the short run. Precise figures feel like competence; ranges feel like hedging.
But an operator who has been in the business long enough has been burned by the confident number before, and has learned to read it as a tell. A band with its sources named is not us being cautious. It is us showing our work — here is what we know, here is exactly how much of it is about you specifically, here is where the line is between research and your result. That line is the product. Couverte is consulting rebuilt as software, and the thing that makes consulting worth paying for has never been the confidence of the number. It is the rigour behind it, and the honesty about its edges.
If someone shows you a precise uplift figure before they have seen your menu, you already know what it is. We would rather show you the range, name the research, and earn the specific number the honest way — by running your audit.