Restaurant value proposition: traditional method vs the Masterestaurant method

Step 1 — Write down the customer before the dish (deliverable: a one-page profile)

Start with the customer, never with the menu: what this first step must produce is a one-page profile naming the segment, the exact hour they eat, the budget per head and the reason they leave home or the office, and it counts as done when you can recite it without looking. That owner in Bogotá who opens this guide had dozens of dishes and no customer written anywhere, with a prime cost far too high that he blamed on suppliers; the menu was the invoice for that vagueness. Sequence matters more than it looks, because a location chosen for a customer who never shows up costs a large amount of misdirected build-out, and advertising will not fix it. Write ONE profile, not four. If four segments come out, you have not decided anything yet.

Step 2 — Draft the hypothesis in eleven words and make it falsifiable

A value proposition works when it fits in eleven words and a measurement can prove it wrong. «Home cooking made with love» can be neither tested nor refuted, so it is not a proposition: it is a wish printed on the facade. «The door-to-door executive lunch for financial-district office workers» is one, and it collapses the day average delivery time stretches too far or monthly repeat visits fall off. Your deliverable here is a sentence in the present tense, with a service verb, a time or a price and a named segment, taped up in the kitchen where the crew reads it. Verify it by asking three employees to repeat it from memory the next day: if two paraphrase it wrong, it is still too long. Diego F. Parra demands it in writing before anyone touches a single supplier.

Step 3 — Cost backwards: set the anchor dish food cost first, then build the recipe

Set the target food cost of the anchor dish first and decide the recipe afterwards, never the reverse. The traditional method prices the plate, looks at what is left at month end and calls that leftover margin; we start from the ceiling, a food cost per dish that is the MAXIMUM tolerable figure, never the goal, and work backwards into portion weights, supplier and acceptable waste. For example, an anchor dish with a moderate food cost and a set price leaves a contribution margin per unit, and from there you can finally calculate how many plates per service pay the rent. The deliverable is one sheet with the standardized recipe, cost per portion and contribution margin; verify it by weighing three random portions during service and confirming the deviation stays under 5%. Payroll and rent never load onto the plate: they live in the break-even calculation.

Step 4 — Cut the menu until every dish defends the hypothesis

Cutting the menu is the physical proof that the proposition exists. Take the profile from step 1 and the sentence from step 2, and run every dish through two dry questions: does the segment I wrote down order this dish? does it share at least two inputs with the anchor dish? Whatever fails both goes. Out of a long menu, only a fraction usually survives, and the cash effect shows up on three fronts at once: fewer purchase references, less spoilage from slow rotation, shorter ticket time in the kitchen. The deliverable is the new menu plus its shared-input matrix; verify it by counting how many distinct ingredients come through the back door each week, before and after. This is where owners dig in, because removing a dish feels like removing a sale, and that instinct is precisely what sustains the 74%.

Step 5 — Pick three control figures and run them on a 30-day board

Three figures and no more: that is the board. Pick repeat customers, average ticket of the target segment and weekly real food cost of the anchor dish, then log them every Monday for thirty days on a sheet that fits on one page. Against that board the step 2 hypothesis is proved or knocked down, which is why all three figures must move because of your decisions, not because of weather or seasonality. The deliverable is the template with four rows of data by month end; it is verified because every cell carries a date and the signature of whoever measured it. If by day 30 real food cost differs by more than three points from the step 3 target, the problem sits in kitchen execution rather than in the proposition, and that distinction saves you from rewriting a hypothesis that was sound to begin with.

The four mistakes that wreck this guide, and how to dodge them

The most expensive mistake is widening the menu when the board reads badly, and almost everyone makes it in week three. Widening is the instinctive answer to a sales dip and it is exactly what pushes prime cost back up, so the hard rule reads: during those thirty days no dish gets added, whatever happens. The second mistake is tracking four or five figures, because nobody fills a five-column board on Monday at seven in the morning. The third is changing prices halfway through the experiment, which destroys the average-ticket reading. And the fourth, quieter one, is leaving the hypothesis inside the owner's head instead of taping it in the kitchen, because a server who cannot recite it cannot sell it either. Four mistakes, four rules: do not widen, do not overmeasure, do not move prices, do not keep the sentence to yourself.

What happens if the hypothesis collapses on day 30?

If the board refutes your hypothesis, celebrate: you just bought information cheaply, with thirty days instead of two years of losses. The underlying statistic is harsh and worth keeping in view:

according to the U.S. Bureau of Labor Statistics (BDM) data show that most restaurants survive their first year, roughly half reach five years and only a minority are still standing at ten, and suppliers do not explain that curve — decisions that were never put to the test do. A fallen hypothesis gets corrected by changing ONE variable —the segment, the hour or the anchor dish, never all three— and then you run another thirty-day cycle. Change all three at once and you land back at the starting line having learned nothing, which is the elegant way to lose a year.

Closing checklist: how to know everything landed

Everything landed when you can tick six boxes without arguing about any of them. One: the segment profile exists on paper and you recite it without reading. Two: the eleven-word sentence hangs in the kitchen and two out of three employees repeat it meaningfully. Three: the anchor dish has a standardized recipe, cost per portion and food cost under 32%. Four: the menu shrank to a small core of references and every dish shares inputs with the anchor. Five: the board holds four weeks of dated, signed data. Six: you decided in writing whether the hypothesis holds, gets adjusted on one variable, or gets discarded. If any box sits half done, do not launch another cycle: finish that one. The discipline of closing before opening is what separates a method from a good intention with a calendar.

The four differences that move the till

SEQUENCE. The traditional method picks the site, then the menu, then the guest; the Masterestaurant method runs exactly backwards, and that inversion is what saves a large share of misdirected build-out, because a site chosen for a guest who never shows up cannot be fixed with marketing. FALSIFIABILITY. A value proposition you cannot disprove is a wish, not a proposition. «Home cooking made with love» cannot be tested; «the 22-minute door-to-door executive lunch for financial-district office workers» can, and it collapses the day average delivery time slides well past that promise and the repeat rate drops below what the operation needs. COSTING IN REVERSE.

The four differences that move the till — in practice

Traditional sets the price and then checks what's left; we fix the anchor dish food cost target first, with a hard ceiling that is never a goal, and derive price from there, loading payroll, rent and utilities onto the break-even point of the business and NEVER onto the plate, which is the most expensive and most common costing mistake in the trade. CHANNEL GRANULARITY. A virtual restaurant, a dark kitchen and a 40-seat dining room share no proposition and no revenue structure even when they share a kitchen: according to the Independent Restaurant Coalition (2025), aggregators charge up to 15-30% per order, which devours a menu designed for table service, which is why the delivery menu carries a short list of transport-resistant references at recalculated prices rather than the full dine-in menu with a surcharge.