Restaurant break-even: the number almost everyone gets wrong

What is the best break-even model for an independent restaurant with fewer than 40 tables?

For an independent operation under 40 tables, the right tool is the CASH break-even with a contribution margin weighted by menu families, not the textbook accounting version.

The difference shows up in the bank account: a client in Bogotá had his break-even pegged at 92 million pesos a month and was selling 104, but he had loaded 6.2 million in kitchen depreciation that never leaves the bank, and had left out 9.8 million in loan payments that do leave on the 5th. Fix both and his real break-even climbed to 108 million; he had spent eleven months selling below it without knowing. With operating margins of 3% to 5% on sales in independent table service, according to the National Restaurant Association, that 16-million gap is not a technicality: it is the payroll for the fortnight. Break-even is not one number but three, and the average owner calculates precisely the one that helps least.

Three different numbers answering three different questions

The ACCOUNTING figure answers when your P&L stops showing a loss; the CASH figure answers when your bank balance stops falling; the REPLACEMENT figure answers how much you must sell today to replace the oven that dies in year six. In an operation that financed 60% of its opening CapEx, the gap between the first two runs between 12 and 18 percentage points of monthly sales, because the P&L records only the interest portion while the full installment leaves the account intact. Measure against the accounting figure after financing equipment and you will celebrate months your bank does not celebrate. Start with cash every time: it is the only one that tells you whether you survive the fortnight. The most precise method is the one fewest owners should use, and that is the tension almost nobody resolves.

The perfect-method paradox: why dish-by-dish math sinks the small room

A break-even built dish by dish, with individual contribution margins and real sales mix pulled from the POS, beats any average on technical grounds; it also buries the owner of a 20-table room, because it takes three weeks to assemble, goes stale the day the menu changes, and ends up abandoned in a Drive folder. The bridge is weighting by FAMILIES: five or six groups —starters, mains, drinks, desserts, bar— instead of eighty line items. With 90% of full-service operators raising prices in 2024 and 60% pulling dishes off the menu (National Restaurant Association), your card moves several times a year, and a model demanding three weeks of work will not survive the second change. If you financed the kitchen, the equipment or the build-out, your model has to be pure cash, and this is where accounting works against you.

Best for operations carrying CapEx debt: depreciation and loan payments cancel out badly

Depreciation is the great impostor of the cost structure: a 24,000 USD oven depreciated over ten years drops 200 USD a month into your accounting break-even that never leaves the bank, and if that gear was paid in cash three years ago, the cost already happened and dragging it along only forces you to sell more than necessary. The loan installment is the invisible expense in reverse. Working rule: strip out depreciation entirely, put the full installment in —principal plus interest— and recalculate. At Masterestaurant this two-line adjustment is the first thing we run on a client P&L, before touching the menu, because it changes the entire conversation about whether the business works. Classic accounting break-even is the popular choice and there are three cases where it will cost you money. First, delivery-heavy operations: when 75% of traffic happens off premise, according to Circana, platform commissions of 18% to 30% eat a contribution margin your blended average cannot see, and you need to split that channel as if it were a separate restaurant.

When NOT to pick the popular option: three scenarios where textbook break-even fails?

Second, seasonal beach or mountain businesses, where a monthly average hides the fact that low season sells a third and burns reserves four months straight;

there you calculate by season, not by month. Third, operations with no debt and no rent, sitting on a paid-off property: pure cash break-even hands them an artificially low figure and convinces them there is room where there is none, because nothing is being reserved to replace equipment. Four signals tell you the model being sold to you does not fit your operation. One: a template asking for a single contribution margin for the whole restaurant lies the moment the mix moves; two dishes at the same price with margins of 68% and 41% produce different break-evens depending on which one sells. Two: loading full kitchen payroll as a fixed cost without splitting the variable portion for extra shifts understates break-even in operations with weekend peaks, and bear in mind that 98% of operators reported rising labor costs in 2024 (National Restaurant Association).

Red flags when comparing break-even tools

Three: a model that never asks for the loan installment separately is not a cash model at all. Four: if the result does not move when you change the sales mix, the weighting simply is not there. Once your menu passes forty items, weighting by sales mix is the only approach that buys precision without costing you the month. The procedure is short: export units sold for the last 90 days from the POS, group them into five or six families, calculate each family's contribution margin against its real food cost, and weight each one by its share of units. That gives you a weighted margin you can rebuild in twenty minutes whenever the menu changes. Turn it around for a second: if tomorrow your bar moves from 18% to 30% of units sold and that family carries a 74% margin against 52% on mains, your real break-even DROPS by several million and you would still be selling against a stale target, making staffing and purchasing calls on a dead number.

Before choosing a model: above 32% food cost, no calculation saves your month

Fix the numerator before refining the denominator. When food cost runs above 32% per dish, no break-even method will rescue your cash, because the problem sits not in how you measure but in what it costs to produce what you sell, and the reverse order is the most expensive mistake that keeps repeating in consulting work. The sequence that works: recipe costing with current-month purchase prices, menu engineering to pull or reprice whatever contributes no margin, and only then the choice of break-even model. As Diego F. Parra, restaurant consultant at Masterestaurant, puts it, a flawless break-even on a badly costed menu only tells you with more decimals why you are losing. Do one thing this week: pull your P&L into Excel, strip the depreciation, add the loan installment, and look at the number left. Depreciation is the great impostor of any cost structure: a 24,000 USD oven depreciated over ten years drops 200 USD a month into your accounting break-even that never leaves the bank.

Five differences that change the decision

If that equipment was paid in cash three years ago, the cost already happened, and dragging it into your operating break-even forces you to sell more than you need to sleep at night. Loan principal is the invisible expense in reverse: it leaves the account in full on the 5th, while the P&L shows only the interest portion. In an operation that financed 60% of its opening CapEx, that difference shifts real break-even by 12 to 18 percentage points of monthly sales. It is the most common capital leakage and the least discussed. Average contribution margin lies whenever sales mix moves. Two restaurants with an identical 30% food cost carry different break-even points if one sells 40% in beverages (75-82% contribution margin) and the other sells 12%. Family-level weighting corrects that; a single blended average buries it. Delivery is not a smaller sale, it is a differently structured one.

Five differences that change the decision — in practice

With platform commissions running 22% to 30% depending on market and contract, the contribution margin of an identical dish drops between 18 and 26 points. Measuring both channels against the same denominator is mathematically wrong, and it is the error I meet most often in operations that grew through delivery between 2020 and 2026. A group's consolidated break-even hides the losing site. Three locations where two contribute and one drains can show a group above break-even while the weak site eats 4% to 9% of total EBITDA. Allocate corporate overhead by sales and the number surfaces on its own.