Masterestaurant Index of Profit per Seat & per m² 2026: the average seat earns $3,180 a year — not what your P&L says

Finding 1 — How much does a single seat actually earn per year?

The average restaurant seat earns $3,180 a year, but that figure hides a brutal spread: across 214 Masterestaurant audits from 2023 to 2026 the range runs from $1,940 to $5,720 depending on segment.

The traditional method averages everything into one P&L and profitability shows up as a single number that blends the table turning six times with the one turning once. Diego F. Parra sees it again and again: the owner knows monthly sales and global food cost, but has no idea what seat 14 or the back-corner square meter earns. That average is comfortable and misleading. The right unit of analysis is not the restaurant or the dish: it is the seat and the square meter. When you push EBITDA down to that granularity, the reassuring 11.4% aggregate breaks apart into seats that quietly subsidize others that are bleeding cash. The traditional P&L average hides the spread, and that is where margin dies.

Finding 2 — The healthy aggregate hiding a seat in the red

Across the 214 Masterestaurant audits, with the exact same menu food cost, the best seat earned $5,720 a year and the worst lost $820. The aggregate read 11.4% EBITDA and everything looked perfectly healthy. That is the mistake Diego F. Parra finds in most venues: the global number works as anesthesia. Nobody fixes a seat losing $820 a year because the P&L never isolates it; it stays buried under the ones turning six times per service. Multiply that seat by ten in a dead zone and you are giving away $8,000 to $12,000 a year without seeing it. The per-seat index turns a soothing average into a heat map: which seat produces, which drains, and exactly how much each one adds to or subtracts from the final EBITDA. Rent is paid per square meter, yet the traditional method never assigns it to the unit that consumes it.

Finding 3 — Rent is paid per square meter, but the P&L never brings it down to the meter

When you push the cost per m² down to each floor zone, the dining-room meter yields $2,410 a year on average, while the cold zone —back of house, near restrooms or the kitchen— yields 46% less and drags the whole result down. It is the same logic retail uses, sales per m², applied to the real cost of running a table with its food cost, its prime cost, and its imputed slice of rent and payroll. A venue pays the same for the dead-corner meter as for the window meter that bills three times more. The Masterestaurant method separates those meters: when you see that 22% of your surface yields half, you stop treating the room as one homogeneous block and start redesigning flow toward where the meter actually pays the rent. Surgical repricing wins margin without touching the average customer's perception, something a blanket menu hike cannot do.

Finding 4 — Surgical repricing instead of an 8% hike across the whole menu

Instead of raising everything 8%, the per-seat index flags which seat and which time slot tolerate price: the two-top by the window on Friday night absorbs a +12% the guest never registers, while the communal midday-weekday table won't tolerate even +3%. Diego F. Parra does it the opposite way from the average consultant: he doesn't raise the menu, he raises contribution margin per seat where demand is rigid. With the menu untouched in the eyes of 70% of guests, margin per seat rises between 4 and 9 points depending on segment. A linear hike, by contrast, punishes the seat already earning and the dead one alike, and usually scares off traffic in exactly the fragile slots you needed to protect. EBITDA per seat is calculated by assigning each seat its real sales, its food cost, its prime cost, and its slice of rent and payroll based on the meters and hours it occupies.

Finding 5 — How EBITDA per seat and per meter is calculated

This is not a napkin estimate: across the 214 Masterestaurant audits from 2023 to 2026 the model crosses tickets per table, turnover by slot, and electrical and staffing consumption by zone. A seat turning six times with a $24 ticket and 30% food cost earns a very different contribution margin than one turning once with a $52 ticket. After subtracting rent imputed by its m² and its slice of floor payroll, the net EBITDA of that seat appears: the $3,180 average, with its $1,940 to $5,720 range. That granularity is what lets you decide: relocate tables, close a zone during dead slots, or reprogram shifts where the meter doesn't pay. A seat losing $820 a year is not scrapped on reflex: it is redesigned or repurposed before you discard it. In the Masterestaurant audits, three levers recover most red seats. First, relocate the flow: turning the dead table toward a higher-visibility circulation lifts its turnover between 20% and 40%.

Finding 6 — What to do with the seats and meters draining capital

Second, reprogram the slot: if the meter only loses during the midday weekday shift, closing that zone in that shift trims imputed payroll without sacrificing sales. Third, change the cold meter's use —from a two-top to a waiting bar or a delivery pickup area that yields per m² without occupying productive floor. Diego F. Parra insists on the order: first you measure seat by seat, then you decide. Applying these three levers, the average audited seat rose from $3,180 to a $3,900 to $4,600 range in twelve months, without touching menu food cost. The traditional P&L average hides the dispersion: across 214 audits, the best seat earned $5,720/year and the worst lost $820, with the same menu food cost. The aggregate showed 11.4% EBITDA and everything looked healthy. Rent is paid per m², but the traditional method never brings it down to the unit.

Finding 7 — What changes when you measure per seat and per m² instead of per site

Imputing $/m² to each zone, the dining-room square meter yields $2,410/year on average, while the cold zone —back, near restrooms or kitchen— yields 46% less and drags the result down. Surgical repricing versus a blanket increase: instead of raising the whole menu 8%, the index flags which seat and which time slot tolerate price. Contribution margin per seat rises without touching the average guest's perception.