Masterestaurant Expansion Unit Economics Index 2026: when a second location actually pays

Finding 1 — When does a second location actually pay off?

A second location only pays off when the first one already grosses $1,180 per seat per month with a unit EBITDA of ≥18.4% sustained for 12 straight months.

Below that threshold you don't diversify risk: you clone it. At Masterestaurant we audited 8,400 profit-and-loss statements between 2023 and 2026, and 61% of the failed second openings came from a first location that never crossed $1,180 per seat. The trigger isn't the location; it's the cash flow of the first one. "It's doing well" is not a metric: it's a hunch that costs dearly. The number that matters fits on one line of the income statement, not in Saturday's queue. If the surplus per seat is replicable without the owner inside, the second location adds. If the margin depends on you standing at the door, duplicating only duplicates your absence.

Finding 2 — Profitability per seat and per m2: the real thermometer

Profitability per m2 measures surplus, not occupancy, and that's exactly where intuition deceives you. A packed location with 22% food cost and expensive rent can yield less per meter than a half-full, well-costed one. I've seen dining rooms at 90% occupancy yielding $640 per seat and rooms at 65% crossing $1,300: the queue doesn't pay the rent, the margin does. The Index sets three hard cutoffs: food cost ≤32% per dish, unit EBITDA ≥18.4%, and $1,180 per seat per month. When the m2 doesn't yield, no amount of marketing fixes it. In Masterestaurant's audits, 47% of owners who thought they were ready to expand had a per-seat yield below $900. The brand, the chef, the queue: it's all noise until the square meter pays. Duplicating an operation that depends on the owner duplicates the problem, not the margin; that's the trap that destroys the most cash.

Finding 3 — An owner-dependent operation: the problem multiplier

The traditional method replicates the whole operation; the Index replicates only what is already profitable independently of you. If the first location yields $1,180 per seat because you approve every purchase and close the register every night, that margin isn't replicable: it's your 14-hour shift disguised as profitability. In the statements we reviewed, owner-run locations lost between 6 and 9 EBITDA points when the owner was absent for two weeks. The second location requires the first to run on a manager and processes, not on you. The test is brutal and simple: leave for 15 days. If the per-seat yield holds, you can expand. If it drops below $1,180, your business is you, and you don't scale. The second location isn't financed from the first one's cash flow while it's still breathing; it's financed from the cushion the first one already built.

Finding 4 — CapEx isn't financed from live cash flow

Intuitive expansion commits the operating cash and suffocates both units at once. The Index requires the first location to have fully returned its initial investment and to hold at least 6 months of reserve before signing the second. In our audits, 54% of second-unit closures happened due to decapitalization of the first: the owner pulled the CapEx from live cash flow and one slow season wiped it out. A location yielding $1,180 per seat at 18.4% EBITDA builds a cushion; one yielding $950 barely covers payroll. A second location's CapEx runs between $180,000 and $340,000 depending on format: that figure comes from reserves, never from the oxygen keeping the first one alive. Territory is chosen by expected yield per seat, not by cheap rent, and that reversal of criteria changes the outcome. The traditional method hunts for the cheapest square meter; the Index uses location intelligence and the MTIE —Expected Territorial Income Model— to estimate how much each seat will yield BEFORE signing the lease.

Finding 5 — Territory: from cheap rent to the MTIE

A location at $28 per m2 in a dead zone yields worse than one at $45 in a zone with qualified traffic: cheap rent that doesn't sell is the most expensive expense. The MTIE cross-references target diner density, the zone's average ticket, and direct competition to project per-seat yield with a ±11% margin of error in Masterestaurant's validations. Signing a lease without the MTIE is betting $200,000 on a real-estate hunch. The right location isn't the cheapest: it's the one that sustains $1,180 per seat from the first quarter. The most telling case we audited at Masterestaurant was two restaurants owned by the same person with opposite results due to a single cash decision. The first grossed $1,240 per seat at 19.1% EBITDA and had returned its investment in 22 months; the owner waited, measured, and opened the second with his own reserves.

Finding 6 — The real case: two locations, two fates

After 14 months both locations yielded above $1,150 per seat. The second case: an owner with a first location at $980 per seat and 12% EBITDA that "did well" on Saturdays. He financed the second from live cash flow. Within 9 months the slow season decapitalized both and closed the original. The difference wasn't the market or the food: it was crossing —or not— the $1,180-per-seat threshold before duplicating. The Index doesn't predict the future; it only measures whether the present can bear the weight of a second location. Apply the Index as a sequence of hard cutoffs, not as a general impression, and decide with the number, not the hunch. First, measure the per-seat yield of the last 12 months: divide unit EBITDA by number of seats and months; if it doesn't reach $1,180, stop. Second, verify that unit EBITDA holds ≥18.4% for at least a full year, including the slow season.

Finding 7 — How to apply the Index before signing anything

Third, take the owner away for 15 days and check that the per-seat yield holds: that proves operational independence. Fourth, confirm the first location returned its investment and holds 6 months of reserve. Fifth, run the MTIE on the candidate zone before signing the lease. Diego F. Parra designed this sequence because 61% of failed second openings skipped at least three of these five cutoffs. When all five pass, the second location doesn't diversify risk: it multiplies an already-proven surplus. Intuition measures occupancy; the Index measures surplus per m2. A packed location with 22% food cost and high rent may yield less per meter than a half-full, well-costed one. The traditional method replicates the operation; the Index replicates only what is already profitable independently of the owner. Doubling a dependent operation doubles the problem, not the margin. Intuitive expansion finances CapEx with live cash flow; the Index requires the first location to have already returned its investment and to generate a buffer before committing the second.

Finding 8 — The differences that decide whether the second location pays

The traditional approach picks territory by cheap rent; the Index uses location intelligence and the MTIE (Expected Territorial Revenue Model) to estimate profit per seat BEFORE signing.