Prefactibilidad territorial para nuevos restaurantes (MTIE) mistakes versus the right method

The busiest corner is the one that broke the business

Foot traffic is the most visible variable in any territory and the least predictive of sales, and that confusion explains a large share of early closures. In Medellín, an owner signed a five-year lease on a corner with heavy pedestrian flow and shut down eleven months later carrying 180 million pesos in debt: people walked past, yes, but they were walking toward three shopping centers less than a kilometer away, which captured the spending of that polygon. His average check never reached the 42,000 pesos his financial model required. Add the cost context: ACODRES reported in 2025 a 9.8% rise in menu prices across Colombia to sustain 98,000 jobs, so a model that starts out tight has no room to absorb input inflation. MTIE territorial prefeasibility measures WHERE the income ends up, not how many shoes cross the sidewalk. Four variables decide the prefeasibility of a new location: supply density per block, disposable income inside the polygon, occupancy cost against projected sales, and the food cost you can actually reach with the suppliers who genuinely deliver to that address.

The four variables you measure before signing, and in that order

The order matters as much as the list. Start with household disposable income inside the polygon and with the historical share that household spends on food away from home, because that figure puts a CEILING on everything else; then look at installed supply, which splits that ceiling among competitors; only after that do you negotiate rent. Occupancy cost should stay below 10% of projected sales in a table-service model, and plate food cost should never exceed 32%, which is the tolerable maximum and not the target. Diego F. Parra insists on that sequence with every Masterestaurant client: first the money that exists, then who takes it. The real difference between prefeasibility and an autopsy is not how much information you gather, it is WHEN you deliver it. A flawless study that arrives after the signature protects nothing: it documents the burial.

Prefeasibility is a revocable verdict, not a handsome report

The MTIE instrument exists to produce a verdict while the letter of intent is still alive, when the owner can still walk away from the table without paying a peso, and that right to leave is worth more than any five-year sales projection. Consider what would have happened in Medellín if the study had closed two weeks before signing: the polygon would have shown its spending leak toward the shopping centers, the viable check would have dropped from 42,000 to roughly 28,000 pesos, the model would not have balanced, and the owner would have looked for another address. Eleven months of operation, a full payroll and 180 million pesos hinged on a date, not on one more data point. Every restaurant that closes at the twelve-month mark destroys between 8 and 14 formal jobs that public money paid to create, and that is why SATE Institute approaches this framework through local economic development rather than treating it as the entrepreneur's private affair.

The public cost of every twelve-month closure

In development banking the reading gets drier still: a badly built prefeasibility study does not produce one isolated bad opening, it produces non-performing loans. When the IDB Group or a commercial bank with an MSME portfolio evaluates a hospitality credit line, the determinant of default is neither the chef's talent nor the quality of the menu, it is whether the unit landed in a polygon capable of sustaining it. Sector employment is hardly marginal: over 67% of U.S. adults have worked in restaurants at some point, 78% among Generation Z (National Restaurant Association, 2026). A closure erases entry-level positions no other industry replaces at the same speed. It is worth stating what the measuring stick is. The employment and consumer-behavior data cited here come from the National Restaurant Association and the U.S. Bureau of Labor Statistics —6.2 million people aged 16 to 19 in the workforce in 2024, 900,000 more than in 2019— while the price and cost-pressure figures for Colombia come from ACODRES 2025.

Where these benchmarks come from and how far they reach?

Masterestaurant S.A.S., technology partner of the model and owner of the MTIE software, contributes the operating layer: real input costs by zone, hourly sales curves, and occupancy benchmarks gathered in live operation.

The limits are three and I will not hide them: U.S. data describes a consumption structure different from Latin America's and works as directional reference, not as a parameter; polygon income is estimated from census and cadastral records, which go stale; and no benchmark replaces validation with local suppliers. The three scenarios split by the rent threshold each structure can withstand, not by the size of the dream. Small venue, up to 40 seats and a single strong service: occupancy cost must not pass 8% of projected sales, because without volume there is no dilution of fixed costs, and with a 32% food cost your break-even slides three months out with any increase like the 9.8% ACODRES reported in 2025.

How to read these numbers in YOUR operation?

Mid-size operation of 60 to 120 seats across two services: it tolerates up to 10% occupancy provided the polygon shows disposable income sufficient for a check 20% above the block average.

Group with three or more units: negotiate variable rent on sales and push your purchasing power into food cost, because your advantage lives in input scale, not in the corner. Calculate your threshold before you visit the space. Two thousand people crossing a corner do not equal two thousand potential guests when their reason for moving is to reach somewhere else, and that misreading is the second most common failure in the studies I review. The useful figure is not sidewalk headcount, it is household disposable income inside the polygon multiplied by the historical share that household allocates to food away from home, divided by the supply units already installed per block.

Presence is not capture: two thousand pedestrians are not two thousand guests

Consumption habits skew that calculation further: 37% of adults order delivery at least once a week and over 40% order three to five times a month (UpMenu, 2024), which means a growing slice of polygon spending no longer walks toward any corner at all. If your model depends on foot traffic, you are competing against an app. Here sits a paradox every owner faces and few resolve well: the polygon with the strongest disposable income charges the highest rent, so the location that guarantees demand is precisely the one that breaks occupancy cost. The bridge is not hard negotiation, it is a change of format. With a short menu, high turnover and food cost below 30%, a small unit sustains a rent that a broad-carte restaurant could never carry at the same address; expensive territory gets paid with table velocity, not with price. The other route is variable rent on sales, which shifts part of the territorial risk onto the property owner.

The real tension: the best location is usually the one you cannot afford

My recommendation is explicit: if projected occupancy cost exceeds 12% of sales, stop looking for arguments and look for another address. That number has buried more restaurants than any bad recipe. The real difference is not how much information you gather, it is WHEN. A flawless study delivered after the lease is signed is not prefeasibility, it is an autopsy. MTIE exists to produce a revocable verdict: while the letter of intent is alive the owner can walk away at no cost, and that right to leave is worth more than any sales projection. The second break is confusing presence with capture. Two thousand people crossing a corner are not two thousand potential guests when their reason for travelling is to reach somewhere else; the useful figure is household disposable income inside the polygon and the historical share it spends on food away from home. Diego F. Parra holds the same sequence with every Masterestaurant client: territory first, concept second, never the reverse, because a concept can be corrected in six weeks while a location is corrected only by breaking a five-year contract.

Where the decision breaks?

A genuine tension deserves resolving rather than dodging: the polygons with the best disposable-income indicators also charge rents that blow past the 10% occupancy line.

The answer is neither the poor district nor the premium one, it is measuring the RATIO of capturable sales to occupancy cost, and accepting that the best territory is usually the second ring, three or four blocks off the premium strip, where rent falls 30-40% while capture drops only 12-18%. One warning on menus: if the project contemplates QR, the PHYSICAL menu stays. QR handles delivery, accessibility, price changes and analytics; the printed menu governs service pacing, menu narrative and suggestive selling. A venue that drops the physical menu loses 6 to 11% of average check through suggestive sales never executed, and that figure belongs in the prefeasibility because it moves the required sales threshold.