SATE Institute analysis of territorial prefeasibility for new restaurants (MTIE) 2026: before vs after: impact and employment

Finding 1 — What does territorial prefeasibility measure that a sales projection does not?

It measures the territory, not the owner's hope. A sales projection starts from what the entrepreneur expects to bill;

territorial prefeasibility starts from public data series that existed before he ever thought about opening, and that is why it is useful for saying no while saying no still costs nothing. The typical credit file for a small restaurant business arrives with three years of projected cash flow and zero measurement of the environment meant to sustain it, and the investment officer, who has never costed a plate, ends up rating a number nobody ever published. MTIE organizes already-published indicators —employment, financial inclusion, food loss— and crosses them against the project's cost structure. The restaurant sector's weight in a territory's economy is a data point about the territory, not the venue, and it should be documented with that market's own source.

Finding 2 — The moment the data arrives decides whether it is prefeasibility or an autopsy

A study delivered after the lease is signed amounts to a forensic report. What separates the before and after of MTIE is not how much information there is —information is already overflowing— but the distance between the document's date and the contract's date, because that gap determines whether the capital can still be pulled back. When the data lands with the venue already taken, the owner reads to justify; when it lands two months earlier, he reads to decide. That is where the reading matters: the ILO reports in its Labour Overview of Latin America and the Caribbean 2024 a regional youth unemployment rate of 13.8%, nearly triple the adult rate, and that pool of available labor completely rewrites the payroll assumption of a neighborhood project. Read it before signing and it is leverage; read it afterward and it is an excuse.

Finding 3 — No figure enters the file without the organization that published it

A number without a named source is not evidence, it is an opinion with decimals. Territorial prefeasibility replaces in-house numbers with verifiable series, and the rule is strict: the ILO for employment, the World Bank for financial inclusion, ReFED and UNEP for food loss and waste, the National Restaurant Association for sector structure. I got this wrong for years, accepting projections built on round assumptions because they were well argued.

Finding 4 — SDG 8: the restaurant as the employability infrastructure nobody built

The sector is the largest labor entry door in the country, and it deserves to be said without corporate-responsibility decoration. The National Restaurant Association (2026) reports that 51% of adults had their first job in a restaurant, and that figure is no charming anecdote: it is the argument that gets a food project into an impact credit line on different terms. The same association documents in 2024 that 46% of restaurant managers in the United States belong to minority groups, the highest share of any sector. On the other side, the ILO records that female informal employment in the region grew 22.8% during 2024 against 15.7% among men, and that asymmetry is precisely what a territorial indicator must capture before the project promises to formalize forty positions its break-even point will never carry.

Finding 5 — SDG 12: the waste you see at the landfill before it shows in the P&L

Food loss and waste is a territorial cost, not merely a kitchen problem. ReFED (2025) calculates that US foodservice generated USD 157 billion in surplus food during 2024, equal to 14% of sector sales, with 12.4 million tons of waste of which 9.73 million —78.4%— ended up in a landfill. UNEP, in its Food Waste Index Report 2024, widens the frame: 290 million tons wasted by food services in 2022, inside a global total of 1.05 billion while 783 million people go hungry. Translate that 14% into your own project and you will see that the food cost variance you chase with standardized recipes has already been measured at country scale. If your territory has no recovery route, that percentage stays inside your numbers.

Finding 6 — What happens if the project gets approved without reading the tipping indicator?

What happens is the payroll model breaks in month four and nobody knows why. Picture a venue projecting front-of-house wages at legal minimum, trusting tips to close the gap:

NELP (2024) documents that tips make up 58.5% of servers' earnings and 54% of bartenders', so more than half the real wage depends on a flow the owner does not control. Now drop the average check 12% because the territory had less office density than the hunch assumed. Tips fall proportionally, the most experienced server walks across the street, service degrades, the check drops again, and by month six you are paying for recruiting and training out of cash you no longer have. That whole spiral started with an indicator that took one afternoon to read.

Finding 7 — The paradox of abundant data and blind decisions

There has never been so much public information available, and never have so many restaurants opened blind. The contradiction resolves once you understand that the owner does not need more data: he needs the four or five figures that move his break-even point, organized at the moment of deciding and translated into his cost structure. That is where MTIE comes in —the Territorial Business Intelligence Model, the platform Masterestaurant S.A.S. contributes as technology partner of the Twin Ecosystem Model— and the expert reading of this synthesis belongs to Diego F. Parra under the Masterestaurant framework. The model does not predict whether a venue will succeed, and anyone promising that is selling smoke. It does something more modest: it returns a viability range by segment over indicators that already exist. The Independent Restaurant Coalition (2024) notes that 36% of US restaurant owners were born abroad, against 19% in other industries.

Finding 8 — What to demand in the file before signing the lease

Demand three indicators with a named source and a date, and send back any file that lacks them. First, real availability and cost of labor inside the polygon, contrasted against the ILO series: regional youth unemployment of 13.8% in 2024 means abundant supply, but also high turnover if the project pays below the neighbor. Second, the area's digital payment penetration, because it sets your point-of-sale investment and your exposure to cash shortfalls. Third, a surplus recovery route, using the 78.4% of foodservice waste ReFED reports going to landfill as the benchmark for what happens when none exists. With those three on the table, the investment officer is rating a territory. Without them, he is rating a well-written hunch.

Finding 9 — What actually changes between the two decisions

The difference is not how much information exists, it is when it arrives. Before MTIE the data shows up once the lease is signed and the capital committed; after, it shows up while saying no is still an option. A study that arrives late is a forensic report, not a prefeasibility. The nature of the figures changes too. Traditional forecasting runs on house numbers —what the owner hopes to sell— while territorial prefeasibility runs on verifiable public series: the ILO for employment, the World Bank for financial inclusion, ReFED and UNEP for food loss and waste, the National Restaurant Association for sector structure. A number with no publishing organization behind it does not enter the file. And the reader changes.

Finding 10 — What actually changes between the two decisions — in practice

A business plan is read by the owner; a territorial prefeasibility is read by the investment officer, the risk analyst and the program's M&E unit. That forces disaggregation by segment —fast casual, full service, QSR— and by size —single unit, 3-10, multi-unit— because a healthy food cost or prime cost range differs sharply between a neighborhood QSR and a white-tablecloth full service. The underlying distinction deserves bluntness: MTIE does not promise the restaurant will work. No tool can. What it does is cut the odds that failure comes from a variable that could have been measured and nobody measured, which across gastronomic MSME portfolios is both the most common cause and the most avoidable one.