Scaling a restaurant: before vs after, measured with 2026 data

When does opening a second location stop being a gamble?

It stops being a gamble when the original location strings together three straight quarters with prime cost below 62% and food cost below 32%, and those numbers come out of a system rather than a spreadsheet rebuilt on the 30th.

Nine continuous months is the minimum needed to separate season from structure, because December hands you one good quarter and a corporate contract expiring in March explains two of them. The International Franchise Association projected roughly 20.000 new franchise units in the United States for 2025, a 2,5% rise to 851.000 locations, and behind every one of them sits a proven unit with a documented series. The Latin American owner who expands on two good months and enthusiasm signs microcredit-rate debt against a margin he never measured. That is where family wealth starts to burn, and it usually takes eighteen months to become visible.

The bottleneck is not capital, it is the data series

An investment officer at the IDB Group does not turn down a restaurant project because the food is weak: the rejection comes because the small business cannot prove, with a continuous series, that the margin of location one repeats in location two. That is the real barrier to scaling in Latin America, and it coexists with a paradox worth naming out loud. The ILO documents in its Panorama Laboral 2025 that informality hovers around 48% of regional employment, with retail, lodging and food services holding much of that volume. An informal business has no credit history; without history no capital arrives, without capital there is no second location, and without a second location there is no formal employment to register. The loop breaks at a single point, and it is not the interest rate: it is the OPERATING DATA the restaurant produces daily and throws away.

What the Twin Ecosystem does that a management course does not?

SATE Institute sets the development agenda and measures impact against SDGs 8, 9 and 12; Masterestaurant S.A.S., technology partner and owner of the software, supplies the platform that turns purchasing, waste, payroll and average ticket into standardized series.

The difference from a traditional training program is one of nature, not quality: a course leaves knowledge in the owner's head, the platform leaves evidence on a server. Once those series reach twelve months, the restaurant no longer requests a loan with a PowerPoint business plan, it files a scoring dossier with verifiable operations. Diego F. Parra insists on an order that many owners reverse: first the system that measures, then the capital that expands. Done backwards, credit reaches a business that still cannot say which dish makes money, and borrowed money accelerates the mistake instead of fixing it. Alternative scoring evaluates real cash flow instead of real estate, and that institutional shift is worth more than any interest-rate subsidy.

Operating data replaces the mortgage collateral you do not have

IDB Lab has spent years pushing instruments of this kind to close the regional MSME financing gap, precisely because the average restaurant owner holds no deed to pledge yet does hold twelve months of purchases, waste and daily cash. Twelve months of continuous series allow you to forecast the new location's cash flow within a tolerable error band, say ±8% on expected sales; with no series, the forecast is wishful thinking dressed in round percentages. In Brazil, where ABRASEL reports an annual food service payroll above 107.000 million reais in 2025, the volume is there and formal credit does not reach it. The gap between those two figures is almost always a problem of operational accounting, not of public policy. Three scenarios, three different readings of the same figures. Small restaurant, one location, sales under 40.000 USD a month: your goal is not the second location but closing nine months with food cost below 32% and logging every waste entry; the useful benchmark is not Wingstop, it is your own March against your own September.

How to read these numbers in YOUR operation?

Mid-size, two or three locations: the question here is whether location two's prime cost converged with location one's inside the first six months, and if it did not, which line item drifted.

Gastronomic group, four locations or more: measure yourself against the expansion cadence of operators that do publish, like Jollibee with its 350-store target for the United States and Canadá, or Raising Cane's aiming at 1.600 locations by the end of the decade. The concrete decision that table yields is how many openings your team can absorb per year, not how many you would like to sign. The expansion figures come from public operator filings and trade press: Yum! Brands' 8-K for the second quarter of 2025 records 565 gross new KFC International units, Restaurant Dive documents Wingstop's global target of 10.000 locations and Chipotle's 7.000-restaurant goal for North America.

Where these benchmarks come from and how far they go?

Labor and informality data come from the ILO and ABRASEL; US franchise data from the International Franchise Association, which reports more than 210.000 new jobs in 2025 and a base above nine million.

Two honest limits. First, nearly everything published belongs to large chains with access to cheap capital, so their pace does not transfer to an independent operator. Second, prime cost averages vary by country, format and tax structure: use them as a reference frame, never as an imported target applied without adjustment. Follow the chain to the end, because the outcome is predictable. Location two opens with eighteen-month debt and a food cost nobody measured; by month three a four-point deviation shows up and the owner blames the learning curve. Month six, that deviation is still there and has already eaten the cash cushion, so the owner starts funding the new operation with location one's cash.

What happens if location two opens without location one's series?

Month nine, location one loses purchasing power, negotiates worse with suppliers and watches its own margin fall. Month twelve, two sick locations instead of one healthy one.

With a data series, the month-three deviation gets caught in week two and fixed with a recipe card, not with a loan. That is the whole difference, and it costs less than an opening campaign. When a poorly measured restaurant blows up during expansion, it destroys jobs that took years to formalize, leaves suppliers unpaid inside short supply chains and pushes workers who had entered the system back into informality. That asymmetry, with social cost running several times above private cost, explains why a GovTech think tank like SATE Institute concerns itself with a metric that looks like a business owner's private matter. Mexico shows the upside when expansion is properly ordered: the AEF counted 101 Spanish franchise networks operating there in 2025 with 1.556 establishments, and Wendy's announced agreements for more than 60 new restaurants in the country.

The social cost of a failed expansion is not paid by the owner alone

Those jobs are registered, they pay contributions and they can be audited. Start this week with the only move that opens the door: load your last ninety days of purchases and waste into the system and look at the margin that actually shows up. The core difference is not available capital, it is the QUALITY of the information behind the decision. A restaurant with twelve months of continuous series can forecast the new unit's cash within a tolerable error band; one without a series is gambling, and gambling with microcredit-rate debt is how family wealth disappears in eighteen months. The second shift is institutional. Once operations produce standardized data, the restaurant stops being an opaque credit subject and becomes assessable through alternative scoring, the very instrument organizations like IDB Lab have pushed to close the regional MSME financing gap. Operating data replaces the mortgage collateral the owner does not have.

What actually changes between the two scenarios?

Third, and this gets overlooked: measured scaling changes job quality, not just job count. A group that knows its prime cost can schedule shifts in advance, offer contracts and pay social security without the margin collapsing;

a group that does not know it adjusts by hiring day laborers. Labor formalization is an arithmetic consequence of predictable margin, which is why SDG 8 is decided on a cost sheet long before it reaches a policy document. Finally, waste. A restaurant measuring loss by critical SKU cuts purchasing without touching the menu, feeding directly into SDG target 12.3 and the IDB's #SinDesperdicio agenda. The same intervention that improves the private margin improves the public indicator — a rare alignment in development policy, and one worth exploiting.