Requirements and permits to open a restaurant: before vs after with Masterestaurant

Which permit route is best if you're opening your first or second location

For the independent operator with fewer than 40 tables and a permit budget under USD 8,000, the best route is territorial PREFEASIBILITY first and permits second, because you spend USD 300 to 900 on a land-use study before committing to a lease that rarely costs less than USD 40,000 in deposit, advance rent and shell work.

The reverse sequence — the one almost everybody follows, sign and then find out — turns a cheap filing into structural risk: relocating a kitchen already built because the exhaust duct was never authorized runs USD 25,000 to 60,000 and halts operations for weeks. According to Mexico's National Chamber of the Restaurant and Seasoned Food Industry (CANIRAC), a formal opening processes between 8 and 12 permits depending on the municipality, with timelines from 30 to 120 calendar days. That 90-day spread is what you negotiate, or what you pay in dead rent.

The five families of requirements, and which one really decides your business model

The requirements and permits to open a restaurant in 2026 fall into five families: incorporation and tax registration, land use plus municipal operating license, health permit with food-handling certification, civil protection and structural safety, and the model-specific ones — liquor license, smoke extraction, sidewalk terrace, registration as a delivery establishment. Of those five, the territorial family is the only one that can void your entire model.

Land use doesn't merely authorize selling food: it sets maximum OCCUPANCY, operating hours, whether liquor sales are allowed, and whether the zone tolerates facade-mounted exhaust ducts. A 60-seat wine bar concept closing at 2 a.m. dies on a land-use record that permits 35 people and cuts off at midnight, and no appeal revives it. In Colombia the Sayco-Acinpro certificate adds another block; in Peru the district municipality's operating license conditions everything else.

The order in which you commit capital: where the real difference sits

The difference between an operator who opens on schedule and one who burns cash isn't which permits the law demands — that's public, free information each municipality publishes — but the ORDER in which you commit capital against those permits. Signing the lease before validating land use turns a USD 450 filing into USD 40,000 of exposure, and that jump never shows up in any Excel model because nobody models it.

At Masterestaurant, Diego F. Parra frames it as a two-orders-of-magnitude asymmetry: correction always costs a hundred times verification. The clause that solves it fits in one paragraph of the contract: lease conditioned on obtaining the land-use ruling and civil-protection clearance within 45 days, with full deposit refund if either is denied. A serious landlord signs it. One who resists is telling you something about the space.

Best for groups opening three or more locations a year: a checklist per site, not per country

If your operation opens three or more locations a year, build the territorial checklist SITE BY SITE rather than by country, because processing time is not a national figure. It varies by municipality, by counter clerk, and by whether the file arrives complete the first time; two locations of the same brand, 14 kilometers apart, can differ by 50 days of licensing.

At expansion pace that becomes material: Chipotle projected 315 to 345 openings for 2025, more than 80% with a drive-thru (Chain Store Age / Chipotle, Q4 2024), and Wingstop added 255 net restaurants in the first half of 2025 alone (Restaurant Dive). Nobody sustains that cadence improvising at government counters. What holds the calendar is a master file — corporate records, powers of attorney, blueprints, engineering calculations, tax certificates — replicated and adapted, plus one named owner per site. A group treating permits as an annex to the lease is financing bureaucracy with working capital.

When NOT to choose the popular option of waiting for the license to build your digital engine?

The popular option — opening the Google Business Profile and the delivery platforms once the license arrives — is wrong in at least three scenarios. First, if your location depends on delivery to get going:

the platforms require tax incorporation, a current health permit and, in several markets, staff food-handling certification, and their verification isn't instant. Second, if you open in a dense competitive zone, because a Google listing needs weeks of reviews before it carries weight, and each additional star moves between 5% and 9% of revenue according to Michael Luca's work at Harvard Business School.

Third, if your model is low ticket and high volume, where one weak opening month eats the cushion. Those 60 to 90 days of paperwork are time you're already paying for in rent: use them for photography, menu, categories and domain registration, not for sitting still.

Red flags when comparing spaces and permit agents

Four concrete signals should stop a negotiation. One: the landlord refuses to include the land-use conditionality clause, or assures you verbally that the previous tenant was already a restaurant — the license is NOT inherited with the property, it's granted to the operator and the declared activity. Two: the permit agent quotes a closed price without having seen the site or the cadastral record; whoever didn't measure can't promise a deadline.

Three: the exhaust duct discharges onto the facade or a shared courtyard with no written neighbor agreement, the most common cause of closure by citizen complaint after the first month of operation. Four: the building has no current structural safety certificate and the landlord proposes handling it «along the way», when civil protection usually demands it before the final ruling. Each of these four gets verified in under a week and before you move a dollar.

What happens if the ruling is denied on day 70? The scenario worth modeling?

Model the ugly scenario before signing, because it's the only one that teaches you anything.

Take a 35-table space at USD 3,500 monthly rent, a two-month deposit, and construction started on day 20 on the assumption that the license lands in 45 days. On day 70 the denial arrives, citing occupancy and ventilation. You're carrying USD 8,200 in dead rent, roughly 18,000 in work that can't be dismantled, and a live lease that forces you to keep paying or negotiate an exit.

That same space, with the prefeasibility ruling requested on day minus fifteen for USD 600, would have told you no before the risk ever existed. That's the tension of this trade: the rush to open and the discipline to verify look opposed, and they aren't — verifying shortens the calendar, because rework is what actually stretches it. Request the land-use ruling this week, with exact address and declared activity, before you look at another blueprint.

Where the decision actually happens?

The difference is not which permits the law demands —that is public, free information— but the ORDER in which you commit capital against them.

Signing the lease before validating zoning turns a 450 USD study into a 40,000 USD exposure, and that jump appears in no spreadsheet because nobody models it. One thing I correct in every expansion session: filing time is NOT a national figure. It varies by municipality, by counter clerk, and by whether the file arrives complete the first time. Two sites of the same brand, 14 kilometers apart, can differ by 50 days of license.

The territorial checklist is therefore built per unit, never per country. The local digital engine gets built DURING the filing, not afterwards.

Where the decision actually happens — in practice?

Those 60 to 90 permit days are free time to verify the Google Business Profile listing, upload photos, lock the primary category, plant the correct Maps pin and prepare geotargeted ads for the 3-kilometer radius.

Whoever waits for the license to open Google opens with zero proximity signals against neighbors holding two years of history. There is a genuine tension between speed and formality, and it does not resolve by picking a side. Opening fast without complete permits invites closure and kills the Google listing the day Maps flags it temporarily closed —a label that buries local visibility for weeks. Yet waiting for a perfect file while rent runs destroys working capital.

Where the decision actually happens — key points?

The bridge is phased opening: internal launch with staff, then a soft opening at reduced occupancy under the license already issued, and delivery only once the health permit is physically in hand.

An accounting detail restaurant investors routinely miss: permit costs and expediter fees are pre-opening expense, not fixed-asset investment, so they never amortize against first-year margin. If your feasibility study for investing in restaurants buried them in CAPEX, your break-even was wrong from day zero.

And about food cost, because the permit shapes the menu: if the authorized exhaust bans a fryer and a charcoal grill, your menu loses its highest contribution-margin categories and ends up pushing dishes whose food cost climbs past the 32% ceiling we hold as the maximum. The permit writes the menu before the chef does.