How to open a restaurant step by step: what changes when the audience arrives before the keys

Best for: a first venue under 15 tables with less than USD 120,000

If you are opening your first venue, under fifteen tables and with a budget below USD 120,000, the right order is to launch a content channel six to eight months before the kitchen, and the sixty-page business plan can wait. The reason is arithmetic, not taste: filling an unknown venue with cold paid media costs between 25 and 40 USD per new guest across Latin American markets, while an audience built with your own short video brings that cost down to 2 to 6 USD, because distribution is decided by interest and not by budget. In cash terms, two hundred new guests a month mean USD 6,000 in paid media or roughly USD 800 in sustained production. With low six-figure capital, that monthly gap of USD 5,200 is the payroll of two cooks through the most fragile half-year the business will face.

Why does discovery start with video and not with search?

Because nearly 40% of young users look for places to eat on TikTok or Instagram before Google, a figure Prabhakar Raghavan, then senior vice president of Knowledge & Information at Google, made public in 2022 and the company has never walked back.

If four out of ten potential guests explore on a platform where you do not yet exist, construction and permits move forward while demand stands still. The contrast with a mature market helps: in the United States, Datassential measured closures below 1,000 in the spring of 2025, the lowest in at least seven years, and in that same market nobody opens without prior presence. In Colombia, ACOGA reported more than 2,700 restaurants closed during the 2024 crisis. Two venues side by side, the same investment, the same kitchen: the one with a line had existed in the neighborhood's head for months. Before signing a lease it pays to measure how much attention the concept can capture in that territory, because a corner with heavy foot traffic and zero digital conversation is an expensive venue running on borrowed demand.

Territorial prefeasibility: rule out three of every four venues from your desk

With public data on population, competitors, traffic and local searches — what the trade calls location intelligence — you rule out three of every four candidates from your desk and visit only the ones that survive the filter. The Masterestaurant team runs that diagnosis in every prefeasibility study, and Diego F. Parra insists on a point that stings the founder in love with a recipe: the cost of landing the FIRST guest is the largest hidden expense of an opening and almost nobody budgets it, while three weeks go into debating oak chairs versus metal ones. It suits you if rent exceeds 8% of projected sales. There are three cases where building an audience before opening is the wrong call. First: you buy an established franchise, where the brand already carries demand and the system charges 8.5% to 11.2% of sales in royalty plus marketing fund, according to Toast; parallel content there competes with the manual and rarely beats it.

When NOT to pick the popular option: three cases where owned content is wrong?

Second: a pure-capture format — airport bar, food court, hospital — where the decision happens two meters from the counter and no video moves that needle.

Third: an operating partner who cannot sustain two posts a week for eight months, because a channel abandoned after ten videos communicates neglect rather than craving. If your case is one of these, the money works harder in the kitchen and in staff. Franchises also close less often: 20 to 25% at five years versus roughly 50% of independents, per SBA data cited in the trade. Four signals make me stop an opening, and all four show up before signing. One: the advisor presents a sales projection without saying how many covers per shift sustain it, because a handsome annual figure divided by 360 days usually demands a full house the dining room cannot physically hold. Two: the lease carries an annual increase tied to inflation with no cap, in a country where ACODRES measured a 24% sales drop in the first half of 2024; the floor falls and the rent climbs.

Red flags when comparing opening options

Three: the franchisor will not hand over the full list of current and closed franchisees — FRANdata tracks more than 4,000 brands and over 200,000 franchisees, so the information exists and hiding it is a decision. Four: the opening budget has no line for content or photography. That zero is not savings. It is debt paid later in media at 30 USD a head. If you run a group with two or three units and you are weighing expansion, the asset to build during opening number three is not the venue: it is the channel that will make number four cheap. Fast-growing chains understand this well. 7 Brew, the fastest-growing chain according to Technomic and Restaurant Business, grew sales 267% and units 350%; Chipotle opened its 4,000th restaurant in December 2025 in Manhattan, Kansas; and thirty chains opened more than 100 locations in 2024, led by Starbucks, Jersey Mike's and Wingstop, per Technomic.

Best for operations already thinking about the second venue

In Colombia, Frisby billed more than 1.21 trillion COP with 12% growth, according to Valora Analitik. None of those openings starts from zero attention. It suits you if your current acquisition cost tops 12 USD per new guest and you want it lower before committing capital to construction. Picture the classic sequence taken to its consequence: idea, plan, investors, venue, permits, construction, staff, opening, and marketing at the end, once the money is gone. The venue opens at 100% installed capacity and 20% real occupancy, so you pay full payroll for an empty room and burn between USD 6,000 and 9,000 a month in cold media trying to fix it. Three months later working capital is exhausted, the 30% discount arrives to fill tables, and that discount re-educates the neighborhood: it learned your restaurant is worth a third less. Break-even moves away precisely when you needed it closest.

What happens if you flip the order and leave marketing for last?

The tension here is real and I will not soften it: pre-opening content does not sell food, it sells time, and time is exactly what an undercapitalized opening lacks.

So you buy it earlier, while it is still cheap. Start from the opening date and count eight months backward: that is where the first video goes, not the first architectural drawing. The Spanish franchise sector hints at which concept scales: fast food accounts for 24.8% of billings and 35.2% of outlets in franchised food service, according to Tormo Franquicias Consulting 2024, because a low check tolerates forecasting errors that fine dining never forgives. During those eight months the goal is not to sell, it is to accumulate local followers at marginal cost and test two or three dishes against real reactions. A channel with 4,000 neighborhood followers matches, in pulling power, the USD 6,000 in paid media you will not spend during month one.

What an opening plan with a prebuilt audience looks like?

Your concrete next move: shoot the first kitchen video this week, even if the venue is still a floor plan. Acquiring the first customer changes nature entirely.

Paying cold ads to fill an unknown venue costs 25 to 40 USD per new diner across Latin American markets; building an owned audience over six months drops that to the 2-to-6 USD range, because short video distributes on interest rather than budget. The arithmetic is brutal: 200 new diners a month cost 6,000 USD in ads or 800 USD in sustained content production. Territorial prefeasibility stops being a hunch. With public data on population, competition, foot traffic and local search volume —what the industry calls location intelligence— you can rule out three of every four candidate locations without leaving your desk, then visit only the ones that survive the filter. I defended the restaurateur's eye over the data for years, and I was wrong: that eye handles the last 20% of the decision, never the first 80%.

Where the difference is actually decided?

The investor conversation flips. A pitch without traction sells a promise, and investors discount promises to zero; a pitch with 8,000 local followers and 1,200 waiting-list emails sells a risk already reduced.

The valuation gap between those two cases can double the equity the founder keeps. Footage shot before opening has a second life almost nobody counts. Menu tests, construction, the oven arriving, the first payroll: all of it publishes across the first year without producing anything new. A restaurant that opens without that archive has to film and operate simultaneously, precisely during the most chaotic months, and the result is an account that posts three times and dies.