The restaurant as the first door to formal youth employment: the numbers almost nobody tracks

1.86 hires per seat: the number that exposes the revolving door

Three venues in Bogotá closed 2025 with 41 hires to sustain 22 positions, that is 1.86 hires per seat, and that figure says more about the health of the business than any line in the income statement. Each replacement cost between 380 and 620 USD counting recruitment, uniforms, supervision hours and learning-curve waste, so the 19 surplus hires took between 7,220 and 11,780 USD out of the year, money nobody budgeted because there is no account to record it in. The owner arrived convinced he had a sales problem. He had a payroll problem in disguise. The decision that comes out of this figure is concrete: open a line called REPLACEMENT COST in your monthly P&L, right beside food cost, and calculate it as hires ÷ positions. If the result goes past 1.4, stop spending on advertising this quarter.

270 million workers and an informality rate that will not come down

Hotels, catering and tourism employ more than 270 million people worldwide, close to 8.2% of the global labour force according to the ILO (2024), and the same sector carries one of the highest informality rates in the entire MSME universe. That is where the tension no public policy has settled lives: the activity that absorbs an inexperienced young worker fastest is also the one that retains him worst. In Colombia, 95% of the food-service market consists of independent establishments according to Acodrés (2024), venues of three to twelve employees with no HR function and no training budget. A young cook starts on Tuesday, learns by watching, and four months later leaves without a document proving anything. The door opens wide and shuts fast, and the sector that generates the most first formal jobs is the one producing the least evidence of what it teaches.

Why a credit officer should ask about turnover before asking about revenue?

Two restaurants billing the same amount carry different risk profiles if one turns over 90% of its staff a year and the other 35%, and that gap belongs to the loan book, not to sentiment.

Payroll volatility reaches cash flow through three measurable channels: overtime while the position sits empty, average-ticket decline from degraded service, and high waste during the learning curve. With replacements running 380 to 620 USD, a 22-seat venue turning over at 90% burns between 7,500 and 12,300 USD a year on rehiring alone. That is EBITDA the bank never saw coming. The reading we apply at Masterestaurant when reviewing a leveraged operation is straightforward: turnover anticipates delinquency more precisely than the sales history does, because sales recover over one good weekend and a trained team takes six months.

Learning-curve waste: food cost that is really turnover cost

A cook with two weeks on the line wastes far more than one with six months of curve behind him, and that waste gets booked as food cost when it is turnover cost under another name. The same kilo of protein hitting the bin appears in SDG 12 on food loss and waste and in your income statement with an identical figure, though nobody connects the two reports. If your target food cost is 30% and you run at 32%, those two points on 400,000 USD of annual sales are worth 8,000 USD, and a share of it that almost never gets measured comes from new hands cutting badly. I got this wrong for years: I kept the waste sheet apart from the staffing sheet, as though they were two problems. They are one. Cross your staff start dates against the weekly waste curve and the step will show up; if it does not, your waste measurement is not daily, and that is the first thing to fix.

How to read these numbers in YOUR operation: three scenarios?

Turnover benchmarks do not apply the same way to an eight-employee venue as to a four-location group, so translate before you decide.

In a SMALL operation of 6 to 10 positions, 70% turnover means 5 to 7 exits a year and 1,900 to 4,300 USD in replacements: the fix is not a programme, it is the owner running induction personally through the first 40 hours. In a MEDIUM operation of 20 to 30 positions, that same rate costs between 5,300 and 18,600 USD, and there it does pay to give the head chef two protected hours a week for training. In a GROUP of 60 positions or more, annual replacement reaches 25,000 to 37,000 USD and the lever changes nature: internal micro-credentials, a visible career ladder, and a wage floor 15% above the local market in the two most critical entry roles. Pick your bracket and put the figure on Monday's agenda.

Where these benchmarks come from and what they do not prove?

It is worth stating the origin of each number before someone takes it into a committee.

The 270 million global jobs and the 8.2% share of the labour force come from the ILO (2024), measured across hotels, catering and tourism together, not restaurants alone. The 15.7 million U.S. restaurant jobs in 2026, projected to reach 17.3 million by 2036, and the 172,500 net jobs added in 2024 come from the National Restaurant Association, which surveys its members and therefore under-represents informal work. The 95% independent share in Colombia is Acodrés (2024), counted on establishments rather than sales. Mexico's 2.9 million tourism jobs growing 3.5% are from INEGI (2024). None of these sources measures youth turnover by position, and the 380 to 620 USD replacement range comes from costing one specific Bogotá operation: treat it as an order of magnitude, not a regional average.

From payslip to micro-credential: what proves the young worker learned

A payroll record proves the young worker was there, never that he learned, and that distinction decides whether a multilateral-funded employability programme gets renewed or shelved. Schools already proved that kitchens build formal employment at scale: school-meal programmes sustain 7.4 million cooking jobs worldwide according to WFP (State of School Feeding Worldwide, 2024), and not one of those posts issues a transferable certificate. That is where the cheap reform sits. A micro-credential covering food handling, mise en place and portion control, assessed with a checklist signed by the chef and sealed at 90 days, costs less than a single 380 USD replacement and turns a kitchen assistant into an employable candidate in any other kitchen in town. Training someone so they can leave in better shape sounds counterintuitive. Whoever sees a ladder stays, and the 73% of women-led firms without access to growth financing reported by UNDP (2024) is the segment gaining most from credentials that require no capital. Start this week with a single position.

Where the gap actually opens?

The first difference is accounting and nearly invisible: the traditional method treats replacing a kitchen assistant as an HR event, while the Masterestaurant model treats it as a variable cost line, sitting next to waste and food cost.

Once that line appears on screen every month the conversation changes tone within a quarter, because an owner staring at 4,800 USD a year evaporated on hiring the same seat three times stops arguing about paying twenty dollars more. Then comes evidence. An employability program financed by a multilateral bank has to prove the young worker learned something transferable, and a payroll sheet proves nothing: it only proves attendance. Micro-credentials change the unit of measure, certifying competency instead of tenure, which lets you build SDG 8 indicators that survive an external audit.

Where the gap actually opens — in practice?

I got this wrong for years, assuming retention was the sovereign indicator; a young worker who leaves at month eight holding three verified badges is a better development outcome than one who spent three years washing lettuce with nothing certified.

The third gap is territorial. Prefeasibility — where to open, with what menu mix, at what price — predetermines how many formal positions a venue can sustain, and most gastronomic MSMEs settle it by neighborhood intuition. With density, traffic and competition data the same seed capital sustains more positions, and that multiplier is exactly what a local economic development fund hunts for and rarely finds. That leaves the waste circuit, which most people file separately out of pure org-chart inertia. Food loss and waste in a high-turnover team is structurally larger, so every point of turnover you shave drags along tenths of food cost and kilos of organic residue. It is the only case I know where SDG 8 and SDG 12 move with the same lever.