A busy restaurant that is not making money almost always has one of five problems: prime cost (food + labor) has drifted above 60-65% of sales, waste and shrink are eating margin nobody measures, the menu is underpriced against today's invoice prices, fixed costs have crept past what the sales volume can carry, or the numbers the owner looks at are too old and too averaged to show any of it. The fix starts with measuring weekly, not reacting monthly.
I have stood in kitchens doing real volume - lines out the door, teams sweating, dish pit slammed - that lost money that same night. Sales were never the problem. The system underneath the sales was. If your restaurant is busy and broke, this article is the checklist I would run on it, in the order I would run it.
First, Know the Math You Are Up Against
Full-service restaurants typically keep a net margin of about 2-6% of sales. That is the whole game. On $100,000 a month in sales, the operation that does everything right keeps $2,000-$6,000. That thin band is why a restaurant can feel successful and still not pay its owner: it only takes 3-4 points of quiet leakage to erase the entire profit.
| If your monthly sales are | A 2-6% net keeps | 3 points of leakage costs |
|---|---|---|
| $50,000 | $1,000 - $3,000 | $1,500 |
| $100,000 | $2,000 - $6,000 | $3,000 |
| $250,000 | $5,000 - $15,000 | $7,500 |
| $500,000 | $10,000 - $30,000 | $15,000 |
Leak 1: Prime Cost Has Drifted and Nobody Noticed
Prime cost - food plus fully loaded labor - is the first place I look, every time. Healthy full-service operations keep it at or below 60% of sales. The trouble is that nobody decides to run a 68% prime cost. It drifts there: a supplier raises chicken 9%, two new hires train slow, overtime covers a no-show, portions creep. Each one is invisible in the week it happens. Together they are the profit.
Run the number tonight. Last full week's food purchases adjusted for inventory, plus every dollar of labor including taxes and benefits, divided by the week's sales. If you cannot produce those three numbers for last week within 30 minutes, that is itself the finding - you are flying a plane with no instruments.
Put your own numbers in and see how far you are from the 60% line - and what closing the gap is worth per month.
Run the prime cost calculatorLeak 2: Waste and Shrink That Never Hit a Report
The NRDC estimates restaurants lose roughly 4-10% of the food they purchase before it ever reaches a guest - trim, spoilage, over-production, remakes, staff grazing, and the quiet walk-out-the-back-door kind. On a $30,000-a-month food spend, that is $1,200-$3,000 a month, and in most independent restaurants not one dollar of it appears on any report. It just shows up as a food cost percentage that is mysteriously 3 points above what the recipes say it should be.
The gap between theoretical food cost (what your recipes predict) and actual food cost (what invoices and inventory prove) is the single most honest number in your restaurant. Under 2 points is tight. Over 4 points means waste, portioning, or theft is being subsidized out of your pocket.
Leak 3: The Menu Is Priced for Last Year's Invoices
Most owners price a dish once - the day it goes on the menu - and never again. Meanwhile every invoice that arrives since has moved. If your case of chicken went from $92 to $104 and the sandwich price did not move, you took a silent pay cut on every single order. Multiply that across a menu and a year, and the restaurant is working harder than ever to earn less than ever.
Recost your top 10 sellers against this month's actual invoice prices. In my experience the top 10 carry most of the food cost, and at least 2 of them are quietly underwater or close to it. Those 2 dishes are often the entire answer to the question this article asks.
Check any dish in 60 seconds - plate cost, current price, and the price that hits your target food cost.
Run the menu item calculatorLeak 4: Fixed Costs the Volume Can No Longer Carry
Rent, salaried payroll, insurance, software, loan payments - none of it cares whether you had a good week. If sales have slipped 15% since the lease was signed but the fixed stack has not moved, your break-even point may simply be above what the room currently does. That is not a hustle problem, and no amount of yelling at food cost fixes it. It is a structural problem with exactly three levers: raise the average check, lower variable cost, or renegotiate the fixed stack.
Know your break-even in dollars per day. Not roughly - exactly. Every operator I have ever met who turned a losing restaurant around could quote that number from memory, because once you know the bar, every shift becomes a pass or a fail instead of a vibe.
Get your break-even in monthly sales, daily sales, and covers per day - from your real fixed costs.
Run the break-even calculatorLeak 5: The Reports Are Too Old and Too Averaged to Catch Any of This
Here is the meta-problem that lets the other four live: most restaurants look at a P&L 3 weeks after the month closes, averaged across the whole period. A monthly average is where problems go to hide. The bad week disappears into the good one, the bleeding dish hides inside the category, and by the time the number looks wrong, you are 6 weeks past the cause. Annual numbers are a eulogy. Monthly numbers are a diagnosis. Weekly numbers are a steering wheel.
- Pick 5 numbers: sales, food cost %, labor cost %, prime cost %, and break-even progress
- Pull them every Monday for the prior week - same format, one page
- Compare against target and against last week, not against memory
- When a number moves more than 2 points, find the cause that week, not at month-end
- Assign every fix one owner and one recheck date
The Bottom Line
A restaurant that is busy but broke is not unlucky and its owner is not lazy. It is leaking in one of five measurable places, and the leak is invisible only because the measurement is missing. Run prime cost for last week, check theoretical vs actual food cost, recost the top 10 sellers, calculate break-even, and put the whole thing on a one-page weekly cadence. The money is usually still in the building - it is just walking out through holes nobody is watching.
Want a second set of eyes? A free MOS audit walks your operation, names which of the 5 leaks is costing you most, and tells you exactly what to fix first - no commitment past the conversation.
Book a free audit