Restaurant Prime Cost: The 60–65% Ceiling That Decides Survival

When an operator tells us their restaurant is losing money, the first number we ask for is not revenue, and it is not the food-cost percentage everyone fixates on. It is prime cost — food and labour, added together, as a share of sales. More restaurants are decided by that one number than by anything else on the profit-and-loss statement, and most owners have never calculated it as a single figure. This is the gauge we read first in a turnaround, and it is the heart of how GGB reads a business.

It is not financial advice, and the ranges here are typical and indicative — your own P&L gives the exact figures. But the discipline is universal: prime cost tells you, faster than anything else, whether you have an operating problem you can fix or a structural one you cannot outrun. It sits above the detail in food-cost control and restaurant profit margins; this piece is the operator’s manual for the master number those two feed into — what it is, the cadence that holds it, the workbook that produces it, and the ways it goes wrong.

What prime cost is — and why food and labour belong together

Prime cost is simply your food (and beverage) cost plus your labour cost, expressed as a percentage of revenue. Two lines, one number. Most operators track them separately — a food-cost percentage here, a wage bill there — and miss the thing that actually governs the model: their sum.

They belong together because they are the two largest controllable costs in the business, and because they trade against each other. A scratch kitchen with everything made in-house runs a lower food cost but needs more skilled labour; a concept built on prepared or bought-in components runs a higher food cost but less labour. Move work from one line to the other and the individual percentages shift — but the constraint that decides profitability is where the two land together. Judge either in isolation and you can talk yourself into believing a kitchen-heavy concept with a fashionable food cost is healthy while its labour quietly sinks it — or the reverse.

The 60–65% ceiling

Here is the number worth keeping. As a working rule, a restaurant wants its prime cost — food plus labour — somewhere in the mid-50s to low-60s as a share of revenue, and it wants to treat the mid-60s as a ceiling. Component-wise, the teaching bands are food at 28–32% and labour at 25–30% — with the sum held inside 60–65% and pushed toward its lower edge; concepts vary in how they compose it, which is exactly why the sum is the honest gauge.

The ceiling exists because of simple arithmetic. Whatever prime cost does not consume is all you have left to cover everything else — rent and occupancy, utilities, marketing, repairs, licences, finance — and to leave a profit. If prime cost is 60%, you have 40% to do all of that and still keep a single-digit-to-mid-teens margin, which is workable. If prime cost is 70%, you have 30% — and in most locations rent and overheads alone will eat it. The restaurant can be full, well-run on the floor, and still structurally unable to make money, because the model was broken before the doors opened.

Read it down the P&L

The reason prime cost is the master gauge is that it sits at the top of the P&L, where the largest, most controllable money moves. Read the statement the way an operator should:

• Revenue — everything that came in.
• Less food and beverage cost — what the product cost.
• Less labour — wages, and the on-costs that come with them: in the GCC that means visas, accommodation, transport and end-of-service, not just the payslip.
• = what’s left after prime cost — the contribution the rest of the business lives on.
• Less rent and occupancy — the fixed line you negotiated once and cannot change after.
• Less the controllable overheads — utilities, marketing, repairs, the rest.
• = the profit that actually reaches you.

Everything below prime cost is real and matters, but it is smaller and slower to move. Prime cost is where the largest sums sit, where weekly discipline pays off most, and where a problem is either caught early or compounds out of reach. That is why we read it first — and why the rest of this piece is about producing that number weekly instead of discovering it monthly.

What breaching the ceiling actually does

A prime cost above the ceiling does something specific and brutal: it makes you work for the landlord and the payroll before yourself. Service can be excellent, the reviews glowing, the room full — and there is still nothing at the bottom, because food and labour are taking the share the rest of the business needed.

The hard part is that volume does not fix it. If the model loses a few fils of margin on every dirham because of a prime-cost overshoot, more covers simply scale the loss. A structural prime-cost problem is not solved by a busier Friday; it is solved by resetting the structure — the recipes and portions on the food side, the schedule and productivity on the labour side, and sometimes the concept or the price itself. This is the line between an operating problem you can fix inside the current four walls and a structural one that needs the model rebuilt. Prime cost is how you tell the two apart.

The weekly cadence: what to pull, when, who owns it

Prime cost only protects you at the frequency you read it. Month-end is an autopsy; weekly is a diagnosis. The cadence that works is unglamorous and specific:

What to pull — five inputs, nothing exotic:

1. Purchases — the week’s supplier invoices and cash buys, logged as they land.
2. Inventory — an opening and closing count. Count the big movers properly — proteins, oil, dairy, beverage — and estimate the tail; a consistent 80% count beats a perfect count that never happens.
3. Sales — the week’s revenue from the POS, by day.
4. Labour — actual hours worked at actual rates, including the on-costs, not the rota you intended.
5. Overtime — flagged separately, because it hides inside “labour” and moves first when control slips.

When: the same morning every week — Monday for the week ended Sunday works for most operations. The value is in the rhythm, not the date; a moved read becomes a skipped read within a month.

Who owns it: the head chef owns the food inputs and signs the count. The manager owns hours and the rota-versus-actual gap. The owner reads one number — this week’s prime cost, sitting beside last week’s and the target. Once the sheet exists, the whole cycle takes well under an hour. Track the two components — the food-cost variance and labour as a share of sales — but watch their sum against your ceiling, because the sum is what survival turns on. Where labour is the heavier half, the Labour Productivity read shows whether the schedule is matched to covers; where food is, the count and the costed recipes do the work.

This cadence is not theory for us: GGB’s discipline descends from a real 2013 multi-outlet prime-cost workbook still held on file — the same columns, kept weekly across multiple sites, purchases to usage to hours to the one number. The instrument has been digitised since; the discipline has not changed. You can see that lineage in the evidence room.

The workbook walk: one week to one number

Here is the whole workbook, walked once with illustrative round numbers — the structure is the point, not the figures.

Say the week’s revenue is AED 100,000.

Food, by movement — never by purchases alone:

• Opening inventory: AED 22,000
• Plus purchases: AED 31,000
• Less closing inventory: AED 23,500
• Usage = 22,000 + 31,000 − 23,500 = AED 29,500 → 29,500 ÷ 100,000 = 29.5% food cost — inside the 28–32% band.

Buying AED 31,000 does not mean consuming AED 31,000; the inventory movement is what turns a purchases pile into a true usage figure. Skip the count and a stock build-up masquerades as a good week — or a drawdown as a bad one.

Labour, fully loaded:

• Wages for hours actually worked, plus the week’s share of on-costs: AED 27,800 → 27.8% — inside the 25–30% band.

The one number:

• Prime cost = 29,500 + 27,800 = AED 57,300 → 57.3%. Against a 60–65% ceiling, that is a healthy week — written on one line, beside last week’s.

Now watch what the weekly rhythm is really for. Suppose three weeks later the same sheet reads food 31.4% and labour 29.1% — prime 60.5%. Still inside the band, but the trend has announced itself. Two weeks after that: 33.2% and 30.4% — 63.6%, pressing the ceiling. A weekly reader saw the drift at 60.5% and went looking — portioning, waste, an over-staffed shoulder shift — while the fix was cheap. A month-end reader meets 63.6% as a surprise, four weeks late.

The classic failure patterns

Every stressed P&L we open shows some mix of the same four patterns — worth naming so you can check your own operation against them:

• Month-end-only reads. The most common and the most expensive. On the illustrative AED 100,000 week above, a two-point prime-cost drift is AED 2,000 a week — AED 8,000 already gone by the time a month-end read sees it, and the habit that caused it four weeks embedded. Frequency is the control; the arithmetic never forgives the calendar.
• Theoretical-versus-actual drift. The costed recipes say food should run at 29%; the workbook says 32%. That three-point gap is the most informative number in the kitchen — it is portioning, trim waste, unrecorded comps and staff meals, spoilage, or theft. Operators who track only the theoretical number are reading the menu, not the business; the workbook’s usage line is what actually left the store.
• Overtime leakage. The rota was built inside the labour band; the actuals are not. Overtime creeps in at a premium rate, shift by shift, approved verbally and reconciled never. It is why the cadence flags OT hours separately — labour percentage can look nearly right while its composition quietly worsens.
• Wages instead of loaded labour. Counting the payslip and forgetting the on-costs — in the GCC, visas, accommodation, transport and end-of-service accruals — understates the true labour line and flatters prime cost by design. Load the labour number fully or the ceiling you are managing to is fiction.

Prime cost by format: where concepts typically sit

The 60–65% ceiling is the universal frame; formats differ in how they compose the number. Typical patterns, in our experience across GCC operations:

• Delivery-led, QSR and cloud kitchens typically run food at the top of the 28–32% band — bought-in components, packaging-adjacent waste — with labour at or below the bottom of the 25–30% band. Product cost up, headcount down.
• Casual full-service typically sits mid-band on both lines — the balanced composition, and the one where small drifts on both sides most easily add up unnoticed.
• Premium scratch kitchens typically run food in the lower half of its band — whole-ingredient purchasing, in-house production — with labour at the top of its band or pressing it, because the skills and hours live in the payroll.

None of these compositions is wrong; each is a deliberate trade. What is wrong is not knowing which trade you have made — or letting the sum drift through the ceiling while each line, read alone, still looks defensible for its format.

Start by seeing it

If revenue is holding but the profit has thinned, prime cost is the first place to look — and most owners have never seen theirs as a single weekly number. Trade between food and labour deliberately, by concept; hold the sum inside the band; read it every week against the target. A restaurant that does that is in control of the one number that decides the rest.

The Restaurant Profit Leak Audit is the two-minute way to see where you stand today: five figures — revenue, food, labour, rent and delivery commission — and it shows your biggest likely leaks, including how food and labour sit together. It is free and confidential, and it is the honest place to begin before a full, P&L-based turnaround that resets prime cost and installs the weekly workbook discipline that keeps it in range.