Restaurant Prime Cost Formula: Where the Group-Level Number Hides a Location Problem

What the Restaurant Prime Cost Formula Actually Calculates

Prime cost is calculated as (total cost of goods sold + total labour cost) / total revenue x 100. It combines the two largest costs a restaurant controls day to day - food and labour - into one percentage that shows how much of every sales dollar goes to running the kitchen and floor before rent, marketing, or profit are even considered.

Rent, insurance, and loan payments are mostly fixed once a lease is signed. Prime cost is the largest set of costs still within a manager's control on a weekly basis, which is why it gets tracked far more often than a full P&L, and why an operator who wants to move margin usually starts here rather than on the fixed side of the ledger.

Take a location doing $120,000 in monthly revenue. Food cost, also called cost of goods sold (COGS), runs $33,600, or 28% of revenue. Labour cost, including wages, payroll taxes, and benefits, runs $40,800, or 34%. Add the two together and prime cost comes to $74,400, or 62% of revenue.

The formula only works if both halves are complete. Food cost should include every ingredient that goes into what's sold, not just the invoice total for the month, which misses waste, spoilage, and free items given as comps. Labour cost should include base wages and overtime, employer payroll taxes, benefits, and any bonuses or commissions, not just the hourly rate on the schedule. Leave either one incomplete and the resulting percentage understates the real cost, which is a common way a location's own number looks better than its true position.

That single number is doing a lot of work. It is usually the first thing an operator checks, ahead of rent, utilities, or marketing spend, because food and labour are the two costs a shift manager can actually influence week to week. But a blended number this convenient has a trap built into it: it looks the same whether every location is running close to that figure, or one location is dragging the average up while the rest sit well below it.

Why a Healthy Group Prime Cost Average Can Still Hide a Failing Location

Multi-location groups usually calculate prime cost once, at group level, and stop there. Blend the numbers across several sites and a group might land at 62% - close to the top of what is typically considered a healthy range for a multi-location group, so nobody flags it. But an average is only ever an average. It can just as easily be three locations running comfortably between 56% and 62%, and one location running hot at 71%, with the group number never showing which is which.

The blending usually happens without anyone deciding to blend it. Most accounting rollups pull revenue and cost totals from every location into a single P&L before anyone calculates a percentage, so the group figure is what gets reported first, and the location-level figure only gets pulled if someone asks for it. If prime cost review happens monthly at the group level and quarterly, or never, at the location level, a site can run 8 or more points hot for an entire quarter before the gap shows up anywhere a manager is actually looking.

Break the same locations out individually and the picture changes. City Centre Branch, Airport Outlet, and Harbour View each sit inside a normal range for their concept. North Branch is 9 to 15 points above the others - a gap large enough to represent real margin loss, not rounding noise. At $120,000 in monthly revenue, every extra point of prime cost is roughly $1,200 a month leaving that one location. A group that only reports the blended figure could keep approving North Branch's numbers as within range for months before anyone looks underneath it.

Where Prime Cost's Two Halves Break in Different Directions

North Branch's 71% is not one problem - it is food cost and labour cost both running high at the same time, which is worse than either issue alone would be. The two halves of prime cost usually drift for different reasons, and checking only the blended number hides which one is actually happening. Food cost drifts from portion inconsistency, ingredient waste, or a supplier price increase that never gets repriced through recipe costing. Labour cost drifts from scheduling that reacts to yesterday's rush instead of today's forecast, or from overstaffing that quietly offsets a food cost problem nobody has time to notice.

Plot food cost against labour cost for the same four locations and a different pattern appears. City Centre Branch and Airport Outlet each keep both numbers low. Harbour View has food cost under control but labour cost running slightly high on its own. North Branch is the only location where both numbers are high at once - food cost and labour cost compounding rather than offsetting each other. That combination is the expensive one: understaffing that causes rushed prep and higher waste, or overstaffing that hides a portion-control problem because nobody has the time to notice it. A blended prime cost of 62% never shows which quadrant a location is actually in.

Which quadrant a location falls into also points to a different fix. A location with high food cost and low labour cost usually has a portioning, waste, or supplier-pricing issue that a labour schedule change will not touch. A location with high labour cost and low food cost usually has a scheduling or staffing-model issue that tightening recipes will not touch. North Branch's combination means both fixes are needed at once, and fixing only one will still leave prime cost elevated, which is exactly why the two costs need to be looked at separately rather than as one blended percentage.

What a Healthy Prime Cost Percentage Looks Like By Restaurant Type

Whether 62% is fine or a problem depends on what kind of restaurant is being measured. Quick-service typically runs 55-60%, since labour is lower and the format is simpler. Casual dining sits at 58-65%, balancing table-service labour against a wider menu. Fine dining runs highest, at 60-70%, because skilled kitchen and floor staff cost more relative to a lower-volume, higher-price menu. Multi-location groups blending across sites typically land at 58-62% - lower than any single concept on its own, because scale purchasing and standardised labour models pull the average down. A single location running the same concept usually sits 2 to 4 points above its group's blended figure, simply because it does not benefit from the same purchasing scale.

That adjustment is not a rounding allowance, it is a real structural difference. Multi-location groups negotiate purchasing volume across every site, and can move labour between locations or share back-of-house staff during slow periods. A single independent location doing the same concept has neither lever, so a 2 to 4 point gap above the group's blended figure is expected rather than a sign of poor management. The distinction matters when comparing a location's number against a benchmark: compare a single-site business to the multi-location range and it will look worse than it actually is.

Against that scale, North Branch's 71% is not borderline. It sits outside even the fine-dining range, in a concept that is not fine dining. That is the number that should trigger a location-level review - not the 62% group average sitting comfortably inside its own band.

The worked number to check first is not the group's blended prime cost - it is the same formula applied one location at a time. If every site sits within 2 to 4 points of the group average, and the group average sits inside its concept's benchmark band, prime cost is doing its job. The moment one location sits 8 or more points above the rest, as with North Branch here, that is the trigger to act: pull that location's food cost and labour cost apart, plot them the way shown above, and find out which side, or both, is driving it.

In practice, that means three checks rather than one: calculate prime cost at group level to see where the number sits against the benchmark for your concept, calculate it again at every individual location to see how far each one sits from that group number, and only then split the outlier's food cost and labour cost apart to find which half, or both, is actually moving. Skipping straight to the group number is what let North Branch run 9 to 15 points hot without anyone noticing.

Getting to that answer quickly depends on whether prime cost is visible by location or only at group level. Multi-site operators who track food cost and labour cost apart, restaurant by restaurant, catch a drifting location while it is still a small number, not months later in a P&L review.