How to lower restaurant prime cost in 90 days

Most independents can take four to eight points out of prime cost in a single quarter without cutting quality, raising prices across the board, or laying anyone off. The reason it is achievable is that the drift is almost never one big leak — it is six small ones that nobody measures weekly. This is the sequence we run, week by week.

First, define the number correctly

Prime cost is cost of goods sold plus total labor, divided by net sales for the same period. Cost of goods is opening inventory plus purchases minus closing inventory — not what you bought. Total labor is every dollar spent on people: hourly wages, salaried managers, payroll taxes, benefits, agency cover and paid training. If your number excludes management salaries, it is understated by three to six points and the plan below will look like it is not working.

Targets by format: full-service independents 60 to 65 percent, fast-casual and counter service 55 to 60, bars with a strong beverage mix 50 to 58, coffee shops 55 to 62. Above 70 in any format, the establishment is not earning its rent.

Weeks 1–2: measure before you touch anything

Count inventory on the same day and at the same time every week — food and beverage separately. Pull sales, hours and labor dollars for the same seven days. Build one page with weekly prime cost, food cost, beverage cost, labor cost and sales per labor hour, and post it where managers see it. Two weeks of clean data is enough to find the leaks; four weeks is better, but do not wait for perfect before acting.

In parallel, export twelve weeks of item-level sales from the POS and cost every recipe on the top fifty sellers. Most operators discover two or three high-volume items running at 45 percent food cost, and one signature dish that is effectively sold at cost.

Weeks 3–4: purchasing and portioning

Three-bid your top ten line items by spend. Not everything — the top ten usually cover 60 to 70 percent of purchases. Ask incumbents to hold price for the quarter in writing; several will, because losing the account costs more than the concession.

Then check receiving. Weigh deliveries against invoices for two weeks straight. Short deliveries and silent price creep on invoices are the most common unrecorded cost in an independent kitchen, and both stop within days of someone visibly checking.

Reset portions on the ten highest-volume items with scoops, scales and a spec sheet with photos. A protein portion running an ounce heavy across 400 covers a week is thousands of dollars a quarter.

Expect two to four points of food cost from this block alone.

Weeks 5–6: menu engineering, not price increases

Sort every item into four groups by contribution margin and popularity. Promote the high-margin, high-volume items into the eye path of the menu and into every server recommendation. Re-cost or re-spec the popular low-margin ones. Delete the items that are neither — they consume prep labor, inventory lines and walk-in space for nothing.

Where price moves, move it on the items with the least price sensitivity and the strongest perceived value, not across the board. A targeted 4 to 6 percent move on a third of the menu is almost never noticed; a blanket increase is.

Cutting menu length also cuts labor, waste and inventory at the same time — it is the single change that touches all three legs of prime cost.

Weeks 7–9: rebuild the schedule from a forecast

Forecast covers and sales by daypart, write the forecast down so it can be wrong in public, then set an hour budget per shift rather than a weekly total. The overrun almost always hides in a soft Tuesday lunch, not on Saturday night.

Manage to sales per labor hour. Most full-service establishments need $45 to $70; counter service and coffee need less per hour but far tighter control at the peak. Review actual against forecast every Monday with the manager who built the schedule in the room.

Attack the structural causes before touching floor hours: an overweight salaried layer, split shifts that trigger minimum-hour payouts, prep scheduled at the wrong hour, and menu complexity that demands a station the volume does not justify. Track overtime as hours, not dollars, so a strong sales week cannot hide it.

Expect two to three points of labor from this block.

Weeks 10–12: waste, theft and holding the gain

Run a waste log for two weeks — spoilage, overproduction, comps and remakes, recorded at the point it happens. Comps and remakes are usually a training or ticket-time issue, not a cost issue, and they show up nowhere else.

Tighten the bar: free pour costs most bars three to six points against jigger or metered pour, and a weekly beverage count with variance by category makes that visible within a month.

Then lock the rhythm in. Weekly inventory, weekly prime cost posted, Monday schedule review, quarterly re-bid on the top ten items, and a menu re-cost every time a major supplier price moves. Gains made in a quarter unwind in two if the measurement stops.

What not to do

Do not cut labor hours to fix a food cost problem — it is the most common and most damaging mistake in this business, and it pays for a two-point food issue with service failures and turnover. Do not cut portion sizes on the items guests recognize you for. Do not discount to chase volume while margin is thin; a fuller room at the wrong check average makes the number worse, not better.

What a realistic result looks like

An establishment starting at 72 percent prime cost typically lands between 64 and 67 by the end of a quarter: two to four points from purchasing, receiving and portioning, one to two from menu work, and two to three from scheduling. On $1.2 million in sales, six points is roughly $72,000 a year that was already in the building.

If you want a benchmarked read on where your prime cost sits against your format before you start, our free margin health check gives you an answer in a few minutes.