Operating Costs vs Menu Prices: The Myth That Costs Your Restaurant 6 EBITDA Points

1. Does raising prices fix a margin that's collapsing?

No. Raising menu prices does not fix a broken cost structure; it only hides the leak for one or two quarters. The number that governs profitability isn't the nominal price, it's prime cost:

food plus labor. In 80% of the units I audit, that prime cost sits 3 to 9 points above target, usually above 60% of sales when it should live between 55% and 58%. When an owner pushes the signature dish from 42,000 to 46,000, cash improves the first month; by the third quarter, waste, over-portioning and overtime have eaten those 4,000 and the margin fell back into the same hole. The price bought time, not health. Diego F. Parra repeats it in board meetings: pricing is cosmetic until the structure underneath can carry the weight. Prime cost is the real lever because it concentrates 55% to 65% of every dollar sold, while the nominal price only moves the numerator.

2. Price is visible; prime cost is what rules

A restaurant can charge high and bleed out, or charge fairly and thrive, depending on the gap between real cost and theoretical cost. That gap is the metric nobody watches: if the recipe says 28% food cost and the register shows 34%, those 6 points are pure leakage, not a pricing problem. In a unit with 120,000 USD in monthly sales, 6 points are 7,200 USD a month lost to waste, pilferage, unstandardized portions or unnegotiated purchasing. Closing that gap yields more than any menu hike, and it scares away zero customers because the diner never sees it. Menu engineering moves more EBITDA than raising prices evenly, because it redirects demand toward the dishes that already have the best cost structure. Cross two axes per item: popularity and contribution margin in dollars, not percentage. A flat 5% increase punishes the star and the dog alike; engineering relocates, rewrites and repromotes.

3. Menu engineering before a linear price hike

On the menus I redesign, moving a single high-margin dish from line 9 to line 1, dropping the currency sign from the price and anchoring it next to a pricier option lifts its mix from 8% to 14% of orders. That move, without touching a single price, usually adds 2 to 4 points of overall contribution margin. The star at 24% food cost should sell more; the dish at 41% should leave or be redesigned, not repriced upward. Cash flow exposes what the P&L disguises: a management statement with accounting profit but negative cash signals that the problem lives in OpEx and the conversion cycle, not in the menu price. I've seen locations with 9% profit on paper and a bank balance falling every week, because they pay suppliers in 15 days and collect card payments in 3, yet stock inventory for 21 days of sales. There the enemy isn't the menu, it's capital trapped in the pantry and mismatched payables.

4. Cash flow doesn't lie even when the P&L smiles

Raising prices in that scenario pushes more revenue into a pipe that already leaks. Before touching a single menu number, I measure inventory days, perishable turnover and the full cash cycle. If conversion takes longer than the supplier finances, the price is irrelevant. Payroll, rent and utilities don't load onto the plate: they belong to the break-even point, and that's where 70% of owners misread the diagnosis. A healthy 30% food cost won't save a restaurant whose rent went from 6% to 11% of sales or whose payroll crossed 35% without the ticket rising. Raising the menu price to cover inflated OpEx is treating a fracture with makeup. The correct order: prime cost first, then labor productivity per man-hour, rent as a percentage of sales, and only at the end pricing. In Masterestaurant audits, attacking overtime, trimming the dead shift and renegotiating the lease recovers 4 to 8 points of margin before the diner notices a single dollar of difference on the menu.

5. Theoretical cost vs real cost: the gap you pay for

The only serious way to know if a price is wrong is to measure the gap between theoretical and real cost per dish, week by week. Theoretical cost comes from the spec sheet: exact grams per ingredient and its current purchase price. Real cost comes from the register: period purchases minus ending inventory, divided by sales. When that gap exceeds 3 points, there's an operational leak —portioning, waste, theft or spoilage— and no price increase closes it, it only finances it. In the units I audit, standardizing recipes and weighing portions cuts real food cost by 2 to 5 points in 60 days, equivalent to a 6% to 9% price hike but with no customer pushback. Price is the last lever, not the first; you touch it when the spec sheet has nothing left to give. Raising the price is correct only when the structure is already clean: prime cost on target, theoretical gap under 3 points and OpEx controlled.

6. When to raise the price (and how to do it without bleeding traffic)

There, a hike captures value rather than covering a hole. The technique I use: never raise the whole menu at once, but the low-elasticity, high-mix dishes —the ones customers order by habit, not by price— in 4% to 7% increments, aligned with psychological prices that don't cross round barriers. A dish from 38,000 to 39,500 barely moves demand; from 38,000 to 42,000 does scare it. On the menus I adjust, a surgical hike on the right 30% of dishes adds 3 to 5 points of margin with traffic loss under 1%. Price, set well on a healthy structure, does multiply; set on a broken one, it only postpones the collapse. Menu price is a visible number; prime cost is the governing number. A restaurant can charge high and bleed, or charge fairly and thrive, depending on its gap between actual and theoretical cost.

7. The difference that decides the quarter

Menu engineering (mix, contribution margin, placement) moves more EBITDA than a linear price increase, because it redirects demand toward the dishes that already have the best cost structure. Cash flow doesn't lie: a managerial P&L with accounting profit but negative cash signals the problem lives in OpEx and the conversion cycle, not in the nominal menu price.