I Ran Restaurants for Years Before I Knew My Real Food Cost

"Just raise your prices."

That is the entire food cost advice industry in four words. Beef up? Raise prices. Tariffs? Raise prices. It is advice that costs nothing to give and solves nothing, because your regulars are not a bottomless wallet, and in a year when customers are already pulling back, "charge more" is how you trade a cost problem for a traffic problem.

I want to give you something better. By 17 I was helping run 11 of my family's restaurants, from takeaways to a brewery, and for the first stretch of those years I could not have told you our real food cost number. We priced the menu off a guess and the guy down the road, and we were profitable by luck. When we finally ran the actual numbers, one of our best-selling dishes was making us almost nothing. It hurt to see. It would have hurt more to keep not seeing it.

So here is the formula, the honest benchmarks, and what I would actually do this year.

The food cost formula

Food cost percentage is the share of every food sale that goes back out the door as ingredients.

The quick version, the one I run every week:

Food cost % = Food purchases ÷ Food sales × 100

You spent $2,500 on food this week and did $10,000 in food sales, so your food cost is 25%.

The more precise version adjusts for what is sitting on your shelves:

Food cost % = (Starting inventory + Purchases - Ending inventory) ÷ Food sales × 100

Say you started the week with $8,000 in inventory, bought $6,000 more, ended with $7,500, and sold $20,000 of food. That is $8,000 + $6,000 - $7,500 = $6,500 used, and $6,500 ÷ $20,000 = 32.5%.

Use the quick version weekly and the inventory version at month end. No software required, just a clipboard and an honest walk-in count.

One trap to avoid: costing a single dish (plate cost) is not the same thing. Plate cost tells you what the recipe should cost, while the period formula tells you what actually happened, including the burnt naan, the over-portioned curry, and the case of tomatoes that died in the back. The gap between those two numbers is where your money is going.

What is a good food cost percentage?

The honest band for most restaurants is 28% to 35%, and where you should sit inside it depends on what you run:

• Pizza and QSR: 25% to 30%. Dough and toppings are cheap, so a pizza shop running 35% has a problem somewhere.
• Casual and full service: 30% to 35%, because more proteins and more perishables mean more ways to leak.
• Steakhouses and seafood: 35% and sometimes higher, which is fine, because you make it back on check size.

And the gap is not small money. Most operators I talk to are sitting closer to 40 or 42 percent, and on a restaurant doing $50,000 a month in food sales, the difference between 30% and 42% is $6,000 a month, or $72,000 a year. Same menu, same tables, same everything.

Here is the part the benchmark charts leave out: your food cost percentage is not a grade, it is a lever connected to everything else. A 35% food cost with strong sales beats a 28% food cost in an empty room, and chasing a low number by shrinking portions is how you end up with a great spreadsheet and no regulars.

Why 2026 is punishing anyone who does not measure

The pressure right now is real, and it is not evenly spread.

Wholesale food prices sat 35% above pre-pandemic levels as of May, per the National Restaurant Association's read of the Producer Price Index. Beef and veal were up 15.9% in a single year, with the US cattle herd at its lowest level in 75 years, USDA expects menu prices across the industry to rise another 3.6% this year, and 68% of operators say tariffs pushed their food and beverage costs higher.

There is one piece of good news: egg prices collapsed 86.5% from their avian-flu peak, so if your breakfast menu survived last year, your reward finally arrived.

But zoom out and the picture is ugly. The NRA's 2026 State of the Industry report found 42% of operators were not profitable last year, and that is not "tight" or "could be better", that is losing money. In that environment, running your menu off a guess is a slow leak you have chosen not to fix.

How to actually lower your food cost

Skip the generic list. Here is what moved the number in my restaurants, in order of effort.

Count weekly, not monthly. A monthly count tells you that you had a problem four weeks ago, and food cost does not drift slowly, it spikes. A supplier quietly raises a price, portions creep on busy nights, a promotion runs on an item nobody margin-checked. A weekly check catches the leak while it is still small, and in my experience that habit alone puts most restaurants back in range within about sixty days.

Cost your top ten sellers first. You do not need all 90 items costed this week, because ten dishes usually carry most of your sales. Cost those, compare plate cost to what the period formula says, and you will find your leak faster than any consultant.

Put a spec on proteins. Beef at these prices does not forgive free-hand portioning, and a scale plus a portion spec on your five most expensive ingredients pays for itself in a week. We found a single curry house was over-portioning chicken by about 15% on busy nights, and nobody was stealing, everyone was rushing.

Requote suppliers quarterly. Loyalty is for customers, not for your produce vendor. Take your top 20 purchased items and get two quotes every quarter. You do not have to switch, because the quote alone changes your current supplier's behaviour.

Run a one-week waste log. A notepad by the bin, where everything that goes in gets a line. It is annoying for seven days, and then it changes how your kitchen thinks about prep quantities permanently.

If you want the bigger financial frame around this, I wrote a full guide on building a restaurant budget that works. And food cost is the first of eight numbers I run in the money formulas playbook, if you want the full set.

The other half of the formula nobody talks about

Look at the formula again: there are two ends to it.

Everyone attacks the top, buying cheaper, wasting less, portioning tighter. Good, do that. But the bottom of the formula is food sales, and it moves the percentage just as hard: same kitchen, same waste, more sales, lower percentage. The maths does not care which end you fix.

Now ask yourself where your easiest lost sales are. They are not hiding in a new marketing campaign, they are the people already calling your restaurant while your team is buried in a Friday service, holding money, getting a busy tone.

There is a formula for that leak too: calls per day, times your missed call rate, times your average order value, times 365. Run it with a normal independent's numbers, say 25 calls a day with a quarter of them ringing out during service at a $40 average order, and you get $87,600 a year that never appears on any report. Not spent, just never captured.

I watched this at Baci Trattoria in Naples, Florida. Frank, the owner, was running five phone lines and still losing six or seven calls a night, and some of those were $200 catering orders. After he put Certus AI's phone agent on the line, the missed calls stopped, and Baci has since recovered over $50,000 in revenue from calls that used to ring out, across more than 1,000 orders taken by the AI. Read the full Baci case study if you want the detail.

That recovered revenue lands in the denominator of your food cost formula, and it is margin you already earned the hard way. Certus runs on flat monthly pricing, sets up the same day, and sends orders straight into Toast, Square, Clover, Skytab, or Aloha NCR. If your food cost percentage is stubborn and your phone rings unanswered, fix the phone first - it is faster than renegotiating beef.

The verdict

• If you have never run the period formula: stop reading and go count your walk-in, because this week's real number beats a perfect system you start next month.
• If you know your number and it is over 35% outside of steak and seafood: cost your top ten sellers this week, since the leak is almost always in three dishes.
• If your number is fine but profit is not: your problem is the denominator, so count your missed calls for one week and then decide if the phone deserves to keep costing you money.

"Just raise your prices" is what people say when they have not looked at your numbers. Look at your numbers.

Want to see what your phone is leaking? Book a demo with Certus AI. Twenty minutes, and you will know.