Getting the Numbers Right in a Commercial Kitchen
Food service math isn't about memorizing formulas. It's about adapting when your supplier delivers four pounds less than the invoice says, or when you need to scale a recipe from 50 covers to 200 and your convection ovens can only fit three pans at a time. Most people think this is basic arithmetic. It is, but the way it applies in a working kitchen is where things get complicated quickly. At its simplest level, you're tracking yield, portion cost, food cost percentage, and conversion factors across recipes and orders. Yield percent is the most misunderstood metric I see. A recipe calling for "one pound of carrots" doesn't mean you use one pound of edible product. You peel them, you trim the ends, you throw away the bruised bits. That one-pound raw input might give you 14 ounces of usable product. If you're costing a menu item and you only factor in the raw weight, your food cost percentage will look better than it actually is, and you'll underprice accordingly. I learned this the hard way about six years into my first real position. We were serving a roasted carrot side that looked profitable on paper at 28% food cost. It was actually running closer to 41% once I did a proper yield test and recalculated. We raised the price by $1.25 and stopped losing money on what we thought was a winner. Portion cost breaks a recipe down to the cost of each individual serving. You take the total recipe cost divided by the number of portions. But the trick is knowing what "portion" means in your operation. A one-cup scoop isn't the same as a heaping scoop, and different servers will be inconsistent unless you standardize with scoops or ladles. I've seen kitchens where the variance between portions cost them an extra 3-4% on food cost every single month because two different people were plating the same dish using completely different scoop sizes.
Scaling Recipes Without Losing Your Mind
Scaling is where applied math hits the wall of reality. The textbook answer is just multiply every ingredient by the same conversion factor. The actual answer depends on whether you're dealing with spices, emulsions, thickening agents, or something that reduces and concentrates over heat. When I scale a sauce from 10 portions to 50, I multiply the liquid and the vegetables and the stock by five. But I don't multiply the salt by five. I multiply it by maybe three and a half, then taste and adjust. Spices and seasonings don't scale linearly because the surface area and concentration dynamics change when you're cooking in a forty-quart pot versus a ten-quart pot. The heat distribution is different. The reduction rate is different. The flavor compounds interact differently at different volumes. Thickeners are another trap. Roux, cornstarch slurry, beurre manié — they don't scale the way you'd expect. If you quadruple a recipe that calls for two tablespoons of cornstarch, you don't just use eight tablespoons. At higher volumes, the thickening power per unit actually decreases slightly because the starch granules have more liquid volume to gelatinize through, and the agitation patterns in a large batch pot are different. I usually start at three-quarters of the scaled amount, bring it to a simmer, and add more in increments until I hit the texture I need. This saves time compared to burning a batch and starting over.
The Actual Workflow I Use Every Day
I keep everything in a spreadsheet with recipe cards, but the real work happens before I even open it. Every new dish I add to the menu goes through a costing process that takes about twenty minutes. I weigh out every ingredient at its purchased form, apply the cost per pound from our current supplier invoice, calculate the yield loss, divide by the number of portions, and then add a markup target based on where that dish needs to sit in our overall food cost range. For a typical week, I'm doing maybe six to eight of these, plus adjusting old recipes when supplier prices shift. Price changes are constant. Last March, our beef supplier increased prices by 18% on ground chuck and brisket. I had to recalculate seventeen menu items that used either of those as a primary ingredient within a single afternoon. If you're doing that by hand on a calculator, it takes about ninety minutes and you'll make mistakes. If your recipes are properly set up in a spreadsheet with formula links to the cost cells, you can update the base ingredient cost once and have every affected dish recalculate automatically. That cuts it down to about fifteen minutes of review time to make sure nothing looks wrong. The one workaround I keep coming back to for edge cases: when you're sourcing from multiple vendors and the same ingredient has three different price points depending on grade and origin, I create a weighted average cost based on what I'm actually purchasing, not the most expensive option on the sheet. Some systems insist on costing at the highest price point to be "conservative." That gives you misleading numbers. If you're consistently buying the mid-tier product and pricing off the premium tier, your menu will be priced too high and you'll lose covers to competitors who are actually costed more accurately.
Get the Full Details
Where This Breaks Down
Applied math for food service assumes a certain level of consistency that doesn't exist in most kitchens. Ingredient weights vary. A "quarter pound" patty can run anywhere from 5.8 to 6.2 ounces depending on how the line cooks it. Portioning tools wear out. Scoop sizes degrade over months of dishwasher cycles. Supplier invoices can be wrong — I've received three separate invoices this year alone with incorrect weights or duplicate charges. If your costing system doesn't account for these variances, your theoretical food cost will drift away from your actual food cost over time, and you won't know by how much until your monthly inventory reconciliation hits. The biggest limitation is that math alone can't tell you when a recipe needs a practical adjustment. A dish might be perfectly costed on paper but take too long to execute during a busy shift. That doesn't show up in any spreadsheet. I've had to pull menu items that were profitable but operationally unsustainable because they required three different prep stations and forty-five seconds of active cooking time per plate during a rush where our tickets were already averaging twenty-two minutes. If you're running a small operation with fewer than twenty menu items, a well-organized spreadsheet is probably all you need. For anything larger, dedicated recipe management software like WebCAT, ChefTalk, or Toast's recipe module will save you significant time, though they come with monthly fees and a learning curve that can eat into your productivity for the first few weeks. There's no perfect tool. The ones that cost less tend to have weaker portion management features. The enterprise platforms are accurate but overkill if you're doing fifty covers a night.
