Food cost percentage for caterers, and what to do when it is wrong
The restaurant answer is 28-32%. For caterers it is a different number, calculated a different way, and the fix is almost never the menu.
The short answer
For a catering business, food cost as a percentage of total event revenue typically runs 12-22%, much lower than the 28-35% restaurants target, because a catering invoice also carries labour, rentals and logistics that a restaurant cheque does not. Measured against the food line of the invoice alone, 28-35% is the right target. Calculate it as (opening inventory + purchases - closing inventory) ÷ food revenue for the same period, and always at real yield, not invoice weight. When the number drifts, the cause is usually portioning, yield or purchase-price drift rather than the menu itself.
Somebody will tell you your food cost should be 30%. They are quoting a restaurant number at you, and it does not survive contact with a catering P&L.
The confusion is about the denominator. A restaurant's cheque is food. A caterer's invoice is food, plus cooking it, plus carrying it somewhere, plus serving it, plus the plates it went on. Divide the same food cost by a much bigger number and you get a much smaller percentage — and that smaller percentage is not a sign you are doing brilliantly.
Two percentages, and you need both
| Measure | Formula | Healthy range |
|---|---|---|
| Food cost against total revenue | Food cost ÷ total event revenue | 12-22% |
| Food cost against the food line | Food cost ÷ food revenue only | 28-35% |
The first tells you whether the business works. The second tells you whether the menu works. They fail for different reasons and they get fixed in different places, which is exactly why tracking only one of them leaves you guessing.
How to calculate it properly
Per period, not per event, and with inventory on both ends:
Food cost = opening inventory + purchases − closing inventory
Then divide by the food revenue for the same period. The inventory count is the part everyone skips, and skipping it is what makes a monthly number swing 8 points for no reason: you bought heavy in the last week of the month for an event in the first week of the next one.
Count it the same day each month. Two consistent counts beat four careful ones taken whenever somebody remembered.
Per-event costing, which is the one that changes behaviour
The monthly number tells you there is a problem. The per-event number tells you which event caused it. You need both, and per-event is the one that actually changes what you do next Tuesday.
That means costing to real yield. A whole salmon side is not fillet. A case of romaine is not usable leaf. Every ingredient gets a yield percentage once, and after that your costing is only as stale as your prices.
| Ingredient | Typical usable yield |
|---|---|
| Whole chicken → meat | 50-55% |
| Beef tenderloin, trimmed | 60-70% |
| Salmon side → portioned fillet | 65-75% |
| Romaine → usable leaf | 70-75% |
| Onions, peeled and trimmed | 88-90% |
| Potatoes, peeled | 78-85% |
| Shrimp, shell-on → peeled | 50-60% |
Treat these as starting points and replace them with your own the first time you weigh trim. Your butcher, your knife skills and your spec all move these numbers.
Six reasons the number drifts, in the order to check them
1. Purchase prices moved and nothing told you
This is the most common cause by a wide margin, and the most invisible. Your menu was costed in March, it is now September, and eleven ingredients have moved. Nobody did anything wrong. The costing just aged.
2. Portions grew
Scoops get generous when the line is under pressure and nobody is watching. A 6 oz portion served at 7 oz is a 17% food cost increase across the whole event, and no one on the line experiences it as a decision.
3. Yields are optimistic
If the recipe assumes 70% and the kitchen gets 60%, every plate is 17% more expensive than the sheet says. Weigh the trim once. It is a genuinely unpleasant twenty minutes and it will pay for itself the same week.
4. Overproduction
Cooking 15% over on every event is a permanent 15% tax that shows up as leftovers and never as a decision anybody made. See the quantity planning post for the numbers to plan against.
5. Waste and spoilage
Prep waste, spoilage, dropped product, and the pans that went home with the crew. Not evil, just uncounted. Count it for one month and you will know whether it matters at your scale.
6. The menu is priced wrong
Last, not first. Almost everybody starts here, and almost nobody's problem is here. Fix the five above before you touch a price, or you will raise the price and watch the number drift right back.
The two-minute monthly habit
- 1Count inventory the same day every month.
- 2Pull purchases from your supplier invoices for the same window.
- 3Divide by food revenue for the window. Write the number down.
- 4Compare to last month. A swing of more than 3 points is a question, not noise.
Four numbers, once a month. It is not sophisticated and it is not supposed to be — the value is entirely in having twelve of them in a row, because a trend tells you things a single month never will.
When the spreadsheet stops being enough
One person, twenty menus, patient with a spreadsheet: it works. The moment it stops working is specific and you will recognise it — a supplier price changes and you cannot answer, quickly and honestly, which of your live proposals are now underpriced.
That is the job Apron Desk does here. Ingredient prices in one place, recipes rolling up to a true plate cost, and a price change propagating to every menu and proposal that used it — so the answer to that question takes a second instead of an evening.
See it with your own events in it.
14 days free, no credit card, $50 of AI usage included every month. If it is not for you, do nothing and it ends by itself.
Start free