How a headcount becomes a shopping list
Catering quantities fail in one of two places, and they are not the same place. The first is the plate: you decided six ounces of chicken per guest and served five, because a hundred guests helped themselves and the first forty took seconds. The second is the invoice: you ordered exactly the weight you needed to serve, forgot that the roast loses about a quarter of its weight in the oven and a further share to bone and trim before it ever gets there, and ran out at guest seventy.
This calculator separates those two problems. It starts from a plated portion — the weight of cooked, edible food you intend each guest to receive — and multiplies by a service-style uplift, because self-service is not portion control. It then works backwards from cooked weight to raw weight to purchase weight, which is the direction the supplier's scale runs.
Every step in that chain is measurable. Cooking loss you can measure by weighing a tray in and out of the oven. Trim yield you measure with a yield test: weigh the case, break it down, weigh what is usable. The food yield percentage calculator runs that test formally and gives you the number to type into the trim yield field here.
The formula, variable by variable
The purchase quantity is one chain of multiplications and two divisions:
AP lb = guests × plated oz × uplift ÷ 16 ÷ (1 − cooking loss) ÷ trim yield
Guests × plated oz gives total cooked ounces at portion-control discipline. Divide by 16 to reach pounds.
Uplift is the honest admission that a buffet is not a plate. When guests serve themselves, average consumption rises, and it rises most on the items people recognise and want. Attended buffets, where a cook holds the serving spoon, land near 1.15. Unattended stations, where guests wield the tongs, sit closer to 1.25. Family style, with platters on the table, is in between. Plated service takes 1.00 because the kitchen decides the portion.
Cooking loss is divided out, not multiplied. If a roast loses 25% of its weight, the cooked weight is 75% of the raw weight, so the raw weight is cooked ÷ 0.75. Dividing by 0.75 raises the number by a third, not a quarter — the most common arithmetic slip in catering, and it always errs short. The USDA Agricultural Research Service publishes measured cooking yields by cut and method; look yours up rather than assuming.
Trim yield is divided out last, for the same reason and in the same direction. It is the edible portion weight as a percentage of the as-purchased weight: a case of whole tenderloins that yields 12 lb of trimmed centre-cut from a 16 lb case has a 75% yield.
The order matters conceptually even though multiplication commutes: cooking loss applies to the trimmed product, and trim applies to what arrives at the dock. Keeping them in that order is what stops you from applying a bone-in cooking yield to a boneless purchase spec.
Worked example: 150 guests, attended buffet, chicken and two sides
You are catering a wedding for 150 guests. Service is an attended buffet. You plan a 6 oz cooked portion of chicken thigh, two sides at 4 oz each, and a one-hour reception with five passed pieces per guest per hour. Your yield test on the case of boneless thighs returned 85% after pulling the excess fat, and weighing the hotel pans in and out of the combi gave a 25% cooking loss.
- Apply the uplift. 6 oz × 1.15 = 6.9 oz of cooked chicken per guest.
- Total cooked weight. 150 × 6.9 = 1,035 oz. Divide by 16: 64.69 lb of cooked chicken has to reach the table.
- Add cooking loss back. 64.69 ÷ (1 − 0.25) = 64.6875 ÷ 0.75 = 86.25 lb of trimmed raw chicken goes into the oven.
- Add trim back. 86.25 ÷ 0.85 = 101.5 lb as purchased. That is the number on the order, or about 46.0 kg.
- Sides. Two sides at 4 oz each is 8 oz per guest plated, × 1.15 = 9.2 oz. 150 × 9.2 ÷ 16 = 86.25 lb cooked across both sides. At a 90% combined yield that is 86.25 ÷ 0.90 = 95.8 lb to purchase.
- Appetizers. 150 × 5 × 1 = 750 pieces, which at a typical six varieties is 125 of each.
Notice the gap between the plated intent and the invoice: you set out to serve 6 oz per guest, which is 56.25 lb of chicken, and you buy 101.5 lb. The purchase weight is 1.80 times the plated weight. That multiple — uplift × 1/(1−shrink) × 1/yield = 1.15 × 1.3333 × 1.1765 — is worth knowing for your own menu, because it is the number that turns a per-guest price into a food cost.
Reading the result and deciding how much slack to leave
Treat the purchase figure as a floor, not a target. It is the quantity that feeds exactly the headcount you entered at exactly the portion you entered, with zero left over and zero waste beyond the trim and shrink you declared. Three things push the real order above it.
Guarantee versus attendance. Most catering contracts set a guaranteed minimum, and most caterers prepare for a few percent above it. Adding those guests to the guest count field is the clean way to model it, because it flows through every category at once.
Pan geometry and case size. You cannot buy 101.5 lb of chicken. You buy whole cases. Round up to the case and let the extra become staff meal or the client's leftovers, and price accordingly — the recipe cost per serving calculator lets you cost the case you actually buy rather than the pound you actually serve.
Variety splits attention unevenly. When you offer two proteins, guests do not split 50/50. Plan roughly 60% of the headcount on each and accept the overlap, which is a 20% over-order in total protein. That is a deliberate choice, not an error in the formula.
On the low side, one sanity check matters: the planned protein per guest after uplift output. Below about 4 oz of cooked protein, guests read the plate as light regardless of how generous the sides are. Above 10 oz you are buying a steakhouse portion, and unless that is the brief, you are giving away margin.
Planning portions per guest
| Category | Plated portion | Notes |
|---|---|---|
| Protein, main course | 5-7 oz | 6 oz is the common default; 4 oz when two proteins are offered |
| Protein, when guests take both | 4 oz each | Order 60% of headcount on each protein |
| Starch (potato, rice, pasta) | 3-4 oz | Measured cooked; dry rice roughly triples and dry pasta roughly doubles |
| Hot vegetable | 3-4 oz | Two vegetables at 3 oz reads more generous than one at 6 oz |
| Salad greens | 1.5-2 oz | Volume, not weight, is what guests judge |
| Bread rolls | 1-2 pieces | 1.5 average for a dinner service |
| Passed appetizers, before a meal | 4-6 pieces/hour | Front-load the first 30 minutes |
| Appetizers as the meal | 10-12 pieces | Over a two-hour reception, across 6-8 varieties |
| Dessert | 1 portion + 15% | Sheet-cake yields are fixed by the pan, so round to cuts |
These are planning conventions used across catering operations, not a code requirement. Your own event history beats any table here; when you have it, use it.
Quantity planning is also a food-safety plan
Ordering enough is only half the job; the other half is holding it safely. The FDA Food Code limits how long time/temperature control for safety food may sit outside temperature control, and states and counties adopt their own versions of it with their own deadlines. Cook in waves sized to the service window rather than producing the entire 101 lb at once, and check the code your jurisdiction has actually adopted rather than the national model.
Mistakes that leave you short
- Multiplying by the yield instead of dividing. A 75% cooking yield means raw = cooked ÷ 0.75, which is 33% more, not 25% more. Multiplying gives you 56% of what you need.
- Applying a bone-in cooking yield to a boneless purchase spec. Decide which product you are buying first, then pull the yield figure for that product.
- Forgetting the staff. A 150-guest wedding with 14 crew is a 164-guest cook. Put them in the guest count.
- Planning appetizers by the platter instead of by the piece. Guests count pieces. Six varieties at 125 pieces each is a plan; "a few trays" is not.
- Assuming a second protein halves the first. It does not. Guests take some of both, so order 60% of the headcount on each.
- Using a raw portion size as a plated portion size. A 6 oz raw chicken breast plates at roughly 4.5 oz. Enter cooked weight in this calculator.
- Ignoring the ends of the line. On an unattended buffet the first third of guests consistently over-serve. That is exactly what the 1.25 uplift is paying for.
Where this fits with costing and yield testing
Quantity planning and cost control are the same arithmetic read in opposite directions. This calculator turns guests into pounds. Costing turns pounds into dollars, and the bridge between them is the yield figure you used here.
Once you know you are buying 101.5 lb of chicken, the edible portion cost calculator tells you what that chicken really costs per usable pound after the same 85% trim, which is the number that belongs on a costing sheet rather than the supplier's price per pound. Feed those edible-portion costs into the recipe cost per serving calculator to get a per-guest food cost, and then check the whole operation with the restaurant prime cost calculator, where catering labour usually matters more than the food.
What this calculator deliberately does not do is decide your menu. It does not know that the client's family will arrive hungry, that the venue's oven holds four pans, or that half the guest list is vegetarian. It gives you a defensible starting quantity from stated assumptions, and it makes every assumption visible so you can argue with it.
Key terms
- AP (as-purchased) weight
- The weight the supplier invoices: bone, fat, skin, peel, packaging weight of the product itself. This is what you order and what you pay for.
- EP (edible portion) weight
- The weight left after trimming and fabrication, before cooking. Yield percentage is EP ÷ AP × 100.
- Cooking loss (shrink)
- Weight lost during cooking to evaporation and rendered fat, expressed as a percentage of the raw trimmed weight.
- Service uplift
- The factor by which self-service raises average consumption above the plated portion size. It is an operational observation, not a physical constant.