Make or Buy: Costing Prep Labor Properly

The decision nobody actually calculates

Every kitchen makes make-or-buy decisions constantly. Stocks, sauces, dressings, pastry, bread, pickles, portioned proteins, prepped vegetables — each one is either produced in-house or bought ready to use, and the choice is usually made once, on instinct, by whoever set the menu up. It is then almost never revisited, even after wages have risen 30% and the supplier has launched a product that did not exist when the decision was made.

When operators do run the numbers, they run them on ingredients. Flour, water, yeast and salt cost far less than a bought-in loaf, so the kitchen bakes. That comparison is not wrong so much as incomplete: it puts your raw goods against a supplier price that already includes their labor, their overhead and their margin. Comparing one side's inputs against the other side's finished price will favour making it every single time, regardless of whether making it is actually the better decision.

The fix is not complicated. Put a defensible number on the hours your own team spends, add it to the ingredients, and compare like with like. What usually happens is that the conclusion stays the same but the margin turns out to be far narrower than anyone assumed — and knowing how narrow it is tells you which decisions are stable and which are one wage review away from reversing.

Why the hour is not free

The standard objection to costing prep labor is that the cook is already there. They are rostered from 8am, the stock takes an hour, and no additional money leaves the business because of it. On a single quiet Tuesday that is true. As a way of running a kitchen it fails, because it treats scheduled hours as unlimited.

Payroll hours are a budget you set in advance. Every task you add to the prep list consumes part of it, and when the list exceeds the budget the difference gets paid for in one of three ways: an extra shift, a longer close, or corners cut somewhere that shows up during service. All three cost money. The cheapest of them is the extra shift, and that is a straightforward hourly cost you can see on the payroll report.

So cost the hour at what an hour actually costs: the wage plus burden. Payroll taxes, workers' compensation, insurance and paid leave typically add 15% to 30% on top of the gross rate. A cook on $18 costs closer to $22 loaded, and using $18 in a comparison biases the answer towards in-house production by roughly a fifth of the labor line. The labor cost formula works the same way at the department level; this is just the same logic applied to one task.

What a real comparison looks like

Take a batch of house sauce base: 90 minutes of hands-on work, $96 of ingredients, 40 portions out the other end. The cook is on $18 with a 22% burden, so $21.96 loaded, and the labor in the batch is $32.94. Total make cost is $128.94, or $3.22 a portion. The commercial equivalent lands at $3.60.

On ingredients alone that read as $2.40 against $3.60 — a third cheaper, obviously worth doing, no discussion required. With labor counted it is a 10% saving of about $15 a batch. Still positive, but now it is a decision with visible sensitivity. If the batch gets halved, the fixed setup time does not halve and the advantage largely disappears. If wages rise 10%, most of the rest goes. If the supplier offers a case discount, it flips.

That sensitivity is the useful output. A 10% advantage means review it annually. A 40% advantage means stop thinking about it. A negative number means you are paying a premium for the house version, which may still be the right call — but you should be making that choice deliberately rather than believing it saves money.

The variables that decide it

Four things move these numbers more than anything else, and three of them are within your control.

• Batch size. Setup, heating and cleanup are fixed per batch. Doubling the batch might add 30% to the minutes while doubling the output, which is the single fastest way to make an in-house item cheaper. Frequency matters as much as technique.
• Who does the task. The same job costed at a commis rate versus a sous chef rate can differ by 50%. Prep that has drifted upwards to expensive staff is a common and invisible cost.
• Yield. The comparison is per sellable portion, not per batch. An item that loses more than expected in cooking or portioning costs more per portion than the recipe card claims. Measure it before you cost it.
• Supplier pricing tiers. Bought-in prices move with volume. A quote at your current usage may look poor while a quote at consolidated volume across two or three items looks very different.

Yield in particular deserves attention, because it distorts both sides of the decision at once. If you are not already adjusting for it, the yield percentage guide covers how to measure it properly.

What the numbers leave out

A per-portion cost is an input, not a verdict. Several things that matter never appear in it.

Consistency is the largest. A bought-in product is identical every delivery. A house product carries the variance of whoever made it that morning, and variance produces remakes, complaints and the quiet erosion of a dish's reputation. If a house item is only marginally cheaper and noticeably more variable, buying it in is usually the better operational decision even at a slightly higher price.

Opportunity cost is the subtlest. If the hour comes out of genuine slack, it is nearly free. If it comes out of a prep list that is already overrunning, the true cost is whatever gets dropped to make room — and what gets dropped is rarely the thing you would have chosen. Resilience matters too: if one person is the only one who can make it, the menu depends on their attendance.

Against all that sits the reason to make things: some house production is why guests choose you. Bread, stocks, pastry, cures and pickles can be part of the identity of the food, and that is worth paying for. The point of costing it is not to eliminate house production but to know what it costs, so you spend that money where guests notice rather than on a dressing nobody could identify in a blind test.

How to run the review

Do it once a year and treat it as a two-hour exercise rather than a project. List every prep item that takes more than fifteen minutes. Rank by weekly minutes consumed, not by how interesting the item is. Take the top ten — they will account for most of your prep hours — and time each one honestly across three occasions and more than one cook.

Cost each at the loaded rate of whoever realistically owns the task, add ingredients at yield-adjusted prices, and get a quote for the equivalent bought-in product. Then sort the list by percentage advantage. Anything with a clear advantage in either direction is settled. Anything inside 10% either way goes on a watch list to be checked after the next wage or price change.

Two cautions. First, remember that switching to bought-in moves cost from payroll into food cost, so your prime cost is the only figure that shows the net result — a food cost percentage rise after a sensible switch is not a problem. Second, do not act on a saving unless the freed hours are actually removed from the roster or redirected to something productive. Buying in a task and leaving the schedule untouched increases your costs without reducing anything, which is the most common way these reviews end up losing money.

Frequently asked questions

How do I put a value on prep time if the cook is already scheduled?

Use their loaded hourly rate anyway. Scheduled hours are a fixed budget, and any task added to the prep list either displaces another task or extends the shift. Treating the hour as free means you can never compare it to a supplier price that includes labor.

Does buying in more items make my food cost percentage worse?

Usually yes, and that on its own is not a problem. Bought-in products carry the supplier's labor inside their price, so cost moves from payroll into food cost. Judge the switch on prime cost, which captures both lines, rather than on food cost percentage alone.

Which prep items should I review first?

Rank by total weekly minutes consumed, not by cost per batch. A ten-minute task done twice a day consumes more payroll than a two-hour task done weekly, and it is the high-frequency items that quietly absorb the roster.

How accurate do the prep times need to be?

Within about ten percent is enough to make a decision. Time each task three times on different days and with more than one cook. Estimates from memory come in low almost every time, usually because setup and cleanup are forgotten.

Should quality ever outweigh a cost advantage?

Yes, and it often should. If guests can tell the difference and the item is part of why they come, a premium is justified. The test is whether the difference is noticeable to a guest rather than to the chef who makes it.

What if the in-house version is cheaper but nobody has time to make it?

Then it is not cheaper. A saving that depends on hours you do not have is either paid for in overtime or in something else being skipped. Either cost the additional hours honestly and see if the advantage survives, or buy it in.

How often does the answer actually change?

More often than most operators expect, because wages have moved faster than supplier prices in recent years. Any item where in-house production wins by less than ten percent should be rechecked after every wage review.

Run the numbers

Use the free Prep Labor Cost Calculator to apply everything above to your own figures.