Pricing baked goods: yield, shrinkage, and waste

Baking has the cleanest scaling arithmetic of any craft in this playbook and the messiest cost per unit. The arithmetic is clean because a recipe is proportional: double the flour and everything else doubles with it, so a batch cost divides neatly by the units it yields. The cost per unit is messy because three things happen between the recipe and the money, and none of them appear in the recipe.

Dough loses weight in the oven. Ovens charge by the load, not by the loaf. And product that does not sell inside its shelf life is a total loss, in a way a mug or an engraved sign never is.

This chapter costs those three properly. The underlying method is unchanged from pricing handmade products: materials, plus labor at a real rate, plus a share of overhead, into a true cost per unit, with your price above it at a margin you choose. The guide on pricing handmade products without underpricing yourself walks that method through one item without a bakery around it, which is a useful thing to read first if the arithmetic below is new.

One thing this chapter is not about. Wedding tiers, celebration cakes, and made-to-order designs are a catering job, quoted per order and booked around a date, and they cost out completely differently from the same trays baked every week. Everything below is about production and wholesale baking.

Ingredient cost, and the weight that disappears

Start from a cost per gram for every ingredient: pack price divided by pack weight, off your current invoice rather than a remembered one. Baker’s percentages then do the rest, because a recipe expressed against flour weight scales to any batch size without re-deriving anything.

The step most bakers skip is bake loss. Dough goes into the oven wet and comes out lighter, and how much lighter depends on the product, the oven, the steam, and the bake time. If you sell by a stated finished weight, your dough has to be scaled up to land there:

Dough weight per unit = target finished weight ÷ (1 − your bake loss).

Measure your own loss rather than borrowing a figure. Weigh the scaled dough for a batch, weigh the same batch after it has cooled, and the difference over the starting weight is your bake loss for that product in your oven. It will differ between a lean loaf and an enriched bun, and it is the reason a recipe costed on raw ingredient weight quietly understates every unit.

Then there is the yield you lose to handling: scaling remainders, the piece that came out short, misshapen units, end slices, and the tail of a batch that did not fill a full tray. Count units that leave the bench sellable, not units the recipe theoretically makes.

The oven charges by the load

An oven has the same cost shape as a kiln. Gas or power for the bake, the oven amortized over a realistic lifetime, and the maintenance that keeps it calibrated. Measure the energy on your own meter across a full bake rather than trusting a nameplate, and multiply by the rate on your own bill.

And the same hard lesson applies: a half-full oven costs what a full one costs. Baking two trays in a deck that holds six triples the bake cost of every unit on them. Scheduling is a pricing decision, and consolidating short runs into full loads is one of the few cost reductions available without changing a single ingredient.

Proofing, retarding, and cooling are not labor, but they are occupancy. Bench space, rack space, and retarder space cap how many batches a day the room can turn, which means your real constraint is usually space and oven hours rather than hands. Knowing which constraint binds tells you which product is expensive: the one that ties up the scarce thing longest.

Labor itself is easy to undercount because mixing is the fast part. Scaling, shaping, panning, docking, glazing, filling, finishing, cooling, packing, and washing down are the slow parts, and they do not shrink when the batch grows the way mixing does. Time a product end to end once and use that number.

Shelf life is the cost that has no equivalent elsewhere

A mug that does not sell this month sells next month. A loaf does not. Everything you bake is on a clock, and product that passes its life is a complete loss of material, labor, oven time, and packaging all at once.

That makes sell-through part of your cost, not part of your sales report:

Cost per unit sold = cost per unit baked ÷ your sell-through rate.

If nine in ten of what you bake sells at full price, every sold unit carries about a ninth again. At four in five, it carries a quarter again. Day-old markdowns soften the loss without erasing it, and a unit sold at a discount still cost exactly what a full-price one cost.

Where that waste lands depends entirely on the channel, which is the real difference between the two halves of this business:

• Retail and market selling means you bake to a forecast and you personally eat everything the forecast got wrong. Waste is your cost, and it varies enormously by day of the week.
• Wholesale means an account orders and you bake to that order, which is why wholesale waste is usually far lower. The exception is worth reading your agreement for: where an account can return or claim credit for unsold product, the waste has simply moved back to you with extra handling attached.

Sampling, staff eating, and donations belong in the same count. Choosing to do all three is reasonable and often good business. Treating them as free is not, because they consumed real ingredients and real oven time. Count them, then decide.

Wholesale pricing runs backward

Direct selling lets you build a price up from cost. Wholesale rarely does. The account has a shelf price in mind and a markup they need, so your wholesale price is capped from above by what their own margin allows.

Which reverses the exercise. Instead of costing up to a price, you are checking whether your cost fits underneath one. If it does not, the answer is a cheaper formulation, a bigger batch, a fuller oven, or declining the account, and none of those are the answer of working faster for less.

Two costs show up on the wholesale side that direct selling never has. Delivery is charged per drop rather than per unit, so a small order carries the entire cost of getting there, which is exactly why minimum order values exist. And payment terms mean you are financing the account: you bought the flour weeks before the invoice is paid, and that gap is working capital coming out of your pocket.

The general wholesale discipline in craft fairs and wholesale applies here unchanged, and it is the section to read before a first stockist conversation.

Where the per-unit cost lies to you

1. Costing raw dough weight rather than finished units. Bake loss makes the recipe look cheaper per unit than it is.
2. Dividing by units baked instead of units sold. Sell-through is a cost, and it is the largest hidden one in this trade.
3. Pricing off an old invoice. Butter, eggs, chocolate, and flour move sharply and seasonally, and a recipe cost is a snapshot rather than a fact. Re-cost from current receipts on a schedule, the way the supplies and sourcing chapter describes.
4. Half-empty ovens and short runs. The load is the unit the oven bills you in.
5. Spreading delivery across units instead of drops. A single-crate delivery is expensive per unit and looks cheap when averaged.
6. Counting mixing as the labor. Finishing and packing usually outweigh it.

If you want to see the stack assembled before you fill in your own figures, the wholesale bakery costing page walks a dozen through ingredients, packaging, and margin.

Measuring your own numbers

Keep a bake log in the spirit of the batch records and consistency chapter: product, batch size, dough weight scaled, units out sellable, units sold at full price, units marked down, units wasted, oven load and bake time. Break it down by day of the week, because the day-of-week pattern is where forecasting waste actually improves.

A month of that gives you a real bake loss, a real yield per batch, a real sell-through per product per day, and a real oven cost per load. Re-cost your ingredients monthly off invoices and the four numbers stay current; a photographed invoice only helps if it can be read, which is what the scan quality article is about.

Those same monthly figures are what a cost of goods sold is built from at the end of the year, and butter and flour moving mid-year is exactly the case that makes the accounting method matter. COGS for makers covers what that number is and how to keep it defensible without a data-entry weekend.

Labeling and food rules are the one area this playbook cannot settle for you, and they depend on where you are, where you sell, and whether you are selling direct or through a shop. The labeling and regulations chapter points at the right authorities rather than guessing; check the requirements for your state and your channel before you sell.

Get the bake loss, the oven load, and the sell-through honest, and the rest of this playbook runs on baked goods the same way it runs on any other craft in it.