Pricing laser and CNC work: sheet yield, consumables, and machine time

A laser or CNC shop has a costing problem that most crafts do not: almost nothing you sell is bought in the unit you sell it in. You buy plywood by the sheet and sell it by the sign. You buy a tube or a box of end mills and consume them by the minute. You spend twenty minutes setting up a job and then run one piece, or two hundred.

Every one of those mismatches is a place a per-unit cost can come out wrong, and they all come out wrong in the same direction. This chapter walks each of them, using the same underlying method as pricing handmade products: materials, plus labor at a real rate, plus a share of overhead, into a true cost per unit. That method on its own, without a machine complicating it, is the subject of pricing handmade products without underpricing yourself.

Buying in one unit and selling in another is the recurring theme below, and it is a bookkeeping problem as much as a pricing one. The unit conversions article covers how a sheet becomes a part and a tube becomes a minute once the numbers stop living on paper.

Sheet yield, not part area

The instinct is to cost a part by its share of the sheet. Take the sheet price, work out what fraction of its area the part covers, and charge that. It is tidy, and it understates you on nearly every job.

The reason is that you paid for the whole sheet. Everything between your parts, the kerf the beam or the bit removes, the clamping margin your machine cannot reach, and the offcut left at the end all came out of the same purchase. The only honest arithmetic is:

Material cost per part = sheet price ÷ parts you actually nest on one sheet.

Count the parts in your real nest, not in an ideal one. That number is also a lever most shops never pull: fitting one more part onto the same sheet lowers the material cost of every part on it, permanently, for no extra spend. Rotating a part, mirroring alternate rows, or tightening spacing to what your machine can hold is worth real money on a repeat product.

Offcuts deserve a hard rule. An offcut is only worth subtracting from a job’s cost if you have a product that uses that size and you genuinely use it. A rack of drops you are keeping “just in case” is not inventory. It is material you already paid for, sitting where you cannot see it, and pretending otherwise makes every job look cheaper than it was.

Engrave time and cut time are not the same cost

Both are machine hours, but they behave differently, and confusing them is how a job gets quoted badly.

Cutting is a path. Its time follows the total length of the cut and the number of passes, so it scales with the perimeter of your parts and the thickness of the stock. Engraving is a raster. Its time follows the area covered and how finely you scan it, so a densely filled panel takes far longer than an outline of the same size, and dropping the line interval or the resolution changes the time dramatically.

The practical consequence is that you cannot quote off the finished size of a piece. A small tag with a solid engraved background can occupy the machine longer than a large sign that is mostly cut lines. Read the estimated job time out of your control software, then verify it against a clock on the first run of any new design, and cost the piece from the time it actually takes.

Your machine-hour rate, and why it may not be one rate

A machine hour has the same four parts here as anywhere: power, the machine amortized over a realistic lifetime, the wear parts consumed while it runs, and the maintenance labor that keeps it running.

What makes laser and CNC distinctive is how much of that sits in the wear parts, and how directly they are consumed by the minute:

• On a laser, the tube or emitter has a rated life in operating hours, and extraction filter media, lenses, mirrors, the bed or honeycomb, and air-assist components all wear as you run. Divide each replacement cost by the hours it lasts.
• On a router or mill, cutting tools are the dominant consumable, and a bit’s life is measured in cutting minutes rather than in parts. Add the spoilboard you resurface, collets, dust-collection filters, and the spindle itself.

Because tool life depends heavily on what you are cutting, a shop that runs several materials may honestly need more than one rate. Abrasive sheet goods and hardwoods retire tooling faster than softer stock, and averaging them hides which of your products is quietly expensive to make. If you cut two materials with very different tool wear, cost them at two rates.

Setup is a fixed cost, so quantity changes the unit cost

This is the point most shops feel and few price for. Fixturing the stock, squaring the bed, running a test cut, focusing, and jigging a second-side operation take the same time whether the run is one piece or a hundred. Which makes your unit cost a formula rather than a figure:

Unit cost = (setup time × your rate ÷ quantity) + material per part + machine time per part + finishing per part.

Run the arithmetic once with a quantity of one and once with a quantity of fifty and the difference will be uncomfortable. A shop with a single per-unit price is subsidizing every small order out of its large ones, and small orders are usually the ones that arrive most often.

Personalization is the same problem wearing a friendlier face. Twenty tumblers with twenty different names are twenty setups, not one run of twenty. Nothing amortizes, because every piece has its own file, its own placement check, and its own chance of a typo that costs you the blank. A personalization charge is not an upsell; it is the setup cost of a run of one, priced honestly.

Where the per-unit cost lies to you

Beyond the four above, three more that catch shops out:

1. Rejects on your own stock. Burn marks, flashback, chip-out on a final pass, a slipped fixture, a mis-registered flip. Each one costs the material, the machine time already spent, and the time to run it again. Losses are a rate, and the good pieces carry them.
2. Customer-supplied blanks. These look cheap because you have no material cost, and they carry a risk the numbers rarely show: if you ruin a customer’s item you are replacing something you did not buy and did not mark up, plus running the job twice. Price that exposure into the job, agree in writing what happens to a spoiled blank, and treat an unfamiliar coating or an unknown plastic as the unknown it is.
3. Finishing, which is nearly all labor. Weeding and pulling transfer tape, masking and unmasking, sanding edges, removing char, oiling or sealing, assembling multi-part pieces, and packing. None of it involves the machine, all of it involves you, and it is routinely the largest labor line in the shop. Time it once per product and put the minutes in the cost.

Measuring your own numbers

Keep a job log, in the spirit of the batch records and consistency chapter. For each job: the product, the material and sheet size, the parts you nested per sheet, engrave minutes, cut minutes, setup minutes, finishing minutes, quantity, and any rejects with the cause.

Separately, log tool changes and tube or emitter hours. Those two logs are the only way to turn a manufacturer’s rated life into your rated life, in your materials, at your feeds and speeds. After a few months you will have a real cost per sheet-part, a real rate per machine hour, and a real setup time per product, all of them yours. Sheets bought this quarter and consumed next is also the case where cost of goods sold stops being obvious, and COGS for makers is the plain-English account of how that lands.

If you want to see the shape of the stack first, the laser and CNC costing page walks a sign through its blank plus amortized machine minutes, which is the same convention described here.

What carries straight across

Nothing else about the business changes. Where to sell applies as written, and this trade leans harder on wholesale and business-to-business work than most, so the margin discipline in craft fairs and wholesale is the section to read before you take on a standing order. Repeat work is where a shop like this makes its money, and repeat work is exactly where a unit cost that was slightly wrong compounds.

Get the sheet yield, the two kinds of machine time, the tooling life, and the setup amortization honest, and the rest of this playbook works on engraved and machined goods the same way it works on any other craft in it.