Every quote starts with a material number, and most shops get that number wrong in the same direction: too low. They price the part by its net weight or its net footprint — the area or volume the drawing actually defines — and quietly forget that you never buy a part, you buy a sheet, a bar, or a plate and throw a chunk of it away. Nesting material yield is the ratio between the material that ends up in finished parts and the material you paid for, and it is the single most common place a machined or fabricated quote leaks margin before the spindle even turns.

This guide treats yield as a number you estimate on purpose, the same way you estimate cycle time. We will separate net from gross, define scrap and drop, put realistic yield ranges on the table by process, and show how to fold the skeleton you throw away into a price you can defend.

Net part vs gross material: where the leak starts

The drawing describes the net part. The invoice describes the stock. The gap between them is yield, and on a lot of jobs it is bigger than estimators expect.

Take a 200 × 150 mm laser-cut bracket. You nest it on a 2500 × 1250 mm sheet, and after the part-to-part gap, the skeleton, and the edge-clamp margin, you get twelve good parts per sheet. What matters for costing is simple: the sheet costs what it costs, and you got twelve parts, so the material cost per part is the sheet price divided by twelve. Quote it instead as a bare 200 × 150 mm rectangle of metal — the net footprint — and you have undercharged for the skeleton on every single part. The yield here is the bracket's footprint against the sheet area it really occupies, and it is never 100%.

The rule is blunt: cost from what you buy, divided by what you get — never from the net part alone.

Yield, the formula

Material yield is a percentage, and the cost adjustment flows straight out of it.

1. Net material per part — the area (sheet) or volume/weight (bar, plate, billet) the drawing defines.
2. Gross material per part — the stock consumed per part: sheet price ÷ parts per sheet, or bar length per part including cutoff and facing.
3. Yield % = net ÷ gross × 100.
4. Material cost per part = gross material per part × material unit price.
5. Effective material cost = net material cost ÷ yield %.

Step 5 is the one estimators skip. If a part's net material is €8 and your realistic yield on that job is 70%, the material genuinely costs €8 ÷ 0.70 = €11.43, not €8. The missing €3.43 is the drop and skeleton you bought and cannot sell as a finished part. Multiply that across a production run and the "small" rounding becomes the difference between a profitable job and a break-even one.

Scrap vs drop: they are not the same money

Estimators use the words interchangeably; the costing treats them differently.

• Scrap is material removed to make the part — the skeleton webbing between nested laser parts, the chips milled off a billet, the kerf vaporised by the cut. It leaves as swarf or offcut and is gone from the job. Some of it has a recycling value (a few cents per kilo for steel, more for aluminium and a lot more for stainless or titanium), but you do not get the stock price back.
• Drop is usable leftover — the unused end of a bar, the remnant edge of a plate, a half-sheet too big to bin and too small for this job. Drop can be returned to inventory and consumed by a future part, which means its cost should not all land on this job.

The mistake is charging full price for material that becomes recoverable drop, or assuming drop will be reused when in practice it sits in the rack until it is scrapped. A defensible quote takes a position on each: scrap is a full cost minus a small recycling credit; drop is either credited to inventory or, if your shop realistically never reuses it, costed as scrap. Pretending all drop gets reused is how shops convince themselves their margins are healthier than they are — one of the recurring themes in reducing quoting errors and improving accuracy.

Typical yield ranges by process

Approximate, real-world starting points to calibrate against your own job history — not fixed constants. Geometry, batch size, and nesting skill move every one of these.

Process / stock	Typical material yield	Main loss
Laser / plasma sheet	60–85%	Skeleton, part-to-part gap, edge margin
Turned parts from bar	60–80%	Chips, facing, parting kerf, bar-end drop
Milled from plate/billet	30–60%	Roughing stock-removal, clamping allowance
Press-brake blanks	70–90%	Nesting gap, bend-relief offcuts
Tube / profile cutting	75–92%	Saw kerf, end drop

Two patterns stand out. Sheet and tube work lives or dies on nesting efficiency — the geometry is mostly perimeter, so the win is fitting more parts per stock unit. Milling from solid is the brutal one: a part that is 40% of its billet means 60% of the stock you bought leaves as chips, and on aluminium or stainless that is real money on the floor. Costing milled parts from net weight instead of billet weight is one of the biggest single quoting leaks, covered alongside the rest of the build-up in how to price CNC machined parts.

Nesting: where sheet yield is actually won

For flat-cut work, yield is mostly a nesting problem, and nesting is mostly geometry plus discipline.

• Rotate and mix parts. A nest of one part orientation wastes the gaps; mixing parts and rotating them fits small parts into the corners big parts leave behind.
• Fill the skeleton. Drop small parts from other jobs into the holes of large parts — common-line cutting and shared edges turn dead skeleton into product.
• Respect the real edge margin and gap. Too tight and parts tab together or the clamp fouls; too loose and you bleed yield. Use the gap your machine and material actually need, not a guess.
• Batch across jobs. Two small jobs in different materials waste two sheets; two small jobs in the same grade and thickness can share one. Nesting across the week's queue, not job by job, is where multi-job shops claw back the most material.

A 5–10 point improvement in sheet yield is not exotic — and on expensive stainless or aluminium that swing can be larger than the entire machine-time cost of the cut. The interaction between yield, cut path, and downstream forming is exactly what dedicated sheet metal quoting software is built to model.

Bar and tube: the drop nobody quotes

On a turned job you buy a 3-metre bar and feed it through the lathe. Each part consumes its turned length plus the parting kerf plus a facing allowance, and at the end of the bar a stub too short to chuck becomes drop. Quote the part by its finished length and you have ignored the kerf on every part and the end-drop on every bar.

Worked example: a part needs 80 mm of finished length, the parting tool eats 3 mm of kerf, and you allow 2 mm facing — so 85 mm of bar per part. A 3000 mm bar with a 100 mm unusable end gives 2900 ÷ 85 ≈ 34 parts. Material cost per part is the bar price ÷ 34, not the bar price ÷ (3000 / 80 = 37). Those three "phantom" parts are the kerf and drop, and they are pure cost. This per-bar arithmetic, and where it sits in the turning quote, is laid out in the CNC turning and lathe quoting guide.

Folding yield into the price

Once yield is a number, the price build-up is clean. Compute net material cost, divide by yield % to get effective material cost, add a recycling credit on scrap and an inventory credit on genuinely reusable drop, then carry the result through your normal margin step.

The order matters. Apply margin to the effective material cost, not the net — otherwise you are marking up a number that was already too low, and the margin percentage hides the fact that you gave the skeleton away at cost. (If margin and markup are not the same thing to you yet, margin vs markup settles it.) And make sure the yield assumption and the machine-time burden are not double-counting or both ignoring the same drop — the burdened-rate logic in machine shop hourly rate calculation keeps the time and material sides honest.

From drawing to yield-aware quote, automatically

The reason yield gets skipped is not that estimators do not understand it — it is that working it out by hand for every part is slow. You have to read the material and stock size off the drawing, estimate the nest, count the kerf, decide what is drop, and only then reach a material number. Under deadline that whole chain collapses into "call it the net plus a bit," and the bit is rarely enough.

QuoteBuddy reads the technical drawing, pulls the material, stock form, and geometry it needs, and builds the material side of the quote from stock size and yield rather than net part alone — so the skeleton, the kerf, and the drop are in the number instead of in your losses. The cost engine applies your yield assumptions, material prices, and target margin the same way on every part, so two estimators quoting the same drawing land on the same defensible figure.

Start a 30-day trial and run a few real drawings through it. Compare the material cost it builds against what you would have quoted from the net part — and see how much of your margin was hiding in the skeleton.