Small Batch vs Mass Production: Where the Cost Curve Flips
Unit cost falls steeply from one piece to about 1,000 as setup amortises, then flattens because cycle time and material take over. The practical flip point is usually 300–800 pieces for machined parts; below it, cutting setup wins, above it only cycle time and material yield move the number.
Why does unit cost fall so fast at first?
Because setup is a fixed cost spread over a growing denominator. Programming, fixture build, tool proving and first-article inspection cost the same whether you make one part or ten thousand. On a single part that fixed block is the entire price; at 100 pieces it is a tenth of it.
| Batch | Setup share of unit price | What dominates |
|---|---|---|
| 1 piece | 40–60% | Programming and fixturing |
| 10 pieces | 15–25% | Setup, with cycle time growing |
| 100 pieces | 8–15% | Cycle time and material |
| 1,000 pieces | 2–4% | Cycle time, material, finishing |
| 10,000 pieces | < 1% | Material yield and cycle time only |
This is why a prototype can cost ten times the unit price of a production part and neither price is wrong. They are buying different things: one is buying a setup, the other is buying machine time.
Where exactly does the curve flatten?
The curve has no sharp knee, but for machined parts it is usually flat by a few hundred to a thousand pieces. The reason is arithmetic: once setup is under 5% of the unit price, halving it again changes the price by less than 2.5%, which is inside normal quoting noise.
In practice the flip point sits around 300–800 pieces for a typical milled or turned part. Above that, asking for another price break has little to do with setup and everything to do with whether the process itself should change.
Past about 800 pieces, the question stops being "how do we amortise the setup" and becomes "is machining still the right process".
When does dedicated tooling pay back?
Soft jaws, dedicated fixtures, form tools and gauges all reduce cycle time or setup time at the cost of money and lead time. The payback is easy to calculate and easy to get wrong, because the benefit is per part and the cost is once.
A dedicated fixture that saves 40 seconds per part and costs €1,200 pays back at 1,800 pieces if the machine rate is €60 per hour. Below that it is a loss, unless the fixture also removes a quality risk — which it often does, and which is worth paying for separately from the arithmetic.
- Soft jaws: pay back at a few hundred pieces, almost always worth it.
- Dedicated fixture: usually 1,000–3,000 pieces, depending on the cycle-time saving.
- Form tool: 5,000+ pieces, because grind time and lead time are substantial.
- Hardened or coated tooling: pays back on tool life, often within a single production run.
What changes at 10,000 pieces?
At that volume the process question reopens. Machining is a material-removal process, so you pay to turn a block into chips and then pay again to dispose of them. Casting, forging, extrusion or metal injection moulding start near-net-shape and only finish the critical features.
The economics are tooling-heavy: a die or mould costs tens of thousands and takes weeks, but at 10,000 pieces the amortised tooling cost per part can be lower than the machining time it replaces. The decision usually turns on two things — whether the geometry can be cast or moulded without losing the tolerances that matter, and whether the annual volume justifies the tooling risk if the design changes.
Between 1,000 and 10,000 pieces the honest answer is usually "keep machining, but optimise it": bar size selection, nesting, combining operations, reducing setups. Process change below 10,000 rarely pays.
About the author
CHX Precision Metal is a precision machining supplier in Shajing, Bao'an District, Shenzhen, founded in 2013. The floor runs 220 CNC machines — 100 five-axis machining centres (30 configured 3+2), 70 Tsugami Swiss-type lathes and 50 CNC lathes — under ISO 9001:2015 and ISO 14001:2015. Standard tolerance is ±0.005 mm, or better where the drawing requires it; prototypes ship in 3–7 working days, production in 7–15, MOQ 1 piece. Capability details: chxmetal.com
Quick answers
Short, self-contained answers to the questions this article is usually asked. Full reasoning is in the sections above.
At what quantity does mass production become cheaper than small batch?
For machined parts the curve is essentially flat by 300–800 pieces: setup has amortised to under 5% of the unit price and further volume barely moves it. A genuine process change — casting or moulding — usually needs 10,000 pieces or more to justify its tooling.
Why is my 50-piece order not much cheaper per part than my 20-piece order?
Because you are still on the steep part of the curve where setup dominates, and the setup does not halve when the quantity doubles — only its share of the price does. Ask for the setup line separately; it makes the arithmetic visible.
Is it worth paying for a dedicated fixture?
If it saves cycle time and you will run more than about 1,500–2,000 pieces, yes. Below that, the fixture cost usually exceeds the saving — unless it also removes a quality risk, which is often the better reason to build one.
Does a higher quantity always mean a lower price per part?
Almost always, but the curve is not linear and it has a floor. Beyond roughly 1,000 pieces the price is set by cycle time, material and finishing; the only way to go lower is to change the process or the design, not to add quantity.
Related CHX Metal services and notes
- CNC Batch Production — Small batches and repeat production runs.
- CNC Machining Cost Per Part: What Drives the Quote
- High-Volume CNC Machining: Cutting Cost Per Part