From prototype to production run — how CNC machining cost changes

A prototype quoted at several hundred zloty and the same part at a fraction of that price in a batch — to many buyers this looks like a calculation error. It is not. CNC machining cost has a structure in which a large share of the money goes not into the cutting itself but into production preparation. And preparation is a fixed cost: it is spread over however many parts you order.

In this post we break the part price down into its components: what exactly sits inside preparation, how setup divides across the batch size, what an indicative unit cost index looks like for 1, 10, 100 and 1000 parts, and what to do so the transition from prototype to production run is as cheap as possible.

What CNC machining cost consists of

The price of a machined part is, simplified, three blocks:

• production preparation: reviewing the documentation, CAM programming, selecting and setting the tools, designing or selecting the fixture, setting up the machine and producing a measured first article,
• machine time: the cycle time of one part multiplied by the number of parts, plus the stock material,
• inspection and production-adjacent activities: measurements, quality documentation, packaging, logistics.

The first block is largely independent of quantity. A program for milling a housing is written once — regardless of whether the machine then produces 1 part or 500. The same applies to turning: tooling up the turret, tool offsets and the measured first article cost the same whatever the batch size.

The second block grows linearly with quantity. The third partially: first-article approval is fixed, while in-process inspection depends on the batch and the requirements. For a fuller review of the price drivers, see the post how much does a CNC part cost.

Setup — a fixed cost divided by the number of parts

Let us run an example. The figures are for illustration only and serve to show the mechanism — they are not a price list.

Assume a milled part where preparation (programming, fixturing, setup, measured first article) takes 4 hours in total, and the machining cycle for one part is 12 minutes.

• at 1 part: you pay for 4 hours of preparation and 12 minutes of cutting — preparation is over 90% of the time,
• at 10 parts: the same 4 hours divide across 10 parts, i.e. 24 minutes of preparation per part,
• at 100 parts: only 2.4 minutes of preparation fall on each part — less than the cycle time itself,
• at 1000 parts: the share of preparation in the unit price is negligible; the cost is decided by cycle optimisation and production organisation.

That is why "how many parts?" is the first question of every quote. Without that information the supplier does not know whether to price the part as a prototype or as a batch — and the difference in unit price can be severalfold.

In practice there is one more fixed cost: changeover between jobs. The more items in a single order share similar tools and fixtures, the less time the machine loses between parts — an extra argument for bundling parts into packages instead of sending each one separately.

Indicative unit cost index: 1, 10, 100 and 1000 parts

The table below shows the typical shape of the cost curve for a part of average complexity. The values are indicative and index-based — a part in a batch of 1 has an index of 100; the others show the relationship, not prices. The actual curve depends on geometry, material and requirements.

Two conclusions from this curve. First, the biggest drop in unit price happens between 1 and 10 parts — after that the curve flattens. Second, in large runs further savings no longer come from spreading the setup but from optimisation: shorter cycles, better tooling, fewer changeovers.

What helps you move cheaply from prototype to production

The transition to a production run is cheap when the work done on the prototype is not thrown away. In practice, the following help:

• a frozen drawing revision — every design change after the prototype means reworking the program, sometimes a new fixture and a new first article,
• preserved process documentation — the CAM program, tool sheet and parameters from the prototype are ready for the production batch,
• repeatable fixturing — vices with datums, dedicated fixtures or soft jaws made once serve all subsequent batches,
• tolerances only where functionally needed — tightening "just in case" raises the cost of every part in the batch, as we explain in the post how much does accuracy cost,
• a commercially available material — an exotic grade in small quantity can cost more than the machining,
• a declared repeat schedule — if the supplier knows the batches will return, they can invest in fixturing and amortise it over a longer horizon.

It is worth saying this outright in the enquiry: "prototype 2 pcs, target batch 200 pcs quarterly". That information changes how the process engineer designs the process — from stock selection to fixturing strategy.

Repeat orders — the second wave of savings

The cost curve does not end with the first batch. On the second and subsequent orders of the same item, preparation is much shorter: the program exists, the process is documented, the fixtures are waiting on the shelf and the parameters were proven on the first run. What remains is setting up the machine and resuming production.

The conditions for this saving to actually work:

• the drawing does not change between batches, and if it does, the revision is clearly marked and communicated,
• the orders return to the supplier who maintains the process documentation from the first batch,
• the gaps between batches are reasonable enough that the fixtures and tools are still available.

This is a strong argument for consolidating parts with one subcontractor instead of re-tendering every batch — the process knowledge stays with the supplier and works in favour of future prices.

When it pays to order 10 parts instead of 2

Since setup dominates at small quantities, the total cost of a 10-piece batch is often only moderately higher than that of a 2-piece batch. For illustration: if 2 parts cost an index total of 210, then 10 parts may cost around 380-450 — five times more parts for less than twice the budget. These are indicative values, but the mechanism is universal.

A larger trial batch makes sense when:

• the design is already stable and changes after testing are unlikely,
• the parts wear out or may be damaged during assembly trials,
• you need parts for several test recipients at once (assembly, laboratory, customer),
• the part is a spare and a stock shortens future downtime.

Conversely: if you expect geometry changes after prototype testing, ordering a stock means freezing money in parts that may end up as scrap. In that case 2 parts and a fast correction loop are the better choice.

If, on the other hand, your batches grow to the point where you start considering your own machine instead of external orders, run the numbers coolly — the post CNC machining outsourcing or your own machine park will help.

Summary

CNC machining cost does not fall with batch size because the supplier "gives a discount" but because the fixed cost of preparation is divided by the number of parts. The biggest jump in economics happens between 1 and 10 parts, and in large runs the price is decided by cycle time, not setup. A stable drawing, repeatable fixturing and clearly declared quantities are the simplest tools for lowering the unit price.

Planning a prototype with a production run in sight? Send the drawing and the expected quantities through the contact form — we will prepare a quote within 48 hours, broken down by quantity variants so you can see how the unit price changes with the batch.