What Actually Drives the Cost of a CNC Machined Part

What is actually in the price

A quotation for a machined part is a sum of components, and knowing them makes the rest of this article make sense.

Component	Nature	Grows with quantity?
Material	Per part, driven by the weight bought	Yes — linearly
Fixturing and programming	One-off	No — spread across the batch
Machine setup	One-off time	No — spread across the batch
Cycle time	Per part	Yes — linearly
Cutting tools	Per part, driven by tool life	Yes — linearly
Finishing and secondary operations	Per part, plus a setup each	Yes, plus one-off per operation
Documentation and packing	Per part, plus one-off	Mostly per part

Two things in that table behave differently from the rest, and they explain almost everything about how a machining quotation moves.

The one-off items — fixturing, programming, setup — are fixed. They do not grow with the batch, so their share of the unit price falls as the batch grows.

The per-part items — material, cycle time, tools — are variable. They apply to every piece, so their share of the price stays constant no matter how many you order.

The whole shape of a machining quotation follows from the interaction of those two.

The quantity curve, and what its shape tells you

Plot the unit price against quantity and you get a curve that falls steeply and then flattens. That shape is not arbitrary — it is the fixed cost being divided by a growing number.

The useful information is where the curve flattens. That is the point at which the fixed costs have become negligible and the price is essentially material plus cycle time. Beyond it, ordering more does not help much, and further savings have to come from changing the part rather than the quantity.

It also tells you what to attack:

• On the steep part of the curve — the answer is almost always quantity, or the removal of a setup. The part’s material and geometry are not the problem; the batch is too small for the work involved.
• On the flat part of the curve — the answer is the part itself. The material, the removed volume, the tolerance and the feature count are now setting the price, and quantity no longer helps.

Most cost-reduction discussions go wrong because neither party has checked which part of the curve the part is on. Asking for material savings on a fifty-piece order, or for quantity savings on a part that is already at a hundred thousand, are both attempts to fix the wrong component.

The mechanics of how quantity is quoted are covered on the bulk order pricing page.

Material: you pay for what you buy, not what you ship

The material line in a quotation is the weight of stock bought, multiplied by the price per kilo for that grade and form. It is not the weight of the finished part.

The difference between the two is the stock allowance plus everything removed to create the shape, and on some parts it is dramatic:

• A round part turned from bar close to size wastes very little — the allowance is a skin.
• A round part milled from a block wastes everything outside the finished cylinder, which on a long part can exceed the volume of the part itself.
• A part with deep pockets wastes the pocket volume.
• A thin, flat part cut from plate wastes the profile.

Three consequences follow, and all three are actionable:

The stock form matters as much as the grade. Choosing the process and the stock so that the bought shape is close to the finished shape can be a larger saving than any negotiation on the material price. This is the argument that sits behind the choice between turning and milling, and behind the choice between hot rolled and cold drawn bar.

On expensive alloys, the waste line becomes a major cost. When you are buying titanium or a high-strength stainless, turning most of it into chips is expensive even after accounting for scrap value. On those parts, designing so that less material has to be removed is a direct saving.

A cheaper grade is not automatically a cheaper part. If the cheaper material cuts more slowly, wears tooling faster or needs extra operations, the machining line can rise by more than the material line falls. The material and the machining have to be compared together — which is why a free-machining grade that costs more per kilo can produce a cheaper finished part.

Setup: the strongest lever, and the most ignored

A setup is one fixturing of the part in one orientation. Every feature that cannot be reached from the current orientation needs another setup.

Setups cost in three ways:

• Time — fixturing, indicating, proving the position, and often a first-piece check
• Accuracy — every setup introduces its own positional error, and tolerances between features cut in different setups become a stack of both errors plus the fixture
• Risk — every time a part is taken out and put back, there is a chance of damage, mis-loading or a chip left under a locating face

This is why the number of setups is the strongest cost lever available, and why the most effective single change to a drawing is almost always removing a face that does not need machining. It does not reduce the material or the cut time much. It removes an entire operation, its fixturing, its handling and its error contribution.

The practical instruction that unlocks this is stating which faces are functional. A drawing that shows every face machined will be quoted that way. A drawing that says which faces are sealing, locating or mating — and leaves the rest as-supplied — is a cheaper part, and the shop can see the saving.

Where a part genuinely needs features on several faces, the setup count is what justifies a multi-axis machine or a different process route, which is covered in the comparison of three-axis, four-axis and five-axis machining.

Cycle time: what actually makes a cut slow

Cycle time is the per-part cost that never goes away, and it is driven by five things.

The volume of material removed. More metal to take off means more time. This is the one that most directly rewards designing a part that is close to the stock shape.

The machinability of the grade. Aluminium cuts fast, brass cuts faster, mild steel is workable, stainless and titanium are slow, and hardened material is slower still. The same geometry in two different materials can differ substantially in cycle time — the grades we run, and how each behaves on the machine, are on the materials page.

The tolerance. Cutting to a general tolerance block is quick. Cutting to ±0.01 mm requires a different approach — lighter finishing passes, more care over tool wear, sometimes a different tool, and often measurement between operations. See tolerances we hold for where the machine capability actually sits.

The surface finish. A fine finish needs additional passes at reduced feed. Where the finish is cosmetic it has to be achieved over the whole visible face, which may be more area than the functional requirement.

Feature difficulty. Deep pockets need longer, more slender tools that must be fed more slowly. Thin walls need lighter cuts to avoid deflection. Small features in hard material need care. These are the drivers that make an apparently simple part slow.

Tools, finishing and secondary operations

Cutting tools appear in the price as tool life: how many parts a cutter makes before it is replaced, and how much it costs. Abrasive materials, hard materials and scale on hot rolled stock all shorten tool life, which is why the same geometry can cost more from a different stock condition.

Finishing is quoted as a separate line because it is usually a separate operation, often through an external partner. Anodizing, plating, powder coating, passivation and the rest each carry their own setup and handling. Where a part is finished, that line can be a meaningful share of the total, and it is worth knowing that a cosmetic finish over a large area costs more than one over a small one.

Secondary operations — deburring, chamfering, tapping, marking, assembly — are usually quoted on the same drawing as the primary operation. Individually small, collectively significant, and often the reason two apparently similar parts quote differently. They are listed on the secondary operations page.

One further line belongs in this list even though it is not a machining cost: the documents and the packing. Every batch ships with a material certificate (MTR) tracing the grade to its mill heat and with pre-shipment photos, and both are part of the order rather than extras. What ships is set out on the quality documents page, and where a programme needs something beyond it, the time to say so is at the quotation stage.

The levers that work, in order of effect

1. Remove a setup. Eliminate a face that does not need machining, or consolidate features so they can be reached from one orientation. Nothing else comes close for the effort involved.
2. Relax a tolerance that does not need to be tight. Take the tight tolerance off features that are not functional and leave it only where it matters. Tolerance is a per-part cost that never goes away.
3. Choose a more machinable grade, if the application allows it. The material and the machining have to be compared together, and the faster-cutting grade frequently wins on total cost.
4. Start from stock closer to the finished shape. Change the process, the stock form or the nesting so that less metal is removed. On expensive alloys this can be the dominant saving.
5. Simplify the geometry where it is free to do so. Larger internal radii, shallower pockets, fewer distinct features. This costs nothing functionally in most parts.
6. Increase the quantity, if the part is on the steep part of the curve and the volume is real rather than hoped for.
7. Accept a standard finish where the finish is not seen or not functional.

The changes that save nothing

Several intuitive cost-reduction moves do not work, and some make things worse.

Shrinking the part. Less material and a shorter cut, but smaller tools, less rigidity, harder handling and often a tighter achievable tolerance. Small parts are frequently more expensive per unit weight, not less.

Switching to a cheaper material. The material line falls and the machining line may rise further. A cheaper material that machines badly is a false economy.

Thickening or thinning a wall to “save material”. The material saved is trivial. The change may make the part harder to hold or introduce a distortion risk, which costs more.

Adding a fifth face to be machined “while you are at it”. Every added face is a setup, and setups are the expensive thing.

Tightening a tolerance “to be safe”. A tighter tolerance that the function does not need buys nothing and costs on every part in the batch. This is one of the most common and most expensive habits in design.

Asking for a price reduction without changing anything. Where a part is on the flat part of the curve, the price is material and time, and there is nothing left to negotiate. The productive conversation is about the part, not the price.

The same part, four versions

It is easier to see the levers at work by walking one part through a series of changes. Take a machined housing, produced in a moderate quantity, and change one thing at a time.

Version 1 — as originally drawn. Every face machined, tight tolerances across the drawing, a fine cosmetic finish, features on three faces. Four setups, a slow material, and a finish applied over the whole visible area.

Version 2 — the same part with the tolerance review done. The tight tolerances stay on the bore, the sealing face and the mating faces. Everything else reverts to the general block. The part is not functionally different, but every feature now cuts at a comfortable tolerance rather than at the edge of what the machine holds, and the finishing passes on the non-critical surfaces can be run faster.

Version 3 — the setup review done as well. Two of the faces never needed machining; one is inside the assembly and one is cosmetic on a surface that is never seen. Both are left as-supplied. Two setups disappear, along with their fixturing, their handling and the positional error they carried. The tolerances between the remaining features also improve, because fewer of them are now cut in different orientations.

Version 4 — the finish review done. The cosmetic finish is limited to the visible face rather than the whole part, and the internal surfaces ship as-machined. The finishing line falls, and so does the handling around it.

None of those four versions is a different part. They have the same function, the same material and the same fit. What changed was how much work the shop was asked to do that the function did not require — and the price falls at every step, with the largest single step being the removal of the two setups.

That progression is worth keeping in mind the next time a quotation looks high. The question that produces most of the saving is not “can you do it cheaper” but “which of the things I have asked for does the part actually need”.

A version five, worth considering where the quantity is real, is the quantity itself — moving from the steep part of the cost curve to the flat part. That is a purchasing decision rather than a design one, and it is covered on the bulk order pricing page.

Why a simple part can be expensive

The last thing worth understanding is that cost follows difficulty, not apparent complexity. This table is the quickest way to check a part against that idea.

Makes a part expensive	Makes a part cheap
A wall thin enough to deflect under the cutter	Every feature reachable from one direction
A pocket deep relative to its width, needing a slender tool fed slowly	Generous internal radii
A tight tolerance on a small feature, or far from the datum	A soft, free-cutting material
A hard or abrasive material	A general tolerance block with only a few tight dimensions
A requirement to machine every face	An as-supplied finish on the non-functional faces
A fine surface finish over a large area	Stock chosen close to the finished shape
Features that must be machined from several directions	Few distinct features, each easy to reach

A part can look complicated and be quick, and a part can look trivial and be slow. That is why a quotation built from a shape rather than a drawing is a guess, and why the drawing details that move the price most are often the ones the drawing omits.

How to get a quotation that means something

The fastest route to a firm price, and the one that produces a price that does not change afterwards, is a drawing or a 3D model together with four statements:

• The quantity, and the annual volume if the part repeats
• The material grade and the stock condition
• Which tolerances are functional and which features they apply to
• The surface finish, and whether it is functional or cosmetic

Those four cover most of the cost drivers listed above. Where the grade or the tolerance is genuinely open, saying so lets the quote come back with an alternative rather than a single number, and where a change would save money without affecting function, that note comes back inside the DFM feedback. Send the drawing through the request a quote form and the reply arrives within 24 hours. Orders start at 50 pieces per style and material, samples from a single piece in 5–10 days, and every batch ships with a material certificate (MTR) tracing the grade to its mill heat along with pre-shipment photos.