Setup is the one cost on a shop quote that behaves completely differently from everything else, and most estimators never treat it that way. Cutting time scales with quantity; material scales with quantity; setup does not. It is a fixed lump of machinist hours spent before the first good part exists — programming the job, building the fixture, loading tools, touching off, cutting a first article — and then it is spread across whatever batch you run. That single property is why setup time reduction changes your per-part cost and your quote more than almost any other lever you can pull, especially on the small and mid-size batches a job shop actually lives on.

This article is about the quoting consequences of setup, not just the lean-manufacturing theory. We will put numbers on how setup amortizes, where the hours actually go, which reductions are real, and — the part most guides skip — what to do with the savings once you have them: pass them to the customer to win the job, or keep them as margin.

Why setup is the line item that hides in plain sight

On a one-off or a batch of five, setup is frequently the largest single block of cost on the quote, larger than the cutting time it precedes. Yet it is the block estimators are most likely to guess at, because it does not come off a feeds-and-speeds calculation. It comes off experience with this machine, this fixture, this material — and when that experience is wrong, the whole quote is wrong before a chip is cut.

The deeper problem is that setup hours are quantity-independent, so they distort your unit price in a way that is invisible if you only look at the total. Quote €600 of setup on a batch of 10 and each part carries €60. Quote the same €600 on a batch of 100 and each part carries €6. Nothing about the part changed; the only thing that moved is the divisor. If you do not show setup as its own line, you cannot defend a quantity break, and you cannot explain to a customer why ten pieces cost what they do.

How setup amortizes into per-part cost

The arithmetic is simple and worth committing to muscle memory:

Per-part cost = (Setup cost ÷ Quantity) + Cycle cost per part + Material per part

Where setup cost is the setup time multiplied by the fully burdened machine-plus-labor rate you built per machine — the same rate the machine shop hourly rate calculation walks through. The table below shows what a single 3-hour setup at €80/hour (€240) does to a part with €12 of cycle and €8 of material as the batch grows:

Quantity	Setup per part	Cycle + material	Total per part
1	€240.00	€20.00	€260.00
10	€24.00	€20.00	€44.00
25	€9.60	€20.00	€29.60
100	€2.40	€20.00	€22.40
500	€0.48	€20.00	€20.48

The curve is steep at the start and flat at the end. Below roughly 25 pieces, setup dominates the price and any reduction you make flows almost directly to per-part cost. Above a few hundred, setup is a rounding error and cutting time is what matters. This is also why a clean split between setup and run time is the foundation of a defensible quantity break — covered from the build-up side in how to price CNC machined parts.

Where setup time actually goes

You cannot reduce what you have not measured. Before chasing tactics, break a typical setup into its components and time them honestly for a week. Most shop setups decompose into the same buckets:

• Programming and proving — writing or editing the CNC program, running it in single block, cutting and inspecting the first article.
• Fixturing and work-holding — finding, mounting, and dialing in vises, soft jaws, chucks, or a custom fixture.
• Tooling — pulling tools from the crib, assembling holders, presetting lengths, loading the carousel or turret.
• Touch-off and offsets — establishing work coordinates, setting tool length and diameter offsets.
• Material staging and first-article inspection — getting stock to the machine, cutting one, measuring it against the print.

A common pattern in small shops: half the setup clock is spent on tooling and fixturing that could have happened off the machine while the previous job was still running. That idle-machine time is the single richest target, and it is the core idea behind SMED — separating work that must happen with the spindle stopped from work that does not.

Setup time reduction tactics that change your quote

Not every reduction is worth chasing, and some cost more in fixturing than they save. The tactics that reliably move per-part cost in a job-shop setting:

1. Externalize what you can. Preset tool lengths offline, kit tooling and fixtures for the next job before the current one finishes, and stage material at the machine. Work done while the spindle still turns is setup time that effectively disappears.
2. Standardize work-holding. A repeatable vise/soft-jaw or quick-change pallet system turns "dial it in" into "drop it on the stop." The fixture costs money once; it saves setup on every future run of the family.
3. Build part families and group similar jobs. Running geometrically similar parts back to back means you tweak a proven program and reuse fixturing instead of starting cold. This is the planning logic behind work plans for multi-step operations.
4. Save and version programs and setup sheets. A documented setup that ran clean last quarter is a 20-minute repeat, not a 3-hour rebuild. The most expensive setup is the one you do twice because no one wrote down the first.
5. Reduce first-article churn. Tight, ambiguous, or over-specified tolerances stretch proving and inspection. Knowing which callouts actually drive cost — see GD&T and tolerances for estimators — lets you set up to the real requirement, not the worst-case reading.

After you cut setup: pass it on or pocket it?

Here is the strategic question almost no quoting guide answers. Once a recurring job's setup drops from 3 hours to 1 hour, the €160 you no longer spend is yours to allocate, and the right move depends on the part's place in the quantity table above.

On small batches, where setup dominates, passing the saving through can be the difference between winning and losing the RFQ — a 30% lower per-part price on a batch of 10 is dramatic and visible to the buyer. On large batches, where setup is already pennies per part, passing it through buys you almost no competitive advantage and simply gives away margin, so keep it. The discipline is to decide deliberately, per quote, rather than letting an outdated standard-setup figure quietly inflate every price you send. Treat the saving as a margin decision, framed by the difference between margin and markup.

There is a lead-time dimension too: less setup means the job ties up the machine for fewer hours, which frees capacity and can shorten the dates you promise. That feeds directly into manufacturing lead time estimation.

A worked example: from 4 hours to 90 minutes

Take a recurring bracket: a 4-hour setup at €80/hour (€320), €9 cycle, €6 material, quoted in batches of 20.

• Before: setup per part = €320 ÷ 20 = €16. Per-part cost = €16 + €9 + €6 = €31.
• After externalizing tool presetting, adding soft jaws, and reusing a saved program, setup falls to 90 minutes (€120): setup per part = €120 ÷ 20 = €6. Per-part cost = €6 + €9 + €6 = €21.

That is a €10, or roughly 32%, drop in per-part cost on a batch of 20 — entirely from setup, with no change to the cutting itself. Win the next ten reorders at the lower price and the soft-jaw fixture has paid for itself many times over. Quote the old number out of habit and you either leave money on the table or lose the job to a shop that did the math.

From a drawing to a setup-aware quote, automatically

The reason setup gets guessed at is that estimating it properly means reading the whole drawing — counting the operations, the work-holding the geometry forces, the tolerances that will stretch first-article proving, the number of distinct setups a part actually needs — and then deciding how to amortize all of it across the batch. Do that by hand on every RFQ and setup is exactly the line that gets rushed.

QuoteBuddy reads the technical drawing and surfaces the features that drive setup — the operations implied by the geometry, the distinct fixturing the part requires, the tolerance callouts that force extra proving — then builds the quote with setup amortized over the batch the same disciplined way every time, so a 10-off and a 200-off are priced from the same honest model instead of the same tired guess.

Start a 30-day trial and run a few of your real recurring jobs through it. See whether the setup-aware per-part price it builds matches the number your best estimator would reach after an unhurried look at the print — on every quote, not just the ones with time to spare.