Most shops cost the spindle time to the minute and then wave a hand at everything that happens after the part comes off the machine. Deburring, tapping, reaming, marking, inspection, kitting, light assembly — these get swept into a round "finishing" number or, worse, absorbed silently into overhead. That habit is exactly why the secondary operations cost is the line that quietly turns a profitable-looking job into a break-even one. The cuts are right, the material is right, and the margin still evaporates at the bench.

The problem is that secondary operations are labour, not machine time, and they scale with feature count and quantity in ways that have nothing to do with cycle time. A part that machines in four minutes can carry six minutes of hand deburring across thirty edges, two minutes of tapping, and a minute of inspection — and on a batch of 200 that bench work outweighs the spindle. This guide breaks secondary operations into line items you can actually estimate, with rates and time standards you can defend, instead of a fudge factor you hope covers it.

Why secondary operations get mispriced

Machining cost lives in your quoting habit because the machine meters it for you: spindle hours, a burdened rate, done. Secondary work has no meter. It happens at a bench, with a deburring tool, a tap wrench, a height gauge, a torque driver — and nobody logs the minutes. So it gets estimated by feel, and "by feel" almost always means underestimated, because the operations look trivial one at a time and only add up across the batch.

The second trap is that this work scales with the wrong variable. Machining scales with how long the cut takes; secondary operations scale with how many features and how many parts. Thirty edges to deburr is thirty edges whether the part took two minutes or twenty to cut. Eight tapped holes is eight tapping cycles per part, every part. When you price these as a percentage of machining cost, you overprice the long-cut parts that happen to be simple and underprice the fast-cut parts that are covered in features. The fix is the same discipline you already apply to setup and programming in how to price CNC machined parts: pull the operation out as its own line and cost the time directly.

Treat each operation as its own line item

A secondary operation is just a labour time multiplied by a labour rate, plus any consumable or outside charge. The estimate is honest when you build it from three numbers per operation:

1. Time per unit of work — minutes per part, or minutes per feature times the feature count.
2. Bench labour rate — your burdened shop-floor or finishing rate per hour, the same way you derive any rate in machine shop hourly rate calculation.
3. Setup or fixed cost — fixture build, first-article inspection, a tapping head change — spread across the batch quantity.

So a single operation costs (time_per_part × quantity × rate) + setup, divided back to a per-part figure for the quote. Do this once per operation and the "finishing" mystery dissolves into a short, defendable list. The same logic is what drives a proper multi-step routing — see work plans for quoting multi-step operations — where every step carries its own time and rate instead of hiding in a lump.

Deburring: the most underestimated minute in the shop

Deburring is the operation estimators feel safest skipping and lose the most on. Every machined edge, every cross-drilled hole intersection, every thread exit raises a burr, and somebody removes it by hand before the part ships or fails inspection. The time is real and it is rarely logged.

Estimate it by edge category rather than guessing a flat number per part. A simple broken edge is seconds; a cross-hole or a thin-wall intersection that needs a deburring blade and a loupe is far more. A rough working table:

Deburring task	Typical time per feature
Break a simple machined edge	5–15 s
Deburr a drilled/tapped hole, one side	10–20 s
Deburr a cross-hole intersection (burr inside bore)	30–90 s
Hand-blend a milled pocket edge to spec	1–3 min
Full hand polish / cosmetic blend	5–20 min

Count the features, multiply by the category time, and you get a deburring minute count you can stand behind. A bracket with twenty edges and four cross-holes is easily eight to ten minutes of bench time — at a €45/h finishing rate that is €6–€7.50 per part, every part, that a flat "two minutes for deburr" would have missed by 4x.

Tapping, reaming and other in-process secondaries

Some secondaries happen at the machine but are still costed separately because they add cycle time the roughing pass does not capture, or because they carry tooling risk. Tapping is the headline case: it is slow, it is where parts get scrapped, and a broken tap in a near-finished part is a total loss plus rework.

Cost tapping per hole, not per part. A rule-of-thumb cycle is a few seconds per hole for small CNC-tapped threads, rising fast with diameter, depth, and material. Add a risk allowance on tough materials or blind holes — a small per-hole scrap reserve is cheaper than discovering the loss after the fact. Reaming, counterboring, back-spotfacing, and thread-milling follow the same pattern: time per feature, times feature count, plus a tooling line if the job needs a dedicated cutter you will not reuse.

A short checklist for in-process secondaries:

• Count tapped, reamed, and bored features from the print, by size.
• Apply a per-feature cycle time, scaled by diameter and depth.
• Add a scrap/risk reserve on blind holes, tough alloys, and fine threads.
• Add a tooling line for any special cutter bought for this job alone.

Assembly, marking and inspection

Anything that joins, identifies, or verifies the part is secondary labour too, and it is increasingly common on jobs that used to ship as loose parts. Light assembly — pressing in dowels or threaded inserts, fitting bearings, installing helicoils, fastening sub-components — is pure bench time, and it carries a fixturing or first-piece setup cost that you spread across the batch. Where assembly shades into welding or fabricated sub-assemblies, the same time-and-rate logic in welding and fabrication cost estimation applies.

Marking and inspection are the two everyone forgets. Laser or dot-peen marking has a per-part cycle plus a one-time programming setup. Inspection is not free: a first-article on a tight-tolerance part can be 20–40 minutes of CMM or hand-gauging, and recurring in-process checks add seconds to minutes per part. On a print loaded with controlled dimensions, that inspection time is driven by the tolerance callouts — the link between tolerances and downstream cost is laid out in GD&T and tolerances for estimators.

Worked example: where the margin actually goes

Take a small aluminium bracket: 4-minute machining cycle, run in a batch of 150, machining rate €90/h, bench rate €45/h. The naive quote adds "10% for finishing." The line-item quote counts the work.

Cost element	Basis	Per part
Machining	4 min @ €90/h	€6.00
Deburring	9 min @ €45/h	€6.75
Tapping (6 holes)	1.2 min @ €45/h	€0.90
Insert press (2)	0.8 min @ €45/h	€0.60
Marking	0.3 min @ €45/h + setup/150	€0.30
Inspection (amortised FAI + in-process)	setup/150 + 0.4 min	€0.55
Secondary subtotal		€9.10

The "10% for finishing" approach would have added €0.60. The real secondary cost is €9.10 — more than the machining itself. Quote the part at machining + 10% and you are underwater before you account for material or margin. This is the single most common way an accurate-feeling quote loses money, and it is the kind of leak that disciplined line-item estimating prevents — the broader case for that discipline is in reducing quoting errors and improving accuracy.

Put secondaries on the quote as their own lines

Once you have costed each operation, show them. A buried "finishing %" tells the customer nothing and tells your future self nothing when you re-quote the part next year. A quote that lists deburring, tapping, assembly, marking, and inspection as separate lines is readable, defendable, and re-usable — which is the whole reason to keep a consistent manufacturing quote template. And because each line carries a markup decision, settle that deliberately the way you handle margin vs markup in manufacturing pricing rather than letting a flat percentage paper over it.

From drawing to a secondary-inclusive quote

The slow, error-prone part of all this is the reading: counting the edges that need deburring, the holes that need tapping, the features that need inspection, the inserts that need pressing — every one of them sitting on the drawing, and every one of them easy to undercount on a hurried scan before a deadline. That undercount is precisely where the secondary operations cost goes missing.

QuoteBuddy reads the technical drawing and surfaces the features that drive secondary work — edges, holes, threads, tolerances, finish and marking callouts — so the estimator counts from a complete picture instead of a fast glance. The cost engine then builds each secondary operation as its own line, by time, rate, and quantity, the same way every time, so two estimators quoting the same part land on the same number and nobody ships a job that is upside-down at the bench.

Start a 30-day trial and run a few real drawings through it — from upload to a complete, itemised quote. Compare the secondary lines it builds against what your "finishing percentage" would have charged, and see where the bench work was quietly eating the job.