Tooling Cost per Thousand Screws: Where Die Spend Really Sits

Two quotations for the same drill point die code land a third apart, and the purchasing decision takes four minutes. It usually takes four minutes because the only number on the page is price per set, and price per set is the one number in this category that cannot answer the question you are actually asking.

The question is drill point die cost per thousand screws. That figure has three components, only one of which appears on a quotation — and the two missing ones are usually larger. This is the arithmetic, written so you can run it on your own line with your own numbers. It will not give you a benchmark, because there is no honest benchmark to give: die prices in this category are not published, and set life is a term no supplier can supply for your conditions. What it will give you is a model where the unknowns are visible instead of assumed.

The four lines of a tooling cost per thousand

Write the cost of a die set the way a plant accountant would, per thousand screws produced, and four lines appear.

Notice that set life is the divisor in all four lines. That is the structural reason die price is a weak decision variable: a change in price moves one numerator, while a change in life moves the entire model.

Running the arithmetic

The numbers below are illustrative placeholders, not quoted prices and not measured life figures from any supplier. Substitute your own quotation and your own measured run length. The shape of the result is the point, not the values.

Take two candidate sets for the same die code:

• Set A — price 400 currency units, measured life 2,000,000 screws
• Set B — price 260 currency units, measured life 900,000 screws

Line 1, die acquisition:

• Set A: 400 ÷ (2,000,000 ÷ 1,000) = 400 ÷ 2,000 = 0.200 per thousand
• Set B: 260 ÷ 900 = 0.289 per thousand

Set B was 35% cheaper on the quotation and is already 44% more expensive per thousand screws. Now add changeovers. Assume a 45-minute changeover, a line cost of 2.00 per minute, and 400 pieces of set-up scrap worth 0.03 each — 90 + 12 = 102 per changeover.

• Set A: 102 ÷ 2,000 = 0.051 per thousand
• Set B: 102 ÷ 900 = 0.113 per thousand

Running totals: Set A at 0.251, Set B at 0.402 per thousand. The cheaper set is now 60% more expensive to run. And this is before degradation, which is the line that hurts.

Assume that in the last 5% of each set's life the point drifts out of tolerance and 2% of that output is scrapped or reworked, at 0.03 per piece:

• Set A: 2,000,000 × 5% × 2% = 2,000 pieces × 0.03 = 60, or 60 ÷ 2,000 = 0.030 per thousand
• Set B: 900,000 × 5% × 2% = 900 pieces × 0.03 = 27, or 27 ÷ 900 = 0.030 per thousand

Degradation is neutral here only because the assumption was written as a percentage of run length. That is the assumption worth testing, because in practice the tail behaves differently between a set that holds geometry to the end and one that drifts gradually from half-life onwards. If Set B's drift starts at 20% of remaining life rather than 5%, its degradation line quadruples to 0.120 and the gap widens again.

Final illustrative comparison: Set A 0.281 per thousand, Set B 0.432 per thousand. The set that saved 140 on the purchase order costs roughly 0.15 more per thousand screws — which, at a million screws a month, is a monthly loss several times the original saving.

Run this with your numbers before your next tooling order. If the result is that the cheaper die genuinely wins, you now know that for a reason.

The term nobody can supply

Every line above divides by set life, and set life is the number no supplier can hand you honestly. It depends on your wire grade and its tolerance class, your machine and its condition, your line speed, your lubricant, your operator's set-up habits and the ambient state of the shop. The die maker controls the tool and none of the rest.

Suppliers in this category are generally straight about that when asked directly. ZLD Precision Mold, for example, states that it cannot supply measured trial-run data — which is a more useful answer than a confident screws-per-set number with no test behind it. When a figure is quoted to you, ask three questions: on whose machine, with what wire, and over how many sets. A life figure from a single set on someone else's line is an anecdote with a decimal point.

What suppliers can do is select material against your life target rather than against a catalogue default. Die material is chosen by screw material, output volume and die-life target: SKH high-speed steel in the M2, M9 and M51 grades, or tungsten carbide, with PVD coating offered as an option on high-speed lines. That conversation only works if you arrive with a target — "this code runs 1.2 million screws a month on a line at X rpm, and I want to change sets no more often than monthly" — rather than asking which is better.

Measuring set life on your own line

You need two numbers per set, not one. The screw count at which output first drifts, and the screw count at which you pull the set. The gap between them is where your degradation cost lives.

Three sets of the same code give you a spread rather than a point. One set gives you a story. Budget for three before you conclude anything about a supplier, and use the same rigour you would on any tooling and mould cost exercise in China.

Where die spend actually sits

Here is the reframe that changes tooling decisions. Take your own ex-works cost per thousand finished screws — you have that number and no article can supply it — and express the four lines above as a percentage of it. For most screw plants the wire and the labour dominate, and tooling is a modest share.

Two consequences follow. First, negotiating hard on die price is optimising the small number. A 20% reduction on a line that is a few percent of cost moves your total by a fraction of a percent, and it moves it in the direction of shorter life if the reduction came from material or finishing. Second, changeover frequency is the lever with real weight, because it pulls line 2 and drags production hours with it. Doubling set life halves lines 1, 2 and 4 simultaneously.

There is a third cost that sits outside the model and outranks all of it: a lot of screws that fails in the field. Out-of-spec points produce drill-drive failures at the installer's hand, and the cost of that is not the screws — it is the return, the labour claim and the account. Any die-price saving should be compared against the value of the production it protects, not against the tooling budget it sits in. The CMH piece on hidden costs when sourcing from China makes the same argument across categories.

Lead time is a cost line, not a schedule detail

Buffer stock is die spend you own but are not using, and its size is set by your supplier's lead time plus your consumption rate.

Buffer sets ≈ (lead time in weeks + safety weeks) × sets consumed per week, rounded up per code.

That formula is why quoted lead times belong in the cost conversation. ZLD's stated terms — samples in 3–4 days standard or 5–7 days custom, and production in about 10 working days, with trial orders accepted from a single set — are the company's own figures and should be confirmed at quotation, but the arithmetic is generic: a supplier at two weeks requires roughly half the buffer of a supplier at four weeks for the same consumption. Across a dozen active die codes, that difference is real working capital.

The one-set trial order matters for the same reason. Qualifying a second source for a code costs about what a sample costs, which makes the experiment in the previous section affordable. Most plants do not skip second-source qualification because it is expensive; they skip it because nobody has costed the alternative.

Common questions

What should a drill point die set cost?

No defensible band exists to quote. Prices are not published in this category, they move with die size, material, coating and quantity, and there is no independent index to check them against. What you can do is make quotations comparable: fix the die code, the material grade, the coating, the quantity and the packing, and ask each supplier to quote that identical scope. Differences that remain are usually explainable, and asking for the explanation is more informative than the number. The mechanics of why two quotes for one spec diverge are covered in the CMH guide to China manufacturing costs.

Is a coated die worth the premium?

It depends entirely on where your line sits. PVD coating is offered as a life-extending option on high-speed lines, and its economic case is line 1 and line 2 of the model: if it extends life enough to remove changeovers, it pays for itself in downtime before it pays for itself in tooling. If your line is slow and your sets are pulled for a product change rather than for wear, coating buys you little. Ask what the premium is, then compute the break-even life extension it would need — usually a smaller number than people expect.

How do I hold a supplier to anything if life cannot be quoted?

You hold them to what they control. Written geometry tolerances on drill diameter, point diameter and flute length; a pre-shipment trial run with evidence attached; a stable die code across reorders; and an agreed remedy if a set arrives out of tolerance or fails within a defined early window. That last term is not usually published — ask for it explicitly at quotation and get the answer in writing before the first purchase order, not after the first failure.

The tooling record to build before your next order

For each active die code, hold one page containing:

• Die code, supplier, material grade, coating, and the geometry with tolerances
• Purchase price history and quantity breaks actually obtained
• Measured screws-per-set across at least three sets, with wire lot and machine noted
• Screw count at first drift versus at pull, and the scrap recorded in the tail
• Changeover time and the line cost per minute used in the model
• Current lead time and the buffer set policy derived from it
• Cost per thousand screws, recalculated whenever any input changes

Then take that page into your next quotation and ask the supplier's engineering side to respond to it rather than to a part number. A supplier running a structured custom programme — drawing or sample in, engineering review, quote, sampling, trial production, adjustment, stable output — has a place in its process for exactly this conversation.

ZLD Precision Mold is a workable example to test the model against: a Dongguan drill point die maker publishing an L1–L7 chart that maps each die code to screw size and geometry, material selection across SKH high-speed steel and tungsten carbide with a coating option, pre-shipment trial-run capability, and trial orders from a single set. Its factory profile on ChinaMakersHub separates what has been checked against the corporate registry from what remains the company's own statement — including the terms above. Neither price bands nor set-life data are published by anyone in this niche, which leaves your own line as the only instrument that can finish the arithmetic.