Custom Manufacturing

Choosing an Automotive Stamping Supplier: Four Scenarios (and the Mistakes That Taught Me to Tell Them Apart)

There's no single best answer for automotive stamping sourcing. A process engineer with nine years of stamping and CNC part orders breaks it into four scenarios — prototype, bridge volume, mass production, and tooling transfer — with the real costs of getting each one wrong.

Maja LindqvistOctober 29, 20265 min read6 views

There Is No Single Best Supplier. There Are Four Situations.

People ask me some version of 'who should I use for automotive stamping?' almost every week. I've stopped giving one answer, because one answer is always wrong for somebody.

Here's my background so you can weigh the rest of this: I've been handling orders for automotive stamping dies, stamped parts, and machined components for nine years. In that time I've personally made, checked, and approved eleven mistakes worth roughly $38,000. Four of them came from the same wrong assumption — that the right process depends on the part. It doesn't. It depends on where the part is in its life.

Before the scenarios, one thing that isn't scenario-dependent: the phrase automotive die covers two different industries, and knowing which one you're buying from saves weeks. A stamping die cuts and forms sheet metal. An automotive mold — injection or die-cast — makes interior trim, lamp housings, bumper fascia. If you search for automotive mold manufacturers while you actually need stamped brackets, you'll spend a week on calls before anyone admits they can't help. In 2020 I did exactly that. Three weeks, two sample payments, zero parts. The shop I needed was a die builder, not an automotive mold maker, and nobody told me because I never asked.

Scenario A: Prototype or low volume (1–500 pieces)

You are here if: the part is still changing, or you need parts to test fit, or you're building a demo vehicle.

Don't buy a die. I know that sounds obvious, and I still got it wrong.

In 2017 I ordered a soft-tooled run — low-grade steel, single station — for 300 mounting brackets on a test build. The tooling quote was $1,900 and it made parts in two weeks. Looked like clever purchasing. Then the brackets wouldn't seat on the frame. The shop had formed them with no springback compensation, because the spec said 1.5 mm mild steel and what I received was 1.5 mm from a different heat. Two hundred and forty of them went into a scrap bin. About $2,100 gone, plus three weeks (ugh).

What I do now for anything under 500 pieces:

  • Laser cut and brake formed blanks for flat or single-curve stamping car parts, when the geometry allows it
  • CNC machining from billet when the part is a bracket, a boss, or a one-off test housing — cnc automotive work is often cheaper than a die at this quantity, which surprises people
  • Printed check fixtures to verify fit before anyone cuts metal

That's the counterintuitive part of this scenario. At low volume, machining beats stamping on total cost even though the unit price looks worse on paper. People searching for cnc cars or cnc car parts are usually looking for exactly this — a machining route through a problem that a stamping shop can't economically solve at 50 pieces.

The lesson I wrote on the wall: the prototype stage validates geometry and assembly, not the production process. Stop trying to make it do both.

Scenario B: Bridge volume (about 500–20,000 per year)

This is the awkward middle, and it's where most of my money went.

You are here if: volume is real but not committed — a program might scale and might not, and nobody has signed a multi-year PO.

You have three routes: a soft tool with single-station dies, a small automotive progressive die if the part justifies it, or staying on CNC and eating the unit cost. Most people pick soft tooling. So did I, in 2019, on a 12,000-piece order.

The tool was aluminum. It ran fine through the first 4,000 strokes. By piece 4,200 the cutting edge had worn enough that burr height went past the print, and the customer's line jammed a locating pin onto a burr. Four thousand pieces needed hand deburring at $4,300 total, plus two weeks of schedule.

The quote wasn't dishonest. I asked for tooling for 12,000 pieces and I got tooling for 12,000 pieces. What I never asked was: what is the design life, at what stroke count does the edge need rework, and what happens at 60% of the order if the drawing changes.

The rule I use now — match tool life to this order plus one more. If you genuinely expect a second and third run, a hardened steel auto stamping die with two or three stations pays back faster than the cheaper aluminum set, and you stop babying the press. If this is truly a one-time run, take the soft tool and put the deburr into the quote where you can see it.

Scenario C: Mass production (100k+ per year)

You are here if: the part is released, the volume is contracted, and the drawing has a longer revision history than your email thread.

At this level automotive stamping means a real automotive progressive die: hardened tool steel, 4 to 8 stations, coil feed, in-die sensing, and a press whose tonnage and bed were chosen around the die rather than the other way round. Automation is standard automotive industry stamping here — you aren't buying parts, you're buying process capability.

The failure usually isn't the die. It's the handoff. In 2023 we approved a supplier's first article on their lab press, skipped the run-at-rate study, and shipped. Parts that were in tolerance on a 25-ton press were out of tolerance in the customer's production setup, because the die had been shimmed for the lab press's ram deflection and not theirs. Sorting and air freight came to about $6,800. The credibility cost more than that.

If you're evaluating an automotive stamping plant for this scenario, ask four things before you ask for price:

  • Is the quality system IATF 16949, and is that certificate for this site or a sister plant?
  • Which press will run my die, and what are its tonnage and bolster dimensions?
  • What PPAP level, and does the run-at-rate happen on the production press?
  • What is your high-strength steel and aluminum experience?

That last one is where I'd spend my own interview time. A shop that's excellent at mild steel brackets is not automatically excellent at 780 MPa rails, and springback on AHSS and 6000-series aluminum is a different discipline.

Scenario D: You already own the tooling

This is the one where my instinct was most wrong, so I'll be blunt: moving existing tooling to a new automotive stamping source is usually harder than building new tooling, and it is almost never the fast option.

In 2021 we moved a set of dies from one automotive stamping plant to a cheaper one, on the theory that the dies already existed and setup was just freight. I assumed 'same drawing, same part.' Didn't verify. The new plant's press had a different shut height. Their bolster slots didn't line up with our clamp pattern. The nitrogen springs in the lower shoe were eight years old and had lost their preload. The die ran — badly — after nine days of shimming, and the ramp-up cost $11,000 plus a customer escalation.

Dies aren't standard parts. They're matched to a press, a feed line, and a material. Before you move one, collect the die's own drawings, the original press tonnage range, shut height, bolster dimensions, the last approved first-article report, and the last 20 parts produced. If a supplier can't review that package before quoting, the quote is fiction.

How to tell which scenario you're actually in

Four questions, in this order:

  1. What is the annual volume, not the first order? Be honest — a first order of 5,000 that never repeats is Scenario A wearing a Scenario B costume.
  2. Will the drawing change in the next twelve months? If yes, cap your tooling spend and accept the higher unit price on purpose.
  3. Do you already own tooling? Then you're in Scenario D, and the comparison isn't die cost versus die cost. It's transfer cost versus new tool cost.
  4. What's the material and how complex is the form? Flat or single-curve aluminum profiles are often better as automotive aluminum extrusions than as stampings. High-stress three-dimensional parts with a grain-flow requirement are automotive forging parts, not stampings, and no amount of die work changes that. Deep draws in steel are where stamping genuinely wins.

I don't have hard data on how often people pick the wrong scenario. I wish I'd tracked it, because my guess is most of the time, but my sample is only my own eleven mistakes. What I can say anecdotally is that in nine years, not one of my expensive failures was a bad supplier. They were all a correct supplier answering the wrong question.

What's changed since I started

What was best practice in 2020 doesn't fully apply now, and two shifts matter.

First, high-strength steel and aluminum are no longer exotic in stamping in car manufacturing. They're the default on anything structural or weight-sensitive. Springback compensation used to be a specialty; now it's table stakes, and a die shop that hasn't invested in forming simulation is a die shop that will iterate on your dime.

Second, small-batch demand has changed the economics. Ten years ago a 2,000-piece run went to a die because there wasn't a credible alternative. Now cnc machine car parts from billet can land close enough on unit cost that a die stops making sense. The fundamentals haven't changed — buy the process that matches the volume and the risk — but the execution has.

The one thing that doesn't depend on your scenario

Verify before anyone cuts metal. Every expensive mistake in this article traces back to a sentence I accepted without a follow-up: tool life, press tonnage, spring preload, run-at-rate.

If you typed stamping automotive or cnc automotive into a search box this week, you're probably one question away from knowing which of the four scenarios you're in. Ask it before the PO goes out.

Price figures here come from quotes and invoices between 2017 and 2024 and are directional only. Die and machining costs vary widely with station count, material, and tooling design — verify current pricing against your own quotes before budgeting.