The 7-Point Checklist I Run Before Approving Any Metal Component Order (After $47K in Mistakes)

I've been handling metal component orders for automotive programs since 2016. In that time, I've personally made — and documented — enough sourcing mistakes to fill a small binder. Rework costs, scrapped lots, air freight I didn't budget for: roughly $47,000 in avoidable losses spread across eight years.

This checklist is what came out of that. Seven points. I run it before releasing any PO for stamping, CNC machining, forging, or die work. It's built for OEM buyers and Tier 1 sourcing people who order metal-formed parts and don't want to learn these lessons the way I did.

Feel free to steal it. That's the point.

Step 1: Verify the Drawing Revision Number — Before Anything Else

This one sounds insultingly basic. It has also cost me more money than every other mistake combined.

In my first year (2017), I released a stamping order against Rev C of a bracket drawing. Rev D had been sitting in our shared drive for two weeks — changed hole positions, different bend radius. Nobody flagged it. The result came back on the dock three weeks later: 1,200 pieces, $4,600 in tooling and material, straight to the scrap bin. That's when I learned to physically open the drawing from the release folder and read the revision block out loud. Every single time.

"Plus — check the revision date against the program's engineering change log. Rev letters lie. Dates don't."

The checklist item: Drawing number + revision letter + revision date, confirmed against the current ECO log. Takes ninety seconds. Saves four figures.

Step 2: Match Every Specification to the Actual Production Process

Here's the mistake almost nobody talks about: the tolerance on your drawing has to be achievable by the process you're buying. If it isn't, your supplier will either (a) quote it and quietly ship out-of-spec parts, or (b) pad the price to cover sorting. Either way, you lose.

If I remember correctly, my worst version of this was a 2019 order for an intercooler end tank stamping — a vertical flow intercooler assembly for a retrofit program. Drawing called for a flatness tolerance that's fine on a CNC-machined surface but borderline impossible on a deep-drawn stamping. The supplier said "no problem." It was, in fact, a problem. We sorted the lot by hand at our own cost.

Same logic applies across part types. An alternator belt and a timing belt look like cousins; they're engineered completely differently, and confusing the tolerance philosophies between them is the kind of thing that gets a whole BOM questioned. Different process, different rules. Ask your supplier to confirm in writing that the process they're quoting can hold the spec.

Step 3: Ask About Material Substitution Policy in Writing

We didn't have a formal material substitution approval process. Cost us when a batch of mounting hubs — a hub assembly similar to the wheel master mt-5000 pattern we'd sourced before — came in with a steel grade one step softer than specified. The supplier had "switched sources." Form, fit, and function? Fine, mostly. Fatigue life? Nobody could tell me.

Now every PO includes one line: No material substitution without written approval from [engineering contact], including grade, mill certificate, and heat number. Not a suggestion. A release condition.

Ask for the mill cert before the parts ship, not after they arrive. You'd be surprised how often it doesn't exist.

Step 4: Map the Secondary Operations (This Is the Step People Skip)

This is the counterintuitive one. Most buyers focus on the primary process — stamping, machining, forging — and treat secondary operations as an afterthought. That's backwards. The secondary steps are where the hidden costs live.

Deburring. Passivation. Anodizing. E-coat. Heat treat. Tumble finish. Every one of these can fail independently, and every one adds lead time you probably didn't model. The most frustrating part of sourcing: you can nail the primary process perfectly and still lose the batch at deburring. You'd think the part is done once it's formed, but it isn't done until it's finished.

Write out every operation the part touches, in sequence, and confirm which supplier owns each one. If your stamping house is outsourcing heat treat to a third party they won't name, that's a risk you need to price in.

Step 5: Confirm Tooling Ownership and Amortization Terms

After the third time I discovered tooling had been amortized into piece price without my knowledge, I was ready to read every quote line-by-line with a highlighter. What finally helped was a one-page tooling schedule attached to every PO.

Three questions, always: Who owns the die or mold? Is tooling paid upfront or amortized? If amortized, at what piece count does it become ours? Get it in writing, because "we always do it this way" is not a contract term.

The difference here can be way bigger than people expect. A progressive die for a mid-complexity stamping can run $25K–$80K depending on stations and tolerances. That number shouldn't be a surprise in month fourteen of a program.

Step 6: Run a First Article Inspection — Not a Sample

Granted, this requires more upfront work. But it saves time later, every single time.

A first article inspection (FAI) means measuring the first piece off the tool against every dimension on the drawing — full balloon report, not three critical dims. If your supplier resists, that's information. Most good shops expect it and build it into the quote.

One more thing worth mentioning: ask for an FAI off the production tool, not a prototype tool. I've watched a program pass FAI beautifully on a bridge tool and then fall apart on the production die. Two different animals. Verify which one you're inspecting.

Step 7: Ask What Your Supplier Doesn't Do Well

This is the most useful question I ever started asking, and it feels like the wrong question right up until the first time it saves you.

I'd rather work with a specialist who knows their limits than a generalist who overpromises. Valéo's public materials, for example, describe their focus around stamping, tooling, and CNC work — not everything with a bolt hole in it. That kind of boundary is a good sign, not a weakness. A vendor who says "this isn't our strength" on one line item has just told you they're being straight about the other thirty.

That said, be fair about this: "what don't you do well" is not the same as "what can't you do." A shop can be excellent at deep-draw stamping and merely competent at tight-tolerance finishing. The answer tells you where to double up your incoming inspection, not where to walk away.

Common Mistakes That Sink Otherwise Good Orders

• Approving the quote instead of the spec package. The quote is a price. The spec package is the contract. Review both, separately.

• Trusting lead time quotes without buffer. Standard delivery windows are usually aspirational. Build in three to five business days on anything deadline-critical.

• Skipping the packaging spec. I've had parts arrive cosmetically ruined because the packing method was never agreed on. Finished parts in a cardboard box with no separators = rework.

• Not naming a single point of contact on both sides. The "ask anyone" approach almost never works. Name one person on your side, one on theirs, put both on the PO.

• Assuming the supplier's QC is your QC. It isn't. Their inspection protects them. Yours protects you. Budget for incoming inspection on the first three lots of any new part.

The best part of finally getting this list systematized: no more 3am worry sessions about whether a batch will land correct. The checklist doesn't eliminate mistakes — but it catches almost all of mine before they become expensive. Take it from someone who paid tuition on all seven points.