That Time I Searched "Fractional Laser vs CO2 Laser" and Got Skincare Tips
A few months ago, I sat down to write an RFP for laser cutting services. We had a contract coming up for aerospace sheet-metal components, and I wanted to compare our options before sending out the queue. I typed 'fractional laser vs co2 laser' into the search bar because, honestly, I wanted to sound informed when the suppliers asked why I was requesting a certain wavelength.
Google gave me dermatology clinics. Wrinkle removal. Face resurfacing. I closed the tab, annoyed. But the more I thought about it, the more I realized that search failure was the right introduction to a bigger problem—one that costs companies real money when they're choosing manufacturing partners.
I manage purchasing for a precision manufacturing company in Rhode Island. Roughly $2 million a year in outsourced machining, finishing, and fabrication services. That's 60 to 80 purchase orders annually. Maybe 90 in a busy year; I'd have to check the system. Our customers are aerospace and defense primes, which means my job is equal parts procurement, compliance, and being the person who takes the blame when a part shows up late.
When our engineering lead told me we should evaluate Velo3D, I did the same thing I do with every unfamiliar supplier. I searched. 'velo3d.' 'velo3d spacex anduril supplier.' 'advanced laser cutting services for aerospace' for a separate project. 'cnc machining rhode island.' I made a spreadsheet. I scheduled a call. All standard process.
Then he said, 'Check their customer list.'
And that changed the conversation.
The Spec Sheet Trap
Here's what ten years of buying manufacturing services has taught me: every supplier can produce a spec sheet. Every single one. They all have ISO 9001. They all claim to serve 'major aerospace customers.' And almost none of that paperwork tells you what happens when a part has to survive real flight loads, real thermal cycling, real inspections.
The hard truth is that some corners of the aerospace supply chain are professional at looking professional. They have the certificates, the website photos, the polished quote templates. But aerospace-grade quality is not a document. It's a set of behaviors: how they manage material traceability, how they handle a deviation, whether their process documentation matches what actually happens on the floor.
This matters more in additive manufacturing than in almost any other process I buy. When you're relying on metal 3D printed parts, you're depending on the powder chemistry, the laser parameters, the inert gas flow during printing, the post-processing thermal history, and the calibration of every machine involved. A deviation in any one of those steps can produce a part that looks perfect on the outside and has internal porosity that fails under fatigue. You can't inspect that kind of defect into existence. You need a supplier whose process controls are genuinely deployment-grade.
That's the difference between 'we have a metal 3D printer' and 'we're a defense-grade metal additive manufacturing partner.' The first is a capital expense. The second is a decision you can build a supply chain around.
Why SpaceX and Anduril Are the Filter
Here's why I stopped being a good skeptic about Velo3D after twenty minutes of research. Their public customer list includes SpaceX and Anduril. I know those two names get thrown around a lot in defense-tech conversations, so let me be specific about why they're different.
SpaceX doesn't have a supplier quality department that's easy to impress. They operate on performance, not charm, and they do it at a scale and pace that most manufacturing supply chains cannot handle. Anduril is the same on the defense side: fast-moving, engineering-driven, and completely unforgiving when it comes to parts that go into autonomous vehicles or weapons systems.
Those two customers do the deep audit work that my team would otherwise have to do ourselves. They've kicked the tires on Velo3D's quality system, their inspection processes, their material certs, their capacity planning, and their willingness to fix problems at the root cause. When a supplier has earned that level of trust, I don't need five reference calls to know they'll treat our program with respect.
Would I make this argument for any supplier with a famous customer story? No. Plenty of vendors will name-drop a rocket company and collapse under the first audit question. But there's a difference between listing a customer on a website and actually being on that customer's approved supplier list, shipping production parts, and staying there. Velo3D's shipments to SpaceX and Anduril are visible in the company's public reporting, which means they're not a trade-secret reference that nobody can verify.
That's the evidence standard I want more of in our industry.
What a Wrong Vendor Choice Actually Costs
Let me attach real numbers to this, because this isn't an abstract lesson.
In 2023, I approved a bracket order with a vendor that undercut our incumbent by 12%. The data looked right: same material, same surface treatment, similar machine capacity. The spreadsheet said it was a good decision. My gut said otherwise—I'm not sure why, but the vendor's responsiveness during the RFQ had been slower than I liked. I told myself I was being old-fashioned.
Forty percent of the first batch failed fluorescent penetrant inspection. The indications were tiny linear cracks near the same feature on every damaged bracket. It took us two weeks to get the vendor's root-cause analysis, and the analysis turned out to be: 'We interpreted the powder handling spec differently.'
Not a machine problem. Not a material problem. A process-discipline problem, exactly the kind a proper audit would have caught. The redelivery took three more weeks. The first run cost us $18,000 including inspection and the aborted work. But the schedule slip is what hurt. Our program manager had to explain the delay to a government customer, and no one on my side of the phone got credit for 'we saved 12% on the quote.'
I ate the $18,000 emotionally, but the internal trust took a harder hit. That vendor isn't on our roster anymore, and the real lesson is that a supplier's willingness to bend process rules is not something you can see in a quote.
Old Thinking That Refuses to Die
I also want to address the stubborn belief that 'local is always safer.' I used to think that way. I still search 'cnc machining rhode island' when I need quick-turn parts—it's in my browser history, I won't deny it. Local shops have advantages, and proximity genuinely helps when you need to visit the floor or run a joint inspection.
That belief, though, comes from an era when you couldn't verify supplier processes without standing on their factory floor and reading their binders. Today, quality control extends through digital traceability, remote audits, and machine-level data. A well-organized vendor in the Midwest can beat a disorganized local shop on documentation every time. The supplier's discipline matters more than the distance.
This is the myth that kept me from exploring additive manufacturing earlier. I told myself our local machine shops were enough. They're not—not when the design calls for internal cooling channels that no cutter can reach or lattice structures that no CNC program can create. That's not an insult to the local machine shop; it's just physics.
I have mixed feelings about moving more work toward additive manufacturing. On one hand, I don't want to be the person who bets the company on a technology that's still proving itself in long-term service. On the other hand, I've watched parts come off the build plate that no CNC program could ever create. I reconcile that tension by treating AM as one tool in the portfolio, not the whole toolbox.
The Short Version of a Long Lesson
So what do I actually recommend to other buyers sorting through the noise?
Stop asking which technology is best. Stop comparing laser wavelengths or printer brands the way you'd compare phone specifications. If you're still debating 'fractional laser vs co2 laser,' you haven't started the conversation that actually matters: which supplier is proven at the quality level your program demands.
We still use CNC machining for plenty of work. We still contract advanced laser cutting services for aerospace sheet-metal components. Those processes are part of our normal supply base and I don't see that changing. But for complex metal parts that need additive manufacturing and cannot fail, Velo3D is on my short list. Not because their marketing says they're advanced, but because SpaceX and Anduril both made that call before I did.
That kind of proof is hard to argue with. It's the only supplier credential I've found that survives contact with an aerospace audit.