The $180K Desktop Metal Experiment: What Our Procurement Team Learned About Hidden Costs

Why we went down this path

It started with a Reuters headline. That's not a joke. I was scrolling through additive manufacturing news today on Reuters back in January 2023, reading about binder jetting adoption rates, when our CEO walked by and said, 'We need to cut outside machining costs by 20% this year.'

I'm a procurement manager at a 48-person manufacturing company in Ohio. I've managed our outside services budget—roughly $180,000 a year—for six years. So I knew the target wasn't crazy. But I also knew the usual tricks: renegotiating with vendors, consolidating orders, delaying non-critical work. The CEO wanted structural change. That meant bringing some metal fabrication in-house.

So I started looking at desktop metal systems. Not the company name alone, but the category. We needed something that could print small metal parts without the footprint of a full-scale additive manufacturing cell. That's how I ended up with a spreadsheet comparing the Desktop Metal Production System P-50 against two binder jetting alternatives, plus a desktop metal CNC machine for post-machining, and—later—a desktop laser engraving machine for metal.

Here's where the story gets interesting.

The first vendor meeting (and the near-miss)

I'll call them Vendor A and Vendor B, though I should add that neither is a direct Desktop Metal competitor—they're resellers who bundle different tools.

Vendor A quoted us the P-50 with a spare parts kit, installation, training for two operators, and a year of remote support. All-in: $112,000. Vendor B quoted what sounded like the same package for $94,000. If I'd been my younger self, I would have signed with B that afternoon.

But I've been burned by hidden fees before. So I broke down the TCO line by line. That's when I had the comparison insight.

Vendor B's quote didn't include the build box replacement, the carrier gas line upgrade, or the software license for their additive manufacturing MES module. They also charged a $3,800 'onboarding fee' that was just a PDF and a two-hour Zoom call. By the time I added everything: $104,300. Still less than A, but the gap was narrowing. And then I asked about service response times. B's contract listed 96-hour response, while A guaranteed 24-hour. For a production system, that difference is huge. If a printer sits idle for three days, we lose about $1,400 in potential output—more if there's a rush order.

So I went with Vendor A. The final TCO difference over three years wasn't $18,000 in B's favor—it was about $2,200 in A's favor, once I factored in downtime risk.

That's when I learned the first lesson: the cheapest quote is just a starting point, not a decision.

The MES gap (my process failure)

Now, I'm not proud of what I'm about to tell you.

We took delivery of the P-50 in June 2023. We had a decent build prep workflow, but we didn't have a formal production tracking system. I'd validated the printer, but not the process around it.

The third time we had a build failure, I noticed a pattern: the failures happened when the same operator ran back-to-back shifts without logging the batch settings correctly. We had no additive manufacturing MES system to force a digital checklist. It cost us $2,400 in wasted powder and a week of scheduling delays before I finally wrote a requirement for a lightweight MES module. The vendor we bought the printer from offered one as an add-on—$9,000 including integration. It hurt to pay that after I'd already negotiated so hard on the hardware, but it now saves us at least that much every year in scrap reduction.

Lesson two: the machine is only as reliable as the data around it.

The laser engraving rabbit hole (and CO2 laser ablation)

Three months later, our quality team asked for a way to mark serial numbers on finished metal parts. They wanted a desktop laser engraving machine for metal. I started researching budget options, and every spec sheet kept throwing around the term 'CO2 laser ablation.'

What is CO2 laser ablation, exactly? In short, it's a process where a CO2 laser vaporizes a thin layer of material to create a mark. CO2 lasers are great for wood, acrylic, glass, and coated metals. But bare, polished metal? Not so much. CO2's wavelength (10.6 µm) is strongly reflected by most metals, so you need way more power to get a mark—and the results often look washed out.

I didn't know that until I asked a specialist. The reseller who sold us the P-50 admitted, 'That's not our strength. You should talk to someone who does fiber laser marking, not CO2.' He even gave me a referral. And he was right. The fiber laser engraver we ended up buying from another company marked the parts beautifully within two days of setup.

Here's the thing: the reseller lost about $4,000 in commission by telling us the truth. But he won our trust for the next three years of consumables and service contracts.

Lesson three: a supplier who knows their boundaries is more valuable than one who claims to do everything.

The actual numbers

Let me show you the cost summary, because that's what I'd want to see if I were in your shoes.

• Desktop Metal Production System P-50 (including installation, training, 1-year support): $112,000
• Desktop metal CNC machine (from a specialty machining vendor): $24,500
• Desktop fiber laser engraving machine for metal (different vendor): $8,900
• Additive manufacturing MES add-on: $9,000
• Permits, ventilation, and facility upgrades: $7,300
• First-year operating costs (powder, binders, tooling): $18,200

Total first-year investment: $179,900. (I'm rounding, but that's within 2% of my actual tracking spreadsheet.)

In 2024, we saved $31,700 in outside machining and marking costs. That puts us on track for a payback period of about 5.7 years—if we count the total investment. But if we isolate the P-50's impact and ignore the support systems, the printer alone pays back in 3.4 years.

What I'd do differently

I'd buy the MES system at the same time as the printer. That $9,000 should have been in the original quote, not a follow-up purchase after process failures. I also would have researched laser marking wavelengths before asking for quotes. I almost bought a CO2 desktop laser engraving machine for metal and would have wasted $6,000 on something that couldn't do the job.

The bigger picture: knowing what not to promise

The 'professional boundary' lesson kept coming up. The best vendors told us exactly what they couldn't do. The ones who said 'we can handle everything' always had a catch—an expensive integration fee or a factory-certified specialist who had to fly in from another state.

Per FTC guidelines (ftc.gov/business-guidance/advertising-marketing), marketing claims must be truthful and not misleading, and they need substantiation. When a supplier says 'this machine replaces all traditional machining,' there's no FTC rule against the phrase, but the claim has to hold up in practice. In our experience, it didn't. The desktop metal CNC machine didn't replace a full-size mill; it just reduced the work that had to go out. That's a fine claim—and it's true.

I also learned something funny about USPS. When we ship small printed samples to prospective customers, we try to fit them in a USPS large envelope (max 12 in × 15 in, thickness up to 3/4 in), because the rate for a large envelope is $1.50 as of January 2025, versus parcel rates that start around $5.50. That's not a huge part of our budget, but it's the kind of detail that makes a procurement manager happy. I've started checking the USPS specs (usps.com/stamps) whenever we design packaging for a new part.

Anyway, this is the long way of saying: boundaries are good. For us, that meant a portfolio of specialized machines from three different vendors—not one 'universal' system that did everything poorly.

Final takeaway

If you're evaluating desktop metal equipment, don't let a single vendor's sales sheet define your roadmap. List the jobs you actually need to do, calculate TCO, and be honest about your own limits. The companies that admit what they don't do? They're the ones who'll get your repeat business.

Now if you'll excuse me, I have a spreadsheet to update. (Should mention: this year's refresh will be the 7th consecutive year of tracking, and the numbers are finally getting smooth.)