Buying a Mazak or Any CNC/Laser Machine: The 7-Point Checklist I Built After $140,000 in Mistakes
- Who this checklist is for
- Step 1 — Pull 90 days of actual job data before you look at a single machine
- Step 2 — Build the cost model on your floor rate, not the brochure speed
- Step 3 — Vet the dealer, not just the OEM
- Step 4 — Do a written utility and site survey, not a walkaround
- Step 5 — Force every exclusion into daylight on the quote
- Step 6 — Run your own parts at the demo, and hold the acceptance test to it
- Step 7 — Lock training and the first 90 days before the truck arrives
- Watch-outs and the mistakes I keep seeing
I handle capital equipment and job quoting for a 34-person metal fabrication shop outside Glens Falls, NY. Eleven years. In that time I've personally signed off on six machine purchases that went sideways in some way, and I've documented every one of them — roughly $140,000 in avoidable cost between rework, redone site prep, expedited freight, and one fiber laser that was just plain the wrong size for our job mix.
Who this checklist is for
This is for the person about to sign a purchase order for a laser cutter, a Mazak CNC turning center, a press brake, or a laser welding cell. Call it $80,000 to $600,000, single machine or a two-machine package.
Seven steps. I run them in this order and I don't skip number three anymore, no matter how good the demo was. If you want the short version: the machine is maybe 40% of the decision. The other 60% is the dealer, the floor it sits on, and the person who has to run it at 6 a.m. on a Tuesday.
Step 1 — Pull 90 days of actual job data before you look at a single machine
Not the parts the sales engineer showed you. Not the free laser engraving projects you downloaded to see if the software would open them. Your jobs, from your ERP or your quote log, for the last 90 days.
I want four columns: material and grade, thickness or diameter, bounding box, and quantity per order. Then I sort by revenue and find the 80% line. That line — not the biggest, ugliest part in the shop — is what you're actually buying a machine for.
In 2019 we sized a laser around a 10 mm stainless job that felt like our whole identity. I finally ran the numbers two years later: 8% of revenue. Meanwhile 71% of what we cut was 3 mm and under. We paid for capability we used for one customer.
Checkpoint: you should be able to write one sentence — "80% of our parts fit inside [X by Y] and [Z] mm thick, with [N] setups per shift." If you can't write that sentence, you're not ready to demo.
Step 2 — Build the cost model on your floor rate, not the brochure speed
Spec-sheet speed and shop-floor throughput are different animals. The brochure number assumes perfect nesting, no changeover, no operator walking to the bathroom.
Model it with your numbers:
- Machine hour rate including power, floor space, and financing
- Setup and changeover time, measured in minutes, not estimates
- Material yield from your actual nesting patterns
- Consumables — nozzles, lenses, shielding glass, tips
- Assist gas if you're not on nitrogen generation
- Preventive maintenance contract
When we sized the Mazak CNC turning center in 2022, we modeled 92% utilization because that's what the old lathe ran during busy season. The new machine landed at 68% in year one — different work envelope meant different jobs came to it. The payback period moved out almost seven months. Nobody lied to us; we just modeled the best month instead of the average one.
Checkpoint: if your model only works at 90%+ utilization, it doesn't work. Build the conservative case at 65–70% and see if the project still clears.
Step 3 — Vet the dealer, not just the OEM
This is the step people skip. The machine brand's reputation and the service you actually receive are two different things, and the gap between them is the dealer.
When we looked at a Mazak dealer covering Glens Falls, NY and the wider North Country, the first question I asked wasn't about spindle speeds. It was: how many field service techs do you have within a two-hour drive of my dock, and what's the name of the one who'd be sent to me?
Ask these, and write the answers down:
- How many field techs, total, and where are they based?
- Average first-response time for a down machine — not the target, the actual last-quarter average
- Do they stock spares locally, or does everything ship from a regional hub? Which parts?
- Who performs warranty work — their techs or a third party?
- Are their techs factory-certified on this specific model generation?
In 2021 we bought from a dealer 4.5 hours away with two field techs covering three states. The machine was fine. The service wasn't. Nine days down — no, eleven, I'd have to pull the log — plus $3,200 in expedited freight and lost machine time. The OEM's name on the badge didn't buy us a single hour back.
Checkpoint: you can name the service manager and the tech who'd be dispatched to you. If the salesperson can't answer that on the spot, that's a red flag.
Step 4 — Do a written utility and site survey, not a walkaround
I knew I should have gotten the rigging path measured in writing. I figured, what are the odds the door's too low? The odds caught up with us in September 2022 on a $9,800 redo that included removing a header and re-pouring a pad section.
Get signatures on a single sheet from your electrician, your rigger, and whoever owns the building. Cover:
- Voltage, phase, amperage available at the disconnect — and transformer headroom
- Compressed air quality: dry, oil-free, dew point you can live with
- Chiller location and heat rejection out of the space
- Exhaust and filtration routing, including where the duct actually goes
- Floor loading and whether you need a pad
- Ceiling height with the door open, and overhead clearance for loading
- Door width and height — measured with the pallet, not the machine
- Rigging path from the street to the final spot, including turns
Laser installations also carry safety obligations that aren't optional: ANSI Z136.1 covers safe use of lasers, NFPA 79 covers the electrical side of industrial machinery, and 29 CFR 1910.147 covers lockout/tagout. Your AHJ and your insurer will have their own requirements — confirm them before the machine ships, not after. This is the part where a phone call in month one saves a shutdown in month three.
Checkpoint: one signed sheet, three signatures, no verbal assumptions.
Step 5 — Force every exclusion into daylight on the quote
Our current quote-review template, last revised January 2025, exists because of a line that said "installation included*" and an asterisk that cost us $4,600.
Make the dealer itemize these, in writing, as included or excluded:
- Freight, and who owns the risk in transit
- Rigging and offloading
- Installation and commissioning
- Training — how many days, on-site or at the showroom, classroom or at the control
- Warranty start date and what voids it
- CAM post-processor and software licenses, including renewal
- Fixturing, chucks, tooling packages
- First-year consumables and a spare lens or nozzle kit
- Calibration and any acceptance documentation
Warranty start date deserves its own line. We once had coverage begin 19 days before the machine was powered on, because the clock started at shipment. Nineteen days of a 12-month warranty, gone before the first part was cut.
Checkpoint: zero asterisks. If a quote needs a footnote to be understood, it's not a quote yet.
Step 6 — Run your own parts at the demo, and hold the acceptance test to it
Bring three parts: your highest-volume one, your tightest-tolerance one, and the ugliest one that always causes trouble. Bring your own material. Bring your own CAM output and watch them post it.
The sample wall — the 3D relief laser engraving tile, the filigree sign, the impossible curved tube cut — is marketing. It's real work, but it's work selected to look good, cut at whatever speed made it look good. It tells you nothing about your 3 mm CRS bracket at 400 pieces a week.
Time the setup, not just the cut. Ask them to change from part A to part B while you watch with a stopwatch. That number is the one that decides whether the machine makes money on small batches.
If you're importing rather than buying domestically — and yes, I've been the guy typing laser welding machine kaufen into a search box at 11 p.m. because a German-built unit looked better on paper — add these before any deposit:
- CE versus UL/NFPA 79 compliance for your jurisdiction, and who signs off
- 400V/50Hz versus 480V/60Hz, and what the transformer costs
- Documentation and HMI in English, contractually
- Spare parts lead time from the factory, in weeks
- Who performs service in your region, and what their travel time looks like
- Duties, customs brokerage, and who handles the paperwork
Checkpoint: you've watched your own part get programmed and cut, and you have the setup time in minutes written on your copy of the quote.
Step 7 — Lock training and the first 90 days before the truck arrives
Three names in a calendar: primary operator, backup operator, programmer. Not "we'll figure it out." Names.
Two rules I now enforce. Training starts the week after commissioning, not the same week — the machine isn't stable yet and nobody learns anything while a tech is still chasing alarms. And the first month of work scheduled on the new machine is lower-stakes work, with the old machine still running for at least two weeks if you can possibly keep it.
Checkpoint: names, dates, and a written plan for what gets cut in weeks 1–4.
Watch-outs and the mistakes I keep seeing
Don't let the sales cycle set your timeline. Quarter-end pressure is real, and it's designed to make a 45-day decision feel like a 5-day one. A machine you buy in a hurry runs for a decade.
Buy for the shop you have. Slightly oversized is fine. Sized for the shop you're planning to become in 2031 is how you end up with a machine running 68% utilization and a payback that never quite arrives.
I have mixed feelings about new versus used. On one hand, a used machine gets you capacity at 55–65% of the capital outlay. On the other, you inherit somebody else's maintenance history and you're outside the dealer's priority queue for service. We buy new for the primary production machine and used for the second-op stuff. That's our split, not a rule.
Honestly, I'm not sure why some dealers consistently beat their quoted response times while others consistently miss. My best guess is it comes down to how many techs they keep on staff versus how many territories they cover. I've never gotten a straight answer on how they set that ratio.
And a boundary on all of this: this checklist worked for us as a mid-size job shop with a two-shift schedule and fairly predictable order patterns. If you're running three shifts on a single cell, or you're seasonal with brutal demand spikes, the utilization math in step two changes a lot. I can only speak to what I've actually had to clean up.
There's something satisfying about the moment a purchase stops being a gamble and becomes a project plan. After enough expensive mistakes, that's what this checklist actually is — a way to stop guessing.
All figures above are our internal shop numbers as of January 2025. Machine pricing, dealer territory coverage, and service terms change — verify current quotes, current coverage maps, and current warranty language directly with the dealer and your AHJ before signing.
Leave a Reply