It was a Tuesday afternoon, and the shop floor was quiet for once. The email queue was the usual mix of vendor quotes and spec sheets. And then there was this: a request from a small machine shop—12 employees, maybe—asking if we could help them interpret a calibration sheet for their Rice Lake weighing system.
I remember thinking, this is going to be one of those nickel-and-dime support calls that eats an hour. I could've pushed it down the queue. But something about the subject line—'new to us, want to verify before we trust it'—made me click.
I'm the quality and brand compliance manager at a metrology services company. I review every calibration certificate and measurement report before it leaves the lab. Roughly 200 unique items per year. Maybe 180, I'd have to check. I've rejected about 12% of first deliveries in 2024 due to documentation gaps or tolerance mishandling. That part's not glamorous. But it's necessary.
The Tuesday email that almost got deprioritized
So I opened the email. The machine shop had just bought a used Hexagon coordinate measuring machine (CMM). They also had a floor scale from Rice Lake and a set of micrometers of unknown heritage. They admitted they'd been burned before by bad calibration claims. So they wanted to understand, before hiring us, whether their equipment was even worth sending in.
Attachments: three pictures of a Rice Lake calibration sheet, a photo of a micrometer, and a note asking, 'how do we read this?'
That last question was refreshingly honest. Most people don't ask. They just send stuff in and trust the process.
I pulled up the sheet on my monitor. Rice Lake calibration sheets list a series of applied weights and the instrument's displayed response. The difference is the error. If the error falls inside the manufacturer's tolerance band—usually printed at the bottom of the table—the system is acceptable.
I was about to type a quick explanation when I noticed something odd. The customer's interpretation, based on their notes in red pen, had the sign of the error flipped. They thought the system was out of tolerance at the 20,000 lb point. It wasn't. The displayed value was lower than applied, which on a compression scale means it's reading light—but the tolerance band allows ±0.1% at that range, and 18 lb on 20,000 is 0.09%. Inside tolerance. Their red pen had marked it 'FAIL' because they'd assumed any deviation was bad.
I typed a reply explaining the sign conventions, the tolerance band, and why a 'negative error' on a Rice Lake sheet can still be a pass. Then I added, 'But if you were relying on that scale for one of your production specs, that error might still be critical for your process capability.'
What the Rice Lake sheet was actually telling them
That opened the floodgates. The machine shop owner called me fifteen minutes later. He said, 'Honestly, we're a small operation. We can't afford to send everything out to a lab. We've got a used Hexagon CMM we picked up at auction, and we're trying to decide if we should invest in calibration.'
We talked for thirty minutes. I explained the difference between a HEPA filter efficiency test and... no, wrong topic. I mean, I explained the basics of CMM verification: measuring a certified artifact (like a ring gauge or step block) versus relying on the machine's self-test.
He said they mainly used their CMM for checking small machined parts. They had a micrometer set that came from a retiring machinist's toolbox—might be accurate, might not. I suggested a simple field check: use a gauge block set (certified, or at least known to be stable) and take repeated readings on a single point to see repeatability.
That's where the industrial multimeter service kit came in. Their CMM's touch probe had been sitting in a damp shed for months, and they weren't sure if the electronics were stable. I recommended checking the probe's electrical continuity and the system's grounding with a multimeter. A service kit like the one many metrology labs use includes test leads, a decent clamp meter, and a resistance measurement function. Not as good as a full calibration, but it catches 90% of setup problems before you waste hours on a precision job.
Then we shifted to the micrometer. The photos they sent showed a clearly usable 0-1 inch mechanical micrometer, but the ratchet was sticky. I said, 'A sticky ratchet can cause erratic readings. But if you're measuring something to ±0.002″, you can probably still trust it after a basic cleaning and a check against a known gauge block.'
The phone call felt like a mini-consult. I didn't charge him for it. That's not something we advertise; it's called being a decent human.
A phone call that did more than email
About a week later, they sent in a small steel cube—a test artifact—along with a request: 'Measure this on your Hexagon CMM and tell us if our CMM is close.'
We did. Our lab's CMM measured the cube's flatness at 0.0008″ and the dimension across two faces at 1.0004″ (±0.0003″). They'd measured the same cube on their used CMM: 1.0012″. That's a 0.0008″ difference—well beyond the 0.0002″ repeatability we'd expect from a decent CMM, but for a machine that cost them $8,000 at auction, it's... acceptable? Not great, but usable for some features.
We sent a clear report with our data, their data, and a plain-language verdict: their CMM was measuring slightly oversized but repeatable. We advised checking the probe qualification routine and doing a more formal calibration if they wanted to hold tolerances tighter than ±0.001″.
They took it. We quoted a calibration at a discount because they'd sent the cube to us first—a small gesture that cost us almost nothing. A year later, they became our tenth-largest account. That started with a $200 (half-hour) email conversation about a Rice Lake sheet.
What it taught me about small customers
Looking back, I should have scheduled that call sooner. At the time, I thought the email was a low-value request. I was wrong. Small doesn't mean unimportant—it means potential. That's not a slogan I'd put on a poster, but it's true in practice.
I have mixed feelings about the discount we gave. Part of me sees it as a smart relationship move. Another part knows we undercut our own labor rate on the calibration. But the lifetime value of that customer, as the business-school types say, justified it. (Honestly, I'm just glad the economics worked out.)
Here's the thing: if a customer asks you to explain how to read a calibration sheet, they're telling you they care. They're not trying to be a pain. They're trying to be precise. Treat that seriously, and you've earned trust. Treat it like a burden, and you've earned a bad review.
Per FTC guidelines, we can't claim the CMM is 'guaranteed accurate to 0.0001 inch'—no calibration is absolute. We're careful to state measurement uncertainty in every report. That discipline, not a marketing slogan, is what keeps the brand credible.
And when we ship a calibrated device back, we use a traceable carrier—USPS Priority Mail, not an unmonitored drop-box (18 U.S. Code § 1708; residential mailboxes are legally for USPS mail only). Nobody wants their calibrated micrometer sitting in a box at a doorway for four hours.
That Tuesday email taught me more than the customer learned. I keep the sender's business card on my desk. It reminds me: every inquiry is a chance to either build trust or build a wall. Walls don't pay the bills.
