The Accuracy Checklist: What I Check Before Trusting Any Measurement Reading

I've spent the past 6 years managing a measurement equipment budget of roughly $180,000 at a 200-person manufacturing company. In that time, I've processed more than 60 calibration orders—maybe 65, I'd have to check the spreadsheet—evaluated 12 different calibration vendors, and built a TCO spreadsheet that our QC manager still refuses to look at. This checklist isn't from a white paper. It's from paying the price for skipping steps.

This list is for anyone who signs off on measurement equipment: purchasing managers, QC leads, maintenance supervisors. If you use a Faro laser tracker, a benchtop multimeter, a UV-Vis spectrophotometer, or a pH meter, these six steps will help you validate accuracy claims before those readings become expensive decisions.

Step 1: Confirm Traceability Before You Trust the Reading

The first thing I check on any instrument certificate is the traceability chain. It sounds obvious, but you'd be surprised how many “calibrated” instruments arrive with a certificate that references nothing. I've seen certificates dated 2024 that simply said “equipment passed” with no mention of a reference standard.

What you want is an ISO 17025 accredited calibration certificate that states the reference standards used, ideally traceable to NIST or an equivalent national standard. I look for the accreditation body's logo and the uncertainty values. If those are missing, I send it back. As of January 2025, that policy has kept two instruments with questionable documentation out of our lab.

Step 2: Read the Accuracy Spec as a Conditional Statement

Here's where I see companies waste money—they buy a Faro laser tracker based on the headline accuracy number, then run it at the edge of its operating range and wonder why parts don't fit. The advertised spec, something like 20 µm + 5 µm/m, means nothing unless the environmental conditions and measurement volume match the spec's conditions.

When I audited our 2023 spending, I found we'd paid for a site calibration on a tracker that was being used at an ambient temperature outside its rated range. The instrument wasn't faulty. It was being run outside spec, and the calibration was pointless. Now every operator has a one-page reference listing the conditions under which our tracker's accuracy spec is valid. I'd rather have a slightly less impressive spec with known conditions than a headline number that means nothing on our shop floor.

Step 3: Ask Whether the Instrument Is Actually Due for Calibration

Everything I'd read about calibration said “annual calibration, no exceptions.” In practice, we cut our calibration spending by roughly 22% once we started assigning intervals based on actual use, not calendar dates.

Here's the logic: a UV-Vis spectrophotometer used twice a week for pass/fail checks doesn't need the same calibration cadence as one used for validated release testing eight hours a day. Same principle applies to a pH meter in a maintenance workshop versus one in a GMP lab.

We now use a risk-based matrix:

• High criticality + high use: annual or more frequent calibration
• Moderate criticality: 12–18 months, based on drift history
• Low criticality, stable environment: up to 24 months, with regular performance checks in between

That doesn't mean skipping calibration. It means making calibration intervals a technical decision, not a habit. A Mettler Toledo pH meter doesn't know what day it is; buffer degradation and electrode drift do.

Step 4: Check the Environment Before You Check the Instrument

This is the step most people skip, and the one that's saved us the most rework. If the measurement environment is outside the instrument's operating conditions, the reading is unreliable—full stop. Calibrating an instrument that's used in bad conditions is like putting new tires on a car with a bent wheel.

A few specifics we've learned the expensive way:

• Laser trackers: temperature gradients in the air distort the laser beam path. Even a “stable” 24 °C can be 24.8 °C near the floor and 23.2 °C near the ceiling.
• UV-Vis spectrophotometers: ambient humidity affects optics and electronics. Keep the sample compartment closed, and don't place the unit near a doorway.
• pH meters: buffer solutions degrade after opening. I checked a “fresh” bottle in our lab once—it had been open for 9 months and its pH was off by 0.3 units. That's enough to lose a batch. We calibrate our Mettler Toledo pH meters daily with fresh buffers using a two-point calibration, and if the slope is below 95%, the electrode gets replaced before any measurements are taken.

If you're fighting calibration drift, stabilize the environment first. In Q2 2024, we added a small air conditioner to our measurement room because it was cheaper than one out-of-spec calibration for the Faro tracker. Our operator taped the temperature range to the door, and that note is still there.

Step 5: Use the Instrument's Own Diagnostics

This step is basically free money. Most modern measurement equipment has built-in self-diagnostics that catch a significant percentage of accuracy problems before they become measurement errors. But most companies never enable them.

For the Faro laser tracker, the faro app and the device's self-check routine run through a diagnostic sequence that catches encoder drift and motor stall issues. We make operators run the self-check at the start of each shift. It takes two minutes, and it gives us a digital trail when a calibration vendor asks why we suspect a problem.

For your multimeter, the same principle applies—except the accessorises are the first place accuracy goes missing. If your multimeter accessories, like test leads, clips, and probes, are damaged or contaminated, the meter itself doesn't matter. We check leads against a known reference resistor—my electrical tech calls it “the 99-cent validation”—and it's caught two bad leads in the past year.

Step 6: Document Everything in a Way an Auditor Can Follow

Documentation isn't the exciting part, but it's the part that saves your budget in an audit. I've been through two ISO 9001 recertification audits and one customer audit, and in every case, the auditors went deeper on instrument traceability than on the actual product records.

For every instrument, we track:

• Instrument ID and manufacturer serial number
• Calibration due date (not just “next due” on a sticker)
• The calibration vendor and certificate number
• The acceptance criteria, so you can see at a glance whether the instrument passed

It took me 4 years and about 60 calibration orders to understand that the certificate itself matters less than the documentation trail around it. The certificate tells you what happened on the day of calibration. Your records tell you whether the instrument was used out of tolerance in between. That gap is what causes product failures and rework.

What I Still Don't Know

My experience is based on roughly 60 calibration orders across our manufacturing floor, QC lab, and field service equipment. If you're running a fully credentialed metrology lab with government-grade requirements, your failure modes are different—and probably more rigorous. I can't speak to those.

But for a typical mid-size manufacturer, this checklist is the difference between trusting readings blindly and knowing exactly what your measurements mean. The common thread across all six steps is that accuracy and calibration aren't about the certificate—they're about the conditions, the use, and the documentation that connects them.

On Buying Certainty When It Matters

One last thing. There will come a moment when you need a calibration done fast—when the Faro tracker is down in the middle of a production run and your customer's inspector arrives next week. In July 2024, we paid a vendor an extra $400—no, $450, I'm mixing it up with the invoice from the previous quarter—for a three-day turnaround on a UV-Vis calibration. It wasn't the 3 days. It was the vendor committing to it in writing, and the cert landing in my inbox the same day the instrument shipped back.

That's the thing about calibration timelines: you're not paying for speed. You're paying for certainty. The “cheaper” vendor with a 10–15 day window might hit 8 days. Or it might miss. The one time we gambled on an “estimated” 7-day turnaround, a critical order went out two days late with an out-of-date certificate, and the customer's response cost us far more than the difference in calibration fees.

Rush calibration fees aren't a waste of money. They're an insurance premium against the uncertainty of the alternative. After getting burned twice by “probably on time” promises, I now budget for guaranteed delivery when the schedule matters. For a mid-size manufacturer, that's the difference between a blip and a pivot.