I Don't Trust Spec Sheets Anymore—and I Approve Them for a Living

I don't trust spec sheets. Not completely. Not anymore.

That's an odd thing to say for someone who approves them for a living. But I've read enough of them under enough real-world conditions that the skepticism is earned.

I'm a quality and brand compliance manager at Evident. Every month, I review roughly 200 product specifications, manuals, and technical claims before they're allowed to reach customers. In Q1 2025, I flagged 11% of first-round deliverables for performance claims that only held up under ideal lab conditions. Those claims weren't lies. They just weren't representative of how most people actually use the products.

Here's what four years in this role taught me: the gap between a spec sheet and your factory floor is where instrument problems are born. And I'd argue the fix isn't better specs. It's better customer education.

The M30 Sensor That "Should Have Worked"

Let me give you a concrete example. A customer ordered a batch of M30 capacitive sensors for fill-level detection. The datasheet said the sensing range was 20 millimeters. Their installation gave them half that before the sensor stopped triggering reliably.

Was the spec wrong? No. The range was measured with a grounded metal target, at room temperature, with a clean power supply and a short cable run. That's how most capacitive sensor ranges are published. It's the industry convention.

But this customer's silo had moist material, a 25-meter cable run bundled with motor cables, and a metal bracket a few inches from the sensor face. Individually, each of those variables shaves a millimeter or two off the effective range. Together, they wiped out the sensing margin entirely.

So the spec sheet was honest. The application just wasn't in the test conditions.

In my opinion, that's not a sensor failure. It's a specification gap. And it's a gap that comes up all the time with capacitive sensors. I've seen it on conveyor lines, tank level monitoring, packaging equipment. Same pattern: published range, real-world mounting, unexpected materials nearby.

The customer's engineer told me the failure cost them a weekend of production and a $22,000 rework. I saw the purchase order for the replacement work myself. The sensor brand didn't matter. The physics didn't care. The spec sheet, read in isolation, led them straight into the gap.

That's why I push our team at Evident to publish the conditions behind every number. If a spec only holds in a specific setup, customers deserve to see that on the same page as the number.

Why C5 Compact Thermal Camera Reviews Aren't Verification

The M30 example is about a spec sheet. But there's a second trap that's arguably more dangerous: online reviews.

Search for C5 compact thermal camera reviews and you'll see a lot of praise. Portability. Thermal resolution. Ease of use. Price. I don't disagree with any of it. I've handled a C5 myself and it's a genuinely capable unit.

But here's what reviewers rarely mention: accurate temperature measurement is the hard part. A thermal camera can produce a beautiful image and still be off by a significant margin if the measurement parameters are wrong.

Emissivity is the classic example. If you point a thermal camera at a polished metal surface with the default emissivity setting, usually 0.95, the camera assumes the surface is a strong emitter. Polished metal is the opposite—it reflects more than it emits. The reading can be off by tens of degrees. That's not a camera defect. That's a setup error.

Yet almost no review you find online discusses emissivity, reflected temperature compensation, or calibration intervals. Reviewers test the ergonomics, the menu structure, the battery life. On day one. In a controlled environment. That's useful information, but it doesn't tell you what the camera will read on day 400, after a year of field use, measuring a stainless steel pipe at 60°C, with the sun hitting it from a specific angle.

I've fielded two "your camera is inaccurate" complaints myself in the last five years. Both times, the camera was within calibration. Both times, the user had missed a measurement parameter. The equipment was fine. The understanding wasn't.

Reviews don't handle that case. There's no conclusion to a forum post that says: "This camera is accurate if you set emissivity correctly, measure the reflected temperature, and recalibrate annually." Verification isn't a review activity. It's a quality activity.

The Underrated Resource: The Manufacturer's Own Content

Here's the advice that surprises people: read the manufacturer's educational material. Not the brochure. Not the datasheet. The application notes, the how-to guides, the installation and calibration documents.

At Evident, we build a lot of this content. The Evident website has a ton of application notes covering measurement principles, installation best practices, uncertainty budgets, and calibration intervals. And the Evident app puts a lot of it directly on your phone, which helps when you're standing next to the instrument on a production floor.

I review this material as part of my job. It's not marketing fluff. The goal is to help you specify, install, and maintain the instrument correctly. A customer who reads it files fewer complaints, gets better results, and comes back for the next project. That's the business working as intended.

And it's not only us. Look at how to use an Eppendorf repeater pipette, for instance. The official guidance covers priming, multi-dispensing technique, and how many pre-rinses you need. If you've ever used a repeater pipette without reading that, you were probably introducing small but real volume errors. The manufacturer wrote that document because they want you to get good results. That's the incentive structure you want in a vendor.

I'm not saying every manufacturer gets this right. Some publish glossy content that avoids the hard details. But the ones that publish genuine technical depth are giving you a gift. Take it.

But I Don't Have Time for That

Okay, I can hear the pushback. "I don't have time to read twelve pages of application notes for every instrument I buy."

You're not wrong. None of us have that kind of time.

I once had to choose a supplier in two hours. The purchase order was moved up, and missing the slot would have doubled the lead time. Normally, I'd get three quotes, check references, request a sample. With the clock running, I went with a vendor I trusted and signed a service contract I kinda knew wasn't in our favor.

The equipment passed. The service contract was a nightmare. Calibration turnaround took twice as long as the alternative vendor had quoted, and the warranty extension had a hidden exclusion. It cost $8,000 to unwind. With another 24 hours, I could have called two references and caught it. In hindsight, I should have pushed back and taken the next production slot. But with the project manager watching, I made the call with incomplete information.

Don't hold me to this, but I've come to believe that time pressure is the single biggest enemy of good equipment decisions. Not budget. Not the vendor. The clock.

Here's my shortcut. You don't need to read everything. Find the one question that matters: under what conditions does this spec hold? Thirty minutes on the manufacturer's application note and one email to their support team will get you the answer. That's less time than the two-hour mistake cost me.

Bottom Line

I'll close with my current view: treat the spec sheet as a hypothesis, the online review as a hint, and the manufacturer's technical documentation as the closest thing to a verified answer that you'll get for free.

I'm not saying this because I work at Evident. I'm saying it because I've seen the consequences when buyers skip the learning step. A $22,000 sensor rejection. An $8,000 service contract mistake. A $3,500 thermal camera sitting in a drawer because the accuracy spec didn't match the application.

I've also seen what happens when informed buyers show up: they ask sharper questions, they specify the right conditions, and their instruments do what the datasheet promised. That's the outcome we're all actually after.

So don't buy on specs alone. Don't trust the review roundup. Read the application note. Ask the calibration question. Check the manufacturer's resources—the Evident website and app are a good place to start, but so are the equivalent resources from any serious manufacturer.

Because the spec sheet is a starting line, not the truth. And I'd rather spend ten minutes explaining that than handle another rework born from a spec sheet read in isolation.

The most dangerous words in our industry are: "The spec says it should work." In my experience, they mean someone skipped the next three pages.