It started with a PDF proof on a Tuesday. Not a dramatic start, but that's how it usually happens. The cover image of a Fujifilm X-T50 camera looked sharp. The spec table for the endoscopy tower was accurate. The explainer under the headline “How Does Fundus Imaging Work?” was clear. The Fujifilm X-M5 camera review next to it? I'd checked every number twice.

And still got the project wrong.

I manage product content for a distributor that sells a wide range of equipment—Fujifilm medical imaging systems, consumer cameras, and a few non-medical lines like mobility scooters. I've been doing this for six years, and in that time I've documented 14 significant mistakes. This one was not the most expensive. But it was the most educational.

The Project That Should Have Been Easy

The client wanted one combined resource: product highlights for the Fujifilm X-T50 camera, a Fujifilm X-M5 camera review, a mobility scooter selection guide, an endoscopy tower datasheet, and a fundus imaging explainer. The idea was that a single PDF would help their customers understand the range.

Big flag. But we said yes.

For two weeks, I checked and rechecked the text. The Fujifilm X-M5 camera review had accurate specs. The Fujifilm X-T50 camera section was correct. The endoscopy tower section listed every major component: processor, light source, monitor, camera head. The fundus imaging article used the phrase “captures the retina” instead of “photographs the back of the eye”—wait, it's actually both. But we made it sound clinical.

Then came the handoff. I exported the PDF. I sent it to the client. The client signed off. And I emailed it to our print vendor.

That's where the story falls apart.

What Actually Went Wrong

The written content was correct. That's the part that still bothers me. The Fujifilm X-M5 camera review was factually clean. The Fujifilm X-T50 camera overview was clean. The endoscopy tower list was clean. The fundus imaging explainer was clean.

What wasn't clean was the file itself.

I had exported the PDF using “web preview” settings to make the file small enough for email. The proof looked fine on-screen. But those settings are not for print. The images in the PDF were effectively 72 DPI, not the commercial print standard of 300 DPI at final size.

Standard print resolution is not a suggestion. It's the industry minimum. If you have a 3000 × 2000 pixel image, the maximum print width at 300 DPI is 10 inches. At 72 DPI, the same image would be about 41 inches wide—which means the file looks crisp on screen but a soft, mushy disaster on the page.

I didn't know that? No, I did. I'd written it in our own style guide. I just didn't check the export settings.

The Color Problem

Then there was the color. The client's brand blue was specified as Pantone 286 C. On our PDF proof, it looked fine. On the printed sample, it was visibly wrong.

Here's the thing about brand colors: Pantone colors don't have exact CMYK equivalents. The conversion for Pantone 286 C is approximately C:100 M:66 Y:0 K:2. But the printed result can vary depending on substrate, press, and calibration. The acceptable tolerance for brand-critical colors is Delta E < 2. Delta E of 2–4 is noticeable to trained observers; above 4, it's visible to most people. Our sample was in the “above 4” range. In other words, everyone could see it. Reference: Pantone Color Matching System guidelines.

It wasn't a Fujifilm problem. It wasn't a color science problem. It was a verification problem. I checked the words and never checked the visual workflow.

An Earlier Clue: The Mobility Scooter Guide

This wasn't my first content mistake. The same month, I wrote a mobility scooter guide. I focused on the obvious things—speed, range, seat width. The question everyone asks is “How fast does it go?” The question they should ask is “Will it turn around in a corridor?”

The mobility scooter had a turning radius that wouldn't fit a standard exam room doorway. A customer service rep caught it before publication, only because she had lived through that exact situation. I didn't know what I didn't know.

Same pattern. We checked the facts we knew to check, and missed the fact that mattered.

Fundus Imaging and the Endoscopy Tower Connection

The fundus imaging article taught me another layer. “How does fundus imaging work?” sounds like a simple explainer. The layperson answer is: a fundus camera sends light through the pupil and captures the reflection from the retina at the back of the eye. But the engineering details—the ring-shaped light path, pupil dilation, focal length—are exactly what a buyer needs to know. The same goes for an endoscopy tower.

An endoscopy tower isn't one product. It's a system. You need the processor, the light source, the display monitor, sometimes a camera control unit, and the cart itself. Most buyers focus on the monitor size and ignore the light source's service interval. The buyer should ask, “What's included in that tower price?”—because the answer changes the budget.

I've learned that the hard way.

The Aftermath and the Checklist

What did this cost? If I remember correctly, the reprint was around $1,400. The embarrassment was more expensive. The client didn't yell, which made it worse. They just said, “This doesn't look like us.”

Did we save money? No. Was it worth the lesson? Jury's still out.

What we did next was better than an apology. We built a print checklist. Not a “check facts” checklist—we already had that. This one is about file specs, color profiles, and physical constraints.

The checklist is boring. That's the point. Fourteen points, including:

  • Confirm export settings: 300 DPI at final size, no web preview presets.
  • Verify all images are embedded and not linked at low resolution.
  • Check brand colors against Pantone reference (Delta E < 2).
  • Ask “what's included?” for every listed system—endoscopy tower, mobility scooter, camera kit.
  • Ask “will this physically fit?” before publishing dimensions.
  • For every “how does X work?” article, have someone with hands-on experience read it.

In the 18 months since, this checklist has caught 47 potential problems. I don't have hard data on industry-wide error rates, but based on our own 14 mistakes, my sense is that at least half are process problems, not knowledge problems. The fix wasn't getting smarter. It was getting systematic. At least, that's been my experience in product content.

The reprint now sits in my office. A little reminder. Five minutes of verification beats five days of correction. Not a catchy motto, but it pays the bills. I think that's the real lesson.

Oh, and the Fujifilm X-M5 camera review and the Fujifilm X-T50 camera guide? They finally ran in print, in the 2025 edition. As of January 2025, the photos were sharp, the colors were right, and the fundus imaging explainer still reads well.

This time, I checked the export settings first.