Print Is Not One Industry: What a Komori Offset Press, a 3D Printer, and Fabric Inkjet Have in Common (and Don't)

Not long after I took over purchasing for a 130-person company, the product team came to me with one word: 'print.' They needed 5,000 expo leaflets, a fabric banner for a trade show booth, and a dozen prototype knobs for manufacturing review. My first instinct was to find one supplier who could handle every job. After a week of quotes that ranged from 'sure, but we don't do fabric' to 'what is an STL file?', I realized what should have been obvious: printing was never one industry. It's at least three, and the only thing they share is the verb.

What tripped me up: every 'print' request lands in a different world

At first, the pain looks like vendor sourcing. Five thousand brochures belong at a commercial printer. A fabric banner feels like print because it has a logo and color. And prototype parts are '3D prints,' but they go through the same purchasing form, the same approval chain, the same hope of one invoice. It doesn't work that way.

The pain isn't terminology either. The pain is that you are not buying a file and a page. You're buying a material transformation. For a high-volume paper job, sheetfed offset is the tool that makes financial and quality sense. A Komori offset press is one of those tools. But if the final item is a functional part with complex geometry, no plate cylinder can grow layers of melted filament. The search term '3d printer stl files' describes a different workflow entirely.

I know this now. It took one wasteful RFP to learn it.

The real problem: the screen makes every output look similar

When I looked at print purchasing through the lens of an office administrator, I kept comparing PDFs and image files. A brochure layout, a banner JPEG, and an STL model all look like 'artwork' on my monitor. That visual similarity hides real process gaps.

Offset lithography transfers an inked plate image onto paper at high speed. A used Komori offset printing machine, if it is maintained, can be a workhorse for 10,000 identical sell sheets. But the question is always the substrate. Paper is not fabric, and neither is hot plastic. If you need 10,000 sell sheets, a fused-filament printer is the wrong speed. If you need 10 product samples, an offset press is the wrong geometry.

And then there is fabric. The exact question 'can i print on fabric with an inkjet printer' appears in our office search history at least once a quarter. The honest answer is yes, for small samples, if the fabric has an inkjet coating. Standard desktop inkjet ink isn't made to survive repeated washing, and that limitation matters. It makes inkjet-on-fabric a proofing or prototyping method, not a production process.

The cost of missing the distinction: rework, rush charges, and awkward phone calls

In 2021, I submitted a request to a local print shop for 'samples in three colors.' The file was an STL model. The sales rep asked a fair question: which process should they use for that file? I had assumed that a printer could 'just print it.' They didn't take the job, which was generous of them. We lost three days before we found a 3D-printing service.

Three days sounds minor until the project has a drop-dead date. We process about 70 purchasing requests every year that get labeled 'print.' Multiply even a one-day routing mistake by that number, and the cost shows up as missed internal deadlines.

There is a financial side too. Public price sheets from major online printers, which I checked in January 2025, still show offset plate setup around $15 to $50 per color. Rush service runs roughly 50-100% over standard. When the wrong process is used first, you pay setup once, scramble through a rush markup, and then pay again for the correct process. One marketing project did exactly that: a $600 sample order turned into a $1,800 problem because nobody asked whether we needed paper, plastic, or cloth at the start.

The clearest lesson arrived in 2023. A designer requested '25 gradient-color samples' and added 'rainbow 3d printer filament' in the notes. The file was an STL model. The request landed with a commercial sheetfed shop that runs a used Komori offset printing machine. They answered that they don't make plastic parts. Engineering had to correct the routing, and finance had to reissue a PO. Everyone stayed professional, but the week wasted was the week we didn't have.

The fix: name material and volume before contacting anyone

The workable solution isn't to make all 'printing' requests converge. It's the opposite. I now stop any request that only says 'print' and ask one follow-up question: What material goes in, and how many units come out?

• Paper or board, high volume, consistent color: Sheetfed offset. In this category, I check whether the vendor has a Komori offset press or similar, and I still ask for a printed sample. The phrase 'komori offset press' tells me about equipment, not about prepress skill.
• Plastic parts or low-volume functional parts: 3D printing. I supply the model, usually as 3d printer stl files, and I specify material. If a visual gradient is the goal, I ask about rainbow 3d printer filament options instead of guessing.
• Fabric, banners, apparel: Textile printing, dye-sublimation, or large-format inkjet. An inkjet printer can handle a coated fabric sample, but a production run belongs with a fabric-focused process.

This is not a technology ranking. Offset, 3D printing, and fabric inkjet are good at different things. I don't expect one supply chain to cover all of them, and I don't want to force it.

For what it's worth, we still buy from multiple specialists. It seems less efficient from a vendor count, but our reorder rate has dropped, and I no longer have to explain STL files to a press technician.

So if someone in your office sends a request with any of these phrases—Komori, used Komori offset printing machine, rainbow 3d printer filament, 3d printer stl files, or can i print on fabric with an inkjet printer—don't start with the machine. Start with the material and volume. Once that's clear, the right process is obvious. The rest is just purchasing.