Inkjet vs. Laser Marking: What I'd Actually Write Into the Contract

The short answer, before the explanation

If you're marking wood, acrylic, leather, glass, or coated metal at under 50,000 units a year, a CO2 laser marking system will almost always give you the lowest cost per mark that survives handling, shipping, and a customer's dishwasher. If you're applying date codes, lot codes, or serialized variable data to curved, flexible, or heat-sensitive packaging at line speed, high-resolution piezoelectric inkjet is still unbeatable — no laser replaces it.

The decision is about the substrate and the failure mode, not the technology. I got that backwards on my first specification, and it cost us $22,000.

What follows is the reasoning, the specifications I now write into contracts, and the parts where the sales literature and the shop floor disagree.

Who's writing this

I'm the quality and brand compliance manager at a small-batch housewares and furniture manufacturer. I review every marked, engraved, or printed item before it ships — around 400 units a month, roughly 4,800 a year once you count the private-label line. In our Q1 2024 quality audit, the rejection rate on first deliveries was 9%, and most of those rejections were mark defects, not product defects. Wrong depth, wrong contrast, drift across a production run.

I've specified coding and marking equipment twice. In 2021 we bought a CO2 laser for our own shop. In 2023 I wrote the marking specification for a contract manufacturer running our private-label line — that's the one that went wrong.

Where I got it wrong

When I first wrote that 2023 spec, I assumed the technology choice was the specification. I wrote "CO2 laser marking, 300 DPI equivalent" and let the supplier work out the rest. Five weeks later we had 6,200 units with marks that were perfectly legible on the sample we approved and washed out on production parts.

The difference? The sample was cut from a different ply of the same laminate. Our substrate varied more than our process did. The vendor claimed it was "within industry standard." We rejected the batch, they redid it at their cost, and we missed the retail window anyway.

Now every contract includes a substrate list with grade numbers, a defined contrast target, and a first-article approval process. (Note to self: this should have been in the supplier onboarding doc from day one.)

The three specifications that actually decide your rework rate

1. Substrate, named precisely

"Wood" is not a specification. Cherry and walnut absorb laser energy differently. Bamboo laminate scorches differently than solid bamboo. If your supply chain switches mills — and it will — an unnamed substrate spec means your marks change and nobody notices until a customer does.

The same applies to plastics, and this is where fiber laser expectations get dangerous. A 1064 nm fiber laser passes straight through clear and natural-coloured thermoplastics — acrylic, polycarbonate, unfilled ABS — leaving nothing behind. You get a mark only when the resin is formulated to react: carbon-black filled ABS, laser-additive grades, some dark polyamides. When a supplier tells you they do fiber laser marking on plastic, the question isn't "does it work?" It's "which specific compound, and is it a laser-grade grade?" Get the resin data sheet before you sign anything.

2. Contrast, measured — not "looks good"

For colour-critical marks, the industry benchmark is Delta E < 2. Delta E of 2–4 is noticeable to a trained observer; above 4 is visible to most people. That's from the Pantone Color Matching System guidelines, and it's the number I put in writing now. I've had suppliers offer "Delta E ≤ 3" as a match. Technically true, commercially useless for a brand-critical mark.

Laser marks aren't measured in Delta E at all — contrast is a luminance difference against the surrounding material. Measure it on the actual substrate with a spectrophotometer and set your own threshold. Ours is 25 units of L* against the unmarked surface.

3. Resolution in the unit that matters

Print resolution conventions are well established: 300 DPI at final size for commercial offset work, 150 DPI acceptable for large-format viewed at distance, 170–200 DPI for newsprint. Those are minimums, not targets.

Here's the catch with inkjet. A high-resolution inkjet printer manufacturer will quote resolution at the print head, which is not the resolution on your substrate. Head height, substrate curvature, ink spread, and cure time all degrade it. Ask for a sample at your maximum line speed on your material — not a bench sample at half speed with a hand-held head.

And for laser, DPI is close to meaningless. Character height and stroke width are what you can actually inspect. We specify a minimum character height of 1.5 mm and a minimum stroke width of 0.15 mm for our smallest engraved text, based on what survives a 50-cycle abrasion test.

Where each technology actually lands

CO2 laser marking (10.6 µm). Absorbed by most organics, glass, acrylic, leather, and coated metals. This is where an industrial laser engraver earns its place: fine detail, permanent marks, and no per-mark consumable beyond electricity and extraction. In our own shop we run a Glowforge Pro — a desktop machine, not an industrial marking cell, which matters for the throughput comparison — and it replaced four out of five of our labelling steps. Not the date coding. The branding.

Fiber laser marking (1064 nm). Excellent on bare and coated metals, engineered plastics, and some ceramics. Fast, precise, low maintenance. It is the wrong tool for most consumer plastics, and I've watched two suppliers learn that the expensive way.

High-resolution piezoelectric inkjet. A piezo print head deforms a crystal to eject drops on demand, which is what makes it precise enough for fine text and tolerant of different ink chemistries. Non-contact, so it handles curved, fragile, and uneven surfaces that no laser jig can hold. It's still the only practical option for multi-colour marks and for high-speed variable data on flexible packaging. Mixed feelings here: consumables are a recurring cost and a recurring headache, but when a customer changes their date-code format on a Tuesday, inkjet responds in ten minutes and a laser needs a new program and a re-approved first article.

What the cost comparison actually looks like

Purchase price is the least interesting number. What matters is cost per acceptable mark, which includes:

• Consumables per mark — ink, solvent, make-up fluid, filters, or nothing but electricity
• Fixturing and jigs. This one surprises people. Laser marking a curved part usually needs a custom jig. Budget $1,500–$4,000 per part family as of January 2025, and verify current pricing with your integrator, because it moves.
• Fume extraction for laser, which is not optional for leather or coated metals, and absolutely not optional for anything containing PVC
• Operator time for head cleaning, ink changes, and focus verification
• Rework — the number that actually decides the ROI

When I ran our own numbers, laser won on cost per mark for everything except date coding. But it took about three weeks — or rather, closer to five once you count the jig build and the operator training — before the per-unit cost beat the inkjet it replaced.

The 12-point first-article checklist I built after that 2023 mess has saved us an estimated $8,000 in potential rework over 18 months. Five minutes of verification beats five days of correction.

Where this advice doesn't apply

A few honest limits.

Above roughly 200,000 units a year, throughput changes the answer. At that volume you're not comparing a desktop unit to a print head — you're comparing automated marking cells, and integration cost is what you're actually buying.

If your marks are regulated — UDI, serialization, traceability requirements — the technology decision sits downstream of validation and record-keeping. Laser has a real edge there, because the mark can't rub off and the program is version-controlled.

If you need colour, or marks on fabric that gets folded and compressed, laser isn't in the conversation.

And if your material mix changes every quarter, don't over-invest in either. Keep the substrate list as a living document, re-run first-article approval every time a grade changes, and accept that some years you'll pay for flexibility instead of cost per mark. That's a defensible decision. Skipping the verification is not.