A plastic card is one of the hardest products to proof flat, because almost everything that goes wrong with a card is invisible on a flat sheet: the chip module, the magnetic stripe, the way the edge reads, and how a finish behaves on a laminated surface. This is our home turf — we make cards, and this list is written from the ones that came back.

Why cards break the flat-proof model

A card looks deceptively simple. It is a rectangle. It is the same rectangle every time: ID-1, 85.60 × 53.98 mm, nominally 0.76 mm thick. A designer can lay one out in ten minutes and a flat proof of it looks complete.

But a finished card is a laminate stack, not a print. It is a printed core, usually PVC or a PVC/PET composite, sandwiched between clear overlay films, fused under heat and pressure, then punched. Sometimes it has a milled cavity with a contact module dropped into it. Sometimes it has an antenna and an inlay buried in the core. Sometimes it has a hot-stamped foil, a signature panel, a magnetic stripe, a lanyard slot, laser-engraved numbering. Every one of those is a physical object with a fixed position, and none of them are meaningfully represented by a rectangle on a PDF.

The customer approving the card is usually looking at a flat artboard and thinking about the design. The things that will bother them in six weeks are all structural.

The chip, the module and the cavity

On a contact or dual-interface card, the module sits in a milled cavity at a standardized position on the front. It is gold or silver, it is reflective, it is about the size of a fingernail, and it is not going anywhere. Two things go wrong at proof stage:

Artwork under the module. A photograph, a face, a logo or a critical line of text placed where the cavity will be milled. On the flat proof it looks fine. On the card, a quarter of it is a metal contact plate.
Nobody told the customer the module would be there. The design is approved with clean white space where the chip goes and the customer is genuinely surprised when a shiny square appears in the middle of their brand.

A proof that shows the module in place — reflective, at true scale, in its real position — turns both of those into a two-minute layout conversation before production instead of a rejected batch after it.

Contactless is the opposite problem

On a pure contactless card, the antenna and inlay are laminated inside the core and are invisible in the finished product. Here the risk runs the other way: the customer sees a plain card and does not believe it is an RFID card, or assumes a chip will be visible and asks why it is missing. On dual-interface and on thin or translucent constructions, the antenna trace can shadow faintly through the surface, especially under a large flat area of a pale solid — which is exactly the kind of thing you want the customer to have seen and accepted in advance rather than discovered on a pallet.

Note also that frequency and chip type are specification decisions, not visual ones. A 13.56 MHz contactless card and a 125 kHz proximity card can be visually identical. The proof cannot show you which one you ordered — that belongs in the job spec, on the proof sheet, in writing, next to the artwork.

The back of the card is a functional surface

Card backs carry more mandatory furniture than almost any other printed product, and each element has a defined position:

Magnetic stripe — a fixed band near the top edge on the reverse. Artwork placed under it is simply gone, and text that runs into it gets cut.
Signature panel — an applied, usually textured white or coloured panel; it will not take your background color.
Regulatory, issuer, helpdesk and return text — small type, frequently squeezed into whatever space is left, frequently below sensible legibility once it is printed at true size.
Barcode, QR, numbering or laser engraving — needs quiet zone and contrast that survives a laminate.
Punch slot — a lanyard slot removes a piece of the card. If it lands on a logo, it lands on a logo.

On a flat proof, the back is a second image the reviewer scrolls past. On a proof they can turn over, the back gets read — which is when someone notices that the support number is last year's.

Edges, thickness and the 0.76 mm problem

This is the one that surprises people most. A card is 0.76 mm thick, which sounds like nothing until you look at a dark card edge-on. If the printed core is white PVC and the design is a deep navy or black, the punched edge shows a pale line all the way around. Designers who have never held a production card do not expect it, and there is no way to see it on a flat artboard.

The fixes are all specification choices — a colored or black core, a full-bleed design tolerance, an edge-to-edge printed core, or simply accepting the line as normal — but they are choices somebody has to make knowingly. Showing the real edge at real thickness on the proof is what makes that a decision instead of a complaint.

Thickness itself carries brand meaning too. A 0.76 mm card feels like a credit card. A thinner card feels like a coupon. If a customer has specified a thin construction to save cost, letting them see and judge the profile before the run is worth doing.

Finishes: where cards are won and lost

Card finishing is unusually visual, and unusually badly served by flat proofs.

Gloss vs matte overlay. The same artwork on a gloss laminate and a matte laminate are two different products. Matte mutes saturation and hides fingerprints; gloss lifts contrast and shows every handling mark. Customers frequently pick from a word rather than an image.
Spot UV over matte. The classic premium card look: a matte body with a glossy raised crest or logo. On a flat proof it is a keyline. In the hand it is the entire reason the card feels expensive. This has to be shown moving under light to mean anything.
Hot foil. Gold, silver, holographic. Foil on plastic behaves differently from foil on board, and it reads as a flash of light rather than a flat metallic color — a flat proof showing "gold" as a mustard fill undersells it badly and misrepresents it at the same time.
Embossing and tipping. Raised characters, sometimes color-tipped, that cast real shadow and show through on the reverse.
Frosted, translucent and metallic cores. Where the substrate itself is the design decision.

One honest caveat: a 3D proof represents these finishes as they appear and behave. It is not a calibrated contract-color proof, and PVC printing has its own color behavior. For a critical brand color on plastic, you still want a printed sample.

Hotel keycards, access cards and other high-volume work

Hotel keycards are their own discipline. They are ordered in large quantities, they are the first physical object a guest is handed, and they get reprinted for reasons that have nothing to do with print quality — a rebrand, a phone number change, a new restaurant name on the back, a franchise standard update. That combination makes approval quality matter more than usual: a mistake is not one bad card, it is fifty thousand of them sitting in a store cupboard.

What we see go wrong on keycards specifically: reverse-side wi-fi and dining text that nobody proofed at real size; a property logo approved on screen and then punched through by a lanyard slot; a matte card specified for a brand that expected gloss; a dark full-bleed design ordered with a white core; and multi-property programs where the back is shared but the front changes, and one property's variant slips through with the wrong front. Every one of those is caught by a customer who can turn the card over and look at the edge.

The same applies to access, membership, loyalty and gift cards. Cards that live in a wallet get looked at closely and for years, which is a higher bar than most printed products ever face.

A practical card proofing checklist

Is the chip module shown, at true position and true scale, over the actual artwork?
Has anything critical been placed under the module, stripe, signature panel or punch slot?
Has the reverse been reviewed at real size, including all small type?
Does the customer know how the punched edge will read against their background color?
Is the laminate finish agreed in a form the customer has actually seen, not just named?
Are spot UV, foil and emboss areas shown behaving, not just outlined?
Is the card technology written in the spec independently of the artwork proof?
Is the approval recorded against a specific version of a specific proof?

Proofing for RFID and smart cards