Security features on a card: what actually deters, and what just looks expensive

Security features get specified the way options get ticked on a car. A hologram here, some microtext there, UV because it sounds serious. Then the card goes into service and nobody ever looks at any of it.

A security feature is not a property of a card. It is a property of a card plus somebody checking, and if the second half is missing the first half is decoration. That is the idea this guide is built around.

It is a more useful idea than it first appears, because it turns an open ended shopping question into a closed one. Instead of asking which features are best, which has no answer, you ask who is going to look at this card, with what, and what they would do if it looked wrong. Those three answers eliminate most of the catalogue in about a minute.

It is written for whoever is specifying: a trade printer, a distributor, an integrator, or a programme owner deciding how much protection a credential actually needs.

01. Start with the threat, not the catalogue

Before choosing a feature, answer two questions. Who would copy this card, and what would they gain?

A gym membership copied by a member's friend is a threat with a low payoff and a casual attacker. A staff credential that opens a stock room is a different problem. A card that carries a stored balance is different again. Each of those calls for a different answer, and the middle one is where most programmes sit.

The second question is who would notice. A card presented to a reader is never looked at by a person. A card presented at a desk always is. Printed security features only matter where a human being is in the loop, and on a pure access control credential the real protection is the chip and its cryptography, not anything on the surface.

There is a third question worth adding, because it is the one that turns a feature into a control: what happens when somebody sees a card that fails the check. If the answer is that a member of staff would not know what to do, the feature is not doing security work. A hologram is only a deterrent where somebody is expected to refuse the card, and that expectation has to exist somewhere in writing.

The two questions that settle the specification

What does a copy get someone, and who is going to look at the card? If the answer to the second is nobody, spend the budget on the chip instead.

02. The three tiers

Every feature belongs to one of three tiers, defined by what it takes to verify it.

03. The five features, in plain terms

UV printing

Ink that is invisible in daylight and fluoresces under a UV lamp. Cheap to add, and genuinely effective against a photocopy or a desktop print, because neither reproduces it. It is useless without a lamp at the point of check, and the lamp is the part that gets forgotten.

Microtext

Text printed so small it reads as a rule or a line at arm's length and resolves into words under magnification. It defeats scanning and photocopying because the resolution is not there. It needs a loupe to verify, which puts it firmly in the covert tier.

Guilloche

Fine interlocking line geometry, the pattern on a banknote. Its value is that it degrades visibly when it is scanned and reprinted: the lines moire, thicken or drop out. It is also the cheapest of the five, because it is simply part of the artwork.

Hologram patch

Applied after lamination, changes appearance with the viewing angle. This is the one a person recognises without being trained, and it is the only feature on this list that a card bureau with good equipment cannot reproduce, because the origination is a specialist process.

Hot foil

Metallic foil pressed into the surface, usually a logo. Tactile, visible and hard to imitate with a printer. In practice it is mostly a brand feature that also happens to be hard to copy, and it is fair to specify it for that reason rather than pretend otherwise.

04. What each one actually resists

Fake cards get made three ways. A photocopy, a desktop printer, or a card bureau that has the same equipment a legitimate printer has. The five features perform very differently against those three.

The column that matters is the one on the right, and it is worth being realistic about how often it applies. A bureau grade forgery takes equipment, artwork files and a reason. For a hotel key card or a gym membership that reason almost never exists, which means the first two columns are the real threat and most of the five features handle them. For a credential that controls something genuinely valuable, the third column is the only one worth reading, and it has one entry.

05. Where they sit on the card

The five features occupy different parts of the card, and two of them compete for the same space as the artwork.

The inlay constraint

A hologram patch and a hot foil block are both pressed. Neither may sit over the chip or the antenna, because the pressure damages the inlay. Where the inlay sits is therefore part of the security conversation, not just the artwork one.

What the designer needs to be told

Three things, and they are cheap to say at the start and expensive to discover at proof stage. Guilloche needs an area of the card where nothing sits on top of it, or the pattern is broken and stops doing its job. UV printing needs an area that is light enough for the fluorescence to show, so it does not go over a dark solid. Microtext needs to be set as a rule in the layout, not as body copy scaled down, because type that small has to be drawn for the process.

06. A sensible default

For a credential that is presented to a person, one overt feature plus one covert feature covers the realistic threat at a sensible cost. In practice that is usually a hologram or a foil mark that anyone can recognise, plus UV printing that a supervisor can check when something looks wrong.

For a credential that is only ever presented to a reader, spend the money on the chip instead. A DESFire credential with diversified keys is worth more than every printed feature on this page combined, and the AES migration on hotel lock estates is the industry acting on exactly that.

One more consideration belongs in the default: whatever is specified has to be reproducible on the next order. A hologram design, a foil die and a guilloche file are all things that have to exist at the factory when batch two runs, and a card that arrives with a slightly different hologram is a card that fails its own check. Freezing the specification matters more on a secured card than on a plain one.

07. Three worked specifications

The same reasoning, applied to three credentials that a distributor might be quoting in the same week.

A hotel guest room card

Nobody inspects it. It is tapped on a lock and returned at checkout, and the only thing standing between a copied card and an open door is the chip. Specify: no printed security features, and spend the budget on moving the estate to an AES credential. A guilloche element in the artwork costs nothing extra and is worth having for the look, which is a legitimate reason to include it as long as nobody is told it is doing security work.

A staff credential with a photograph

Checked by a human being, sometimes by a supervisor who has been told what to look for. Specify: a hologram patch over the photograph edge, plus UV printing somewhere a supervisor can find it, plus a lamp behind the security desk. The hologram makes an improvised copy obvious at arm's length, and putting it across the photograph edge means a swapped photograph breaks it visibly.

A membership card that carries value

Presented at a counter, and a copy is worth money. Specify: a custom hologram rather than a generic one, a guilloche background, microtext in a rule, and a sequential number that the system can check against a record. The number is the part that does the real work: it turns every presentation into something that can be verified against a database rather than against an opinion.

The pattern across all three is the same. Decide who checks, decide what they do when the check fails, and let those two answers choose the features rather than the price list.

08. Where Kaway fits

We manufacture for the trade and do not sell to your customers. UV printing, microtext, guilloche, hologram application and hot stamping are run on our own floor, on PVC and recycled PVC and, for engraving and grain effects, on wood and bamboo.

Minimum order is 500 units, standard production thirteen calendar days from artwork approval to dispatch, and quotes come back the same working day, including the ones where the honest answer is that a feature is not worth what it costs you.

09. Frequently asked questions

Which single feature gives the most protection?

A hologram patch, because it is the only one on the list that cannot be reproduced with printing equipment and the only one an untrained person recognises immediately.

Is UV printing worth it?

Only if somebody has a lamp and uses it. It is cheap, so it is often specified anyway, but a UV mark nobody ever looks at protects nothing.

Can security features go on a wooden card?

UV, foil and print based features, yes. Microtext is harder because the grain interferes at that scale, and a hologram on wood is possible but rarely what a wooden card is chosen for.

Do these features affect the chip?

Only through pressure and position. Foil and hologram application must avoid the inlay area, which is a layout question settled at the quote.

What about serialisation as a security measure?

A unique number is an audit tool rather than a deterrent: it tells you which card was used, not whether it is genuine. It works well combined with an overt feature, and how to specify numbering properly is covered separately.

Can we have a custom hologram rather than a generic one?

Yes, and it is worth the difference where the card matters. A generic hologram is available to anybody, so it proves only that somebody bought a hologram. A custom origination carries your own design, which is what makes it hard to obtain. It is a one off cost that makes sense across a programme rather than a single order.

Do security features make a card less durable?

Not meaningfully. UV ink, guilloche and microtext are part of the print and sit under the overlay like everything else. A hologram and a foil block sit on or near the surface and can show wear over years, which is a consideration on a credential carried daily rather than one kept in a wallet.

How many features is too many?

More than two on a card nobody inspects closely is usually money spent on the wrong half of the problem. Two well chosen features that staff actually know how to check beat five that nobody has been told about.

Do security features work on a fob or a wristband?

Barely. There is almost no flat printable area on a moulded fob, and a wristband is not inspected the way a card is. On those bodies the equivalent controls are a laser engraved serial and a closure that cannot be transferred.