How to Spot a Fake ID at the Door: The Checklist That Actually Works

Ask ten door supervisors how to catch a fake ID and nine will describe the same ritual: tilt it for the hologram, run a thumb over the laminate, hold it up to the light. That advice is not wrong. It is just twenty years out of date, and it points your attention at the one part of the card a forger has already perfected.

The counterfeit market has industrialised. A card ordered online today will typically have a convincing hologram, correct card stock, laser-engraved text, and a UV layer that glows in roughly the right places. What it very often does not have is a correctly encoded barcode — because getting the front right is a printing problem, and getting the back right is a data problem, and the two require completely different skills.

That asymmetry is your edge. Here is how to use it.

The eight-second reality

Before the checklist, an honest constraint. On a Friday at 23:40 with forty people in the queue, nobody performs a twelve-point forensic inspection. You have maybe eight seconds per person before the queue becomes a problem and the door team starts cutting corners.

So the checklist below is ordered by catch rate per second spent, not by thoroughness. Do the first three on everyone. Do the rest only when something in the first three snags.

1. Watch the person, not the card

This is first because it is free — it happens while you are still reaching for the ID.

Genuine customers hand over a card and look around. People presenting a fake are running a script, and scripts have tells: the ID already out and gripped tightly, an unprompted explanation (“I just got it renewed”), eye contact that is either absent or unnaturally sustained, a friend hovering slightly too close, or a group that goes quiet as their turn arrives.

None of this is evidence. All of it is a reason to spend the next four seconds more carefully.

Two things are worth training your team to look at specifically.

Compare the bones, not the look. Hair colour, makeup, weight and facial hair all change; jawline, ear shape, nose width and the spacing between the eyes do not. Door staff who reject on “she looks different from the photo” refuse a lot of genuine customers and still miss borrowed cards. Someone who will not remove sunglasses or a cap when asked has told you something.

Read the group, not just the person. Underage groups tend to buy together, from the same site, in the same order. Four cards from four different states with birthdays in the same fortnight is not a coincidence; neither is four cards from one state that all look immaculate and all belong to people who have never been to that state. If the first card in a group snags, check the rest of the group properly rather than waving them through behind their friend.

2. Ask one question that a script does not cover

Not “how old are you” — they rehearsed that. Ask something that requires them to actually know the document:

• “What’s your zip code?” then glance at it on the card
• “What’s your middle name?”
• “What’s your star sign?” (works because they memorised the date, not the implication)

Someone using a borrowed real ID — which is far more common than a manufactured fake, and much harder to detect physically — usually fails this. The card is genuine, so no amount of inspecting it helps. Only the person fails.

3. Turn it over and read the barcode

This is the highest-value ten seconds in door work, and almost nobody does it manually because the barcode is not human-readable.

On every US and Canadian licence there is a PDF417 barcode on the back. It is not a serial number — it is a compressed text file containing the cardholder’s details as structured fields. The AAMVA standard that defines it assigns a three-letter code to each one:

Code	Field
DCS	Last name
DAC	First name
DBB	Date of birth (YYYYMMDD)
DBA	Expiry date
DBD	Issue date
DAQ	Licence number
DAG / DAI / DAJ / DAK	Street, city, state, ZIP
DBC	Sex
DAU	Height

The check is simple: does the data in the barcode match the data printed on the front?

Three things go wrong on fakes, in rough order of frequency:

The barcode is decorative. Cheap fakes print a barcode that encodes nothing, or encodes the manufacturer’s template data. It scans as garbage or does not decode at all.

The barcode disagrees with the front. The forger reused a template and edited the printed side without re-encoding the back. Name matches, date of birth does not. This is the single most common failure on mid-tier fakes, and it is invisible to the naked eye.

The header is wrong. The payload starts with a compliance indicator, ANSI , an Issuer Identification Number for the issuing state, and a version number. Forgers frequently get the IIN wrong for the state on the front, or encode an AAMVA version that the issuing state never used in the year the card was supposedly issued.

You cannot see any of this by looking. You need something that reads it.

The scannable fake — and why a green light is not an answer

There is a fourth case, and it is the one that catches out venues who believe the problem is solved because they bought a scanner.

Encoding a PDF417 barcode is not hard. The format is public, the AAMVA field codes are published, and free encoders will generate a valid barcode from whatever text you feed them. So the better counterfeit operations now print a barcode that decodes cleanly and agrees perfectly with the front of the card — same name, same address, same date of birth, all of it saying 24 years old.

Point a basic scanner at that card and it does exactly what it was built to do: pulls the date of birth out, subtracts it from today, shows green. The door waves them through with more confidence than if there had been no scanner at all, which is the genuinely dangerous part.

What separates a scanner that catches this from one that does not is whether it reads the barcode’s structure or merely its contents:

• Does the Issuer Identification Number match the state printed on the front?
• Is the AAMVA version one that state was actually issuing in the year on the card?
• Are the field delimiters, header length and licence-class formatting right for that issuer, or subtly generic?
• Do the dates cohere — an issue date before the expiry, an expiry consistent with that state’s renewal cycle, a date of birth that fits the licence class?

Counterfeiters get the visible payload right because that is what they are checking their own work against. They get the envelope around it wrong because nobody in the chain is looking at it. That is the same front-versus-back asymmetry as before, one layer deeper.

It is worth knowing what your existing kit does. Plenty of door scanners — and most phone apps built on a generic barcode library — are age calculators with a camera. Ask the vendor directly whether they validate issuer formatting, and treat a vague answer as a no.

4. Check the maths, not the date

Door staff misread dates constantly at 1am — it is genuinely hard under bad light with a queue building. Two habits help:

• Know today’s cutoff date and write it on the podium. If the drinking age is 21, the cutoff is today’s date minus 21 years. Anyone born after that date is refused. You are now comparing two dates instead of doing arithmetic.
• Check the expiry, every time. Expired documents are refused in most jurisdictions regardless of the holder’s age, and an expired card is a strong hint the person is not its owner.

5. Then, and only then, the physical checks

If something snagged in steps 1–4, now it is worth slowing down. Server-training programmes — RAMP, TABC and most seller-server courses — teach this as three passes, and the sequence is worth keeping because each one uses a different sense and takes about two seconds.

Feel it. Genuine cards are polycarbonate or PVC layers fused under heat into a single slab. Counterfeits are usually laminated or glued, so the edge is where they give themselves away: run a thumbnail around it and feel for a seam, a burr, or film beginning to lift at a corner. Most states also laser-engrave some text into the surface — often the date of birth or licence number — so a card that is uniformly smooth where it should have texture, or oddly raised around the photo, deserves the next two passes. Compare against a known-good card from the same state; keep a reference set behind the bar.

Tilt it. Real holograms and optically variable ink shift colour and image progressively as the angle changes. A counterfeit’s overlay is typically a static printed sticker under gloss: it flares in the light but the image underneath does not move. While it is tilted, check the portrait — genuine cards laser-engrave it, so it looks slightly grey and sits in the card, whereas inkjet and dye-sublimation prints look stuck on and catch the light at a different sheen from the surface around them.

Light it. Hold a torch behind the card. Solid polycarbonate diffuses light evenly; a glued fake shows uneven density, pinholes, or dark patches where adhesive pooled. If the door has a UV lamp, genuine cards carry state emblems in the laminate that fluoresce in a known pattern — cheap fakes have none at all, and mid-tier ones tend to have something that glows in roughly the right place but with the wrong shape or the wrong colour.

Three more, when you have the time and the card is still bothering you:

• Ghost image. Most modern licences carry a secondary translucent portrait. Forgers often skip it or place it incorrectly.
• Microprint. Text that looks like a solid line to the eye but resolves into words under a loupe. Very rarely reproduced correctly.
• Font consistency. States use specific typefaces at specific sizes. A wrong 1 or a differently shaped 4 in the date of birth is a common tell, because forgers hand-edit those fields.

The thing nobody tells you

You will not catch every fake, and you are not supposed to.

A genuinely well-made counterfeit — the kind that costs real money and comes with correctly encoded data — will get past you, past a scanner, and past most trained inspectors. Any product or training course claiming a 100% catch rate is selling you a feeling.

What actually protects your licence in most jurisdictions is not perfection. It is demonstrable diligence: evidence that your venue had a policy, that staff were trained in it, and that the policy was followed on the night in question. A refusals log and a scan record are what turn “we check everything” from a claim into a record.

That distinction matters enormously when an authority is deciding whether you were negligent or unlucky. Requirements differ by state and country, so confirm what yours expects — but the shape of the answer is nearly always the same: show me what you did.

The five-minute version for your staff

Print this and tape it inside the podium:

1. Watch the person while you take the card. Tense, rehearsed, over-explaining? Slow down.
2. Ask one off-script question. Zip code, middle name, star sign.
3. Scan the barcode. Compare it to the front. Mismatch = refuse.
4. Compare against the cutoff date. Not mental arithmetic. Check the expiry too.
5. Only if something snagged: feel the edge, tilt for the hologram, light it from behind. Then ghost image, microprint, fonts.
6. Log every refusal. Time, reason, description. This is the part that protects you.

Why we are building a scanner

Step 3 is the one that catches the most and gets skipped the most, purely because reading a PDF417 barcode by hand is impossible and dedicated hardware costs several hundred dollars per door.

That is the gap we are closing: barcode reading, front-to-back comparison, and a timestamped refusals log, on the phone your door staff already have. If that is useful to you, the early access list below is where to say so — and to tell us what your door actually needs, while that still shapes what gets built.