Skip to content
Field notes
OPERATIONS 8 min read

How to Cut Door Wait Times Without Loosening the Age Check

The queue is costing you drinks and drawing complaints, and the obvious fix — check faster — is the one that loses licences. Here is where the seconds actually go at a door, and how to take them back without checking less.


There is a tension at every busy door that nobody solves by being told to try harder.

A long queue costs you money in a way that is easy to underestimate: peak trading is short, and a customer standing on the pavement at 23:30 is not buying anything, may leave for the place down the road, and is arriving at your rope already irritated — which is where a disproportionate share of door incidents begin. Slow queues also spill onto the street, and pavement crowds are what generate the residents’ complaints that put you in front of a licensing committee for reasons unconnected to anything happening inside.

The obvious response is to speed up the check. That is the one option that is genuinely dangerous, because the failure it produces is invisible until it is a citation.

The good news is that the check is not usually where the time goes.

Where the seconds actually go

Stand behind your own door for twenty minutes with a stopwatch — genuinely worth doing once — and the time divides roughly into three, of which only one is verification.

Handling. The card comes out of a wallet sleeve, a phone case, the bottom of a bag. It gets handed over, angled toward the light, turned, handed back, put away. On a cold night with gloves, this is easily ten seconds and it is almost entirely outside your staff’s control.

Deciding. The date is read under bad light and the arithmetic happens in someone’s head against tonight’s date, at 1am, while a queue builds. This is where errors come from as well as delay, and it is the part that varies most between a fresh doorman and one six hours into a shift.

Everything that is not verification. Guest list lookups, ticket checks, arguments about dress code, the group of six where one person is still finding their ID, the friend who wants to talk about why last week’s refusal was unfair.

Most venues attack the middle bucket, which is real but is not usually the biggest. The largest wins are almost always in the third, and the middle one can be more or less eliminated outright.

Take the arithmetic out first

The single highest-leverage change costs nothing.

Write tonight’s cutoff date on the podium. If the age is 21, it is today’s date minus 21 years. Anyone born after that date is refused. Your staff are now comparing two dates instead of calculating an age — a recognition task rather than a maths task, which is roughly the difference between reading a word and spelling it backwards.

Do this tonight, before you buy anything. It removes a class of error and a couple of seconds per patron, and every venue that adopts it wonders why they did not sooner.

Scanning takes the same work further: the device does the comparison, the result is a colour rather than a number, and it is the same at 02:00 as at 21:00. But the podium note captures a good share of the benefit for the price of a marker pen.

Move the check off the door

This is the structural change, and it is where the halving actually comes from.

A door is a single point through which everyone must pass sequentially. Every second spent there is multiplied by the length of the queue. But verification does not have to happen at the point of entry — it only has to happen before entry.

So send staff down the line. Someone works the queue from the back with a phone, checking IDs while people are standing there doing nothing anyway. By the time a patron reaches the rope, they have been verified; the door becomes a formality rather than a checkpoint.

The arithmetic is straightforward. If the door takes fifteen seconds per person, forty people is ten minutes of throughput. Move ten of those seconds into time the customer was going to spend queuing regardless, and the same forty people clear in a third of the time — without a single check being skipped or shortened.

This is the one strategy that genuinely changes the shape of the problem rather than shaving the constant. It also works better with phones than with dedicated hardware, for the simple reason that you can put a phone in the hand of whoever is free tonight without buying another unit.

Two things it needs:

  • A way to mark someone as cleared that survives them walking twenty metres — a wristband, a stamp, a hand signal to the rope. Get this wrong and you have added a step rather than moved one.
  • Devices that share state, so the person at the rope knows what the person down the line found. Otherwise you have two independent checks and no way to reconcile them.

Sort the queue before it reaches you

The third bucket — everything that is not verification — is mostly a sorting problem.

Split the line. Pre-paid tickets, guest list, table service and general entry all have different processing costs, and mixing them means everyone waits at the speed of the slowest transaction. A guest-list lookup on a phone while a general-entry queue moves past is invisible; the same lookup at the front of a single line stops everything behind it.

The same logic applies to the group of eight where one person is digging for their ID. Step them aside and keep the line moving. It feels rude and it is the single most effective thing a door supervisor does on a busy night.

And count with the scan rather than with a clicker, if you can. Occupancy tracking that updates automatically removes a person from the door whose entire job was arithmetic, and it makes the number defensible if a fire officer asks — which a clicker never quite is.

What to look for in a scanner, if you buy one

Judged purely on queue speed, the features that matter are not the ones on the front of the brochure.

No trigger, no tap. The card should read the moment it crosses the frame. A device requiring a button press or a tap-to-focus has added a deliberate action per patron, and at four hundred patrons that is real time.

Readable at a glance, in the dark. The result needs to be a high-contrast colour state legible in peripheral vision, because your doorman is looking at the person, not the screen. If staff have to read text to learn the outcome, the interface has cost you the second the scan saved.

Works offline. Basements, thick walls, three hundred people saturating the venue Wi-Fi. If verification depends on a network round trip, the queue stops when the signal does — and it will, on your busiest night, because that is when the building is fullest. Parsing and age calculation should happen on the device with no connection at all.

Fast enough is fast enough. Any competent scanner reads a barcode in a fraction of a second, and the difference between one implementation and another is imperceptible next to the ten seconds someone spent getting the card out of their wallet. Vendors compete on this number because it is easy to measure. It is not where your queue time is.

What not to do

Do not stop scanning when the line gets long. This is the actual failure mode, and it happens informally without anyone deciding it — the queue builds, the pressure rises, and the back half of a group gets waved through. That moment is exactly when your process needed to hold, and it is the moment a compliance check is designed to find.

Do not skip the person to save time on the card. Automation handles the document. It does not look at the face, and a borrowed genuine ID passes every digital check there is. The reason to speed up the mechanical part is to give your staff more attention for the part only they can do, not less.

Do not let a passback alert become noise. If your system flags repeat scans, set the window so that smokers re-entering do not trigger it constantly — more on that here. Staff who learn to dismiss alerts reflexively have a slower door and a worse one.

The order to do this in

  1. Write the cutoff date on the podium. Tonight. Free.
  2. Split the queue by transaction type. This week. Free.
  3. Step aside anyone who is not ready. Free, and it needs a supervisor willing to enforce it.
  4. Stopwatch your own door for twenty minutes and find out where the time actually goes before spending anything.
  5. Then, if verification is genuinely the bottleneck, scan — and put a scanner down the line rather than only at the rope.

Four of those five cost nothing, and most venues have not done them.

Where we fit

We are building the phone version, and line-busting is the reason the phone form factor matters: extra doors and extra staff on a peak night should be a login, not a purchase order. Offline-first, because a queue that stalls when the Wi-Fi does is worse than no scanner.

Not shipping yet. If you run a door where the queue is the constraint, the early access list below is where to tell us how yours actually works.