Checking in 500 people without a queue

Registration desks jam for predictable reasons, and almost none of them are the software. Here is what actually determines whether five hundred people are inside in twenty minutes or ninety.

Lanes beat speed

One desk processing three seconds per person is twenty-five minutes for five hundred people — in theory. In practice one person in twenty has a problem, and every one of those blocks the entire queue behind them.

Two changes fix most of it:

Split by first letter of surname, with signs people can read from the back of the queue. Three lanes are three times the throughput and, more importantly, three independent failure points instead of one.

Add a "problems" desk and move anybody who is not a clean scan to it immediately. No ticket, wrong name, paid-but-not-confirmed, walked up without registering — all of it goes sideways rather than holding up the line. This single change matters more than anything else on this page.

Everything that can happen before the day, should

Send the ticket twice. Once at booking, once the evening before. The second one is the one people actually have open when they arrive.

Tell them to screenshot it. Venue Wi-Fi at 8:45am is a coin toss and mobile data inside a conference hall is often worse.

Print the badges in advance, sorted the same way your lanes are split. Handing over a pre-printed badge is a second; printing one is thirty.

What the scanner actually needs to be

A phone camera is enough, and it avoids buying hardware you use twice a year. What matters is not the device but the behaviour:

The answer must be readable at arm's length. The volunteer is watching the queue, not the screen. The person's name should be the biggest thing on it, with the verdict as a coloured badge beside it.

A repeat scan must look calm. Somebody stepping out for tea and back is the commonest scan of the day. If that shows a red cross, volunteers learn to ignore red crosses — and then a genuinely invalid badge walks straight through.

A sound helps more than you expect. One tone for admitted, a different one for refused. The volunteer can keep their eyes on the queue and still know what happened.

Staff it properly

The rule of thumb that works: one lane per 150 expected attendees for the first hour, then collapse to fewer as the rush passes. Most events staff for the average and then watch a queue form in the first twenty minutes, because arrivals are not evenly distributed — they cluster before the opening session.

Brief your volunteers on the three outcomes they will actually see: admitted, already checked in, and not-found. Everything else goes to the problems desk.

Plan for the network dropping

Be honest with yourself about this one. A scanner that needs a connection stops working when the venue Wi-Fi does, and it will happen at the worst possible moment.

Agree the fallback before the day: a printed delegate list, a pen, and a rule that anybody ticked off on paper gets reconciled into the system afterwards. The important part is that this is a decision made in advance, not an improvisation at 9am by whoever is holding the phone.

A system that quietly pretends a scan was recorded when it was not is worse than one that says plainly that it failed — at least the second one tells you to reach for the paper.

The one thing people forget

Have somebody at the door whose only job is to look at the queue. Not scanning — watching. They spot the family that needs three badges, the speaker who should not be queuing at all, and the moment lane two has stalled. That person is worth two extra scanners.