Check-in is the first thing every attendee experiences and the last thing most teams plan. It gets a table, a printer and some volunteers, and then on the morning of the event 600 people arrive in the half hour before the keynote. The queue that follows is not a failure of effort. It is a failure of arithmetic, and the arithmetic is simple enough to do on the back of a badge.
The queue is arithmetic
A queue grows whenever people arrive faster than you can process them, and it keeps growing, every minute, until arrivals slow down. That is the whole model. Two numbers drive it: how fast your kiosks can check people in, and how fast people arrive at the peak.
The first number is published. A realistic figure for a self-service kiosk that prints badges is around 4 badges per minute per station, once you allow for the human part of the interaction, and the matching rule of thumb is one kiosk per 230 people arriving in your busiest hour.1 The two are consistent: 4 a minute is 240 an hour, and 230 leaves a little slack. If your kiosks only scan and do not print, expect them to run faster, and treat these as conservative planning numbers.
The second number is yours to estimate, and it is the one people get wrong, because they plan for total attendance instead of the peak. Here is a worked example. Say 600 people arrive in the 30 minutes before the opening keynote. That is 20 people a minute.
Look at the middle column, because it is the trap. Five kiosks match the arrival rate exactly, which looks efficient on a spreadsheet. In a real lobby it means any single slow guest starts a queue that never catches up. Headroom is not waste. It is what absorbs the guest who cannot find their email.
Work out your own numbers
The worked example is one scenario. Put in your own: how many people arrive in your busiest stretch, how long that stretch lasts, and how fast your kiosks run.
People arrive at 20 a minute. You need 6 kiosks.
| Kiosks | Capacity | Queue at end of peak | Last arrival waits | Verdict |
|---|---|---|---|---|
| 3 | 12/min | ~240 | ~20 min | Falls behind |
| 4 | 16/min | ~120 | ~7.5 min | Falls behind |
| 5 | 20/min | 0 on paper | No slack | No margin |
| 6 recommended | 24/min | 0 | None | Keeps up |
| 7 | 28/min | 0 | None | Keeps up |
The recommendation applies the published rule of one kiosk per 230 people in the busiest hour,1 scaled to the kiosk speed you enter. It assumes arrivals are spread evenly across the peak. Real arrivals bunch, so real queues run longer than this, which is one more reason to keep the headroom.
Designing it to actually be self-service
Kiosk count gets the attention. Three design decisions matter as much.
The QR code arrives before the attendee does. Put it in the confirmation email, then send it again in a reminder the day before, so it is at the top of their inbox when they reach the door. Self-service fails first when people have to search.
One action per screen. Scan, confirm, collect badge. Every extra question at the kiosk, a survey or a session choice, is time multiplied by every person in the queue behind.
Exceptions go somewhere else. This is the one that decides whether the plan survives. Walk-ins, misspelt names, lost codes and changed companies all take minutes, and at a self-service kiosk they take minutes of everyone's time. Give them a separate staffed desk, clearly signed, so the fast lanes stay fast. The same guidance that gives you the kiosk numbers is plain that running on-site badge printing takes more than one person.1 Put those people on the exceptions.
The wait people feel
Some waiting will remain, and here the research is old, well-tested and useful. David Maister's 1985 paper on the psychology of waiting lines sets out principles that still hold:2 occupied time feels shorter than unoccupied time; waits before a process starts feel longer than waits once it has begun; anxiety, uncertainty and unexplained delays all make a wait feel longer; unfair waits feel longer than fair ones; and waiting alone feels longer than waiting in a group.
Each one translates into something you can do in a lobby. Get people scanning quickly, because the wait before the scan is the one that hurts, and the badge printing that follows feels like progress. Show the expected wait on a sign, because a known ten minutes beats an unknown five. Keep the line visibly moving, even if slowly. And keep exceptions out of the main lane, because nothing feels less fair than watching the person in front spend six minutes spelling their surname.
When staffed desks still win
Self-service is the right default for most conferences, but not for every event. Staffed desks still win when check-in involves checking identity documents, when guests are protocol-heavy, such as government delegations and award nominees who expect to be greeted by name, and when your audience is unlikely to arrive with a confirmation on their phone. The honest answer for many events is both: kiosks for the majority, people for the cases that need judgement.
If your guest list still lives in a spreadsheet, check-in is where that will hurt first, because the door works from a copy. We covered that in running event registration on a spreadsheet.
Where Aurentex sits
Check-in on Aurentex reads the same live list registration writes to. Guests get a QR ticket in their confirmation email, the door checks them in by scan or by name search, walk-ins are added on the spot, and the count updates live for everyone watching. Badges print on arrival from the same record, so there is no separate badge file to fall out of date; we set out the hardware costs in on-site badge printing without an agency. On Platinum, check-in keeps working when the venue Wi-Fi drops, with scans syncing once the connection returns.
What we cannot do is change your arrival curve. Nobody can. What we can do is make the arithmetic above the only thing you have to plan for. Your first event is free, so you can try it on real arrivals before you commit to anything.
Common questions
How many self check-in kiosks do I need?
One per 230 people arriving in your busiest hour, at around 4 badge check-ins per minute per kiosk.1 Size to the peak, not the total. For 600 arrivals in 30 minutes, that is 6.
Why do queues build even with kiosks?
Because arrivals outpace capacity. With three kiosks for 600 people in 30 minutes, the queue grows by 8 people a minute to about 240, and the last person waits around 20 minutes. Exceptions in the main lane make it worse.
Is self-service better than staffed desks?
For most conferences, yes, with a separate staffed desk for exceptions. Staffed desks win for ID checks, protocol-heavy guests, and audiences without confirmations on their phones.
How do I make the queue feel shorter?
Keep it moving, show the expected wait, get people scanning fast, and keep it fair by routing exceptions elsewhere. Maister's principles explain why each one works.2