The kiosk question always arrives as a gut feeling. Someone looks at the guest list, decides three sounds about right, and orders three. Sometimes that is correct. Often it is one too few, and the queue at twenty to nine tells everyone so, or one too many, with a stand sitting idle and an invoice that did not need to exist. The annoying thing is that this is not a judgement call. It is arithmetic, and the arithmetic is simple enough to do on the back of the door brief.
What you need is two numbers: how fast people will arrive at the busiest moment, and how long one person takes at a kiosk. Everything else follows. The reason it feels hard is that most people guess at both numbers instead of reasoning about them, and a guess multiplied by a guess produces a confident wrong answer.
The two numbers that decide it
Start with the peak arrival rate, because the peak is what causes queues, not the average. A 600-capacity conference does not arrive evenly across an hour. The bulk shows up in a tight window before the keynote, and that window is what your kiosks have to survive. If you have run the event before, your check-in timestamps will show you the real curve. If you have not, a reasonable rule is that a large share of guests arrive in the fifteen or twenty minutes before the published start.
The second number is service time: how long one person occupies one kiosk, from walking up to walking away. A self check-in that has been designed well, with a guest scanning their own pass, is fast, but it is not instant. People read the screen, find their pass, occasionally mistype. Counting roughly fifteen to twenty seconds per person at a kiosk is honest. CheckInHub's check-in averages around eight seconds, but that is the scan itself; the human walking up, orienting and walking off is the rest, and you size for the whole interaction.
Size your kiosks for the worst fifteen minutes, not the comfortable average. The average never queues; the peak always does.
Doing the sum
Say your 600 guests arrive with 300 of them landing in the busiest twenty minutes. That is 15 arrivals a minute at the peak. If each person takes 20 seconds at a kiosk, one kiosk clears 3 people a minute. To keep up with 15 a minute, you need five kiosks just to break even, and breaking even means a queue that never shrinks. To actually clear it comfortably, you want a little headroom above that.
Laid out, the relationship is easy to read:
| Peak arrivals/min | Kiosks (20s/person) | Result |
|---|---|---|
| 6 | 1 | Keeps pace, no slack |
| 6 | 2 | Comfortable, short waits |
| 12 | 3 | Keeps pace, a queue forms |
| 12 | 4 | Comfortable |
| 15 | 5 | Bare minimum |
| 15 | 6 | Comfortable |
The lesson hiding in the table is that the jump from one kiosk to two does more for a small event than the jump from five to six does for a large one. The first spare kiosk removes the single point of failure and absorbs the natural unevenness of arrivals. That is often the most valuable unit you will deploy, which is part of when self check-in kiosks pay off in the first place.
Where the simple maths breaks
The arithmetic gives you a floor, not a final answer, because real doors are messier than the model. A few things push your number up.
- Mixed traffic — if some guests need help, a name change or a printed badge, those interactions take far longer and clog a kiosk; consider a staffed lane for them so the self-service kiosks stay fast
- Layout — two kiosks crammed together share a single approach and queue as if they were one; spread them out so each pulls its own line, which is the heart of where to place kiosks so the queue keeps moving
- First-timers — an audience new to self check-in is slower for the first few minutes; a greeter pointing people to free kiosks recovers most of that
- Failure — if you have exactly enough kiosks and one drops off the wifi, you are now under-provisioned at the worst moment; a spare is cheaper than the queue
That last point is worth dwelling on. Sizing to the exact minimum assumes everything works, and doors are precisely where things do not. One unit of headroom turns a fragile plan into a calm one.
A quick way to sanity-check the order
Before you confirm the hire, run the numbers backwards. Take the kiosk count you were about to order, work out how many people a minute it clears, and ask whether that keeps up with your real peak. If the honest answer is "only if everyone is quick and nothing breaks", add one. If it is "easily, with room to spare", you may be over-ordering, and the spend is better put into a greeter who keeps the lines balanced.
The point of the maths is not precision to the second. It is to replace a gut feeling with a defensible number, so that when the queue does or does not form, you know why, and next year you adjust from evidence rather than ordering three again out of habit. A door that was sized on arithmetic is a door that stays calm at twenty to nine, which is the only test that matters.
