Plan the front door around the average number of arrivals per minute and you will be understaffed for the only twenty minutes that matter. Almost no event sees a steady stream. People come in waves: a coach unloads, a session lets out next door, a talk is about to start, and suddenly half the day's guests want to be inside in the same ten minutes. The door that copes is the one built for that peak, not the gentle figure you get when you divide the headcount by the length of the day.
It is worth getting right because the queue is what people remember first. In ScanQueue's State of Queuing 2026, 67 per cent of attendees said long queues were their single biggest frustration at live events, ahead of weather, parking and food. The arrival is the first thing your guest does and the easiest thing to get wrong, and a slow door colours everything that comes after it.
Plan for the peak, not the average
Start by being honest about the shape of the day. A 600 person conference with a 9 am keynote does not have 600 arrivals spread across the morning. It has perhaps 400 of them between 8.30 and 9.00, because nobody wants to walk in late to the opening. That is your real number: not 600 across three hours, but 400 across thirty minutes, which is more than thirteen people every minute through whatever doors you have open.
Once you know the peak, the staffing question answers itself. Work out how long one check-in actually takes on your setup, multiply by the people arriving in the busiest minute, and that tells you how many lanes you need open at the front. We pulled the maths apart in cutting the queue at the registration desk, but the principle is simple: a desk sized for the average will fall over at the peak, every time.
Make each check-in shorter
The fastest queue is the one that barely forms, and you get there by shaving seconds off the single transaction. Every second you remove from one check-in is multiplied by everyone behind that person. A scan that takes eight seconds rather than a name looked up on a printed list is not a small gain when four hundred people are waiting on it. We explained where that figure comes from in the eight second check-in explained.
The same logic favours a pass over a search. A guest holding a QR code on their phone is checked in by pointing a scanner at it. A guest whose name has to be found, spelled and confirmed on a list is a conversation, and conversations do not scale to a wall of people. Send the pass before the day, make it easy to find in the confirmation email, and the door becomes a scan rather than an interview.
A queue is rarely a people problem. It is an arithmetic problem you can see.
Add lanes that do not need a person
Staff are the expensive way to add throughput, and on a tight morning you may not have enough of them. Self check-in kiosks are the other lever. A tablet on a stand lets a guest who knows their own details check themselves in, which pulls the confident, straightforward arrivals out of the staffed queue and leaves your crew free for the people who genuinely need help. We set out when that pays off in when self check-in kiosks pay off.
Where you put them matters as much as how many you have. A kiosk in the wrong place creates a second queue that tangles with the first. The aim is to split the flow cleanly the moment people come through the entrance, which we covered in where to place kiosks so the queue keeps moving.
Have a plan for the walk-ups
The peak gets worse when people who never registered turn up anyway, because each one is a slower transaction at exactly the wrong moment. You will always have some. The fix is not to turn them away but to take them out of the main flow: a separate lane for new registrations, so the eight second scans are not stuck behind someone typing in their details for the first time. We wrote up how to keep that orderly in walk-up registration without the chaos.
Brief the crew on the surge
A door is only as fast as its slowest decision, and most of those decisions are made by the person on the lane. Before doors open, the crew should know where to send walk-ups, what to do when a code will not scan, who to call for a problem they cannot solve in ten seconds, and that the priority during the rush is to keep the line moving rather than solve every edge case at the front. A two minute brief is the cheapest throughput you will buy all day. We set out a version of it in briefing your door crew before doors open, and the wider shape of the morning in a realistic run-of-show for event day.
Watch the count, not the clock
The last piece is knowing, in the moment, whether the door is winning or losing. A live count of who has checked in against who is registered tells you whether the queue is clearing or growing while you can still do something about it. If the gap is widening, you open another lane or wave a kiosk forward. If it is closing, you hold. The number is there to make the call for you, rather than leaving you to read the length of the line and guess.
The arrival rush is not a problem you remove. It is a peak you plan for. Size the door to the busiest minute, make each check-in a scan rather than a search, add lanes that do not need a person, and keep one eye on the live count, and the wall of people that arrives at nine becomes a wall that is inside and seated by five past, with nobody remembering the queue because there barely was one. CheckInHub runs the whole front door, the scans, the kiosks and the live count, so the peak is something you watch handle itself rather than something you survive.
