The 90-second check-in: designing the desk for a queue

What actually slows down check-in?

It is rarely typing speed. The common delays are: searching for the reservation, discovering the room is not ready, re-keying payment details, capturing identity data, and explaining a rate or deposit the guest did not expect.

Each of these is a decision that arrived at the desk too late. The reservation search can be pre-empted by an arrival list. The room assignment can be done before arrival. The payment authorisation can be requested in advance. The rate explanation can be in the confirmation.

Which decisions can move before arrival?

Anything that does not require the guest to be physically present is a candidate. The goal is to leave the desk with identity verification, a warm greeting and the key.

• Online check-in with contact confirmation and estimated arrival time
• Pre-authorisation of the card on file
• Room assignment by preference and availability
• Upgrade offers before arrival, not at the counter
• Registration details completed in advance where local law allows
• Group rooming lists loaded and keys pre-made

What should the check-in screen show?

One screen, five blocks, in the order the conversation happens: who the guest is, what they booked, which room, how it is paid, and what the team needs to know. Everything else is a click away, not on the critical path.

How do groups change the design?

A group is not a queue of individuals. It is a rooming list, a master folio and a set of exceptions. The worst design is to process each guest through the individual flow while the coach waits.

The group flow should load the rooming list in advance, pre-make keys, check in in batches, and route incidentals separately from the master account. A dedicated group desk or a pre-arrival session removes the queue from the lobby.

What about identity capture and compliance?

Identity capture is a legal requirement in many jurisdictions, and it is also a delay. The design goal is to capture once, attach it to the profile and never ask again on a repeat stay where the law allows.

Scanning should be fast, should validate the document type, and should not require the guest to hand over a document for longer than necessary. Retention rules should be configured, not left to the desk.

How do you measure whether it worked?

Measure the median and the 90th percentile, not the average. A fast average with a long tail is still a queue. Track arrival-to-key time, queue length at peak, and the share of arrivals completed with pre-registration.

• Median and 90th-percentile check-in time
• Queue length at the 15-minute peak
• Pre-registration share of arrivals
• Repeat-guest recognition rate
• Check-in exceptions per hundred arrivals

How should check-in be designed at different property sizes?

At 10–60 rooms the desk is often one position and the person behind it also answers the phone, handles luggage and sets up breakfast. A queue of four arrivals is a peak. The design should remove work rather than add channels: a printed or on-screen arrival list, rooms assigned before arrival, a card pre-authorisation requested in advance where the guest agrees, and registration details captured once and reused. A full online check-in flow may be less valuable here than a clean list and a key that is ready. The after-hours arrival deserves its own written procedure, because the desk may close at 23:00 and a night porter may handle the handover. Identity capture still has to happen where the law requires it, so the process should say who scans the document and where it is stored. The measure is simple: how long a guest waits at 18:00, and whether the room was ready when they reached it. The arrangement should be written down so a new team member can follow it.

Between 60 and 250 rooms, arrivals spread across a longer day and a 15-minute peak can put eight or ten people in the lobby. Several desk positions and three shifts make consistency the main problem: the same guest should get the same flow at 07:00 and at 23:00. Pre-registration and pre-authorisation absorb much of the counter work, and the arrival list should be reviewed by a supervisor before the peak rather than during it. Groups need a separate flow with a loaded rooming list, pre-made keys and a dedicated position or pre-arrival session, not a faster version of the individual process. VIP and repeat-guest recognition should be handled in room assignment and the pre-arrival brief, so the greeting is the only thing left at the desk. Measure the median and the 90th percentile, and review exceptions weekly. A flow that works only when the queue is short is not a design. The same arrival list should carry the group cut-off times and the rooms that are still not clean at that point.

Above 250 rooms, and in resorts with several lobbies, check-in is a queue-management and staffing problem as much as a screen problem. Coach arrivals, event groups and multiple buildings mean the desk needs batch processes, pre-made keys and a way to separate group traffic from individual guests. Self-service kiosks or mobile key can help at high volume, but only after the underlying flow is clean; otherwise they add a second queue. Identity capture at this scale needs a documented process per jurisdiction and a retention rule configured in the system, not left to individual agents. Language coverage and clear signage matter as much as the software. The front office manager needs arrival forecasts by hour, not just by day, to roster the desk. At this scale the check-in design is judged on the worst 15 minutes of the week, not on a quiet Tuesday morning demonstration. A group running several properties should also agree which parts of the arrival flow are standard and which stay local, so a guest does not meet a different process at each hotel. A short daily review of the previous day’s arrivals, with the time each guest reached the desk, is more useful than a monthly report.

Decide what must happen before arrival, what must happen at the desk and what can happen after. Identity verification and key handover are hard to remove at the desk, though the data capture behind them can move earlier. Payment authorisation, room assignment, upgrade offers and registration details are candidates for pre-arrival. Then decide how groups are separated from individual arrivals, and who owns the arrival list in the last two hours before the peak. Set the measurement before changing the screen: median and 90th-percentile time, queue length at peak, pre-registration share and exceptions per hundred arrivals. The boundary conditions are legal identity capture, room readiness and the after-hours procedure. A property that cannot guarantee a clean room by the promised time should fix housekeeping coordination before buying kiosks. The desk can only be as fast as the work that reaches it. If the queue is caused by room readiness, the fix belongs in housekeeping; if it is caused by payment, the fix belongs in pre-arrival, not in a new screen.

• Which steps can move before arrival without weakening identity checks?
• How are group arrivals separated from individual queues?
• Who reviews the arrival list in the two hours before peak?
• What is the after-hours arrival procedure, and who owns the key?
• Is the target measured at the 15-minute peak or in a demo?