Overbooking

A hotel has 100 rooms for a night and accepts 103 reservations because it expects cancellations, and if all 103 guests arrive, at least three cannot receive their booked room. The extra revenue opportunity comes with a real service obligation and possible relocation cost.

Cloudbeds discusses deliberate hotel overbooking strategies, while Booking.com's partner guidance addresses double bookings and handling them on its platform, but those sources do not establish one policy for every hotel, airline, country or booking contract. Distinguish deliberate overbooking from an inventory-sync failure, since both can leave too many promises but the fixes differ.

Use reliable historical data on cancellations, no-shows and arrival timing, forecast by date, lead time, customer group and cancellation terms, and remember that a broad average may hide a special event or holiday, that promotions can alter cancellation and arrival patterns, and that a last-minute non-refundable booking has a different risk profile. Keep a conservative buffer when data is thin or events are unusual, because a forecast with wide uncertainty should not be treated as precise.

Consider room types or service classes, not only total capacity, since a property may have enough rooms overall but none of the promised type. Check reservations across direct and third-party channels, because slow synchronisation can create simultaneous sales of the last unit, and a channel manager can help but software failures and incorrect settings can still create conflicts.

Define the booking cut-off and capacity controls so staff know when to stop selling or seek approval for an exception, and limit access to override capacity so that every exception has a reason and an accountable owner. Estimate expected benefit and potential harm, including alternative accommodation, transport, refunds, support time and lost trust, and do not assume a refund alone makes the customer whole, since a late-night arrival or essential trip can make relocation especially disruptive.

Review platform and customer terms before deciding how to handle a conflict, because specific obligations vary by channel and place, and where a customer must be moved, find a suitable alternative and communicate clearly without claiming an equivalent room before checking key needs. Keep a record of the affected booking, choices offered and expenses, as it supports follow-up and dispute resolution, and do not promise the same unit to two customers while hiding the conflict.

For airlines, denied-boarding rules and compensation depend on jurisdiction and circumstances, so hotel practices should not be copied onto flights. For healthcare or appointments, capacity may be constrained by staff and safety, not just room count, so more bookings are not always safe.

A simple overbooking rate is accepted bookings above capacity divided by capacity, but it does not quantify customer impact, so monitor the rate of customers actually turned away or moved and separate voluntary changes from involuntary denial, since a guest choosing a better alternative differs from one left without service. Train front-desk or service staff on escalation, because the person facing the customer needs authority to find a practical remedy, and track complaints and repeat booking behaviour after incidents, since short-term room yield is not the only outcome.

After a double booking, audit data quality to find whether the cause was an intentional rule, a manual override or a delayed channel update, and compare forecast against actual arrivals, updating cautiously because a run of lucky nights does not prove the strategy is safe. Review whether an overbooking decision is worth the reputational risk for repeat customers or a small local business, since overbooking is a controlled trade-off only when the business can honour its obligations or deal fairly with the consequences.